EVOLUTION AND
POPULATION GENETICS
OF PROCHLOROCOCCUSMARINUS
by
Ena Urbach
B.S. Molecular Biophysics and Biochemistry
Yale University, 1980
Submitted in partial fulfillment of the
requirements of the
DEGREE OF DOCTOR OF PHILOSOPHY
in Biology
and Civil and Environmental Engineering
at the
Massachusetts Institute of Technology
February 1995
@ Massachusetts Institute of Technology 1995
Signature of Author ,_ _
_
Departments of Biology and
Civil and Environmental Engineering
Certified by
Sallie W. Chisholm, Professor
BiQlogv and Civil and Environmental Engineering
Accepted b)
Departmental Committee on Graduate Studies
Civil and Environmental Engineering
Joseph M.Sussman
Accepted b3
I A- gl
'I.
tal
Departm
r..'I :
i
i
)"
Committee on Graduate Stuaies
Biology
Jo-Ann Murray
EVOLUTION AND POPULATION GENETICS
OF PROCHLOROCOCCUSMARINUS
by
Ena Urbach
Submitted to the Departments of Biology and
Civil and Environmental Engineering
February, 1995, in partial fulfillment of the
requirements for the Degree of Doctor of Philosophy in
Biology and Civil and Environmental Engineering
ABSTRACT
Prochlorococcusmarinus is a tiny, photosynthetic marine prokaryote containing
the unusual photosynthetic pigments divinyl chlorophylls a and b, and lacking
phycobilisomes, the light harvesting apparatus found in most cyanobacteria. The unique
pigments of this organism suggest a phylogenetic affinity with the Prochlorophyta, a
recently described group of (monovinyl) chlorophyll a and b-containing prokaryotes
which also lack phycobilisomes. The Prochlorophyta have been proposed as a
monophyletic group sharing a common ancestry with "green" chloroplasts in algae and
higher plants.
Phylogenetic analysis of relationships among P. marinus,the prochlorophytes
Prochloronsp. and Prochlorothrixhollandica,and other members of the cyanobacterial
radiation indicates that the distribution of prochlorophytes within the cyanobacterial
radiation is polyphyletic, and that none of the known chlorophyll b-containing
prokaryotes is specifically related to chloroplasts. This result, inferred from 16S
ribosomal RNA (rRNA) sequences, has been confirmed by a parallel, independent study
using rpoC gene sequences. The polyphyletic distribution of prochlorophytes and green
chloroplasts among the cyanobacteria implies that photosynthetic systems employing
(monovinyl and divinyl) chlorophyll b arose several times during the evolution of
photosynthetic organisms. However, since the publication of these results it has been
pointed out that they are also consistent with the existence of an ancestral
cyanobacterium containing both chlorophyll b and phycobilisomes, with subsequent loss
of one or the other light harvesting system in all known surviving lineages.
P.marinus is a major constituent of the photosynthetic picoplankton across wide
regions of the tropical and subtropical open oceans, with cell densities of 105 cells/ml
frequently encountered in both the Sargasso Sea and the Pacific. Cultured strains of this
ecologically important organism isolated from Sargasso Sea, north Atlantic,
Mediterranean and Pacific waters were found by phylogenetic analysis of 16S rRNA,
psbB and petB and D sequences to belong to a single lineage within the cyanobacteria.
P. marinusshares this lineage with strains of marine A Synechococcus, a planktonic
cyanobacterium which also shares its open ocean habitat.
An investigation into the genetic structure of P. marinus field populations in
depth profiles from the Sargasso Sea and the Gulf Stream revealed a high degree of
genetic heterogeneity within water samples, detected by partial sequencing of cloned
PCR products containing portions of the single-copy genes petB and D and their
intergenic region, amplified from flow cytometrically sorted cells. Populations within
water samples contained a minumum of six alleles recovered from a sample of 23 clones.
The presence of numerous additional alleles were implied by rarefaction analyses, which
indicated that diversity was incompletely sampled in datasets containing 19 to 28 clones
per water sample. Overlapping sets of alleles were recovered from the two water
columns, from different depths within each water column and from flow cytometrically
distinguishable subpopulations within water samples, suggesting that each of these
populations drew its membership from a single gene pool.
Thesis Subervisor: Prof. Sallie W. Chisholm
Title: Professor of Biology and Civil and Environmental Engineering
ACKNOWLEDGEMENTS
I would like to thank my thesis advisor, Penny Chisholm, for allowing me an
unusual degree of intellectual freedom in pursuing what I wanted to study. Penny's
intellectual courage has allowed her to develop a laboratory with an extraordinary range
of skills and interests. I would also like to thank the members of my thesis committee,
past and present, who have helped me with sharp critical eyes and large amount of
patience: Alan Grossman and Ethan Signer of the Biology Department, Bill Thilly of the
Department of Civil and Environmental Engineering, Bruce Levin of the Zoology
Departement at University of Massachusetts at Amherst (and currently at Emory
University), John Waterbury of the Woods Hole Oceanographic Institution and Mitchel
Sogin of the Center for Molecular Evolution at the Marine Biological Laboratories.
Special thanks to Bill Thilly for his generous contributions of laboratory space,
equipment and supplies and for his intellectual input and to Mitchel Sogin for generous
contributions of his phylogenetic experitise and computer time.
Over the years at MIT I have been fortunate to learn from a number of talented
individuals who have been generous in sharing their technical expertise. Brian Binder,
Phouthone Keohavong, Jasper Reese, Jason Seaman, Michael Way and Sheila Frankel
have all served as both teacher and friend, and without them I would never have been
able to complete this thesis. Samantha Roberts has been a great companion and source of
information as the other "molecular person" in the Parsons lab. I would like to thank
Cremilda Dias for her excellent technical help in obtaining four of the 16S ribosomal
gene seuqneces used in Chapter Two.
I have also had the good fortune to interact with the unflaggingly friendly and
stimulating members of the Chisholm laboratory: Ginger Armbrust, Lana Arras, Kent
Barres, Brian Binder, Jim Bowen, Jeff Dusenberry, Sheila Frankel, Mark Gerath, Karina
Gin, Liz Mann and Lisa Moore.
Finally, I would like to acknowledge the financial support provided by NSF
(BSR-9020254) and by NIH through a training grant to the Department of Biology.
TABLE OF CONTENTS
Page
A BSTRACT ............................................ ................................................ 2
ACKNOWLEDGEMENTS ...................................................... 4
INTRODUCTION ......................................................................................
References ..............................................
.. 11
......................................................
16
CHAPTER ONE:
MULTIPLE EVOLUTIONARY ORIGINS OF ........
PROCHLOROPHYTES WITHIN THE
CYANOBACTERIAL RADIATION
CHAPTER TWO:
PHYLOGENETIC RELATIONSHIPS AMONG ............ 25
CULTURED STRAINS OF
PROCHLOROCOCCUSMARINUS ISOLATED
FROM DIVERSE OCEANIC PROVINCES
20
Abstract ...................................................... ..................................................
Introduction ................................................
.....................................................
Background ......................................................................... ........................
M ethods ..................................
............
.......... .................
Results ...............................................................................................................
Discussion .................................................... ................................................
References ................................................... ................................................
26
26
28
35
37
58
61
CHAPTER THREE: GENETIC DIVERSITY IN NATURAL .....................
POPULATIONS OF PROCHLOROCOCCUS
MARINUS IN THE SARGASSO SEA AND THE
GULF STREAM
67
Abstract ...................................................... .................................................. 68
Introduction ................................................... ............................................... 68
M ethods ...............................................................
.............. ................ 73
Results .......................................................................... ................................ 82
Discussion........................................................................................................ 108
References ................................................
113
CHAPTER FOUR:
EPILOGUE: RESPONSES IN THE LITERATURE ... 116
TO THE PUBLICATION OF CHAPTER ONE:
MULTIPLE ORIGINS OF PROCHLOROPHYTES
WITHIN THE CYANOBACTERIAL
RADIATION
References ....................................................................................................... 120
APPENDIX I:
Page
PRELIMINARY ANALYSIS OF GENETIC ............... 122
DIVERSITY IN GULF STREAM POPULATIONS
OF PROCHLOROCOCCUSMARINUS BY
DIRECT SEQUENCING OF PCR PRODUCTS
AMPLIFIED FROM FLOW CYTOMETRICALLY
SORTED CELLS
Reference .........................................................................................................
APPENDIX II:
130
JUSTIFICATION FOR THE • 3.5% SEQUENCE ......
DIFFERENCE CRITERION FOR SEQUENCES
ASSIGNED TO A SINGLE ALLELE
References ........................................
APPENDIX III:
............ 147
COMPARISON OF NUCLEOTIDE . ......................... 148
DIFFERENCES AMONG CLONES ASSIGNED
TO ALLELE 1 TO KNOWN PATTERNS OF
NUCLEOTIDE SUBSTITUTION ERROR BY
TAQ POLYMERASE: FOR DATA OF
CHAPTER THREE
References ...............................................
APPENDIX IV:
129
.......................................
...... 152
ALIGNED SEQUENCE DATA................................. 153
LIST OF FIGURES
Page
CHAPTER ONE
Figure 1.
Nucleic-acid sequence phylogenies, inferred from 16S .............. 20
rRNA data, illustrating the apparent independent
appearance of chlorophyll b-containing organisms in the
cyanobacterial lineage.
Figure 2.
Phylogenetic tree illustrating relationships among ...................
prochlorophytes, members of the Synechococcus group and
shotgun-cloned sequences from Sargasso Sea
plankton(SAR) and Pacific Ocean picoplankton (ALO).
21
CHAPTER TWO
Figure 1.
Phylogenetic relationships among cultured strains of .........
P. marinus,cloned sequences from Sargasso Sea
picoplankton and other members of the cyanobacterial
lineage, inferred from 16S rRNA gene sequences using
Agrobacteriumtumefaciens as an outgroup.
38
Figure 2.
Phylogenetic relationships among cultured strains of ..............
P. marinus and other members of the cyanobacterial
lineage inferred using psbB sequences.
40
Figure 3.
Phylogenetic relationships among cultured strains of .............. 42
P. marinus and other members of the cyanobacterial
lineage inferred using petB/D sequences.
Figure 4.
Phylogenetic relationships among P. marinus cultured .............. 50
strains and diverse cyanobacterial taxa, inferred by
neighbor joining using 16S rRNA gene sequences.
Figure 5.
Comparison of petB/D intergenic region sequences for .............. 56
cultured strains of P. marinus and other members of the
oxygenic photosynthetic radiation.
CHAPTER THREE
Figure 1.
Depth profiles of temperature and in vivo chlorophyll .............. 74
fluorescence at the Sargasso Sea and Gulf Stream sampling
sites.
Figure 2.
Contour plots for flow cytometric distributions of....................77
P. marinus sorted from Sargasso Sea and Gulf Stream field
populations.
Page
Intergenic region sequences for 71 petBID alleles recovered ...... 86
from the Sargasso Sea and Gulf Stream field samples,
cultured P. marinusstrains Med4, SS120, FP5, MIT9107
and MIT9313, Synechococcus strains WH8103 and
Figure 3.
PCC7002, cyanobacteria Prochlorothrixhollandicaand
Nostoc PCC7906 and chloroplasts from Chlorella
protothecoides,Marchantiapolymorpha and Zea maize.
Figure 4.
Frequency distributions for nucleotide mismatch in all . ........... 91
pairwise comparisons of allele prototype sequences.
Figure 5.
Phylogenetic relationships among 20 petBID alleles cloned ....... 96
from flow cytometrically sorted populations in the
Sargasso Sea and the Gulf Stream, cultured strains of P.
marinus and other members of the oxygenic phototroph
radiation, inferred by neighbor joining using 163
nucleotides at the first two codon positions of petB and
petD.
Figure 6.
Rarefaction curves for clones recovered from a) individual ...... 102
sorted samples, b) lumped samples containing both
members of flow cytometric double populations or
constituents of entire water columns and c) a lumped
sample containing the entire dataset.
Figure 7.
Frequency of Allele 1-like alleles in sorted samples. ............... 111
APPENDIX I
Figure 1.
Autoradiograms from direct sequencing of petB/D PCR ....... 127
products amplified from flow cytometrically sorted P.
marinus from natural populations in the Gulf Stream and
from unsorted, cultured P. marinus MIT9313, isolated
from the 135 m Gulf Stream water sample.
APPENDIX II
Figure 1.
Ratio of third codon position mismatch to all codon .............. 139
position mismatch, plotted against total mismatch for
comparisons of clones with alignable intergenic regions in
the dataset of Chapter 3.
Figure 2.
Optimization of the criterion identifying the upper bound ........ 142
for total sequence mismatch considered indistinguishable
from PCR+SD error.
Figure 3.
Data from Figure 1 divided into two groups by the criterion .... 144
drawn at 3.5% total mismatch.
APPENDIX IV
Page
Figure 1.
16S rRNA sequence data used to infer trees in Chapter ............ 154
Two, Figure 1.
Figure 2.
psbB sequence data used to infer trees in Chapter Two. ........... 162
Figure 3.
petB/D sequence data used to infer trees in Chapter Two. ....... 170
Figure 4.
petB/D sequence data for alleles cloned from flow .............
cytometrically sorted cells from the Gulf Stream and the
Sargasso Sea and from cultured cells, used for analyses in
Chapter Three.
178
LIST OF TABLES
Page
CHAPTER ONE
Table 1.
Evolutionary distance and fractional similarity matrix for .......... 20
16S rRNA sequences from A. tumefaciens and members of
the prokaryotic oxygenic phototroph radiation
Table 2.
Evolutionary distance and fractional similarity matrix for.......21
16S rRNA sequences from A. tumefaciens, members of the
prokaryotic oxygenic phototroph radiation and shogun
clones from marine plankton communities.
CHAPTER TWO
Table 1.
Prochlorococcusmarinus strains used in this study. .................. 30
Table 2a.
Genetic distance and fractional similarity for pairwise ............... 44
comparisons of 16S rRNA sequences.
Table 2b.
Evolutionary distance and fractional similarity at all, first .......... 45
two and third codon positions for pairwise comparisons of
psbB sequences.
Table 2c.
Evolutionary distance and fractional similarity at all, first .......... 47
two and third codon positions for pairwise comparisons of
petBID sequences.
Table 3.
G+C base compositions of sequences used in these ................. 55
analyses.
CHAPTER THREE
Table 1.
Frequency of alleles among cloned amplification ....................
products from flow cytometrically sorted samples.
Table 2.
Pairwise comparisons among alleles identified by ................... 84
subdivision of sets of clones with similar intergenic
regions.
Table 3.
Fractional mismatch at first two codon positions for all ........... 100
alleles not included in the phylogenetic tree with sequences
from cultured organisms, chloroplasts and selected alleles
included in the tree.
Table 4.
Presence of alleles found in multiple samples. ..................... 107
83
APPENDIX II
Table 1.
Theoretical statistics for error in PCR product sequences.......... 135
for the range of reported Taq polymerase error rates and an
estimated range of sequence determination errors.
Page
APPENDIX III
Table 1.
Nucleotide mismatches between clones assigned to Allelel ..... 151
and the Allele 1 prototype sequence, excluding G->A at
position 1243.
INTRODUCTION
G. Ledyard Stebbins has pointed out that "the ultimate aim of scientific endeavor
in any discipline is to obtain facts by reductionist methods and use them to synthesize
broadly integrated theories that provide new insights into the world of nature" (Stebbins
1982). The evolutionary and population genetic studies that comprise this thesis have
been designed to obtain, by the reductionist methods of DNA sequencing and
phylogenetic analysis, information leading to new theories about the evolution of light
harvesting pigments in oxygenic photosynthetic organisms, and about the population
structure of a prokaryotic phytoplankter which inhabits the largest and possibly bestmixed environment on earth, the tropical and subtropical open oceans.
Procholorococcusmarinus is a eubacterial photosynthetic plankter less than 0.8
pm in diameter which contains the unusual photosynthetic pigments divinyl chlorophylls
a and b (chl a2 and b 2), and lacks phycobilisomes, multicomponent protein and pigment
complexes used for light harvesting by most cyanobacteria (Chisholm et al 1988, 1992,
Goericke and Repeta 1992). Upon its discovery in 1988, the unusual pigments found in
P. marinus suggested a taxonomic affinity with the newly recognized Prochlorophyta
(Lewin 1977), photosynthetic prokaryotes containing "normal" chlorophylls a and b (chl
al and bl) and lacking phycobilisomes (Chisholm et al. 1988).
According to the endosymbiotic hypothesis proposed by Mereschkowsky (1905,
1910), plastids of higher plants and green algae originated in a free living prokaryote
containing green pigments, and which came to live endosymbiotically within the
cytoplasm of a eukaryote. Over the course of the symbiosis, the chl a and b-containing
prokaryote degenerated into an organelle incapable of independent survival (Margulis
1970, 1981, Raven 1970). At the time research for this thesis began, the evolutionary
origin of photosynthetic organelles within the cyanobacteria had become well established
by molecular phylogenetic analyses (Fox et al. 1980, Woese 1987, Giovannoni et al.
13
1988, Van den Eynde et al. 1988, Witt and Stackebrandt 1988, Ludwig et al. 1990,
Valentin and Zeitsche 1990a, b), but whether the chlorophyll b photosynthetic antenna
had been "invented" by an intracellular prokaryote containing phycobilisomes or arose by
endosymbiosis of a free-living, chl b-containing prokaryote remained to be established
(Cavalier-Smith 1982). The recently discovered chlorophyll b-containing prokaryotes
Prochlorondidemni (Lewin 1977, Lewin and Cheng 1989), and Prochlorothrix
hollandica(Burger-Wiersma et al. 1986) had been proposed as descendants of the
putative free-living, chlorophyll b-containing ancestor to green chloroplasts, but P.
hollandicahad subsequently been shown to branch separately from cholorplasts in 16S
rRNA and psbA phylogenetic trees (Turner et al. 1989, Morden and Golden 1989a, b).
Chapter One of this thesis contributed to the formulation of new hypotheses about
the origin of chlorophyll b-containing photosynthetic antennae by providing a 16S rRNA
phylogenetic analysis of relationships among the prochlorophytes P. marinus,
Prochloronsp., P. hollandica,chloroplasts and other cyanobacteria (Urbach et al. 1992).
The results were congruent with those of an independent study using rpoC sequences
(Palenik and Haselkorn 1992). Chapter Four is an Epilogue discussing responses in the
literature to this publication.
Chapter Two of this thesis is a phylogenetic study which infers evolutionary
relationships among P. marinus cultures isolated from diverse ocean provinces,
expanding on the previous work. Phylogenetic patterns inferred from three sequences,
16S rRNA, psbB and petBID, are used to address questions of whether oceanic
prochlorophytes derive from multiple lineages dispersed among the cyanobacteria,
whether phylogenetic patterns for the three genes are consistent, and whether they reflect
geographic relationships among sites of P. marinus culture isolation. In addition, this
study provides sequence data which can be used to link the properties of P. marinus in
14
culture to those of populations in the sea and a phylogenetic framework for interpretation
of the evolution of characteristics among P. marinuscultures.
P. marinusis a major constituent of the picophytoplankton in tropical and
subtropical regions of the open ocean, with depth integrated cell abundances in the
Sargasso Sea and north Pacific of 5 x 108 and 2 x 109 cells cm -2 , respectively (Olson et
al. 1990, Cambell and Vaulot 1993) and a global population of approximately 1026 cells
(according to a back of the envelope calculation which assumes a tropical and subtropical
open ocean area of 2 x 1014 m2 , an average P. marinus cell density of 104 cells/ml and a
euphotic zone depth of 150 m). Experiments with cloned, cultured isolates have
established that individual P. marinus genotypes can grow over wide ranges of light
intensity and temperature, but differ in their growth potential at environmentally relevant
extremes of high and low light (Partensky et al. 1993, Moore et al. 1994). In order to
enable future workers to interpret the patterns of distribution and abundance in P.
marinus field populations and to use the properties of P. marinusin culture to predict the
growth and succession of populations in the wild, it is necessary to assess the patterns of
distribution of P. marinus genotypes in natural populations (Wood 1988, Wood and
Leatham 1992, Weisse 1993). Specifically, it is necessary to answer the questions, are P.
marinusnatural populations genetically heterogeneous, and what is the temporal and
spatial scale of genetic variability among populations?
Chapter Three is a direct analysis of P. marinus populations in which the
questions of genetic heterogeneity and differences among populations at different depths
and in different water masses are addressed by sequencing cloned petBID PCR products
amplified from natural populations sorted by flow cytometry. With the basic outline of
the population structure of P. marinus known, it becomes possible to formulate improved
hypotheses about properties and processes involving this important component of the
marine food web.
REFERENCES
Bryant, D.A. (1992). Puzzles of chloroplast ancestry. Curr. Biol. 2:240-242.
Burger-Wiersma, T., Veenhuis, M., Korthals, J., Van der Weil, C.C.M. and Mur, L
(1986). A new prokaryote containing chlorophylls a and b. Nature 320:262-294.
Campbell, L., and Vaulot, D. (1993). Photosynthetic picoplankton community structure
in the subtropical North Pacific Ocean near Hawii (station ALOHA). Deep-Sea
Res. 40:2043-2060.
Chisholm, S.W., R.J. Olson, E.R. Zettler, R. Goericke, J. Waterbury, and N.
Welschmeyer. (1988). A novel free-living prochlorophyte abundant in the
oceanic euphotic zone. Nature, 334(6180):340-343.
Chisholm, S.W., Frankel, S.L., Goericke, R., Olson, R.J., Palenik, B., Waterbury, J.B.,
West-Johnsrud, L. and Zettler, E.R. (1992). Prochlorococcus marinus nov. gen
nov. sp.: a marine prokaryote containing divinyl chlorophyll a and b. Arch.
Microbiol. 157:297-300.
Cavalier-Smith, T. (1982). The origins of plastids. Biol. J. Linn. Soc. 17:289-306>
Fox, G., Stackebrandt, E., Hespell, R., Gibson, J., Maniloff, J., Dyer, T., Wolf, R.,
Balch, W., Tanner, R., Magrum, L., Zablen, L., Blekemore, R., Gupta, R.,
Bonen, L., Lewis, B., Stahl, D., Luehrsen, K., Chen, K. and Woese, C. (1980).
The phylogeny of prokaryotes. Science 209:457-463.
Giovannoni, S.J., Turner, S., Olsen, G.J., Barnes, S., Lane, D.J. and Pace, N.R. (1988).
Evolutionary relationships among cyanobacteria and green chloroplasts. J.
Bacteriol. 170:3584-3592.
Goericke, R. and Repeta, D.J. (1992). The pigments of Procholorcoccus marinus: the
presence of divinyl chlorophyll a and b in a marine prokaryote. Limnol.
Oceanogr. 37:425-433.
Lewin, R.A. (1977). Prochloron, type genus of the Prochlorophyta. Phycologia 16:217.
Lewin, R.A. (1981). Prochloron and the theory of symbiogenesis. Ann. N.Y. Acad. Sci.
361:325-329
Lewin, R.A. and Cheng, L. (1989). Introduction. In Prochloron: a microbial enigma.
Lewin, R.A., and L. Cheng, eds. Chapman and Hall.
Ludwig, W., Weizenegger, D., Betzel, D., Leidel, E., Lenz, T., Ludvigsen, D.,
Mollenhoff, D., Wenzig, P. and Schleifer, K.H. (1990). Complete nucleotide
sequences of seven eubacterial genes coding for the elongation factor Tu:
functional, structural and phylogenetic evaluations. Arch. Microbiol. 153:241247.
Margulis, L. (1970). Origin of eukaryotic cells. New Haven, Yale University Press.
Margulis, L. (1981). Symbiosis in Cell Evolution. Freeman.
Martin, W., Somerville, C.C. and Loiseaux-de Goer, S. (1992). Molecular phylogenies
of plastid origins and algal evolution. J. Mol. Evol. 35:385-404.
Mereschkowsky, C. (1905). Uber Natur und Ursprung der Chromatophoren im
Pflanzenreiche. Biol. Centralbl 25:593-604.
Mereschkowsky, C. (1910). Theorie der zwei Plasmaarten als Grundlage der
Symbiogenesis, einer neuen Lehre von der Entstehung der Organismen.
Biologische Centralblatt 30:278-303; 321-347; 353-367.
Moore, L.R., Goericke, R., and Chisholm, S.W. (1994). The comparative physiology of
Synechococcus and Prochlorococcus: influence of light and temperature on
growth, pigments, flourescence and absorptive properties. Mar. Ecol. Prog. Ser.,
in press.
Morden, C.W. and Golden, S.S. (1989a). psbA genes indicate common ancestry of
prochlorophytes and chloroplasts. Nature 337:382-385.
Morden, C.W. and Golden, S.S. (1989b). Corrigendum: psbA genes indicate common
ancestry of prochlorophytes and chloroplasts. Nature 339:400.
Olson, R.J., Chisholm, S.W., Zettler, E.R., Altabet, M.A., and Dusenberry, J.A. (1990)
Spatial and temporal distributions of prochlorophyte picoplankton in the North
Atlantic Ocean. Deep Sea Res. 37, 1033-1051.
Palenik, B. and Haselkorn, R. (1992). Multiple evolutionary origins of prochlorophytes,
the chlorophyll b-containing prokaryotes. Nature 355:265-267.
Partensky, R., Hoepffner, N., Li, W.K.W., Ulloa, 0. and Vaulot, D. (1993).
Photoacclimation of Prochlorococcus sp. (Prochlorophyta) strains isolated from
the north Atlantic and the Mediterranean Sea. Plant Physiol. 101:285-296.
Raven, P. (1970). A multiple origin for plastids and mitochondria. Science 169:641646.
Stebbins, G.L. (1982). Modal themes: a new framework for evolutionary syntheses. In
Perspectives on Evolution (R. Milkman, ed.). Sinaur pp. 1-14.
Turner, S., Burger.Wiersma, T., Giovannoni, S.J., Mur, L.R., and N.R. Pace. (1989).
The relationship of a prochlorophyte Prochlorothrixhollandicato green
chloroplasts. Nature 337:380-382.
Urbach, E., Robertson, D. and Chisholm, S.W. (1992). Multiple evolutionary origins of
prochlorophytes within the cyanobacterial radiation. Nature 355:267-269.
Valentin, K. and Zetsche, K. (1990a). Rubisco genes indicate a close phylogenetic
relationship between the plastids of Chromophyta and Rhodophyta. Plant Mol.
Biol. 15:575-584.
Valentin, K. and Zetsche, K. (1990b). Structure of the Rubisco operon from the
unicellular red alga Cyanidium caldarium: evidence for a polyphyletic origin of
plastids. Mol. Gen. Genet. 222:425-430.
Van den Eynde, H., De Baere, R., De Roeck, E., Van de Peer, Y., Vandenberghe, A.,
Willekens, P. and De Wachter, R. (1988). The 5S ribosomal RNA sequences of a
red algal chloroplast and a gymnosperm chloroplast. Implications for the
evolution of plastids and cyanobacteria. J. Mol. Evol. 27:126-132.
Witt, D. and Stackebrandt, E. (1988). Disproving the hypothesis of a common ancestry
for the Ochromonas danicachrysoplast and Heliobacteriumchlorum. Arch.
Microbiol. 150:244-248.
Woese, C.R. (1987). Bacterial Evolution. Microbiol. Rev. 51:221-271.
Weisse, T. (1993). Dynamics of autotrophic picoplankton in marine and freshwater
ecosystems. Adv. Microbial. Ecol. 13:327-370.
Wood, A.M. (1988). Molecular biology, single cell analysis and quantitative genetics:
new evolutionary genetic approaches in phytoplankton ecology. In
Immunochemical Approaches to Coastal,Estuarineand Oceanographic
Questions (C.M. Yentsch, F.C. Mague and P.K. Horan, eds.) Springer-Verlag pp.
41-73.
Wood, A.M., and Leatham, T. (1992). The species concept in phytoplankton ecology. J.
Phycol. 28:723-729.
Chapter One
MULTIPLE EVOLUTIONARY ORIGINS OF
PROCHLOROPHYTES WITHIN THE
CYANOBACTERIAL RADIATION
© Nature 1992. Reprinted by permission.
Multiple evolutionary origins of
prochlorophytes within the
cyanobacterial radiation
Ena Urbach, Deborah L. Robertson*
& Sale W. Chisholm
Ralph M.Parsons Laboratory,
48-425 Massachusetts Institute of Technology,
Cambridge. Massachusetts 02139. USA
* Department of Molecular Genetics and Cellular Biology,
University of Chicago. Chicago, Illinois 60637, USA
THE taxonomic group Prochlorales (Lewin 1977) BurgerWiersma, Stal and Mur 1989 was established to accommodate a
set of prokaryotic oxygenic phototrophs which, like plant, green
algal and euglenoid chloroplasts, contain chlorophyll b instead of
phycobiliproteins. Prochlorophytes were originally proposed (with
concomitant scepticism' - ) to be a monophyletic group sharing a
common ancestry with these 'green' chloroplasts". Results from
molecular sequence phylogenies, however, have suggested that
Prochlorothrixr oliandica7 is not on a lineage that leads to plastids s - 2. Our results from 16S ribosomal RNA sequence comparisons, which Include new sequences from the marine picoplankter Prockhorococmn
marinus13"
and the Lissocdiam patella
s
symbiont Prochloron sp." , indicate that prochiorophytes are
polyphyletic within the cyanobacterial radiation, and suggest that
none of the known species is specifically related to chloroplasts.
This implies that the three prochlorophytes and the green chloroplast ancestor acquired chlorophyll b and its associated structural
proteins in convergent evolutionary events. We report further that
the 16S rRNA gene sequence from Procklorococcus is very similar
to those of open ocean Synechococcus strains (marine cluster A
(ref. 16)), and to a family of 16S rRNA genes shotgun-cloned
from plankton in the north Atlantic and Pacific Oceanso '".7
The order Prochlorales subsumes two formally described gen20
era: Prochloron, a symbiont of marine ascidians , and Prochlorothrix, a free-living, filamentous, fresh-water plankter 7 . Like
green chloroplasts, these prokaryotes have appressed thylakoid
membranes containing chlorophylls a and b and lacking phycobilisomes. The recently discovered Prochlorococcus, a freeliving, tiny unicell that is widespread and abundant in the
oceans", shares these traits with its 'relatives', but differs in that
it contains divinyl chlorophylls a and b (exclusively), and arather than 9-carotene'4" 2 .
NATURE
VOL 355
1~N
.'UARY1992
LETTERS TO NATURE
TABLE 1 Evolutionary distance and fractional similarity matrix for 16S rRNA sequences from A. tumefaciens and members of the prokaryotic oxygenic
Dhototroph radiation
1
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
797
885
886
886
910
888
890
884
896
893
910
897
905
883
Agrobacternum tumefaciens
SARIO0
Synechooccus PCC6301
Prochlorococcus mannus
Prochloron sp.
Synechocystis PCC6308
Synechocystis PCC6906
Gloeobacter PCC7421
Gloeothece PCC6501
Myxosarcina PCC7312
Lyngbya PCC7419
OscillatoriaPCC6304
Anabaena PCC7122
Prochlorothrix hollanoica
Cyanophora cyanelle
Marchantia chloroplast
Synechococcus WH7805
Synechococcus WH8103
127
133
130
29
similarity values (xl.000. above) for pairs of 16S riRNAsequences for
Estimated evolutionary distances (xl.000. below the diagona, and fractional
22
Prochlorothrix hollandica.e other prochlorophytes (this work). cyanooacteria 30 (D.L. Distel and J. B. Waterbury, manuscript in preparation). photosynthetic
3
l
22".
tree
, a full-length, shotgun-cloned sequence from the Sargasso Sea' . and A. tumefaciens . used in the construction of the phylogenetic
organelles
23,
in Fig. la. Sequences were aligned according to their conserved primary and secondary structures ano compared by the computer program of Olsen using
788 nucleotides of sequence unambiguously aligned for all organisms considered. Gap values were set at 0.5 nucleotide substitution. Marchantia Marchantia
(isolated from the Sargasso Sea, 30 May 1988, 28* 58.9' N.64" 21.5' W.
polymorpha: Cyanophora, Cyanophora paradoxa. Genomic DNA from Prochlorococcus
33
at a depth of 120 m: ref. 14) was extracted using standard tecnniques from a dilution culture inoculated with an average cell density of one cell per
culture (isolate SSW5. 58% probability of clonality). Prochloron cetis were collected in Fpbruary 1990 from reef flats in the Kamori Channel. Koror. Palau.
West Caroline Islands (7" 25'N. 134* 30'E). Collection and method of isolating the symbiont from the host, Lissoclinum patella are descnribed in ref. 34.
5
genes were
DNA was extracted from frozen cells using standard protocols' . except that the extraction buffer was adjusted to 0.25 M EDTA. 16S rRNJA
amplified from genomic DNA of Prochlorococcususing the polymerase chain reaction (PCR) with the primers PLG1.1 (ACGGGTGAGTAACGCGTRA) and PLG2.1
(CTTATGCAGGCGAGTTGCAGC). designed to be specific for members of the oxygenic phototroph lineage and thus select against sequences from heterotrophic
contaminants in the culture. Prochloron sequences were amplifea using the primers P1 (AGAGTTTGATCCTGGCTCAG) and P2 (CTTGTTACGACTTCACCCC).
which amplify 16S rRNA sequences from all eubacteria. as there was no significant contamination from heterotrophs in this preparation PCRreactions
contained 10 ng genomic DNA. 100 nM each primer, 200 pM eacn dNTP. 2.25 mM MgCI 2 . 50 mM KCI.10 mM Tris (pH 8.4), and 5 units Tao polymerase
covered with 100 pl sterile mineral oil. Cycle parameters were: 94 TCfor 1 min. Tafor i min. and 72 "C
(Perkin Elmer Cetus). Reaction volumes were 200 L1.
for 2 min. Samples were processed for 40 to 50 cycles. T. was 66 'C for the PLG1.1/PLG2.1 primer pair. and 58 *Cfor the P1/P2 pair. PCR products were
purified from preparative polyacrylamide gels by electro-elution (D-gel. EpiGene), and reamplified asymmetrically using 2 nM of one limiting primer. Asymmetric
PCR products from both coding and complementary strands were ethanol-precipitated twice with ammonium acetate and used for dideoxy-nucleotide
sequencing with Sequenase version 2 (US Biochemical Corporatioa, in parallel reactions containing OGTP and dlTP. plus and minus single-strand binding
protein. 1.147 bases of continuous sequence (Escherichia coli 16S rRNA nucleotide positions 128 to 1.312) were determined for Prochlorococcus and
1.377 bases (E coli positions 28 to 1.452) for Prochloron. Sequences are available through GenBank. accession numbers X63140 and X63141.
a
Agrobacter/um
Glosobacter PCC7421
-Synechococcus WH7805
SAR100
Synechoccus WH85103
F
lat_
Agrobacterlum
Gloeobacter PCC7421
Cyanophora cyanelle
Marchantia chloroplast
Syneechococcus WH7805
SARIOG
Synechococcus WH8103
PROCHLOROCOCCUS
Synechooccus PCC6301
___
PROCHLOROTHRIX
PROCHLOROCOCCUS
SAnabaena PCC7122
--
Cyanophora cyanelle
Marchantia chloroplast
Lyngbya PCC7419
Oscillatoria PCC6304
Gloeothece PCC6501
Synechocystis PCC6906
PROCHLORON
Synechocystis PCC6308
hMyxosarcina PCC7312
0
a
nlo
a0s15
Fixed point mutations per sequence position
FIG.1 Nucleic-acid sequence phylogenies, inferred from 16S rRA data,
illustrating the apparent independent appearance of chlorophyll b-cortaining
organisms in the cyanobacterial lineage. The phylogenetic position of a
19
full-length, shotgun-cloned 16S rRNA gene from Sargasso Sea plarKton
(SAR100) is also shown. A sequence for the freshwater Synecnaococus
strain PCC6307 (D L.Distel and J.B. Waterbury. manuscript in precarationl
Synechococcus PCC6301
1- PROCHLOROTHRIX
I--
-
4
Lyngbya PCC7419
Oscillatoria PCC6304
,Synechocyotis PCC6906
.
L_
Gloeothece PCC6501
Anabaena PCC7122
PROCHLORON
LSyechocystas PCCC6308
-
Myxzosarclna PCC7312
branches immediately to the left of Prochlorococcus with 100% bootstrap
confirmation in both distance matrix and parsimony trees (not shown).
Numbers inoicate per cent confirmation by bootstrap analysis, summarizing
the tree output from 100 resampled datasets, of the phylogenetic group to
the right. Only values greater than 50% are labelled. a Distance matrix
analysis. Horizontal distances are proportional to evolutionary distances
analysis of the same data using the heuristic tree
(Table 1).b. Parsimony
I
search option of PAUP". An equally parsimonious tree reverses the relative
positions of Prochlorococcus and Synechococcus WH7805.
NATURE - VOL 355 • 16 JANUARY 1992
LETTERS TO NATURE
TABLE 2 Evolutionary distance and fractional similarity matrix for 16S rRNA sequences from A. tumefaciens. memoers of the prokaryotic oxygenic
phototroph radiation and shotgun clones from marine plankton communities
1
. Agrobactenum tumefacens
2
782
3
808
967
4
821
954
987
5
797
918
925
931
6
812
964
977
977
935
7
805
912
919
932
974
942
2. SAR6
3 SAR7
4.SAR100
5. Synec~ococcus PCC6301
6. Prochlorococcusmarinus
,.Prochlorothrixhollandica
258
222
205
236
217
226
33
47
87
37
94
13
79
23
86
72
23
72
68
27
61
8. Synechococcus WH7805
234
51
37
37
83
13
75
9. Synechococcus WH8103
3. AL037
217
226
51
30
16
30
16
30
90
68
20
7
82
61
8
799
951
964
964
922
987
929
9
812
951
984
984
915
981
922
981
20
20
10
805
971
971
971
935
994
942
981
974
26
Estimated evolutionary distances (x1.000, below the diagonal) and fractional similarity values (x1,000. above) for pairs of 16S rRNlA
sequences for
ProchNorothrixg.
other prochloroohytes (this workl. cyanobacteriaX 1D.L. Distel and J. 8. Waterbury, manuscript in preparation), shotgun-cloned 16S rRNA
geres" 7- 19.and Agrooacterium 2, used in the construction of the phylogenetic tree in Fig. 2. The 186 bases of aligned sequence corresponding to position
306 to 514 in the E coli 16S r•NA sequence were analysed as aescribed in the legend to Table 1.
We compared 16S rRNA gene sequences for both Pro.
chlorococcus and Prochloron with sequences from the oxygenic
phototroph database8s 22 iD. L. Distel and J. B. Waterbury,
manuscript in preparationI using both distance matrix-' (Fig.
Ia: Table 11 and parsimony2 • (Fig. I b) algorithms. Bootstrap
resampling: 5 was used to estimate reliability of the inferred trees
(Fig. la, b). For simplicity we refer only to the results of the
distance matrix analysis in our discussion, although results from
parsimony analysis are in essential agreement.
The 16S rRNA tree strongly supports the derivation of both
Prochlorococcus and green chloroplasts (represented in our
analysis by the Marchantia polymorpha chloroplasti from
different, phycobilisome-containing ancestors among the
cyanobacteria (Fig. la). Prochlorococcus forms a shallowly
branching cluster with marine A Synechococcus strains WH7805
and WH8103 (ref. 16) (confirmed in 100% of bootstrap resamplings), before which branch the cyanobacteria Svnechococcus
PCC6307 (D. L. Distel and J. B. Waterbury, manuscript in
preparationi (not shown: confirmed in 100% of bootstrap
resamplings) and Synechococcus PCC6301 (confirmed in 7 30 /9
of bootstrap resamplings .The Marchantia chloroplast branches
Agrobactertum
Synechococcus PCC6301
.----.
cPROCHLROTHR
I
SARS
-PROCHLOROCOCCUS
-ALo37
Syneehococcus WH7Is05
- SARiMOO
- SAR7
_68
--
-
Synechococcus WH8103
with the Cyanophora paradoxa cyanelle (95% bootstrap
confirmation , in accord with evidence that green chloroplasts
and the cvanobacterium-like cyanelle descend from a common
ancestor ' : . Branching patterns for Prochioron and Prochlorothrix are less than 95% confirmed, thus their descent from separate,
phycobilisome-containing ancestors. though likely, is less certain than for Prochlorococcus and the chloroplasts.
Prochlorococcusis closely related not only to the open-ocean
strains, Synechococcus WH7805 and WH8103, but also to a
diversity of shotgun-cloned 16S rRNA gene sequences from the
Sargasso Sea" ' ' 9 and the north Pacific Ocean' g (Figs 1 and 2;
Tables I and 2). Individual samplings from these sites, where
both Prochlorococcus and Svnechococcus are common, have
routinely netted a puzzling multiplicity of sequences. exhibiting
greater than 95% similarity, interpreted as belonging to open'
ocean Synechococcus "".
This diversity can now be partially
attributed to sympatric populations of Synechococcus and Prochlorococcus. which, despite their sequence similarity, are separate species by ecological as well as phenotypic criteria 2'.
We conclude that the phycobilisome-/chlorophyll b÷ (green
chloroplast Iphenotype seems to have evolved de now at least
three, and possibly four times during the evolution of cyanobacteria: once in the ancestry of the green chloroplast lineage, again
in that of Prochlorococcus,and once or twice in the ancestries
of Prochloron and Prochlorothrix. the mutual independence of
which has not positively been demonstrated. The shallowness
of the Prochlorococcus cluster suggests a particularly recent
origin for this organism's pigment phenotype. In the light of
these results, and those of Palenik and Haselkonms , the order
Prochlorales can no longer be justified as a natural grouping.
The taxon should therefore be abandoned, and prochlorophytes
reclassified in the cyanobacteria.
O
Recewed27 Augustacceted 18 October
1991.
1. Cavaier-Smiti 7 "o i Linn.Soc. LOn 17. 289-306 (1962).
2. Whitton.
B. A &Carr.N.G.in TheBiol~y of we Cyanonacene
• edsCar.N.G &Whitton.a. A.)
1-8 (Unnvers,.of California.
Benrely,
1982).
&
S0.f
004
006
O.8
010
0.12
F xed point mutations per sequence position
F:G 2 Phylogenetic tree illustrating relationships among prochloroDnytes.
rre-ioers of tne Synecnococcas group ano shotg.n-clonea sequences from
71
Sargasso Sea planktone'
(SAR) and Pacific Ocean picoplankton: a iALO).
The tree was constructed using distance matrix analysis and 16S rCRNA
seoJence data (Table 2). Differences in the branching order between this
tree and the one shown in Fig. la are presumaoay due to the short length
of :he shotgun-cloneo sequences. which limits the resolution of this analysis.
Bootstrap analysis indicates that Procnlorococcus. Synechococcus strains
WH8101. WH7805. and the cloned sequences form a coherent phylogenetic
c!.s:er distinct from Prochloromrix ano Synechococcus PCC6301. out the
tr'•icning oroer in tris cluster Is Indeterminate.
1;'1URE - VOL 355 - 16 JANUARY 1992
3 Lewmn
R A C-eng. L in Proch•oron a Atcromw Engnma(eds Lewin. R.A. & Cr',w L
(Chapan an Ill.New York1989)
4. Va Valen. L M. & Maorna V C. Nature 2. 248-250 (191O).
5. Lewm.Ra Arnn1Y Y..Aaa Sc,31 325-329 ;1981)
1-7
6. Margull.L Snr•osis in CellEvolution (Freeman.
Sa Francisco.191).
7. Burger.Wersma T et a lnt J Syst Bacterol 3. 250-257 (1989).
8
9
10.
11.
Turner,S e: a %aure337. 380-382 11989).
Morden.C n ,. Golden. S S 'ature 337. 382-3e5 1989)
MorOen.C W Goldoen.S 5 Natune 33. 400 1989.
Morden.C W & Golden. S S. 1 mosec Evoa
l 3. 379-395 (1991).
12i Kishlno.
H..Myata. T.&Hasegawa.
M.J molec Evc 31a151-160 (1990).
13. Ctisholm. S W. e at Nature 334 340-343 (1988
14. Chisholm.S WN t at. Arolv ecrobiolt
(in the ress).
15. Lewin.R. A & Withers.
N.W Nature 26 735-737 (1975).
16. Wateury. . B. a.
& oa. R.ain Beraey' s Manual of SystematicBacteriloy v 3 (3ds Starley.
J. T. el al. 1728-1746 (Wilolimsand WIens. Balm•re. 1989).
17 GiovaninonS
et af Nature 341. 60-62 (1990-.
18 Schmidtt.
T M DeLon E. F & Pace. N R. 1 Bacr I73. 4371-4378 1991)
19 BrItschg7T B . Govannon, S J AO. enrw M~oao. 57. 1707-1713 U199
20 Lewin R A 'rcoroga1. 217 !1977)
LETTERS TO NATURE
21
22.
23.
24
25
26
27
28.
29
30.
31.
32.
33.
34.
35.
ln
Goecke. R & Repeta, D Limnol. Oceanogr. the press)
Giovannon. S Jert ai. J Bact 170. 3584-3592 (1988,.
Olsen.G. J. Meth Enzym 164, 793-838 (1988).
Swofford.D L. Paup Version 3.0 (Illinois Natural tstory Survey Champaign.Illinois.1989).
Felsenstein.J Evolution 39. 783-791 (19851.
Douglas. S. E. & urner. S . moec. Evol 33, 267-273 (1991).
Olson R. J..Chisholm.S W..Zettler. E. R.. Altabet.M A & Dusernerry. J A. Deep SeaRes. 37,
1033-1051 (1990)
Palenk B & Haselkorn.R Nature 358. 265-267 (1992)
Jukes T. H & Cantor. C.R in Mammalian Protein Metabolism led. Munro. H N! 21 tAcademic.
NewYork.
1969)
Tomioka.N. & Sugiura.M. Molec gen Genet 191. 45-50 (1983)
OhyamaK at . Pt molec S•/ Reporter 4. 148-175 (1986)
Yang.D. et a Proc nam. Acao Sc. U.SA 58I 4443-4447 (1985)
Samorook.J. Frntsch. E F & Manlatis.T Molecular Conrrnga LaboratoryManus IColdSpring
Harbor LaboratoryPress.ColdSpring Harbor.New York 1989)
Swift. H & Robertson. 0 L. Symbrosas 10. 95-113 (1991)
DzelkalmS, V. A. Szekeres M & Mulligan B. J. in Plant MalecularBology le. Shaw C H.)
277-299 (IRL. WashingtonDC 1988).
ACKNOWLEDGEMENTS
We thankM L. SogBnfor assistance with phylogenetic analyses 0 L. Distel
andJ.B. Waterbury
forunpulished
sequences. and W.Thdly
and P.Keohavong fortecrncal support.
ThisworkwassuDporteo
bygrantsfrom the NSF.EPAanoOffceof NavalResearch.
D.L.Rwas
supported by a training grant from the NIH
NATURE - VOL 355 - 16 JANUARY 1992
Chapter Two
PHYLOGENETIC RELATIONSHIPS AMONG
CULTURED STRAINS OF PROCHLOROCOCCUSMARINUS
ISOLATED FROM DIVERSE OCEANIC PROVINCES
ABSTRACT
Based on relationships inferred from 16S rRNA, psbB and pet B and D sequences,
cultured strains of Prochlorococcusmarinus were found to belong to a single lineage
within the cyanobacteria, which they share with strains of marine A Synechococcus.
Branching patterns inferred for P. marinus strains using 16S rRNA, psbB and pet B and
D sequences were consistent, and revealed no correlation between genetic distance and
the geographic distance between sites of culture isolation, with the most closely related
strains originating in surface waters of the Mediterranean and the Pacific Oceans. The
branching order of deeply diverging lineages within the P. marinuslmarineA
Synechococcus cluster, including one strain of P. marinus and two strains of marine A
Synechococcus, appear inconsistent in 16S rRNA gene trees inferred by neighbor-joining,
parsimony and maximum likelihood methods of analysis. This result suggests a nearsimultaneous radiation of P. marinus and marine A Synechococcus lineages.
INTRODUCTION
Prochlorococcusmarinus is a tiny, photosynthetic marine prokaryote containing
divinyl chlorophylls a and b (chl a2 and chl b 2 ) and which recently has been recognized
as a major constitutent of the photosynthetic picoplankton in wide regions of the Atlantic,
Pacific, Mediterranean and Red Seas and in the Banda Sea of Indonesia (Chisholm et al.
1988, Gieskes et al. 1988, Li and Wood 1988, Olson et al. 1990, Vaulot et al. 1990, Li et
al. 1992, Vaulot and Partensky 1992, Cambell and Vaulot 1993, Goericke and
Welschmeyer 1993, Goericke and Repeta 1993, Lantoine and Neveux 1993, Veldhuis
and Kraay 1993, Vaulot et al. 1994). P. marinus comprises 35% of the seasonally
averaged chl a at north Pacific station ALOHA (Letelier et al. 1993) and 30% at Sargasso
Sea station OFP (Goericke and Welschmeyer 1993). The ecological importance of this
ubiquitous and unusual organism has sparked efforts to characterize its physiological
capabilities in pure culture (Partensky et al. 1993, Morel et al. 1993, Moore et al. 1994).
Phylogenetic analysis using 16S ribosomal RNA gene sequences has established
that a strain of P. marinus isolated from the Sargasso Sea (strain SSW5, descended from
SARG, the original P. marinusisolate) is closely related to the marine A Synechococcus
(Waterbury and Rippka 1989), a group of cyanobacterial strains including isolates from
open ocean habitats in which P. marinusalso is found (Urbach et al. 1992).
Cyanobacterial 16S rRNA gene sequences cloned from natural populations in the
Sargasso Sea and north Pacific (Giovannoni et al. 1990, Britschgi and Giovannoni 1991,
Schmidt et al. 1991) also fall into this cluster (Urbach et al. 1992). An independent study
using rpoC gene sequences found that the genetic distance between SARG and a second
P. marinusisolate originating in the Mediterranean Sea (MED), was greater than the
distance between two heterocyst-forming cyanobacteria assigned to different genera
(Palenik and Haselkorn 1992), suggesting that differences between P. marinus isolates
may be significant. rpoC analyses also cluster P. marinus with Synechococcus WH8103
(Swift and Palenik 1992). The 16S rRNA and rpoC studies each reported, in addition,
that P. marinusis unrelated to the "normal" chlorophyll a and b (chl al and bl) containing prokaryotes Prochloronsp. and Prochlorothrixhollandica,and to green plant
and algal chloroplasts. These conclusions have been cast into some doubt, however, by
the discovery that base substitution biases at rapidly evolving nucleotide sites are
correlated with branching patterns in these studies' phylogenetic trees (Lockhart and
Penny 1992).
To help resolve these controversial relationships and examine isolates of P.
marinus from additional oceanographic provinces, we present here the results of an
investigation into the phylogenetic relationships among cultured isolates of P. marinus,
addressed by comparisons of three gene sequences and including marine A
Synechococcus and cloned 16S rRNA gene sequences from the Sargasso Sea. Questions
to be addressed are (1) whether oceanic prochlorophytes represent more than one
independent lineage of chlorophyll b2 -containing prokaryotes dispersed among the
cyanobacteria, (2) whether the inferred phylogenetic patterns for each of the three genes
are consistent, (3) whether a different phylogenetic pattern is inferred by methods
suggested as insensitive to nucleotide substitution bias, and (4) whether evolutionary
relationships reflect geographic relationships among sites of P. marinus culture isolation.
Detailed knowledge of the evolutionary relationships among P. marinus cultured strains
is, in addition, useful for organizing information on the physiological differences among
these strains. Also, if gene transfer does not play a major role in the evolution of P.
marinus,then knowledge of the phylogenetic relationships will provide a link between
phenotypes of cultured cells and sequences recovered by molecular cloning experiments
exploring genetic diversity in field populations (Giovannoni et al. 1990, Schmidt et al.
1991, Britschgi and Giovannoni 1991, DeLong et al 1993, Fuhrman et al. 1993, this
thesis Chapter Three).
BACKGROUND
Characteristics of P. marinus strains. Cultures have been established for P.
marinus isolated from a variety of geographic and hydrographic regimes (Chisholm et al.
1992, Partensky et al. 1993, Moore et al. 1994, L. Moore, pers. comm.). For this study
we selected cultures drawn from widely separated locales within P. marinus'geographic
range: the Sargasso Sea, Mediterranean Ocean, north Atlantic and south Pacific Oceans.
The P. marinus type strain, Sargasso Sea clone SS120 (designated CCMP-1375 at the
Center for the Culture of Marine Phytoplankton, Bigelow Laboratory for Ocean Sciences,
West Boothbay Harbort, ME) originated in the deep euphotic zone during spring
stratification and is descended from SARG, the original P. marinus isolate (Chisholm et
al. 1992). The Sargasso Sea isolate used for 16S rRNA analysis, SSW5, was isolated
from SARG by dilution to a mean cell density of one cell per culture tube, and is
assumed to be identical to SS120 1. Mediterranean clone Med4 (CCMP1378) originated
in mixed surface waters of the Mediterranean in winter and is descended from strain
MED (Chisholm et al 1992, Partensky et al. 1993). North Atlantic isolate FP5 is an
uncloned culture collected from 30 m in the North Atlantic by F. Partensky, and Pacific
strain MIT9107, also not cloned, came from 25m in the mixed surface layer of the
oligotrophic south Pacific, and was isolated by J. Dusenberry and M. DuRand. In
addition to these strains, Sargasso Sea culture MIT9303, isolated from 100 m in the
Sargasso Sea during summer stratification by L. Moore, was included in some of the
analyses. Synechococcus clone WH8103, a high phycourobilin (PUB) strain isolated
from the surface of the Sargasso Sea, and the reference culture for marine A
Synechococcus (Waterbury and Rippka 1989), was selected to represent marine A
Synechococcus (Table 1).
Physiological studies have characterized P. marinus clones SS120 and Med4 (or
their parent cultures) according to pigment content and respose to variation in
temperature and illumination (Partensky et al. 1993, Morel et al. 1993, Moore et al.
1994). These studies confirm the presence in both strains of divinyl chlorophylls a and b
(chl a 2 and b2 ) as well as accessory pigments zeaxanthin, a-carotene and an unknown
pigment which elutes with chlorophyll cl by HPLC (Chisholm et al. 1988, Chisholm et
al. 1992, Goericke and Repeta 1992). In addition, it is now known that SS120 (and
1In light of the results of Chapter Three, which indicate that P. marinusfield populations are genetically
heterogeneous, the possibility exists that the primary isolate SARG could have contained more than one
genetic variant. Since SSW5 and SS 120 were independently isolated from SARG, it therefore is possible
that they are not genetically identical. Omitting the SSW5 sequence from the dataset does not change the
conclusions of this study.
?ý
ilý
NN
N
OO
0 0
00
OV) •
0
mt \
4
m
?0
W)\
0•
e1
1-4 O3
F0
C)
tf
C
.N
SARG) grown at high light (>20 giE m- 2 s-1) contains chl b1 , which may comprise up to
55% of total chlorophyll b (chl b, equal to chl b1 + chl b 2 ) (Partensky et al. 1993, Morel
et al. 1993, Moore et al. 1994). In contrast, clone Med4 (and MED) contains no
detectable chl bl under any growth conditions tested. Both strains are capable of
photoadaptation by changing the ratio of their chl b to chl a2 , but chl b/a 2 ratios for Med
4 are one tenth those of SS120 at all growth irradiances (Partensky et al. 1993, Morel et
al. 1993, Moore et al. 1994).
It has been hypothesized that the different chl b/a2 ratios for SS 120 and Med4 are
genetic adaptations to the low and high light environments from which the two clones
were collected, respectively. The low light SS120 clone is capable of acclimating to
conditions at the bottom of the euphotic zone in oligotrophic waters by elaborating
extensive, chl b-containing photosynthetic antennae capable of absorbing blue light,
while the high light-adapted Med4 fails to elaborate an extensive antenna (Partensky et
al. 1993, Moore at al. 1994). This hypothesis is consistent with the results of growth
experiments which have shown that SS120 is capable of growth at low light intensities (2
to 6 tE m- 2 s-1) at which no growth was observed for Med4, and that Med4 is capable of
growth at high light intensities (>100 igE m- 2 s- 1) at which SS120 fails to grow. The
compensation light intensity (Icomp), which predicts the minimum illumination for cell
survival, was significantly lower for SS120 than for Med4 (Moore et al. 1994), again
consistent with the hypothesis. Other cultures included in the phylogenetic study are
much less well characterized than SS120 and Med4.
Genetic loci. For this study, evolutionary relationships among the four P.
marinuscultures SS120, Med4, FP5 and MIT9107 and Synechococcus WH8103 were
examined using three sequences exhibiting varying degrees of evolutionary conservation:
the 16S ribosomal RNA genes, psbB, and petB and D. Strain MIT9303 was evaluated
using only 16S rRNA, as psbB and petB/D sequences were unavailable for this culture
due to time limitations. Future work after the completion of this thesis may include psbB
and petB/D sequences for this strain.
16S ribosomal RNA (rRNA) genes are the standard tool for phylogenetic
reconstructions (Hillis and Dixon 1991, Olsen and Woese 1993) and have provided
phylogenetic information for many diverse taxa (e.g. Woese 1987, Sogin et al. 1989,
Medlin et al. 1993). 16S rRNA analyses have the advantage of compatibility with a large
and well-characterized database, containing sequences from hundreds of organisms and
organelles, and a dedicated computerized clearinghouse which provides automated
sequence alignment and similarity searching (Larsen et al. 1993). Several cloning
experiments exploring the diversity of microbial communities in the sea have provided
16S rRNA gene sequences, some of which are likely to derive from wild P. marinus
(Giovannoni et al. 1990, Schmidt et al. 1991, Britschgi and Giovannoni 1991, DeLong et
al. 1993, Fuhrman et al., 1993).
psbB and petB and D encode proteins of the photosynthetic apparatus and are
much less constrained evolutionarily than 16S rRNA. These loci were selected to infer
phylogenies sensitive to evolutionary differences on a finer scale. Each has the
additional advantage of being a single-copy sequence unique to photosynthetic organisms
(Vermaas and Ikeuchi 1991), and so is easily amplified from P. marinus cultures, which
contain heterotrophic bacteria. psbB encodes the chlorophyll a-binding antenna protein
CP47 and petB and D encode subunits of the photosynthetic b6/fcomplex responsible for
transferring electrons between Photosystem II and Photosystem I (Widger and Cramer
1991). petB and D are neighboring cistrons with a consistent tandem orientation in
oxygenic photosynthetic prokaryotes and organelles (Vermaas and Ikeuchi 1991, Greer
and Golden 1992) and provide PCR priming sites which permit amplification of a
fragment containing the 3' end of petB and the 5'end of petD, plus the highly variable
intergenic region between them (the amplified locus will be referred to as "petB/D").
The psbB and petB/D loci are not close to each other on cyanobacterial chromosomes
(Vermaas and Ikeuchi 1991).
Problems posed for molecular phylogenetic analysis by nucleotide
substitution biases. An unusal feature of cyanobacterial and chloroplast sequences is the
wide variation in their DNA base compositions (Herdman et al. 1979, Lockhart et al.
1992a), which is especially evident at rapidly evolving sites such as noncoding regions
and silent third codon positions (Lockhart and Penny 1992, Lockhart et al. 1992a, b). An
additional property of the cyanobacterial phylogeny is the presence of many deeply
branching lineages which apparently originated within a short period of evolutionary
history (Giovannoni et al. 1988). The juxtaposition of these two unusual features creates
a situation in which the inferred branching pattern is susceptible to the artefactual
clustering of sequences converging towards similar G+C content (Lockhart et al 1992a,
b, c, Lockhart and Penny 1992, Lockhart et al. 1993), and there is as yet no accepted
method for assessing the relative contributions of nucleotide substitution bias "noise" and
true phylogenetic "signal" to trees inferred from such data (Lockhart et al. 1993).
Bootstrap resampling analyses, which are frequently used to assess the sensitivity of
phylogenetic inferences to sampling error in the nucleotide positions chosen for the
analysis (Felsenstein 1988), cannot detect systematic distortions in branching patterns
due to nucleotide substitution biases (Lockhart et al. 1992a).
Lockhart et al. (1992a, b) used G+C content at third codon positions to
characterize nucleotide substitution biases in the evolution of different taxa, noting that
branching patterns inferred from protein-encoding sequences grouped organisms having
similar nucleotide substitution biases. They suggest that in instances when nucleotide
substitution biases differ among lineages, standard phylogenetic methods will infer
incorrect phylogenetic relationships. Lockhart et al. (1992a) further assert that the
influence of nucleotide substitution bias detected at third codon positions in protein
encoding genes may determine branching patterns inferred for the same organisms using
16S rRNA genes, even though the 16S rRNA sequences may exhibit little variation in
G+C content (c.f. Table 3).
Several molecular phylogenetic methods have been suggested as being relatively
immune to nucleotide substitution biases, although none are yet generally accepted as
such (Sogin et al. 1989, 1993). Woese et al. (1991) suggested that distance and
parsimony methods applied to transversion mismatches would have generally reduced
sensitivity, and so would be less susceptible to these artefacts. Hasegawa and Hashimoto
(1993) suggested that phylgenetic analysis of amino acid sequences would have this
property as well (see also Hashimoto et al. 1994), and amino acid parsimony has been
applied to plastid and cyanobacterial data (Morden et al. 1992). Most recently, two
similar, new algorithms have been devised which are reported to be insensitive to
variation in G+C bias among lineages, LogDet (Lockhart et al. 1994) and paralinear
distances (Lake 1994). However, computer programs which impliment these algorithms
are not yet publicly available.
For our analysis of P. marinus phylogenetic relationships we have employed
standard methods of phylogenetic inference: neighbor-joining, parsimony and maximum
likelihood. In addition, we explored branching patterns inferred by transversion distance
analysis and amino acid parsimony to assess whether these methods, suggested as
relatively insensitive to nucleotide substitution biases, infer a phylogeneic pattern
consistently different from those inferred by standard analyses.
METHODS
Cell culture and DNA isolation. P. marinus clones and isolates were grown in
modified K/10 - Cu medium as previously reported (Chisholm et al. 1992). DNA was
prepared from dense cultures as described (Urbach et al. 1992), or by a modification of
the alkaline lysis protocol of Li et al. (1991). For this protocol, 1 ml of culture containing
106 to 108 P. marinus was concentrated by spin filtration at 2,000 x g using 0.2 Apm pore
Ultrafree-MC filter units (Millipore) and washed twice with cell suspension buffer (0.5 M
NaCl, 10 mM Tris [pH 8.0] 10 mM EDTA). Cells were resuspended in a final volume of
100 pl cell suspension buffer and lysed by addition of 12 .l 0.5 M DTIT and 6 .l 10 M
NaOH followed by a 10 minute incubation at 650 C. After addition of 120 ptl
neutralization buffer (90 mM Tris [pH 8.0] plus 0.5 mol/1 HC1), DNA was precipitated by
addition of 1 pl 20 mg/ml glycogen (Boehringer) and 0.6 ml cold 100% ethanol and
storage overnight at -20 0 C. Precipitates were collected by centrifugation at 16,000 x g
for 30 min at 40 C, washed with 70% ethanol and resuspended in TE (10 mM Tris [pH
8.0] 1 mM EDTA). DNA was stored at -200 C.
Synechococcus WH8103 was grown in S/N Medium according to Waterbury et
al. (1986). DNA was prepared by CsCl density gradient centrifugation (Sambrook et al.
1989).
PCR and DNA sequencing. 16S rRNA genes from P. marinus Med4, FP5,
MIT9107 and MIT9303 were amplified using oxygenic phototroph-specific primers and
protocols as described (Urbach et al. 1992), except that one primer in each PCR reaction
was labelled with biotin (Hultman et al. 1989). psbB and petBID were amplified using
biotinlyated/unbiotinylated primer pairs containing inosine (Knoth et al. 1988) under the
following conditions: psbB primers PPSBB 1353 (5'GTIGCIGGIACIATGTGGTA) and
PPSBB 1928R (5'GCRTGICCRAAIGTRAACCA), 2.25 mM MgC12 , 2 min. 94 0 C,
followed by 1 min 94 0 C, 1 min 510 C, 1 min 720 C. (5 cycles), 1 min 94 0 C, 1 min 560 C, 1
min 72 0 C (10 cycles), 1 min 94 0 C, 1 min 62 0 C, 1 min 720 C (25 cycles), followed by 10
min 72 0 C; petB/D primers PPETBD314 (5'PPETBD314
(5'ATGATGGTIYTIATGATGAT) and PPPETBD 1160R
(5'CCRTARTARTTRTGICCCAT), 1.5 mM MgCl 2 , 2 min 940 C followed by 1 min
940 C, 1 min 45 0 C, 1 min 72 0 C (5 cycles), 1 min 94 0 C, 1 min 500 C, 1 min 72 0 C (10
cycles), 1 min 94 0 C, 1 min 570 C, 1 min 720 C (25 cycles), followed by 10 min 720 C.
100 Wl PCR reactions contained MgCI2 concentrations as indicated, plus lx Mg-free Taq
polymerase buffer (Promega), 200 gM each dNTP, 100 nM each primer and 5 units Taq
polymerase (Promega). PCR products from three or more replicate reactions were
pooled and separated by electrophoresis in 1% agarose gels. Bands were purified using
GeneClean (Bio 101) and single stranded sequencing templates prepared using
Dynabeads (Dynal) according to the manufacturers instructions.
DNA sequences were determined using Sequenase (United States Biochemical),
according to the manufacturer's protocol. 16S rRNA gene sequences were determined
from both strands using amplification primers and internal primers as described (Urbach
et al. 1992). 90.5% of 16S rRNA gene sequences were determined on both strands. psbB
and petB/D sequences were determined using amplification primers and internal primers
PPSBB1527 (5'ITTYTAYGAYTAYGTIGG), PPSBB 1508R
(5'CCIAGRTARTGRTARAAIGC), PPSBB 1898R (5'CGAAAIACICCRTC), and
PPETBD532R (5'CCIACRCTYTCICCICC). psbB sequences were redundantly
determined over 92% of their length, with 60% of nucleotides determined on both
strands. petB/D sequences were determined from one strand only, as the biotinylated
version of PPETBD1160R did not function well as a PCR primer. 94% of petB/D
sequences were redundantly determined.
Sequence alignment and phylogenetic analysis. Sequences were aligned
according to conserved regions of 16S rRNA primary and secondary structure or
according to conserved regions in amino acid translations (Lang and Haselkorn 1989,
Widger and Cramer 1991) using the Olsen sequence alignment editor (Olsen 1990).
Mismatch frequency and base compositions were calculated using the program of Olsen
(1988, 1990). Neighbor-joining (Saitou and Nei 1987) and protein parsimony were
performed using Kimura two-parameter genetic distance estimates (2:1 transition
transversion ratio) and the DNADIST, NEIGHBOR and PROTPARS programs in
PHYLIP verision 3.4 (Felsenstein 1991). DNA parsimony was performed with PAUP's
branch and bound and MULPARS options (Swofford 1991). DNA maximum likelihood
calculations were done using fastDNAml (Olsen et al. 1992), and likelihood comparisons
between different phylogenetic trees were performed using PHYLIP's DNAML program.
Log likelihood values and standard deviations are from DNAML comparisons.
RESULTS
Relationships inferred by standard methods: molecular phylogenetic
branching patterns and bootstrap calculations. Relationships among cultured strains
of P. marinus,marine A Synechococcus and other members of the oxygenic
photosynthetic radiation were investigated using three standard methods for phylogenetic
analysis: neighbor-joining (Saitou and Nei 1987), parsimony, and maximum likelihood
(Felsenstein 1988). All strains of P. marinus and marine A Synechococcus fell into a
single lineage in trees inferred by all analyses in this study (Figures 1, 2, 3). This
Figure 1. Phylogenetic relationships among cultured strains of P. marinus,
shotgun cloned sequences from Sargasso Sea picoplankton ("SAR" sequences) and other
members of the cyanobacterial lineage, inferred from 16S rRNA gene sequences using
Agrobacteriumtumefaciens as an outgroup. (a), Distance matrix tree inferred by
neighbor-joining with Kimura two parameter genetic distance estimates (Kimura 1980)
(Table 2a). Numbers indicate the percent of trees inferred from 100 bootstrap datasets
which contained the phylogenetic group to the right, with numbers above the lines from
neighbor-joining analysis and below from parsimony. Bootstrap values below 50% are
omitted. (b), The most parsimonious tree (668 steps) inferred by parsimony. (c),
Maximum likelihood tree (In L = -4767.84277), not significantly better than the tree
inferred by neighbor-joining (In L = -4800.28223, S.D. 19.0879, Aln L = -32.43945).
Transversion analysis inferred a branching pattern identical to that in the maximum
likelihood tree.
Agrobacterium tumefaciens
P.marinusMed4
SAR6
P.marinusMIT9107
P.marinusFP5
narinus SSW5
nus MIT9303
echococcus WH8 103
R139
;AR7
:hococcus WH7805
PCC6301
rena PCC7120
rochloron sp.
fnrrhnntin yvnnlvmnrnhn
chln
Y/··~V·y·rrr
rr-rry·
1.~,.,.,.,I,,
Genetic distance
'
Agrobacterium twnefaciens
P.marinusMed4
SAR6
P.marinus MIT9107
,marinus FP5
P.marinusSSW5
•amejaclenz
ut
mefaciens
Synechococcus WH8103
ymorpha chlp.
.
,.•
,
a r,.
SAR139
SAR7
ynechococcu W17805
P.marinusMIT9303
Synechcoccus PCC6301
AnabaenaPCC7120
Marchantiapolymorphachip.
Prochloronsp.
Figure 2. Phylogenetic relationships among cultured strains of P. marinus and
other members of the cyanobacterial lineage inferred using psbB sequences. Trees are
arbitrarily rooted to Anabaena PCC7120 to facilitate comparison to those inferred from
16S rRNA gene sequences (Figure 1). (a), Distance matrix tree inferred by neighborjoining with Kimura two parameter genetic distance estimates (Table 2b). Numbers in
parentheses indicate G+C base composition at third codon positions; other numbers are
as in Figure la. Identical branching patterns were inferred by parsimony, maximum
likelihood (In L = -1925.90128) and transversion neighbor-joining analyses, with some
deep branches rearranged in the transversion tree (not shown). (b), Consensus for the
four most parsimonious trees (267 steps) inferred by protein parsimony.
r,
_-.
,,,,~,,,~,·An,~A
rtrUFLFLUs IVILCU'L+
/~
~r\
•L.ZL)
P. marinus MIT9107 (0.27)
'. marinus FP5 (0.26)
. marinus SS120 (0.34)
us WH8103 (0.66)
tis PCC6803 (0.52)
Vrix hollandica(.58)
PCC7942 (0.61)
>last (0.26)
AIAf "'Uc',54
I
" AX
I
J.Y
X
I
I
0.2
0.4
Genetic distance
a
P. marinus Med4
P. marinus MIT9107
- P. marinus FP5
Synechococcus WH8 103
Zea maize chloroplast
Synechococcus PCC7942
"'----'----'
---
_-WJ
Figure 3. Phylogenetic relationships among cultured strains of P. marinus and
other members of the cyanobacterial lineage inferred using petB/D sequences. Trees are
arbitrarily rooted to Nostoc PCC7906 to facilitate comparison to those inferred from 16S
rRNA gene sequences (Figure 1). (a), Distance matrix tree inferred by neighbor-joining
with Kimura two parameter genetic distance estimates (Table 2c). Numbers in
parentheses indicate G+C base composition at third codon positions; other numbers are
as in Figure la. An identical branching pattern was inferred by protein parsimony, and
for P. marinus and Synechococcus WH8103 by neighbor-joining analysis of transversion
distances, which inferred a different pattern for more deeply branching lineages (not
shown). (b), Consensus of the two best trees inferred by DNA parsimony (129 steps).
(c), Tree inferred by maximum likelihood (In L = -1031.58031), which was not
significantly better than the tree inferred by neighbor-joining (In L = -1038.59363, S.D. =
6.7154, Aln L = -7.01331 ).
P. marinus Med4 (0.18)
P. marinus MIT9107 (0.21)
P. marinus FP5 (0.29)
SL-
P.marinus S 20 (0.32)
Synechococcus WH8103 (0.74
Synechococcus PCC7002 (0.55)
Prochlorothrixhollandica (0.67)
rarchantiapolymorphachlp. (0.11)
!amaize chlp. (0.29)
-·------
--
·---
\-· · //
I
I
0.05
0.10
I
Irm
I
m
0.15
Genetic distance
rSS1•D
awdupdymopAacdp.
amarschlp
I
0
I
0.05
I
0.10
ck..I dkm
43
0.15
"
0
m
1D n C> %u m rC
CDI m m
-4 -4
m-mIm.-
S
CDOD
0-4 000
00
m
H
H
HH
H
-4-
-4 m
-4
4
mODco0
0
ON
-
HA H
H-H
Na
SammmmmCmem
" m
-
m
"IT I
C'-H
I
'
U)
O
0-
IN
u0
O
U
4
9:
H
a,
0
amqr
womoooooooo
0 Coooo
H
00
0rmmm
oo
r
m
m
i
mCC) 0C m
o4mcNmcN
i
a'qaq
-4 000000000Cq
00
O
lmm
c
moooooo
mC C)
C N
HCo
C
-tzINONC
O 4mmN4-4-4
D C
-4l
O mmv
C)
m
C">
000000
a
o a'')"zI
m
N
'
IOD0DLfl-VMr
aam
U0m
1
U)
Cmmm.
AA
mMma
ON
0
04r
OZ(
E
0
S
.11
.0
0Cu
raU4
>4
n
0-0
**0O
4
a
H
0
0
aWLO
0
0
-:
·:3:5.L
CC
J
0
0
M
U
*-0 0 I:
.w
.04
(-n
C
00O
a,
U1)0
-i
a0
:$
00
-H
u
E40U0)4
M~~0
0-H
00
CD-H
H-H
04-.0
10
QH
C
0 0
0 m
am
U) H
-H
00 40)
mc
0 0(1
-O 0 E
000
map
U
\
0'
00Ic)
MC
)
54CU
0U 0
O
-1-40
Cu
04
M
0
(1
0-H
w
-- 1
a
rn
4-
4- O 0I
E
00C
aC
OO
*4-
4)
0
0
0
0-H 0
0a
*H
a'
a
P
C-4H
0
CO
0
CuH
m
Immamm
Ho
U) a
I
0
C
Ow
0
(5
Os
*4 H
U)C0
04-4
C :
%o
Cu
-
3b
S
m
U) 0
-1
0
-40CC)
*H
0
-H-H
a'-m
•
4
E
4
M
C
m>C>C>>mI
0
M
W
Z
0 -i 00
0
0
0- -H
-H4)
ý4
-1
0
0
a) ) 4 0H'r0
w
0 0
Cu4
Q0
rZ
0
.-IW
0
00
·r
Y
0
'
000DI•0
0)-O
0.0
U
4
00
0 v
:
0 U)
0
U)
0
Cu
0
0
0
--
o
U) X
)
(
U)\
C
O-4a'a'm
4
C 00000
4)
0
M
'
0
E
0
ý4
0
Q
0 9
rU -H
ý
4O 1a
0 0 0: a ( a0 fa 0 a0-E
=%C
O Cu
Cu Cu 0
4a
U
4a
E -1
->00U) U)-H
.R
0
0*000
0
0
0U)0
4-i.
0
40000
0to
-4
CD
,14H
a
m
' Oa' O
a
-H
0a
-H
Cu
Cu 00 Cu
04
0,
o0o
ON
c C
'
00
4 CH:-HU)
0Cu
0)
'CoC)u0I
m
O
m m0
>4
0)
4-)
~
(Z
-j
Cu 0
)I-V
Dn
O
"
O
41Cu
ýc
-4;
Q
.0
0Cu-1
04040-H
a
.E
04)U
04
0 -H
NO -OH 0
0 -i0
4-4
00
W
0
0
-CH
(Vý4wuu
0
0)Cu.
U) 0
0 0X
Cu
-ý4H
r
r
5
a,
4-4 U Cu
-H
?L
-4
1
>
-H -0
3
m
0
0
Cu-H
0
0
H
·
.)
e
e
'
@HOC
'
,
HJ 0
U)
OH
Oa
4-4r
A-) 00
4-i
4-1
-4
C
U) a)
00Or
E0O 44
Cu.0- 0-1'0
0.ý
0
000)
4
rn
-H
H
4-)
4-4 Cu 0
04- CuH0 -H -04.
U)·
U) x u
0E 0 000
0
0>
QO
roC
'
in
CuH
OHC
0
0 -4U 14..
0a)
0 )0
Cu ) 0
0
EOu
U) 0) 0 H
1-H
a'0O U)0
r
to
-<
a)
O
mm
a)
a
ra 8010
oaHaO
H
X
0
4-
Cu
)
r0
r
Cu
U)
H'O
'
E;
E-4 L
0
-4
U CH
Eu
r
0
W
E4
0
0
00
::I
00
-UCC
He
-n'
m
U-
*-4
U,
-H
4-4
tn
OC0CiL
O'0
a' (1
uo0E
a k
Co-
0
-a
tH
0
-H
a-C
4'00
CU
-4
0
N'
a'
0
*C
O0
Cu
H
'n
0IOem
H Cu
4 a'
1E)HU)U4 Ea
u:3
0
0
4-i
-H
O
u
-H
4-
0
0O 0000
C4--H:)
0d
E,
0,
o
mmo m
U 0
(
-4 4 -1 -A
Q
0 >
00
(*
000
0O
9
co
)-
p4
C* -( -Ha'>4H
0)
00-H
-04OH0Cu -
0
-4
to
0
m4
oO C
m
O
44
-a
-H04
0
Oa
'cr
LfDlOD
'q,
('4r'4
a,
O
0
10
U
0
NN
I
4
Ln ch L('
rl
-
m-
(
.0~
qoo
l
Qeq
c)I.'
U
Lf)-4WC')
I.)
m
rl)
uL
4-4o
nto
m
'-I
CDl
LIOD
(
'U
'e
4-4
0
00ho
'oCe
q,
Mt
0
r(1 0(7
C
OD
o
m
m
0o U)
I
al4o
',0 OD
rn
kot- r)
ILfCCIa
e
nr-
en N
C4O
O0
r-
aDU)
)(
',
DC'
#
~
L
'
m
co 1
-
mOD(
%
r P
NIrkD
q
ý,otý rn
Nc o c q0
D OD '
ý.otý r-)
T"t
I4
c0cI
-
aO
Ci
N
n1
m L
QO
_4(C
4-1
-44
0
Lo0
-4
-4
Nto
.0 -H
.00)
('4 o
-4S
.0-.
m SN
m
N
a
o
N
I'Dý (1)
-
e
O
Q
-,o t, rn
L
~~~9 aCCr
e~ ~ ~~ r
cyn 7 D qo) ()i ,t r -I
m
c-
-4
-4
eODN
- en - D
r- O
'.P
n cT
r-I
0I
.J.
04
mGo
r- D n
0
-A
U]
r.
0
0
tU
-4
a)
*H0-4
4J
O4
ý4
- 4
O v-e-4
-L)
0-4I
0
Ow
to
C
k
4-4
0 04-0
0
COMu
0 O)
OOc'
10 -4
"
00
cU
O O
Uoc)
Z
(Uk
0
-4
0
QCL
0r~
Z
4
04
O
Sa)
4-J
a 0
~0
-4
m
-c0
0O
4-4 to
4 O
c
4O 4
cci
0
O
O
E
uO
4-j
' -(0
Q)
E(0 4.J
O
U)
0
(L)
Q
-H
Qua,
:3
00,
X
U
:)j
cm
0U)
a a, .0
0
0)
0
-a
O4
-A
E,
-4
a4
C
0
0)
0
-
3,CC
(0
4
.Oc
cciH
4) -4 00
OO
4- 0
-4 -4
co
cciOU
U
dO
H
0
O0
(dC
0
U
-F0
O-4o
4-4
cci
0 to
0:*
ci
cci kD
C14 a
0
-4
(0( '1l
U]C
(cc>
0oa, a
>l 1 W
rl
U,
0
-4
an
U4-1
OO
-4
u
ro
&
04.)
(0c
z
,C
4W
z
C
0 -4
4'0j -A
-0-40
I-H
Q
ci
u)
0
do
C jo
o:30
0Z QC0
'0c
04O) 00
U]
-4
4- 0
-4
0
>
OO
En
.0 *)
COa1
OU
- 4U
4
*En
0 - 'H
ci)
4
40-4
-4
u4-
4
-
rci
EOw
$04
Eca
04-4~
w004-4
to 4
004
x0
ý
>1
Ul OD 0
U>1
'0 04-0
Lo 0j
L)
040 4
004-4
41
04-)
0 0
04
.CO 0 0
O ci 0r
0 -4a)
(U)
ý
U)
ci
.C
(0440
ci 44
.00
ci
0
(N
0
1-4
oL)
Nr
-4
OD
-4
N
r-
0
(N
ro o
in
C Lx
-1
rC
-)
I
-4OO
%a "
r-OD
1%
q
o
%
D'-i
r0
4 .-
4-1
OU
(N
-4
NQ kCN 1D.0-0
rl)
%D
r-O cM 1 O
o IV
~ODr
-4
m0DQC
) .0
en0O
cO
\ome
men(
.i)
I'MD r
(,
r-
-14'
Ir4
k
-4m'T
0%U
r ODmrn
r-OD
NO4'
z % 0)
NQm
K
e-4r,
Ilp
NC>
eIý
VM~
om
CCN
O.C
-1
.0
r-
1
-"4)
"N04
0OD
Oat
rN
00<co LI)
UE
0
%
D
b~
O~
aG
weN
em~
men~~
O
0
•4
)0
In
0
0
0-•
00
.q
ill
a4rc0
-4
U
r0
IiOo
I) rI
r- at 'IT
-1
ODC m
om N
%D
%D
~
I
~
0am
.
m
C
I 0%%-0
OD()
0
r- atI
::NQrODq,
l
an
~
m
mIOo
N%0
CoQ"n
cD
q Chm
4 nme kmT
Oo
OD
L,- c)
omf
iM
, )
. CIN r4
CD
In
N
(N4
N
CN
No4
OD
l -,
am~~ mam
nODco
o
'.0
kDy
e
co
mob
o N
Dz
O0
~
(n
" s0
ca
rODV
r- OD
m OD
OD
lo-0
N0e
mh
c%
Lt
N
%nMr
a
zrn 0 me(
t- 0
'.0
cq C4"
o 0e
m
0a0
Q)
on
4 q
~ON
'oTN
ob4
0> .Ol
OIC) 0N
LJ
Nr'- 4 0
eq"
0 4"
maI
0N
<aN
oatc
'0
-H
r4w
C
0>
U)
1-4 mn1
m- r.
co atLt
0 as
n
1NOWN
r-tot
C
N
0c)at Lfl
-a
v r,
*:
(1y4C'
- %r
7%-4*
k-0a.14
'
o4a%
rN40
r-(
NO0%'(N
1,0Ot
G Ln
rD ')
'0
Ný
GoN
-4wf
rOD 'L)0
ac..eN
DDLI)
1
N OD
w4
o 1:O)
OI%
c
.D OD ()
ý4O
Lfl-0
WnCo
rDU)
oN- qLi
'q
:
4
r-
co'I
CO
-0
4-H(
CI
Z
S4
'
U)
(X
In
0
-4
-
'A
4-4
0
z0
ci)
-H
ý4)
to 0,
O
0
-4(
Qi)
a
0
LO
0
El
04
ý44
0
ro
a 0
0
ý4)
a)
44
c)
0
4-1
c:
14o0
L9
0
r0
0
O
a0
to4
:JF
U
0-4O -1
'H
0
L)
0
to
E;
ci)
u
u
0>
O4rýci)
0
0o
u
to,
'ci
410
4. 4-) 4'ci
k
ci)
C-
0,H
0
4-)0
4.
00
0
U)
-4
vi)
>o
0z
44
(15
Lo
O
-4
::s
cU
U
~cci
4-.
to
to
m
0
(-4
'OD
0,
U
0
tp
ti1
-14
0C:
O
to
a
a
0)
(c
-4
'0
ci
to
0
a)
U 4-3
to
a,
.ci
u)-4
OO
*cc 1
'ci
S0
ci)
-H
EO4r4
z 0o'
0
044
o to
04
0
-ro
U)
O U)
W
aU
0
C0 c
c)
'0
to
'H
-4
-to H
4-)
H
0 H
O
00)
4-) )
4-1 q
CO
to
O4 0)
4.)
" ý4
w
-0 O
4-104
SC
L)
d
-H
O
-H
O
.0
ci)
kci
ci)
0
toO
I
00
04
0
Ql -1c
0
-H
clustering was consistent in neighbor-joining and parsimony analyses, whether or not
chloroplast sequences were included in the datasets, and persisted in a 16S rRNA gene
tree which included diverse cyanobacterial taxa (Figure 4).
P. marinus strains Med4, MIT9107, FP5 and SS120 (the "P.marinus four culture
clade") formed a phylogenetic group distinct from Synecyococcus WH8103 in all
analyses. However, the phylogenetic position of the deeply branching P. marinus
MIT9303 was different in 16S rRNA trees inferred by different methods.
Identical branching patterns were inferred for relationships among members of
the P. marinus four culture clade and Synechococcus WH8103 for all three genes using
the neighbor-joining method, and for 16S rRNA and psbB genes using the parsimony and
maximum likelihood methods as well (Tables 2a, b, c, Figures la, b, c, 2a, 3a). For
petB/D sequences, however, relationships among the four P. marinus cultures could not
be resolved using parsimony (Figure 3b), and the maximum likelihood tree, though not
significantly more likely than the neighbor-joining tree, exhibited a slightly altered
branching order (Figure 3c).
Bootstrap analysis (100 resampled datasets each for neighbor-joining and
parsimony calculations) strongly supported the pattern inferred by all three methods for
16S and psbB sequences for branching order among members of the P. marinus four
culture clade and Synechococcus WH8103, except that the relative positions of strains
FP5 and SS120 received varying levels of support (Figures la, 2a). Consistent with the
generally lower resolving power in the petBiD dataset, bootstrap values were lower in
trees inferred from these sequences (Figure 3a). The lower resolution in petB/D trees
may be due to the shorter length of petB/D sequences; after intergenic regions and third
Figure 4. Phylogenetic relationships among P. marinus cultured strains and
diverse cyanobacterial taxa, inferred by neighbor joining using 16S rRNA sequences.
Differences in the branching pattern between this tree and the tree in Figure la are
presumably due to the smaller number of nucleotides used in the analysis, due to short
length and sequence ambiguities of some sequences.
SAgrobacterium tumefaciens
P. marinus Med4
P. marinus MIT9107
P. marinus FP5
P. marinus Sargasso Sea isolate
P. marinus MIT9303
Synechococcus WH7805
Synechococcus WH8103
Synechococcus PCC6301
Synechococcus WH7335
PhormidiumPCC7375
Prochlorothrixhollandica
DL1,,l
P rochorortNp.
U
II
- Synechocystis PCC6308
Myxosarcina PCC7312
l1oeothece PCC6501
,yngbyaPCC7419
)scillatoriaPCC6304
us 10mfx
ora paradoxacyanelle
archantiapolymorpha chip.
'hlamydomonasreinhardiichip.
Euglena gracilischip.
Cyanidium caldarium 14-1-1 chlp.
4
-
LP
rhromrnonna Annira chln
Plectonema PCC73110
PseudanabaenaCCC OL-75-PS
Anabaena PCC7120
FischerellaPCC7414
-- GloeobacterPCC7421
codon positions were removed from the analysis, petBID phylogenetic inferences were
drawn from 258 nucleotide positions, while psbB and 16S rRNA gene analyses
considered 333 and 1085 nucleotide postitions, respectively.
The phylogenetic position of P. marinus strain MIT9303 was ambiguous. While
16S rRNA gene trees inferred by the three standard methods consistently recovered the
P. marinusfour culture clade as a coherent group, they were inconsistent in their
depictions of the relative branching order of five more deeply branching lineages in the
P. marinus/marineA Synechococcus lineage, including P. marinus MIT9303, the four
culture clade, the two marine A Synechococcus strains and SAR7, a cloned sequence
from the Sargasso Sea (Giovannoni et al. 1990) (Figure la, b, c). In bootstrap
calculations, the branching order of these deeply branching lineages received less than
50% support (Figure la). The discordance in the branching order of P. marinus
MIT9303 and marine A Synechococcus lineages inferred by different phylogenetic
methods and their low bootstrap values suggests a near-simultaneous radiation of P.
marinusand marine A Synechococcus lineages during evolutionary history (Hoelzer and
Melnick 1994).
Effects of nucleotide substitution biases. We explored the possible effects of
nucleotide substitution bias on our phylogenetic trees by labelling the psbB and petB/D
neighbor-joining trees according to their third codon position G+C content (Lockhart and
Penny 1992, Lockhart et al. 1992b) (Figures 2a, 3a). It is immediately apparent from
inspection of these labelled trees that P. marinus isolates fail to cluster with chloroplasts
despite similarities in their third codon position G+C content (Figures 2, 3), and that
nucleotide substitution bias is therefore not the sole organizing factor in these trees. 16S
rRNA gene trees also fail to group P. marinus with chloroplasts (Figure 1). However,
within the P. marinuslmarineA Synechococcus cluster the phylogenetic branching
pattern inferred by the majority of analyses is correlated with G+C content at third codon
positions, with the Med4/MIT9107 phylogenetic pair showing considerably lower third
codon position G+C content (0.18 to 0.27) than SS120 (0.32 to 0.34) and Synechococcus
WH8103 (0.66 to 0.72). Although this clustering may reflect actual phylogenetic
relationships in which substitution biases simply differ among lineages, it is also
consistent with a systematic error in phylogenetic inference (Lockhart et al. 1992a).
Again, there is no accepted molecular criterion for determining which hypothesis is
correct.
To further explore the possible effects of nucleotide substitution biases,
phylogenetic branching patterns were assessed using two methods of analysis thought to
be insensitive to them (Woese 1991, Hasegawa and Hashimoto 1993): transversion
analysis with neighbor-joining and, for the two protein-encoding sequences, amino acid
parsimony. Transversion analysis inferred an identical branching pattern for
Synechococcus WH8103 and the members of the P. marinus four culture clade as was
found by standard neighbor-joining for all three gene sequences, although the relative
branching order for deep branches in the P. marinus/marineA Synechococcus clade
agreed with the maximum likelihood calculation in the 16S rRNA transversion tree
(Figures 1, 2, 3). Relationships inferred by protein parsimony were also largely
congruent with trees inferred by standard methods, except that P. marinus strains SS120
and FP5 were joined to form a sister clade to P. marinus Med4 and MIT9107 in the psbB
protein parsimony tree (Figure 2b).
In sum, transversion and protein parsimony analyses do not suggest a consistent,
alternative phylogenetic branching pattern, but instead reinforce the conclusions of the
standard neighbor-joining, parsimony and maximum likelihood analyses: P. marinus
Med4 is most closely related to P. marinus MIT9107, below which diverge P. marinus
strains FP5 and SS120 in somewhat indeterminate order, with the most deeply branching
culture being P. marinusMIT9303. The branching order of five deeply branching
lineages in the P. marinuslmarineA Synechococcus cluster is not resolved.
Intergenic region and third codon position mismatches. Intergenic "nonsense"
regions and third codon (silent) positions in protein-encoding genes are rapidly evolving,
and therefore sensitive indicators of small amounts of evolutionary change. However,
when compared to the first two codon positions in protein-encoding genes or to 16S
rRNA genes, intergenic regions and third codon positions become rapidly saturated with
nucleotide substitutions and (in intergenic regions) insertions and deletions. These tend
to obscure evolutionary relationships when homologous nucleotide positions cannot be
aligned or when the proportion of superimposed substitutions becomes large.
Comparison of aligned sequences of petB/D and psbB from cultured strains of P.
marinus,Synechococcus WH8103, chloroplasts and other cyanobacteria illustrates this
point. In comparisons of P. marinuspetB/D intergenic regions, which range from 34 to
90 basepairs in length, homologous nucleotide positions could be identified only for a
portion of the intergenic regions of strains Med4 and MIT9107 (Figure 5). Third codon
position mismatches ranged from 41.1% to 64.3% (similarity 0.589 to 0.357) for psbB
and from 39.5 to 66.1 percent (similarity 0.65 to 0.339) for petB/D, and were considered
too large to give reliable phylogenetic information (Tables 2b, c) (for a similar analysis,
see Bhattacharya et al. 1991). In addition, third codon positions were highly
heterogeneous in G+C content (Table 3), which may bias phylogenetic inferences.
Intergenic regions and third codon positions were therefore omitted from phlyogenetic
analyses.
Table 3. G+C Base compositions of sequences used in these
analyses.
DetB/D
P. marinus Med4
P. marinus MIT9107
P. marinus FP5
P. marinus SS120
Synechococcus WH8103
Prochlorothrixhollandica
Synechococcus PCC7002
Nostoc PCC7906
Marchantia chloroplast
Zea maize chloroplast
DSbB
P. marinus Med4
P. marinus MIT9107
P. marinus FP5
P. marinus SS120
Synechococcus WH8103
Synechocystis PCC6803
Prochlorothrixhollandica
Synechococcus PCC7942
Anabaena PCC7120
Zea maize chloroplast
Codon Dosition(s)
All
1st2
3rd
0.37
0.38
0.40
0.41
0.57
0.54
0.50
0.49
0.32
0.41
0.47
0.46
0.45
0.46
0.49
0.47
0.48
0.49
0.43
0.47
0.18
0.21
0.29
0.32
0.72
0.67
0.55
0.49
0.11
0.29
Codon position(s)
All
ist2
3rd
0.39
0.40
0.41
0.45
0.58
0.51
0.56
0.55
0.48
0.42
16S rRNA
Agrobacterium tumefaciens
P. marinus Med4
P. marinus Pac7
P. marinus FP5
P. marinus SSW5
P. marinus MIT9303
Synechococcus WH8103
Synechococcus WH7805
SAR6
SAR7
SAR139
Synechococcus PCC6301
Prochloron sp.
Anabaena PCC7120
Marchantia polymorpha chloroplast
0.47
0.47
0.48
0.51
0.54
0.50
0.55
0.52
0.53
0.49
0.25
0.27
0.26
0.34
0.66
0.52
0.58
0.61
0.39
0.26
G+C composition
0.54
0.53
0.53
0.54
0.54
0.54
0.54
0.54
0.53
0.54
0.54
0.54
0.52
0.53
0.54
Figure 5. Comparison of petB/D intergenic region sequences for cultured strains
of P. marinus and other members of the oxygenic photosynthetic radiation. Sequences
are aligned according to amino acid translations in coding regions, and, with the
exception of the sequence for P. marinus MIT9107, intergenic regions are arbitrarily
represented as continuous with petB. An alignment gap has been inserted into the P.
marinusMIT9107 intergenic region sequence in order to accentuate the similarity
between the 5' end of this sequence and that from P. marinus Med4. Intergenic region
sequences from Prochlorothrixhollandica,Nostoc PCC7906 and Marchantia
polymorpha and Zea maize chloroplasts have been truncated. Sequence data for P.
marinus MIT9303 are from this work, other attributions are as in Table 2c.
rI
I
(II
i
II
I
g
I
U
;I IC
I
E-
L E,)
iI
I
I~c
U
I3I
5I t
I
PI u
:
E~5
I
I
I
u~~
e
I
I
I
I
I
i
IU
I$
Eg-
ElE
4
r
I
(iI
EU):
14
U:
I U
C, U
A
ou
E-.
C,
tnt
z
;I
8
0
£I
0
U
.t
U
IIIU U7
0~
3
**
~EI.
to
CII 0
'-
j
E%
0)
£;U
'-
2
).
V
0
12
0
1
)
'Ell
to
0"
0
-q1
0N
U
C
I
a
a
DISCUSSION
Chlorophyll a2 and b 2 -containingmarine prokaryotes arise from a single
lineage within the cyanobacteria. The cultured P. marinus strains included in this study
are phylogenetically restricted to a single lineage within the cyanobacteria, which they
share with marine A Synechococcus strains WH8103 and WH7805. Neighbor-joining,
parsimony and maximum likelihood analyses of nucleotide sequences, parsimony
analysis of amino acid sequences, and analysis of transversions by neighbor-joining, all
replicate this result, which links strains of P. marinus and marine A Synechococcus
having very different G+C substitution biases, despite the presence in the dataset of
chloroplast sequences having third codon position G+C content similar to those of P.
marinus. The consistency of these results argues that the P. marinuslmarineA
Synechococcus phylogenetic grouping represents a true evolutionary lineage, and not an
artefactual cluster.
Branching patterns among four cultured isolates of P. marinus and
Syenchococcus WH8103 are highly consistent. The branching order inferred for P.
marinus Med4, MIT9107, FP5 and SS120 and Synechococcus WH8103 in phylogenies
inferred from three gene sequences was consistent. It is therefore likely that gene
transfer does not play a major role in the evolution of these organisms and that P.
marinus exhibits clonal inheritence similar to that in Escherichiacoli (Selander et al.
1987). This finding buttresses the assumption that prokaryotic organisms exhibiting
closely related sequences at a single locus are likely to be genetically similar at most loci,
and hence phenotypically similar. This assumption is implicit in most analyses of
sequences cloned from natural populations (Young 1989).
Phylogenetic relationships among P.marinus isolates do not correlate with
geography. P. marinus Mediterranean strain Med4 and Pacific strain MIT9107, both
isolated from mixed surface waters, were found to be closely related, while north Atlantic
strain FP5 and Sargasso Sea deep chlorophyll maximum cultures SS120 and and
MIT9303 fell into progressively more deeply branching lineages. There was thus no
correlation between genetic distance and the geographic distance between sites of culture
isolation, with the two Sargasso Sea isolates exhibiting less sequence similarity than the
Med4/MIT9107 pair.
The different possible branching positions for P.marinus MIT9303 suggest
different evolutionary scenarios for P. marinus and marine A Synechococcus
pigments. The ambiguity in the branching position of P. marinus MIT9303 leaves open
the question of the origins of P. marinus and marine A Synechococcus pigments. If the
correct placement of P. marinus MIT9303 were as the deepest branch of the P.
marinuslmarineA Synechococcus cluster, as indicated by the 16S rRNA gene parsimony
and maximum likelihood trees (Figures lb, c), this would suggest that the chl a 2 and b2
P. marinus phenotype was ancestral to the whole P. marinus/marineA Synechococcus
cluster, implying further that cells capable of this phenotype arose from an ancestor
shared with Synechococcus PCC6301 and gave rise to Synechococcus WH8103 (both
strains being chl a l and phycobilisome-containing cyanobacteria, neither of which
contains chl a2 or chl b (Rudiger and Schoch 1988, Waterbury et al. 1979, L. Moore,
pers. comm.). If, on the other hand, the correct placement of P. marinus MIT9303 were
at the base of a unitary P. marinus lineage, as inferred by neighbor-joining using 16S
rRNA gene sequences (Figure la), then the phylogeny would be consistent with a single
origin of the chl a 2 and b2 phenotype without "reversion" to the chl al and
phycobilisome-containing phenotype. The characterization of the P. marinus
photosynthetic apparatus and its genes, with comparison to those of marine A
Synechococcus is likely to yield the most conclusive data pertaining to this question.
Ambiguity in the branching order of deeply diverging lineages in the P.
marinuslmarineA Synechococcus cluster precludes identification of some cloned
sequences from natural populations. Most studies investigating the phylogenetic
diversity of uncultured microorganisms in the sea have employed 16S rRNA gene
sequences, many of which fall into the P. marinuslmarineA Synechococcus cluster
(Giovannoni et al. 1990, Schmidt et al. 1991, Britschgi and Giovannoni 1991, DeLong et
al., 1993, Fuhrman et al, 1993). While some of these sequences form close phylogenetic
associations with cultured P. marinus or marine A Synechococcus strains (e.g. SAR6 and
SAR139), others (e.g. SAR7) form deep branches in the P. marinuslmarineA
Synechococcus cluster and are not specifically related to sequences from characterized
cells (Figure 1). Because the branching order of P. marinus and marine A
Synechococcus lineages in the cluster cannot at present be resolved, these cloned
sequences cannot confidently be assigned to either taxonomic group. Assignment of
phenotypes to these sequences must await the discovery of closely related sequences in
phenotypically characterized cells, results of in situ hybridization experiments, or future
refinements in the methods of phylogenetic inference.
REFERENCES
Bhattacharya, D., Stickel, S.K. and Sogin, M.L. (1991). Molecular phylogenetic analysis
of actin genic regions from Achlya bisexualis (Oomycota) and Costariacostata
(Chromophyta). J. Mol. Evol. 33:525-536.
Brand, S.N., Tan, X. and Widger, W.R. (1992). Cloning and sequencing of the petBD
operon from the cyanobacterium Synechococcus sp. PCC7002. Plant Mol. Biol.
20:481-491.
Britschgi, T.B. and Giovannoni, S.J. (1991) Phylogenetic analysis of a natural marine
bacterioplankton population by rRNA gene cloning and sequencing. Appl.
Environ. Microbiol. 57:1707-1713.
Campbell, L., and Vaulot, D. (1993). Photosynthetic picoplankton community structure
in the subtropical North Pacific Ocean near Hawii (station ALOHA). Deep-Sea
Res. 40:2043-2060.
Chisholm, S.W., R.J. Olson, E.R. Zettler, R. Goericke, J. Waterbury, and N.
Welschmeyer. (1988). A novel free-living prochlorophyte abundant in the
oceanic euphotic zone. Nature, 334(6180):340-343.
Chisholm, S.W., Frankel, S.L., Goericke, R., Olson, R.J., Palenik, B., Waterbury, J.B.,
West-Johnsrud, L. and Zettler, E.R. (1992). Prochlorococcusmarinusnov. gen
nov. sp.: a marine prokaryote containing divinyl chlorophyll a and b. Arch.
Microbiol. 157:297-300.
DeLong, E.F., Franks, D.G. and Alldredge, A.L. (1993) Phylogenetic diversity of
aggregate-attatched vs. free-living marine bacterial assemblages. Limnol.
Oceanogr. 38:924-934.
Felsenstein, J. (1988). Phylogenies from molecular sequences: inference and reliability.
Annu. Rev. Genet. 22:521-65.
Felsenstein, J. (1991). Phylip version 3.4 (University of Washington, Seattle, WA).
Fitch. W.M. and Margoliash, E. (1967). Construction of phylogenetic trees. Science
155:279-284.
Fuhrman, J.A., McCallum, K. and Davis, A.A. (1993). Phylogenetic diversity of
subsurface marine microbial communities from the Atlantic and Pacific Oceans.
Appl. Environ. Microbiol. 59:1294-1302.
Geiskes, W.W.C., G.W. Kraay, A. Nontji, D. Setiapermana, and Sutomo. 1988.
Monsoonal alternation of a mixed and a layered structure in the phytoplankton of
the euphotic zone of the Banda Sea (Indonesia): A mathematical analysis of algal
pigment fingerprints. Neth. J. Sea Res. 22:(2):123-137.
Giovannoni, S.J., Turner, S., Olsen, G.J., Barnes, S., Lane, D.J. and Pace, N.R. (1988).
Evolutionary relationships among cyanobacteria and green chloroplasts. J.
Bacteriol. 170:3584-3592.
Giovannoni, S.J., Britschgi, T.B., Moyer, C.L., and K.G. Field. (1990). Genetic
diversity in Sargasso Sea bacterioplankton. Nature 345:60-62.
Goericke, R. and Repeta, D.J. (1992). The pigments of Procholorcoccusmarinus: the
presence of divinyl chlorophyll a and b in a marine prokaryote. Limnol.
Oceanogr. 37:425-433.
Goericke, R. and Repeta, D.J. (1993). Chromatographic analysis of divinyl-chlorophylls
a and b in samples from the subtropical north Atlantic Ocean. Mar. Ecol. Prog.
Ser.
Goericke, R. and Welschmeyer, N.A. (1993). The marine prochlorophyte
Prochlorococcuscontributes significantly to phytoplankton biomass and primary
production in the Sargasso Sea. Deep-Sea Res. 40:2283-2294.
Greer, K.L. and Golden, S.S. (1992). Conserved relationship between psbH and petBD
genes: presence of a shared upstream element in Prochlorothrixhollandica.
Plant Mol. Biol. 19:355-365.
Hasegawa, M. and Hashimoto, T. (1993). Ribosomal RNA trees misleading? Nature
361:23.
Hashimoto, T., Nakamura, Y., Nakamura, F., Shirakura, T. Adachi, J., Goto, N.,
Okamoto, K-i. and Hasegawa, M. (1994) Mol. Biol. Evol. 11:65-71.
Herdman, M., Janvier, M., Waterbury, J.B., Rippka, R. and Stanier, R.Y. (1979)
Deoxyribonucleic acid base composition of cyanobacteria. J. Gen Microbiol.
111:63-71.
Hillis, D.M., and Dixon, M.T. (1991). Ribosomal DNA: Molecular evolution and
phylogenetic inference. Quarterly Rev. Biol. 66:411-453.
Hoelzer, G.A and Melnick, D.J. (1994) Patterns of speciation and limits to phylogenetic
resolution. Trends Ecol. Evol. 9:104-107.
Hultman, T., Stahl, S., Homes, E. and Uhlen, M. (1989). Nucl. Acids Res. 17:49374946.
Kallas, T., Spiller, S. and Malkin, R.C. (1988). Characterization of two operons
encoding the cytochrome b6-f complex of the cyanobacterium Nostoc PCC7906
and highly conserved sequences but different gene organization than in
chloroplasts. J. Biol. Chem. 263:14334-14342.
Kimura, M. (1980). A simple method for estimating evolutionary rate of base
substitutions through comparative studies of nucleotide sequences. J. Mol. Evol.
16:111-120.
Knoth, K., Roberds, S., Poteet, C. and Tamkun, M. (1988). Highly degenerate, inosinecontaining primers specifically amplify rare cDNA using the polymerase chain
reaction. Nuleic Acids Res. 16:10371.
Lake, J.A. (1994). Reconstructing evolutionary trees from DNA and protein sequences:
paralinear distances. Proc. Natl. Acad. Sci. USA 91:1455-1459.
Lang, J.D., and Haselkom, R. (1989). Isolation, sequence and transcription of the gene
encoding the phtotsystem II chlorophyll-binding protein, CP-47, in the
cyanobacterium Anabaena 7120. Plant Mol. Biol. 13:441-456.
Lantoine, F. and Neveux, J. (1993). Spectrofluorometric analysis of photosynthetic
pigments in the Gulf of Lions (Mediterranean Sea) with an emphasis on the
autotrophic picoplankton. Ann. Inst. Oceanogr., Paris 69:173-175.
Larsen, N., Olsen, G.J., Maidak, B.L., McCaughey, M.J., Overbeek, R., Macke, T.J.,
Marsh, T.L., and Woese, C.R. (1993) The ribosomal database project. Nucleic
Acids Res. 21 Supplement:3021-3023.
Letelier, R.M., Bidigare, R.R., Hebel, D.V., Ondrusek, M., Winn, C.D. and Karl, D.M.
(1993). Temporal variability of phytoplankton community structure based on
pigment analysis. Limnol. Oceanogr. 38:1420-1437.
Li, H., Cui, X. and Arnheim, A. (1991). Analysis of DNA sequence variation in single
cells. Methods 2:49-59.
Li, W.K.W., Dickie, P.M., Irwin, B.D., Wood, A.M. (1992). Biomass of bacteria,
cyanobacteria, prochlorophytes and photosynthetic eukaryotes in the Sargasso
Sea. Deep-Sea Res. 39:501-519.
Li, W.K.W., and M. Wood. (1988). Vertical distribution of North Atlantic
ultraphytoplankton: analysis by flow cytometry and epifluorescence microscopy.
Deep Sea Res. 35:1615-1638.
Lignon, P.J.B., Meyer, K.G., Martin, J.A. and Curtis, S.E. (1991). Nucl. Acids Res.
19:4553.
Lockhart, P.J., Penny, D. (1992). The problem of GC content, evolutionary trees and the
origins of chl-a/b photosynthetic organelles: are the prochlorophytes a
eubacterial model for higher plant photosynthesis? in Research in Photosynthesis,
VoL III (N. Murata, ed.). Kluwer Academic, pp. 499-505.
Lockhart, P.J., Howe, C.J., Bryant, D.A., Beanland, T.J. and Larkum, A.W.D. (1992a).
Substitutional bias confounds inference of cyanelle origins from sequence data. J.
Mol. Evol 34:153-162.
Lockhart, P.J., Penny, D., Hendy, M.D., Howe, C.J., Beanland, T.J. and Larkum, A.W.D.
(1992b). Controversy on chloroplast origins. FEBS Lett. 301:127-131.
Lockhart, P.J., Beanland, Howe, C.J. and Larkum, A.W.D. (1992c). Sequence of
Prochloron didemni atpBE and the inference of chloroplast origins. Proc. Natl.
Acad. Sci. USA 89:2742-2746.
Lockhart, P.J., Penny, D., Hendy, M.D., and Larkum, A.D. (1993). Is Prochlorothrix
hollandicathe best choice as a prokaryotic model for higher plant chl a/b
photosynthesis? Photosynthesis Res. 37:61-68.
Lockhart, P.J., Steel, M.A., Hendy, M.D., and Penny, D. (1994). Recovering
evolutionary trees under a more realistic model of sequence evolution. Mol. Biol.
Evol. 11:605-612.
Medlin, L.K., Williams, D.M. and P.A. Sims (1993). The evolution of the diatoms
(Bacillariophyta). I. Origin of the group and assessment of the monophyly of its
major divisions. Eur. J. Phycol. 28:261-275.
Moore, L.R., Goericke, R., and Chisholm, S.W. (1994). The comparative physiology of
Synechococcus and Prochlorococcus: influence of light and temperature on
growth, pigments, flourescence and absorptive properties. Mar. Ecol. Prog. Ser.,
in press.
Morden, C.W., Delwiche, C.F., Kuhsel, M., and Palmer, J.D. (1992). Gene phylogenies
and the endosymbiotic origin of plastids. Biosystems 28:75-90.
Morel, A., Ahn, Y.-H., Partensky, F., Vaulot, D. and Claustre, H. (1993). J. Mar. Res.
51:617-649.
Neveux, J., Vaulot, D., Courties, C., and E. Fukai. (1989). Green photosynthetic bacteria
associated with the deep chlorophyll maximum of the Sargasso Sea. C.R. Acad.
Sci. Paris 308:9-14.
Ohyama, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano, T., Sano, S., Umesono, K.,
Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi, H. and Ozeki, H. (1986).
Chloroplast gene organization deduced from complete sequence of liverwort
Marchantiapolymorpha chloroplast DNA. Nature 322:572-574.
Olsen, G.J. (1988). Phylogenetic analysis using ribosomal RNA. Methods Enzymol.
164:793-838.
Olsen, G.J. (1990). Sequence editor and analysis package (University of Illinois, Urbana,
IL).
Olsen, G.J., Matsuda, H., Hagstrom, R and Overbeek, R. (1992). FastDNAml (Argonne
National Laboratory, Argonne, IL).
Olsen, G.J. and Woese, C.R (1993). Ribosomal RNA: a key to phylogeny. FASEB J.
7:113-123.
Olson, R.J., Chisholm, S.W., Zettler, E.R., Altabet, M.A., and Dusenberry, J.A. (1990)
Spatial and temporal distributions of prochlorophyte picoplankton in the North
Atlantic Ocean. Deep Sea Res. 37, 1033-1051.
Palenik, B. and Haselkorn, R. (1992). Multiple evolutionary origins of prochlorophytes,
the chlorophyll b-containing prokaryotes. Nature 355:265-267.
Partensky, R., Hoepffner, N., Li, W.K.W., Ulloa, 0. and Vaulot, D. (1993).
Photoacclimation of Prochlorococcussp. (Prochlorophyta) strains isolated from
the north Atlantic and the Mediterranean Sea. Plant Physiol. 101:285-296.
Rock, C.D., Barkan, A. and Taylor, W.C. (1987). The maize plastid psbB-psbF-petBpetD RNAs encode alternative products. Curr. Genet 12:69-77.
Rudiger, W., and S. Schoch. (1988). Chlorophylls. In Plant Pigments. T.W. Goodwin,
ed. Academic Press. 1-60.
Saitou, N. and Nei, M. (1987). The neighbor-joining method: a new method for
reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406-425.
Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: a laboratory
manual. Cold Spring Harbor Laboratory Press, New York.
Schmidt, T.M., DeLong, E.F., and Pace, N.R. (1991). Analysis of a marine picoplankton
community by 16SrRNA gene cloning and sequencing. J. Bacteriol. 173:43714378.
Selander, R.K., Caugant, D.A., and Whittam, T.S. (1987). Genetic structure and
variation in natural populations of Escherichiacoli. In Escherichia coli and
Salmonella typhimunium: cellularandmolecular biology (F.C. Neidhardt, ed.).
Am. Soc. Microbiol. pp. 1625-1648.
Sogin, M.L., Gunderson, J.H., Elwood, H.L., Alonso, R.A. and Peattie, D.A. (1989).
Phylogenetic meaning of the kingdom concept: an unusual ribosomal RNA from
Giardialamblia. Science 243:75-77.
Sogin, M.L., Hinkle, G., Leipe, D.D. (1993). Universal tree of life. Nature 362:795.
Swift, H. and Palenik, B. (1992). Prochlorophyte evolution and the origin of
chloroplasts: morphological and molecular evidence. In Originsof Plastids
(R.A. Lewin, ed.) Chapman and Hall, pp. 123-139.
Swofford, D.L. (1991). PAUP version 3.0 (Illinous Natural History Survey, Champaign,
IL).
Tomioka, N. and Sugiura, M. (1983). The complete nucleotide sequence of a 16S
ribosomal RNa gene from a blue-green alga, Anacystis nidulans. Mol. Gen.
Genet. 191:45-50.
Turner, S., Burger.Wiersma, T., Giovannoni, S.J., Mur, L.R., and N.R. Pace. (1989). The
relationship of a prochlorophyte Prochlorothrixhollandica to green chloroplasts.
Nature 337:380-382.
Urbach, E., Robertson, D. and Chisholm, S.W. (1992). Multiple evolutionary origins of
prochlorophytes within the cyanobacterial radiation. Nature 355:267-269.
Vaulot, D., Marie, D., Olson, R.J. and Chisholm, S.W. (1994). Prochlorococcusis
highly synchronized to the diel cycle in the equatorial Pacific and divides up to
once a day. submitted.
Vaulot, D., and Partensky, F. (1992). Cell cycle distributions of prochlorophytes in the
northwestern Mediterranean Sea. Deep-Sea Res. 39:727-742.
Vaulot, D., Partensky, F., Neveux, J., Mantoura, R.F.C. and C. Llewellyn. (1990).
Wintertime presence of prochlorophytes in surface waters of the North-Western
Mediterranean Sea. Limnol. Oceanogr. 35:1156-1164.
Veldhuis, M.J.W. and Kraay, G.W. (1993). Cell abundance and Fluorescence of
picoplakton in relation to growth irradience and nitrogen availability in the Red
Sea. Netherlands J. Sea Res. 31:135-145.
Veldhuis, M.J.W. and Kraay, G.W. (1990). Vertical distribution and pigment
composition of a picoplanktonic prochlorophyte in the subtropical N. Atlantic: a
study of HPLC-analysis of pigments and flow cytometry. Mar. Ecol. Prog. Seq.
68:121-127.
Vermaas, W.F., Williams, J.G. and Arntzen, C.J. (1978). Sequencing and modification
of psbB, the gene encoding the CP-47 protein of photosystem II in the
cyanobacterium Synechocustis 6803. Plant Mol. Biol. 8:317-329.
Vermaas, W.F.J., and Ikeuchi, M. (1991). Photosystem II. in The Photosynthetic
Apparatus: molecular biology and operation (L. Bogorad and I.K. Vasil, eds.).
Academic Press pp. 26-111.
Waterbury, J.B. and Rippka, R. (1991). Subsection 1. Order CroococcalesWettstein
1924, emend. Rippka et al., 1979. in Bergey's Manual of Systematic
Bacteriology, Vol. 3. J.T. Stanley, et al., eds. Williams and Wilkins.
Waterbury, J.B., Valois, F.S., Franks, D.G. (1986). Biological and ecological
characterization of the marine unicellular cyanobacterium Synechococcus. in
PhotosyntheticPicoplankton(T. Platt and W.K.W. Li, eds.) Can. Bull. Fish
Aquatic Sci. 214:71-120.
Waterbury, J.B., Watson, S.W., Guillard, R.R.L. and Brand, L.E. (1979) Widespread
occurrence of a unicellular, marine, planktonic, cyanobacterium. Nature 277:293294.
Widger, W.R. and Cramer, W.A. (1991). The b6/f complex. in The Photosynthetic
Apparatus: molecularbiology and operation (L. Bogorad and I.K. Vasil, eds.).
Academic Press pp. 149-224.
Woese, C.R. 1987. Bacterial Evolution. Microbiol. Rev. 51:221-271.
Woese, C.R., Achenbach, L., Rouviere, P. and Mandelco, L. 1991. Archaeal phylogeny:
reexamination of the phylogenetic position of Archaeolobusfulgidus in light of
certain composition-induced artifacts. System. Appl. Microbiol. 14:364-371.
Yang, D., Oyaizu, Y., Oyaizu, H., Olsen, G.J., Woese, C.R. (1985). Mitochondrial
origins. Proc. Natl. Acad. Sci. USA 82:4443-4447.
Young, J.P.W. (1989). The population genetics of bacteria. in Genetics of Bacterial
Diversity (D.A. Hopwood and K.F. Chater, eds.). Academic Press pp. 415-438.
Chapter Three
GENETIC DIVERSITY IN NATURAL POPULATIONS
OF PROCHLOROCOCCUSMARINUS IN THE
SARGASSO SEA AND THE GULF STREAM
ABSTRACT
An investigation into the genetic structure of Prochlorococcusmarinus field
populations in depth profiles from the Sargasso Sea and the Gulf Stream revealed a high
degree of genetic heterogeneity within water samples, detected by partial sequencing of
cloned PCR products amplified from flow cytometrically sorted cells. Overlapping sets
of alleles were recovered from the two water columns, from different depths within each
water column and from flow cytometrically distinguishable subpopulations within water
samples, suggesting that each of these populations drew its membership from a single
gene pool.
INTRODUCTION
Prochlorococcusmarinus is a unicellular prokaryote which makes up a high
proportion of the photosynthetic picoplankton throughout wide regions of the tropical
and subtropical marine environment (Chisholm et al 1988, Giskes et al. 1988, Olson et al.
1990, Vaulot et al. 1990, Li et al. 1993, Campbell and Vaulot 1993, Veldhuis and Kraay
1993). Sometimes referred to as a prochlorophyte (Lewin 1981, Chisholm et al. 1988),
P. marinus is recognized by its tiny size (0.6 to 0.8 ptm) and unique set of photosynthetic
pigments, which include divinyl chlorophylls a and b (chl a2 and b2 ) and specifically do
not include phycobilipigments, typical of most other cyanobacteria.
In efforts to understand the role of this organism in the marine microbial
community, a number of recent studies have characterized the water column distribution
of P. marinusin terms of cell number, biomass, productivity, chlorophyll fluorescence
and DNA content at several locations (Giskes et al. 1988, Li and Wood 1988, Olson et al.
1990, Vaulot et al. 1990, Li et al. 1992, Vaulot and Partensky, 1992, Cambell and Vaulot
1993, Goericke and Repeta 1993, Goericke and Welschmeyer 1993, Vaulot et al. 1994)
and over time (Olson et al. 1990, Cambell and Vaulot 1993). These studies have
established that P.marinus is present throughout the euphotic zone at all times of the
year and numerically dominates the picophytoplankton at oligotrophic, deep chlorophyll
maxima during thermal stratification. Other, parallel studies have analyzed physiological
and pigment differences among cultured P. marinusisolated from different depths and
locations (Morel et al. 1993, Partensky et al. 1993, Moore et al. 1994), identifying
differences in their growth rate response to varying levels of illumination which suggest
that some genetic variants may have relatively greater growth rates in brightly lit surface
waters, while others may have growth advantages at dimly lit depths (Moore et al. 1994).
Med4 and SS120, a pair of cloned isolates from 5 m and 120 m in the in the
Mediterranean and the Sargasso Sea, respectively, are the most extensively characterized
P. marinusstrains (Partensky et al. 1993, Morel et al. 1993, Moore et al. 1994). Each can
grow over a wide range of light intensities, but consistent with having been isolated at the
surface, Med 4 is able to grow at high light intensities (>100 kE m- 2 s- 1) at which SS120
is unable to grow, and SS120, isolated from a deep chlorophyll maximum (DCM) at 120
m, grows at low light intensities (2 - 6 gE m- 2 s-1) at which Med4 shows no growth
(Moore et al. 1994). The two isolates differ in their compliment of photosynthetic
pigments as well, with SS120 elaborating "normal" chlorophyll b (chl bl) at light
intensities above 20 pE m-2 s- 1 and exhibiting a tenfold higher chlorophyl b (chl b1 +
chl b2 ) to chl a2 ratio (chl b/a2 ratio) than Med4 over the entire range of growth light
intensities (0.05 to 0.15 for Med4 and 0.4 - 2.4 for SS120) (Morel et al. 1993, Partensky
et al. 1993, Moore et al. 1994). Med4 contains no detectable chl b1 under any conditions
tested (Morel et al. 1993, Partensky et al. 1993, Moore et al. 1994).
Phylogenetic analyses using four gene sequences, 16S ribosomal RNA (rRNA),
psbB, petB/D and rpoC, have shown that cultured strains of P. marinus fall into a single
phylogenetic cluster which they share with the marine A Synechococcus, phycobilisomecontaining cyanobacteria with which they also share their marine habitat (Swift and
Palenik 1992, this thesis Chapter Two). P. marinus Mediterranean strain (Med4),
derived from strain MED (Chisholm et al. 1992, Partensky et al. 1993) and Pacific strain
(MIT9107), both isolated from mixed surface waters (Chisholm et al. 1992, J.
Dusenberry, pers. comm.), were found to be closely related, while DCM culture SS120
and others fell into more deeply branching lineages (this thesis Chapter Two). There was
no correlation between genetic distance and the geographic distance between sites of
culture isolation, with two Sargasso Sea isolates exhibiting less sequence similarity than
the Mediterranean and Pacific pair. Fractional sequence similarity among four P.
marinusisolates ranged from 0.994 to 0.989 for 16S rRNA and from 0.973 to 0.934 at
the first two codon positions of petB and D (0.838 to 0.788 at all codon positions).
Intergenic region sequences between petB and D were similar for Med4 and MIT9107,
while sequences from other cultures were so divergent that homologous nucleotide
positions could not be identified (this thesis Chapter Two).
Data from HPLC and flow cytometric studies suggest that genetic differences
may distinguish P. marinus populations at different depths in stratified water columns,
and that different genetic variants may coexist within water samples. At sites in the
Atlantic, Pacific and in the Red Sea, ratios of chl b/a2 were found to be lower near the
surface than at the DCM (Veldhuis and Kraay 1990, 1993, Cambell and Vaulot 1993,
Goericke and Repeta 1993), with the range of chl b/a2 ratios within a water column (0.15
to 2.9 in the Sargasso Sea) being greater than those observed in individual cultures (0.05
to 0.15 for Med4 and 0.4 - 2.4 for SS120) (Goericke and Repeta 1993). HPLC analysis
of samples from from DCM's in the subtropical Pacific has shown that cells passing
through a 0.65 gtm filter exhibit 1.7-fold lower ratios of chl b2/a2 than cells retained by
the filter, suggesting that subpopulations with different pigment ratios may coexist in the
water sample (Letelier et al. 1993). While the flow cytometric study provided data on
fluorescence ratios for individual P. marinus cells excited at 457 and 488 nm, exploiting
the different fluorescence excitation properties of chl a2 and b at these wavelengths to
estimate pigment ratios (Cambell and Vaulot 1993), the HPLC studies examined bulk
chlorophyll, and so could have included chl b from cells other than P. marinus.
Differences in chl b/a2 ratios in these HPLC reports are therefore not direct evidence of
genetic heterogeneity between P. marinus at different depths.
Flow cytometry of field samples from the Pacific, the Red Sea and the Sargasso
Sea have occasionally identified "multiple populations," consisting of discrete
distributions of chlorophyll autofluorescence and light scatter from individual cells
(Cambell and Vaulot 1993, Veldhuis and Kraay 1993, B. Binder, R. Olson, J.
Dusenberry, E. Zettler, unpublished observations). These flow cytometric
subpopulations may derive from genetically distinct subpopulations which express
different pigment phenotypes when exposed to uniform conditions, but they could also
derive from samples containing recently mixed, genetically similar populations
acclimated to different conditions of light or nutrient supply (Dusenberry and Chisholm
in preparation).
Data consistent with the possibility of genetic heterogeneity within P. marinus
populations at discrete depths have also come from molecular cloning experiments
(Giovannoni et al. 1990, Britchgi and Giovannoni 1991, Schmidt et al. 1991, DeLong et
al. 1993, Fuhrman et al. 1993). Investigations in which 16S ribosomal RNA gene
sequences were cloned from uncharacterized marine picoplankton have recovered a
number of sequences which fall into the P. marinuslmarineA Synechococcus cluster.
While some of these sequences (e.g. SAR6 and SAR139) can be phylogenetically linked
to cultured P. marinus or marine A Synechococcus, others (e.g. SAR7) form deep
branches that are not robustly affiliated with either taxon, and so cannot be identified as
P. marinusor Synechococcus with the data at hand (Urbach et al. 1992, this thesis
Chapter Two). In order to assess the the genetic structure of P. marinuspopulations (as
defined by their flow cytometric signature), therefore, it is necessary to physically
separate P. marinus from other members of the community.
We report the results of a direct analysis of P. marinus populations using cells
sorted by flow cytometry as starting material. P. marinuswere functionally defined for
this study as cells exhibiting the P. marinus flow cytometric "signature:" characteristic
values for red fluorescence and forward light scatter (measured relative to standard 0.57
pm beads), in the absence of orange fluorescence, when excited by 488 nm laser light
(Chisholm et al. 1988). Genotypes were sampled from sorted P. marinusby PCR
amplification of the petBID locus 1 with cloning and single run, partial sequencing of the
resulting clones (Adams et al. 1991, Adams et al. 1992, Kahn et al 1992). Preliminary
analysis ofpetB/D amplification products by direct sequencing revealed the presence of
superimposed sequences differing mostly in the intergenic region, suggesting the
presence of multiple petB/D alleles (Appendix I). 21 to 28 clones were analyzed for each
of eight sorted samples from different depths in the Sargasso Sea and the Gulf Stream,
and including subpopulations of two flow cytometrically defined "double populations" in
the Gulf Stream depth profile. Questions addressed are whether P. marinusfield
populations at discrete depths are genetically homo- or heterogeneous, and whether
genetic differences can be detected between populations at different depths, between
1The
petB/D locus includes the 3' end of petB, an intergenic region, and the 5' end of petD, (this thesis
Chapter Two). These genes encode subunits of the photosynthetic bff complex and are single-copy,
linked cistrons with a consistent orientation in all photosynthetic prokaryotes and organelles in the
literature (Vermaas and Ikeuchi 1991, Greer and Golden 1992).
subpopulations in multiple populations (as differentiated by their fluorescence and
forward light scatter per cell), and between water columns in adjacent, but
hydrographically distinct water masses. Gene sequences obtained from field populations
are compared to sequences from cultured cells.
METHODS
Characteristics of sampling sites and flow cytometric signatures. P. marinus
were collected from depth profiles in the Sargasso Sea (330 58.65'N, 660 6.63'W, 10:00
local time, 11 July 1993) and the Gulf Stream (370 30.68'N, 680 13.69'W, 08:00 local
time, 17 July 1993) during cruise 9306 of RV Columbus Iselin. The Sargasso Sea
sampling site exhibited a typical summer hydrographic profile, with a chlorophyll-poor,
mixed upper layer overlying thermally stratified waters and a pronounced DCM (as
indicated by in vivo chlorophyll fluorescence measurements) at 70 m. The Gulf Stream
profile was more complex, with relatively low concentrations of chlorophyll at the
surface and a pair of subsurface chlorophyll maxima at 85 m and 150 m (Figure 1).
Samples were collected at depths predicted to contain the most different P. marinus
populations, avoiding samples at the immediate surface which are difficult to visualize
using the sorting flow cell. Sargasso Sea samples were collected at 40 m, 70 m (the
DCM) and below the DCM at 120 m. Gulf Stream samples were collected at 50 m, 85 m
(the shallower DCM) and at 135 m, between the two DCM's (Figure 1).
Flow cytometric characteristics of P. marinus populations varied with depth
according to patterns familiar from previous investigations (Olson et al. 1990): at each
depth in the Sargasso Sea P. marinus populations formed a single cluster on two
dimensional scatter plots of chlorophyll fluorescence versus forward light scatter for
Figure 1. Depth profiles of temperature (thin lines) and in vivo chlorophyll
fluorescence (heavy lines) at the Sargasso Sea (330 58.65'N, 660 6.63'W, 10:00 local
time, 11 July 1993) and the Gulf Stream (370 30.68'N, 680 13.69'W, 08:00 local time, 17
July 1993) sampling sites. Open squares on the in vivo cholrophyll fluorescence plot
mark depths at which samples were collected.
0oo
-
e,4
-. )
;-
*-~
c
"
0I
a0
o
I-0
0
Q
0
0
o
c
bQ
ct
I-
U
0
00
r-.
0
oo
(u) qldaG
individual cells, with the distribution of values for each parameter approximating a
lognormal distribution. Modes for chlorophyll fluorescence and forward light scatter
showed an increase with depth, reflecting photoacclimative increase in chlorophyll per
cell, in addition to possible genetic differences between populations at different depths.
In the Gulf Stream, the P. marinus surface population again showed a unimodal
distribution, but populations at 85m and 135m were double ("multiple") populations
(Figure 2).
Sample concentration and flow cytometric sorting. Samples were collected
using Niskin (30 1)or Go-Flo (10 or 30 1)bottles. Subsamples (300 to 750 ml) were
concentrated to approximately 7 ml over 0.45 gpm Durapore filters (Millipore) under light
vacuum (5" Hg) and cells adhering to the filter resuspended by vigorous pipetting. P.
marinus cell recoveries as measured by flow cytometry in similar concentrates were
approximately 50%. Concentrates were immediately sorted by flow cytometry aboard
ship using an Epics V flow cytometer (Coulter Electronics) or frozen in liquid nitrogen
for sorting in the laboratory using an EPICS 753. Frozen cells were sorted over a period
of about about 20 minutes immediately after thawing to avoid loss of autofluorescence,
which can start to occur at about 30 minutes (Vaulot and Xiuren 1988, J. Dusenberry and
R. Olson, unpublished observations). P. marinus populations were flow cytometrically
defined according to their red fluorescence (660-700 nm) in the absence of orange
fluorescence (540-630 nm) and by their characteristic forward light scatter (relative to
0.57 gim beads) when illuminated by blue laser light (488 nm) (Chisholm et al. 1988).
Bitmap sort windows were drawn to separate P. marinus from other constituents of the
photosynthetic planktonic community when P. marinuspopulations were unimodal, and
to isolate members of flow cytometric subpopulations when multiple populations were
observed. Approximately 106 P. marinus cells were collected in each sorted sample.
Figure 2. Contour plots for flow cytometric distributions of P. marinus sorted
from Sargasso Sea (a-c) and Gulf Stream (d-e) field samples. Sorted (sub)populations
used for genetic analysis are outlined. Red Fluorescence signals of the dimmly
fluorescing cells in the 40 m Sargasso Sea sample (a) (processed in the laboratory) were
enhanced by increasing the laser power to 1.2 watts PMT voltage to 1500 volts. Other
samples (processed at sea) were processed using laser power of 1.0 watts and a PMT
voltage of 1100 volts. To minimize the possibility of contaminaton standard beads were
not added to samples used for sorting; distributions (e) and (f) are from reference
samples. P. marinus cells were identified by their characteristic red vs. FALS signature
and absence of orange phycoerythrin fluorescence (orange fluorescence data not shown).
Sargasso Sea samples: (a) 40 m, (b) 70 m, (c) 120 m. Gulf Stream samples: (d) 50 m,
(e) 85 m, (f) 135 m. Red Fluorescence: in vivo fluorescence of chlorophyll excited by
488 nm blue laser light. Forward Angle Light Scatter: forward angle light scatter of 488
nm laser light, a proxy for cell size.
Gulf Stream
Sargasso Sea
d rm
1-i
a
c
"
l
i
; I,
1
I?
I I
I !1 1'
II1 1 -
-1T1IC~
1
e
1r
1
I
0.57 gm beads
l '
fII
l
tllltl11ll1lllI11lf!
l
!11
I | ......................
log Forward Angle Light Scatter
log Forward Angle Light Scatter
Isolation of genomic DNA. A modification of the protocol of Li et al. (1991)
was used to prepare DNA from sorted cells. Cells from the flow cytometric sort were
concentrated above 0.2 gm pore size Durapore membranes using Ultrafree-MC filter
units (Millipore) spun at 2000 x g and washed twice with cell suspension buffer (0.5 M
NaCl, 10 mM Tris [pH 8.0] 10 mM EDTA). Cells were resuspended in a final volume of
100 pl cell suspension buffer and lysed by addition of 12 tl 0.5 M DTF and 6 pl 10 M
NaOH followed by a 10 minute incubation at 65 0 C. After addition of 120 pl
neutralization buffer (90 mM Tris [pH 8.0] plus 0.5 mol/l HCI), DNA was precipitated by
addition of 1 tl 20 mg/ml glycogen (Boehringer) and 0.6 ml cold 100% ethanol.
Precipitates were stored at -20 0 C for several months. In the laboratory, crude DNA
precipitates were collected by centrifugation at 16,000 x g for 30 min at 40 C, washed
with 70% ethanol and resuspended in TE (10 mM Tris [pH 8.0] 1 mM EDTA). DNA
was stored at -20 0 C.
Construction of the petB/D clone library. Portions of the petB and D genes and
their intergenic region were amplified using degenerate oligonucleotide primers
containing inosine (Knoth et al. 1988). Priming sites correspond to highly conserved
regions in petB and D amino acid sequences, with forward primer PPETBD314
(5'ATGATGGTIYTIATGATGAT) corresponding to nucleotide positions 274 to 293 in
the Nostoc PCC7906 petB coding sequence (Kallas et al. 1988), and reverse primer
PPPETBD1160R (5'-CCRTARTARTTRTGICCCAT) corresponding to nucleotide
positions 64 to 83 in the Nostoc PCC7906 petD sequence (Kallas et al. 1988). 100 pl
PCR reactions contained DNA isolated from approximately 105 sorted cells, 5 U of Taq
DNA polymerase (Promega) and a PCR reaction cocktail containing 1 x Mg-free PCR
reaction buffer (50 mM KCI, 10 mM Tris-HCl [pH 9.0 at 25 0 C], 0.1% [wt/vol] Triton X100), 1.5 mM MgC12 , 200 pM (each) dATP, dCTP, dGTP and dTTP, 0.5 pM
PPETBD314 and 1 gM PPETBDI 160R. After initial denaturation for 2 min at 94 0 C,
thermal cycle parameters were 1 min 94 0 C, 1 min 47 0 C, 1 min 720 C (5 cycles), 1 min
94 0 C, 1 min 520 C, 1 min 72 0 C (10 cycles), I min 940 C, 1 min 570 C, 1 min 720 C (25
cycles), followed by 10 min 720 C. PCR products from pairs of replicate reactions were
pooled and separated by electrophoresis in a 1% agarose gel. Products in the 400 to 800
basepair region of the gel were excised and purified with GeneClean (Bio 101). 1/30th to
1/1000th of each primary amplification product was reamplified for 30 cycles of 1 min
94 0 C, 1 min 57 0 C, 1 min 72 0 C, followed by 10 min 720 C, in reaction cocktail with 5 U
Taq polymerase added during the first 570 C incubation. Control reactions lacking
template DNA gave no amplication products for either primary amplifications or
reamplifications, as judged from agarose gels.
Pooled reamplification products from pairs of replicate reactions were excised
from a 1% agarose gel, purified with GeneClean and half to all of each product was
inserted into pCRII vector (TA Cloning Kit, Invitrogen) according to the manufacturer's
instructions. OneShot Competent cells were transformed using half of each ligation
mixture. Both white and blue colonies were picked from LB plates (0.5% tryptone, 1%
yeast extract, 0.5% NaCl, 1.5% Bacto-agar) containing 40 gtg/ml X-gal and 50 gg/ml of
either ampicillin or kanamycin. Plasmid DNA was isolated by InstaPrep (5prime3prime) or standard alkaline lysis (Sambrook et al. 1989). Plasmids were digested with
restriction endonuclease EcoRI and the digests analyzed by agarose gel electrophoresis.
Clones containing 400 to 800 basepair inserts were chosen for sequence analysis.
DNA sequencing and database entry. Double-stranded plasmids containing
PCR fragment inserts in random orientation were sequenced using Sequenase version 2.0
(USB) and primer PTASP6 (5'GATCCACTAGTAACGGCCG), complimentary to vector
sequence flanking the pCRII cloning site. DNA sequences were entered into a sequence
alignment editor (Olsen 1990), translated into amino acid sequences and aligned with a
petBID sequence database according to conserved regions in the amino acid sequence
(Widger and Cramer 1991, Greer and Golden 1992). In order to obtain sequences across
the intergenic region for all clones recovered, clones giving sequence at the 5'
(PPETBD314) end of the amplified petBID fragment, and which did not match sequences
already in the database, were sequenced from the opposite side of the pCRII cloning site
using PTAT7 (5'GAGCGGCCGCCAGTGTGA), which was closer to the intergenic
region. This procedure yielded some sequences spanning the entire length of the cloned
fragment, while others covered only portions of the 3' end of petB, the intergenic region
and the 5' end of petD. Sequences, which were determined in a single run and in one
direction only (Adams et al 1991, Adams et al. 1992, Kahn et al. 1992), ranged from 71
to 562 basepairs in length. Clones with sequences which did not align with petB or D
were eliminated from the analysis.
Sequence comparisons. Fractional sequence mismatch values were computed
using the computer program of Olsen (1988, 1990). The phylogenetic tree was inferred
using the Phylip version 3.4 neighbor-joining program (Felsenstein 1991) and Kimura
genetic distance calculations (2:1 transition:transversion ratio) (Kimura 1980). To
minimize effects of sequence determination and PCR errors, alignment gaps were
omitted from sequence comparisons.
Population genetics calculations. Values for E(Sn) were calculated using
COMPAH90 (E.D. Gallagher, University of Massachusetts/Boston).
RESULTS
Sequence diversity in the combined dataset. The combined dataset for all eight
sorted samples contained 191 clones with inserts homologous to petB and D. Among the
191 clone sequences, 71 alleles were identified as deriving from genetically distinct
individuals in the sampled populations (Table 1).
The 71 alleles were identified according to a two step process, employing both
qualitative and quantitative criteria to distinguish true genetic differences from sequence
mismatches arising from PCR and sequence determination errors. In the first, qualitative
step, sequence comparisons were used to identify 59 sets of clones among which
intergenic regions varied in length from 19 to 91 basepairs, and in sequence to such a
degree that homologous nucleotides could not be identified for use in sequence
alignment. Differences among these 59 sets were deemed qualitatively different from
PCR and sequence determination errors (which were nonetheless undoubtedly present).
Most of the 59 sets of clones having unalignable intergenic regions contained single or
identical clones, or clones exhibiting minimal differences attributable to error, but three
sets contained clones having sufficient mismatch (up to 19.1%) to suggest that they had
been amplified from genetically distinct, though related individuals. In the second,
quantitative step, a criterion of 3.5% mismatch was applied as an upper bound for
sequence differences considered indistinguishable from PCR plus sequence
determination errors in order to subdivide the three sets of clones and add an additional
Table 1: Frequency of genetic alleles among cloned amplification products from flow
cytometrically sorted samples.
Allele
/ Sample
P. marinus
1 Allelel
2 Allele2
3 Allele3
4 Allele4
5 Allele5
6 Allele6
7 Allele7
8 AlleleS
9 Allele9
10 Allelell
11 Allelel2
12 Allelel3
13 Allelel4
14 AllelelS
15 Allelel6
16 Allelel7
17 Allele19
18 Allele21
19 Allele22
20 Allele23
21 Allele25
22 Allele26
23 Allele27
24 Allele28
25 Allele29
26 Allele30
27 Allele31
28 Allele32
29 Allele33
30 Allele35
31 Allele36
32 Allele37
33 Allele38
34 Allele39
35 Allele41l
36 Allele42
37 Allele43
38 Allele44
39 Allele45
40 Allele46
41 Allele47
42 Allele48
43 Allele50
44 Allele51
45 Allele52
46 Allele54
47 Allele56
48 Allele57
49 Allele58
50 Allele59
51 Allele60O
52 Allele61l
53 Allele62
54 Allele63
55 Allele64
56 Allele65
57 Allele66
58 Allele67
59 Allele68
60 Allele69
61 Allele71
62 Allele72
63 Allele73
64 Allele74
65 Allele75
66 Allele76
67 Allele77
68 Allele78
G50
G85Br
G85D
G135Br
G135D
S40
S70
12
16
5120
TOTAL
14
5
1
1
3
1
1
3
2
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
2
1
1
1
1
2
1
1
1
1
4
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
P. marinus Total
No. different Alleles
E(S19)
23
23
14
6
5.3 12.1
24
6
5.2
23
6
5.4
23
16
13.7
19
8
8.0
28
15
10.5
24
21
16.8
187
67
10.9
Chloroplast-like
69 Allele20
70 Allele24
71 Allele40
TOTAL
23
23
25
24
23
21
28
24
191
Abbreviations: G50, 50m; G85Br and G85D, Bright and Dim clouds of 85m double population;
G135Br and G135D, Bright and Dim clouds of 135m double population; all from the Gulf Stream
depth profile. S40, 40m; S70, 70m; S120, 120m; all from the Sargasso Sea depth profile.
E(S19), estimated number of different alleles for a constant sample size of 19 clones.
83
12 alleles to the dataset 2 . The subdivision of these three sets of clones according to the
quantitative criterion resulted in the total of 71 alleles (Table 2).
Pairwise comparisons among alleles identified by subdivision of sets of clones
having similar intergenic regions show a mean fraction of sequence mismatch at third
codon positions (fractional mismatch at third codon positions/3 x fractional mismatch at
all codon positions) of 0.82, significantly different from the expectations of random PCR
and sequence determination errors (this thesis Appendix II).
A prototype sequence (the longest and least ambiguous) was chosen to represent
each allele in sequence comparisons and aligned with other sequences in the database
according to conserved regions in amino acid translations (Figure 3, this thesis Appendix
IV). Translations of prototype sequences were highly homologous to petB and D
sequences already in the database, with all translations containing phylogenetically
conserved amino acid residues, including histidines at positions 186 and 201 in the
Synechococcus PCC7002 sequence, hypothesized to participate in heme binding (Carter
et al. 1981, Widger and Cramer 1991, Hope 1993), when sequence was available for
these positions. The most striking differences among alleles were in the intergenic
region, which differered in length and sequence among alleles. In coding regions
pairwise nucleotide comparisons among the 71 alleles showed that they differed by an
average of 24.1% ±4.9% (Figure 4a), with most differences occurring at third codon
positions which differed by 51.4% ±9.5% (Figure 4b). The more highly conserved first
two codon postions differed by only 10.8% ±6.8% (Figure 4c). Since the overwhelming
2The
3.5% upper limit criterion for sequence mismatch among clones assigned to a single Allele was
arrived at by an optimization protocol which considers the fraction of sequence mismatch at third (silent)
codon positions for sequence comparisons along a scale of total sequence mismatch. Within the bounds of
theoretical limits, the criterion divides the set of pairwise sequence comparisons within the three sets of
alleles having similar intergenic regions into the two subsets having the greatest possible difference in their
mean third codon position to all codon position mismatch ratios (this thesis Appendix II).
Table 2a. Fractional sequence mismatch at all sequence positions, including intergenic
regions (x1000, below the diagonal) and fraction of sequence mismatch located at third
codon positions (fractional mismatch at third positions/3 x fractional mismatch at all
codon positions, for coding regions only) (x100, above the diagonal) for pairwise
comparisons of prototype sequences for Alleles having intergenic regions alignable with
Allele 1.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Allel
Alle2
Alle3
Alle69
Alle71
Alle72
Alle73
Alle74
Alle75
Alle76
1
2
3
4
5
6
7
8
9
10
126
158
057
080
059
073
148
081
119
082
124
133
167
148
156
166
157
154
089
092
160
173
134
169
190
155
191
094
083
091
095
051
070
153
077
129
085
079
085
084
057
077
181
072
180
100
080
086
100
100
076
154
048
147
090
078
084
083
079
077
143
086
142
090
080
079
085
080
080
085
155
142
081
078
089
083
071
071
060
071
164
080
073
080
078
072
059
078
084
057
-
Table 2b. Fractional sequence mismatch at all sequence positions, including intergenic
regions (x1000, below the diagonal) and fraction
of sequence mismatch located at third codon positions (fractional mismatch at third
positions/3 x fractional mismatch at all codon positions, for coding regions only) (x100,
above the diagonal) for pairwise comparisons of prototype sequences for Alleles having
intergenic regions alignable with Allele 4.
1.
2.
3.
4.
Alle4
Alle41
Alle77
Alle78
1
2
3
4
132
042
045
102
138
160
101
102
060
079
091
079
-
Table 2c. Fractional sequence mismatch at all sequence positions, including intergenic
regions (x1000, below the diagonal) and fraction of sequence mismatch located at third
codon positions (fractional mismatch at third positions/3 x fractional mismatch at all
codon positions, for coding regions only) (x100, above the diagonal) for pairwise
comparisons of prototype sequences for Alleles having intergenic regions alignable with
Allele 17.
1. ALLE17
2. ALLE31
1
2
075
071
-
Figure 3. Intergenic region sequences for 71 petB/D alleles recovered from the
Sargasso Sea and Gulf Stream field samples (this work, Genbank accession numbers ),
cultured P. marinusstrains Med4, SS120, FP5, MIT9107 (this thesis Chapter Two) and
MIT9313 (this work, Genbank accession number), Synechococcus strains WH8103 (this
thesis Chapter II) and PCC7002 (Brand et al. 1992), cyanobacteria Prochlorothrix
hollandica(Greer and Golden 1991) and Nostoc PCC7906 (Kallas et al. 1988) and
chloroplasts from Chlorellaprotothecoides (Reimann and Kueck 1989), Marchantia
polymorpha (Ohyama et al. 1986) and Zea maize (Rock et al. 1987). A potion of the
sequence alignment is shown, with the final six codons of petB at left and the initial eight
codons of petD at right. Intergenic region sequences are arbitrarily represented as
contiguous with petB, with alignment gaps inserted between the end of most intergenic
region sequences and the beginning of petD. Intergenic region sequences for P.
hollandica,Nostoc PCC7906, C. protothecoides,M. polymorpha Z maize are truncated.
Designations on nucleotide lines are clone names. Allele designations are indicated on
the translation lines below.
-4 Eý Eý
if -
"-4
4)
r-) E
E-4iE-4i iE- i
- 1E-
E
CC
E-
I
0i
64
14 UE-
H
E4
i
E-4i E4
z
IM.
C,
E-
i0iZ
5. 1CDMC
I14 8
b4
4 I
eel
>U
1-4UE4
-4
0
U
-1 -4>
X
H
o
E4
0,
CA
I
U
-1E4
~
t
2
iL)E-4 i
L)(d$
4
I
II
0
$4(d
C 4
E
(i
41 Ty
4-4
1
L~
4
U
~I~
f<
U
I
4L
4t4
ua
ER 1
E41
~U
4 JJ
14
04
-
R,
p
.
0
0
"%C44
0
OD
M
m'
1
4
IA
vi
0
m
4t'.
0
0 w0IV
-4 0 .-4ý
0-1M
0 r 0-0
toM4 La PC4LA .4 M~ 4
Lo .4
.4
4
0
CR
E-4 I
P
1 Cl
1
E-
4
l--IE
f4
II
1
I
E4-4E
-1
1
'
1
6u0 uIuu M ..
u4I
.
U~.~t.~l
IU
1u
U I
1U4U4U1U4U4V
4UUUU4U4U:.:U
u
V1
0
E
14
a.ua.ci L
0
0
10L-
.1F
U
4U4
.-e4
1
f
%D
0
%0
In
.4
00
.4
0%
4 .-
C4
41
N
U
I
E-
rto
400%0
·
E-
~ý
00
.U:
UE
I44
.
4E-u4vl
W
I
-
1
E-4E44u a4 i m
0.-
0
0.-q
06
U
--E--I,
4
I
I.
E-i 0
4E-
U:
EE E.
iU P 5U * '4.U:4
U
U
U1
-4
4
4 34
P.
E-
4
Uo
.U:U.U
WI.)R
u
,-4
4
U
I
I
U
U
4-)
I4
oas
E-
00
0 (a
010.- n
0.0
4
944
L
U
ti
)
1
-
-H0(
.U:
U
run
U
U
1
C
0
-
.
V
1
4
I4 U
c
4 f--4
OE 4P
1
i
UU
4.
4
4
4
U
0
U
4
4
U
U
r.4
Cl)
0
44.4
L
E- 4
IU4
X~2CE-4l
I
i
4
-
I
4,
U
*2~
ERI I
1
41
4-f 4
04
.*~2~
E4.
V
(9 iU
I
,4I4-
4~
464
4
I
-4
44 to
044 4
V
0t
("
1
l
0
r3
E414~
r,
1
.
E-E4~
n
tq
cq
NO
40
'U
~)C)CI
cn
LA
0%
LA
0ý
LA
1-4
4c
-4
qLn iD1 - 04G
4
o000, -4 eq eq
ml
Ln w D O 41-o0O
e
C'l) W444
eq
('I
('C4
cq
0 m
.- 4-4-4
.-4
.q
14
.4
14
-4 t
C30 %C0
4
000000Cowo
)0
04
0W
0W4) W4)W40) 4 4)M40)140L4)W0
0
0000000M
W 0
)4
0W
4-r 0M-4 0-4
-44aM -0 40M-4CO-i 40M-4 i) '-4 Q) 4 ('4-4 LA r-i LA r4 V) ,4 LiA4LA -i LA r-4
M 40r-40 M
-4 0 4M
00 ,4 GO O O
4 .4 OD .4 OA
CD 4
D.
-4
D44- 00
4CO 00-4- CO4 OD4 -4-4
ao4
ml
-4
-4 PC-4
-4 14
1-4( - 4
.- 4
eq
a
-
0
ý4
ý4ý
87
E4tE4i E-4 i E-4i P
E-4 1EI41-
E-4
'
E1
E
a
-4,
E-4 i
.
,E
4
a ý
UI
jI
I
cI
I
L)
I
I)
IL)
s
IjI
E
I)
'E,
) E-40- i
-
-
IOII
_-,E
a-
'E- '0
4lt
I
.
I
I
) iE-
U
I)I
C.)
I0
)I
*I
g
I4
Id
b4,I
I S4
If
x
b
I
.I
.4
ý
I
b
I444
.4
:i
w
I~
w
I b,
It
I
x
E, I
I
M, I
I
,
1 (4
.11
1I9
E
'I
I E4 g
I1
r
I
l
11
I
E114
b44~
blI
IEý
w
a
I
0h
'10
I~I
-4
EEa)
E4
v
L)
EI
i
lcll~~
10000
Pp
IIIUIII0LOL
n
c
4
~
e
eq c~1
00100
~E
e~l jI)
U).,4 Un.-IU)- UU
a-1 -I1 -Ij)
a4M -
-
0
L
-4
.I
e
.- -410~-4N
0 e 01I~01U)
-4
jI)
I)
I)
-IU'
.IU)
aIa- -IaI
I
E88
C
E-44~C
2
E41
L)
N0
I
00
e
0
0
-4~- r~ e
~ ~ E)E~aI
aa-
.I
-
-I0
I
~~)
-40 -
-4 N aI
~0 0 ~I100
0
)
)
0 a- 0 .l
-IN
a-
a4 N
NC1010
0
.I0
-
0
-I0
N aI
e
0
U
2
.4
O
- 0 a-IN
a-
~eqeq
CItC
I-
r-)
L)54-
4 ).. 0
a a- -I
4
0
p
I4,
t
u
4
~0 ~0E4
'
0
0
i1 0 ()
U)
U')
U)
')E)
1
oo
o
o
o
o
o
q -4eq.I q - e ,I q~ieqaI q))
-Iai
-
.I.4
aI
-
ai.-
-I.I
4.-
-
0-1010-41
L)I
E I
II
01
E-
tjI
C)I
r)I
I
I
I
1
I
-
IPI
WI~~~~~~
)I
IdM 2
4
I4
L)I
1
LI
I
W
L)I
I
I
Id
10
-iE4iE4iH
0
0
luI
r)I
LI
II
1g
0
I
iE4iH
0,0
IL
Ilu
I
I
Ir)
tU
IU
0
UI
iH
0
iH
UI
I
d
)I
I
hd
I
I
I
IA
c
F
I
IIf
I
0
cI
EjI
0
UI
I
I
0
0
0
tjI
uI
U I
ED
a
E9
0
.
I
UI
W,
X
E4
I
WM
iE4iPApE40
I
b4
I
I.
~ ~
eP
0
0
I
I
I
0
I9
I X
1'
tU
0
Iý
V
.4
3
a4
EIE
CA
II
IA
I IA I0 IF
I'
I
W
A
(iI
II
4
tI
I
I
I L)
I
I
U
I,
4
E-I
V~~
'
4
P
)
I
E44
U
0U~
t
4
E4
1
E'
E4'
4IEI
E
E4
Ii
E44
P.0 0
0
0
a
1,
0
0
0~
0
0
.5
E--Iý
89
riOo
0~~'4
4
U
a
I
'4
041N
to
mo
8=
• r)41.4
0 8 -•
-4
-, i
el o)
E4 0.. 0
0-
E-4
>
N N~
E-ii
-4 to ( C'J
13
0 .
Figure 4. Frequency distributions for nucleotide mismatch in all pairwise
comparisons of allele prototype sequences. The number of nucleotides in each
comparison varies according to the lengths of the different sequences. a) all codon
positions, mean 24.1%, std. dev. 4.9%. b) third codon positions, mean 51.4%, std. dev.
9.5%. c) first two codon positions, mean 10.8%, std dev. 6.8%.
o
0
0
&
0
92
S
0
0*
o
I/1
CV
0d
0Z
0 o0
0
0n
0
0
Z0
trn
suospJIdimaoa uanbaS as!miied jo aqtunN
QU)
0
-
0
0
0
I
majority of sequence differences among alleles are located at evolutionarily
unconstrained sites, most mismatches among alleles are likely to be due to evolutionary
processes. Thus, even though some prototype sequences undoubtedly contain errors, the
alleles named in this study serve the purpose of identifying genotypes present in the
natural populations.
Phylogenetic affinities of the cloned sequences. Neighbor-joining analysis
including twenty of the longest prototype sequences revealed that most alleles belong to
the P. marinuslmarineA Synechococcus cluster, within which P. marinus and marine A
Synechococcus lineages are not well resolved (Figure 5) (this thesis Chapter Two). The
presumption that alleles branching with Synechococcus WH8103 in the tree actually
derive from correctly sorted P. marinus cells, and not from contaminating
Synechococcus, is reinforced by the observation that Type 6, an allele which branches
close to Synechococcus WH8103, is highly similar to petB/D sequences from cultured P.
marinus MIT9313 (99.7% sequence identity over 260 bp) and P. marinusMIT9303
(98.3% identity over 272 bp) (Figure 5b). The possibility remains, however, that some
sequences recovered in this study do, in fact, come from mis-sorted Synechococcus. To
reflect this possibility, alleles which fall into the P. marinus/marineA Synechococcus
lineage should be considered as presumptive P. marinus alleles.
Three alleles, numbers 20, 24 and 40, do not cluster with P. marinus/marineA
Synechococcus, but instead form a phylogenetic group branching below the divergence of
chlorophyte and land plant chloroplasts (Figure 5). The position of this cluster is
reminiscent of the phylogenetic branch point for rhodophyte, pheophyte and some green
algal chloroplasts in 5S rRNA, 16S rRNA and psbA phylogenies (Van den Eynde et al.
1988, Maid et al. 1989, Douglas and Turner 1991, Markowicz and Loiseaux-de Goer
1991, Oyaizu et al. 1993). Thus, these alleles may have originated in plastids of
Figure 5. Phylogenetic relationships among 20 petB/D alleles cloned from flow
cytometrically sorted populations in the Sargasso Sea and the Gulf Stream, cultured
strains of P. marinus and other members of the oxygenic phototroph radiation. (a) tree
inferred by neighbor joining using 163 nucleotides at the first two codon positions of
petB and petD. Alleles 20, 24 and 40 form a cluster allied with green plant and algal
chloroplasts. The remaining clones cluster with P. marinus and marine A Synechococcus
WH8103. (b) tree inferred from 77 nucleotides at first two codon positions in petB and D
to include short sequences from P. marinus cultured strains MIT9303 and MIT9313.
Allele 6 is 99.7% identical to P. marinus MIT9313 and 98.3% identical to P. marinus
MIT9303 for all sequence positions.
-C
Cl
Oso
0
.t U
ta
LmII
=J1
97
c
d
OCO
i-7
6
U
6i
0
U
COO
0
-
W4
I
v
cryptophytes or coccolithophorids, both of which were present at the Gulf Stream and
Sargasso Sea stations (R. Olsen, pers. comm.), or perhaps in plastids of the newly
discovered marine chlorophyte Ostreococcus tauri (Courties et al. 1994), which may
have been liberated during sample concentration. Liberated choroplasts lacking
phycoerythrin would match the flow cytometric definition of P. marinus used for cell
sorting and could therefore have been included in the sorted samples without machine
error. The three chloroplast-like alleles were omitted from the population genetic
analysis, leaving 68 P. marinus alleles in the dataset..
Phylogenetic affiliations of alleles not included in the neighbor-joining tree
(because of short sequences) were assessed by comparison of mismatch frequencies
between these sequences and selected sequences used to infer the tree (Table 3). All
alleles not included in the tree were more similar to at least one member of the P.
marinus/marine A Synechococcus cluster than to sequences outside the cluster and so
were included in the population genetic analysis.
Allele counts for P. marinus in sorted samples. Clones recovered from each of
the eight sorted samples contained between 6 and 21 P. marinuspetB/D alleles (Table 1).
Correcting for unequal sample size by the method of Hurlbert (1971), the estimated
number of alleles present in a sample size of 19 clones for each sorted sample varied
from 5.16 to 16.83 (Table 1).
To assess the completeness of sampling for genetic diversity, rarefaction curves
were generated for each sorted sample and for lumped samples containing both members
of flow cytometric double populations, constituents of entire water columns or the entire
dataset (Figure 6). For a sample which contains a good representation of the genetic
diversity in its parent population, a plot of E(Sn) (the estimated number of genetic
w
0
mcc m
H
oN
0
H
.m
nm
"N H
.-
H
D Co
%D l
Co
l
-4
k
c
O
H
(N
H
r
ko 0 0
n) Ho
H
H
k
0c
cc
m
ON
H4 (7%LA
H H4
N
CY) N-
N-
HHH
H
M (0
m
H
H
0
"
m
"N
H HH
m
H
HrC)H
HO
n
(N
cc C)
C)H
0
(N
C)>
Hm
1
C
c0
LA
co
kD
cc
m
cc
p
0
c0
Nr-f
c4
c,40
H
H
cc
cc
cc
L)
NH4 C)
H
H
LA
m
H
m
-
H H
m
cc
c)
C
cc
0
H
0
H
H
P)
-V'
m
N-
H4
0
00
N
o
H
N
H
c
"
H-
m
LA
(N
H
m
m
LA
H- w
HCo C)
H
"
H H H
LA
N
--
cc cc
qv
LA
C)
LA
)
H
cc k cc 4
H1 %D 00
H C) C)
LA
NC)
H-
H
m
cc
H %
H4 C)
LA
HH
LA
cco cc
C>
C)
LA
rN
C)
cc cc
LA 0)
U)
H C)
(y
(N
C)
N
(N
C)
LA
"N
C)
cc
)
C)
I
r)
H
N
LA
C)
"N
H
N
LA
LA
mc
N
HOHH
-N c
Ch cc
)
C)
LA
C)
H
Hi
LA
C)
M
C)
N
C>
H
H
M
C)
"
-v
H
LA
mc cc cc
OHOH
cm
H
H
)
v
c0 cc 0 c0( 0
C)
)
mL
C)
"
H
HNv
)
N-
(NN
H
N-
(N
c
H
cc
c
cc
H
cc
c
cc
(N (N
"
cc (N cc
Co C>
cc
N-
Ln
(N
(" (
Cc "
C) C)
"
LA
C)
C)
O
H
N
(N
H
C)
H
ý
cc
O
C)
N
cc
H
CD
H H
H
c
cc (N
m
N
c C)
H
cc
C)
H
H
c>
N
N
HO C
0
q
m
C)
OO C)
c)
C)
c ý-c
-v
"
LA
c
C)
C)
ccc
HH
LA
cc
C)
c
cc
C)
,)
c> ccw
H1 (N cc cc
H
HC)HC) C)
C)
cc
C)
m
H
H
H
H
0
N
H
H
H
0
LA
T
co
H
(N
0L
0L
U
H
LA
m~ cc
H
0
0
q
H
cc
0
0
a
0c
)H
m
C)
H
C) LA IT
C) C)
Z)
C)
NC)H
cc
n
H
m-
%
cc)
LA
C)
C)
C) C) C) C) H H C)
cc
c
ccm ( cq ch HNr
H
C)
H
r-c)
Nm
4H C
cc
(N
H
cc
N
c)
cc )
r
H
cc
C)
O
v
c
LA n
C
H
cc
H
H
0cc
C)
Nc)
oc
cc
C)
LA
C)
O
.
0)
c N T
o
Ln
C) C)
mT
H
C)
CD 00
C)
cc)
O
cO
" nm
H- C)
H H
c
cc
C)
c
HC)
q
cc k
N
N- c
Nr L
L
ko
cO
LA HOO
Hq CD m C)
mc rr HOHOH
o N- cc cc OH
T HAH LA
O
H
-A P) LA
1 N c
-4 c
-1 HL
H
(N
o) c>
m
P)P
m
P)
)
rm
C)
cc N
LA
m
cc
C)
H
c
NC)
0
cc
C)
w
T
LA cc
H4 H
H
c
C)
H
H4
-4
cc o
q(3)
y
H
H
H
H
H
N
0m
Hk c
H c 0
v
c> c)
C) (O) N- c
H
HOC HH
H H H C
cc
c
C)
H
H
0
cc ma
cm
r
LA 00 N- C)
C) C)
Cc
) H
-( C)
H
c mcq (N
N
CD 'T~
o
C)
N-
c
:)
(N
b0
N
rcc
HO
C
N-
H
H
0
0cc
(D
r-)
T
H
(N
L
H
0N
H
N
cc (N
C)H
c)
c
T
0
H
c)
H
H
H
H
Hm
LA
r-
<D
7N
H
N
HH
oc
:
k0
N
o
A
A N
H
H H H H
-,:N
(N
H
H
cc
mY)
H (N
H
H
"
r~
rH
m
m
0
H
u
H
rH
cc
m
cc
cc
oo
H
m m
HOHHOHH
N
cc cc
100
H
0cc
0
HC)H
cc LA
LA cc
C)
HC>)
Hi
cc cc
cc0 ýZ IV cc
cN HO LA
(N
c
k-
CN
H
C)
cc
(N
(D )
c
N
H
C)
C)
ON
>
(N
C>
O% C(N
N- 0
H
H H
N
04
C
cc
L
-N
H
'
0
C)
(N
cc
ýT
z
-
L
u
LA cc
q
cc
cc
)
ON cc
N
N
O
(N
H
H
H
H-H
m c
r
H
H
Ho mn
rN
H Hco HLA
(N
CD C)
H H
Cm
H
H
c
T
LA
H
D
LA
N
H
cc
cc cc
H
(,4
0
c
r-ACC)1
H
cc
o ) o
c
CD C)
H
LA
C)
N
c
c0)
c)
w
N9
0N
0 c 0
Ncc cc m
H
(N C) H
C) C)
c)
cc
m~
N
H
LA
c
C)H
m)
c0
C)
c
H H
"
N
W H
op
m
c0
c
m
O
)
O
%D cc
••••0
0 LA
m 0 cc H
cc )
(N
cc cc cc LA H
H H
C) CH
cc
r-
0c
H
N
C)
C)
LA
NC)
(N
0
w
c)
N
0
N
m
0
cc
H
N
m
(N
m
0 LAL0
L0
0N
cc
C)
C)C
cc
N
H
N
"
m
(NH
m
c)
No
m
0c
H
H
r
HC
H
c
CH
N0
4
w
H
H
H(N
A
OccCr-
L
"
m
q
C)
HO C)H
LA
m
Pm
H
0
m
N
H HOC)OC)4 H
cc
H
N
N
c
m
m
N
)
0k
C
(NO C
0N
Nr-
C)
N
H
N
0
0
H
c)
0c
C)
c
N
r- m
-v
m
H
LA
c)
C)
(N
L
m
m
H
m
c)
rN
mm
m
(N"
"
(N H
H1
C)
C)
H
H H
N
W
C)
m
H
rN N
00 m
H H H H
0
H
0c
H
4
C
-q
r
0N rN
H
H
M0••00
q
C•
H
H
0
rn "N
CN C)
H
Ln
m
H
C)
P) mm
m)
C)
H-
H
c)
H
L0
cc
C)
LA
In
r-
(n
In
W
H
(N
0
H
zr
0
'0
ma c)
O
m
co ro to
(NH 0
r) In
m
•
H
rn
H
-v
cN
m
ý
0
N
c
M
H
N-
N0
H
1-4
c)
H
H
M
In
c>
ui
C)
%D) -
c>
o)
NIn
In
m
In
m'
m
In
H
m
(N
H
o
-H
0 H
0%D (N 0mH
H H H
C
HH
M
C)
H
)
a)
C)
00
H
NH
In
ay)
-T
H
M
H
H
C)0
rI)
H
H
H
(N
H
" (D
a)
In
H
H
a)
m)
rO
H
-z
(N
i0
W
N
H
w0
0
H
a
c>
0
n
H 4N
N--
H4
a)
(N
C)
m'
H
Hi
-0
In
H•
c,4
H
•
"
H
N
H
H
(N
H
(
NC)
(
CD C)
H H
(rN
,o
H
D
H
C)ý
o
H
-A C
ra)
Q
•
H
H•
m~ In
0
H
In
Hi
0
H
c)
ýz
H
c(4
o)
'ý)
r-
c)
,o
Nn
m
H
H
NH
In
0
N
0r-)
(N N
H
0
C)k
0
a)N I-al
C)
H C)
,T
N-
<
4- %0
HH
ýD
Nl
C)
H
HH
-ý
In
H r0 HIn Hm H
(N
C)
H
'C
m
ON -:I
C)
H
C)
(N
H
N
m•
(N
(N
Cn
H
U)
N
SC
H i
H
H
N
(N C)
HH
N
H
In
ko
C)
(
'
'
N
'
C) H C
H-
N
C)
HN
C)
H
a
N
H
)
M'
H
M
H
m
a)
'
In
)
In
H
H
- Hv (N
H
H
H
H
H
H
(N
H
(
C)
H
N
OD
C)
0
N
H
tlj
"N--
m
CC) NO m
a)
N
H
H
C) C)
C)
m
(N4 'IT
H0
N
In
H
r-
)
C') a)
H C
In
')m
CD H
C) C
N
0
a)
N
H
It)
a)
m'
O
C)
(N
(NH
C)
)
H
C)
a)
H
C)
00
H
a)
wt
H
N
O.
H
(N
a)
H
c)
(N
(N
T
C)
(N
(N
C)
(N
ý
aNL)
In
H
H
H
H
(
O)
a)
H
C)
OON T
IC) N
H
H
m
wt
N
q
C
v
Sa0)
H- C)
-zC'm)
c4
C)
rn
cc)
-v
c,4
C)
H-
C
H-
a
C)
C)
H-
H
H-
C)
O
C)
)
C)
H
N
"
C)
a)
0 0
C) C)
a)
0
C)
In
C)
H-
H
H
H-
N
0
C)
I)
C)
a)
e
N
Hl
)
04
O4
0U4J
H
H
C)
H
(
C)
C)
(N
a3) c)
m'
C)
C)
a)
HA I)
kC
q)
rn
a)
C)
C)
N
(N )
%C) a)
H C) C
(N
I
C)
C)
.
C)
C)
a)
0)
m
C
N- mO
HA N(N
A
H
H
H
H
C
C)
O
H
H
C)
C)
H
C)
I)
c)
N-
--;
I)
In
H-
a)% (N
In
a)
C)
(N
C)
N (H 4 0C)0C)0C)0C)0C)0n
n C)
M
a)
C)C
(NH4C)
H
C)0(H CH C)C)H C>
r-.4
4-4
oC)
a)
H,4 C)
qz'
(N
C)
m
H-
NN
C)
(N
CO
N-
a)
C
C)
N
-I
(N
-
IC)
N-
CO a)
H
NC)
oT
H
%C) m'
N N
H-
kC
0(
a)
C)
N
a)
m
C)
(N
C)
-H
It)
C)
H
(
a)
m
C)
H H
H H H
w
m'
CD "
H
N
m
-
C)
C)
N
m
C)
C)
C)
H
C
c)
-C'
C)
H
rN
r-H
N
C)
In
-:
H
-T
m
C)
C)
I
C)
(N
N
N
C)
In
C)
kC)m
H a)
r-N
O)
H-
n It
CF) NH
(N
kt)
N-
C)
n
0
0 •)oH
O~
H
9
maN
H
"e
m'
'
N
.-
(N
C4
-)z N I
a)N HA N
(N
H
0
c•.a
() o
C)
M
N-
a)0
a)
H
n
0)
A oe
Q4 ý
101
Figure 6. Rarefaction curves for clones recovered from a) individual sorted
samples, b) lumped samples containing both members of flow cytometric double
populations or constituents of entire water columns and c) a lumped sample containing
the entire dataset. The upward trend of all curves indicates that the number of alleles in
all populations are incompletely sampled. E(Sn) (the estimated number of alleles for n,
the number of clones in a random subsample (Hurlbert 1971).
102
IE
e4
CE
kE
Co
vE
Eh
C=
C
_
(us)3
(us)•
CI
Cn
co
C
6
W4,"
omI
An
2k.
ECW
C
ir
C
%n
z
oC
(us)3
(us)g
(us)3
t,
E=
0
0
0
N
LE
E
o
%
0 0.
00
0
0
U,
.
r-
V
In
V
M
f4
-
C
V- N
u qaC
s),
(US)3
103
I
1-V In V
(uS)a
U,
0
IMr4 C
U-
'0
0~
cE
z
(US)a
(us)O
0
s
T
0
0u
e
In
0 0
0
(us)a
104
0
o
I
I
I
I
S
C/)
f4 -
0
0 if
rt)
-
•
t •,•
•
0
0
(US)t',
0
105
0
D
variants for subsample size n) versus n approaches horizontal as n approaches the sample
size (Hurlbert 1971) 3 . Curves for each of the eight sorted samples, and for lumped
samples show upward trends of E(Sn) with n, and no indication of curve flattening. It is
therefore apparent that many more genetic variants were present in the Sargasso Sea and
Gulf Stream populations than were recovered in this study.
Distribution of alleles among sorted samples. Of the 68 P. marinus alleles in
the dataset, only 16 were found more than once, with 12 appearing in more than one
sample. Since rarefaction curves indicate that populations in the different sorted samples
were not exhaustively sampled, the absence of shared alleles is not considered evidence
of dissimilarity between samples. However, the shared presence of alleles is positive
evidence of similarity. Scoring alleles according to their presence in the different
samples reveals no consistent pattern of shared genotypes, but instead suggests that all of
the sorted populations drew membership from a common gene pool (Table 4). Of
particular interest is the finding that eleven alleles were shared between the Gulf Stream
and Sargasso Sea depth profiles and that two alleles were shared by members of the
double population from 85 m in the Gulf Stream, and one by the pair recovered from 135
m.
3 This
assumes a large parent population with a relatively small number of genotypes. It is possible to
imagine a parent population in which every individual has a different genotype, in which case rarefaction
curves would never flatten out, even for samples approaching the size of the parent population. However,
even for this imagined situation an upward trend in the rarefaction curve for a finite sample would indicate
that unsampled variants were present in the parent population.
106
0
8
E
**
*
*
o
*
*
0o
o
'o
so
w
00
-4o
0
EI
*
*
*
*
*
*
*
*
czo C
*
40
*
*
-
co
p04
4)
107
**
*
*
DISCUSSION
Field populations of P. marinus are genetically heterogeneous. We find that
field populations of P. marinus are genetically heterogeneous, according to the
distribution of "presumptive" P. marinus petB/D alleles cloned out of flow cytometrically
sorted field samples. The P. marinus alleles must be considered only presumptive
because they fall into a lineage containing both P. marinus and marine A Synechococcus
cultured isolates, which fail to form distinct phylogenetic clusters (Figure 5, this thesis
Chapter Two). Thus the P. marinus identity of these alleles, presumed because of the
flow cytometric purification, cannot be confirmed by phylogenetic analysis, and it is
possible that some may have originated in marine A Synechococcus mistakenly included
in the P. marinus sorted samples.
Between 6 and 21 alleles were present in the set of clones recovered from each of
eight flow cytometrically sorted samples, with the presence of large numbers of
additional alleles implied by rarefaction analysis (Figure 6). The set of alleles recovered
within a water column is phylogenetically diverse within the P. marinus/marineA
Synechococcus cluster, with both Sargasso Sea and Gulf Stream water columns
containing alleles which form phylogenetic clusters with distantly related cultures
isolated from the Pacific and the Sargasso Sea (compare Table 1 and Figure 5). The
structure of P. marinus populations implied by these results is consistent with the detailed
findings of Selander and colleagues for the population genetics of Escherichiacoli, in
which most of the genetic diversity of the species is found within local populations, and
in which cell lineages are globally distributed (Selander et al. 1987).
Oceanographic implications of genetic diversity within P. marinus populations
are twofold. First, the physiologic activities of P. marinus in the water column are not
108
simply described by the properties of one or a small number of cultures isolated from the
same geographic area, but by the combined activities of a phylogenetically diverse set of
cells, some of which are similar to isolates from distant parts of the world. Like SS 120
and Med4, these genetic variants may exhibit physiological differences which would give
each a selective advantage under different environmental conditions. The second
oceanographic implication is that adaptive changes in P. marinuspopulations over the
annual cycle and with depth are as likely to be due to shifts in their genetic composition,
due to differences in growth rates of indigenous genetic variants under varying
environmental conditions, as to intracellular adaptive responses (Wood 1988, Wood and
Leatham 1992).
P. marinuspopulations recovered from the Gulf Stream and the Sargasso Sea
water columns share eleven out of the 12 alleles which were found in more than one
sorted sample (Table 4), despite the large scale hydrographic differences which
distinguish these two water bodies. The large number of these shared alleles suggests,
again, a similarity to the population structures of E. coli, in which populations in distant
locations share alleles.
Flow cytometrically defined subpopulations from two "double populations"
recovered from the Gulf Stream exhibited overlapping sets of petB/D alleles in our
analysis, with the pair of subpopulations at 85 m sharing two alleles and the pair at 135 m
sharing one. This result implies that these flow cytometric subpopulations are not
genetically different, but may owe their distinct properties to processes such as
synchronized cell division (Vaulot et al 1994) or mixing of cells acclimated to conditions
in different water masses. Since the subpopulations of the two double populations were
incompletely resolved, however, it is also possible that the shared alleles could result
from the presence in either sort window of cells in the "tail" of the other subpopulation's
109
distribution. Also, it should be noted that multiple populations observed at DCM's in the
Atlantic are transient phenomena (unpublished observations of R. Olsen), whereas
multiple populations at Pacific station ALOHA are a stable feature of a permanent DCM
(Cambell and Vaulot 1993). Multiple populations at ALOHA may therefore have a
different origin from those examined in our study, and may indeed contain genetically
different subpopulations.
Genotype frequencies among PCR clones derived from natural populations using
degenerate, inosine-containing primers reflect both gene frequencies in the sample
population and PCR amplification bias for or against specific alleles. It is therefore
speculative to draw inferences from gene frequencies in datasets such as the one at hand.
With this caveat in mind, it nonetheless bears mentioning that Allele 1 and Allele 1-like
sequences predominate in clones recovered from the shallow mixed layer at both
sampling sites (Figure 7), while at DCM's Allele 1-like sequences are present but do not
predominate. Since HPLC and flow cytometric studies find that surface waters under
stratified conditions contain low chl b2 /a2 ratios (Goericke and Repeta 1993, Cambell
and Vaulot 1993) similar to those found in P. marinus Med4 (Goericke and Repeta
1993), and since Allele 1 belongs to the phylogenetic cluster containing MIT9107 and
Med4, these results may be correlated. The results of this study are therefore consistent
with the hypothesis that populations at different depths in stratified watercolumns draw
their membership from a single gene pool, but that gene frequencies vary at different
depths, resulting in populations exhibiting different chl b/a2 pigment ratios.
110
Figure 7. Frequency of Allele 1-like alleles in sorted samples. The frequency of
alleles having intergenic regions similar to Allele 1 (Alleles 1, 2, 3, 69, 71, 72, 73, 74, 75
and 76) is scored for samples at different depths in the two water columns, lumping
Bright and Dim flow cytometric subpopulations at two depths in the Gulf Stream profile.
Although this inference must be considered speculative, the high frequency of recovery
of Allele 1-like sequences in surface water samples is consistent with a predominance of
these alleles in P. marinus populations in the surface mixed layer.
111
4
U)
0)
r)
4(U
0 r
E
0
c
0
0
C)
CO
0
LO
C
c-
0
0
0
0
0
T-
T
o-
°i0)
OO 0
60
-C
0)
4-0)
C_)
<C
O
0i-
CCT
U,-
0-
0)
L_
Li.
0
O
U
o
o
LO
If)
'0
0
o0
LiO
0o
0
(wU) L44da]
0a
LI.1
-C
a0
0
-C
0
112
REFERENCES
Brand, S.N., Tan, X. and Widger, W.R. (1992). Cloning and sequencing of the petBD
operon from the cyanobacterium Synechococcus sp. PCC7002. Plant Mol. Biol.
20:481-491.
Britschgi, T.B. and Giovannoni, S.J. (1991) Phylogenetic analysis of a natural marine
bacterioplankton population by rRNA gene cloning and sequencing. Appl.
Environ. Microbiol. 57:1707-1713.
Campbell, L., and Vaulot, D. (1993). Photosynthetic picoplankton community structure
in the subtropical North Pacific Ocean near Hawii (station ALOHA). Deep-Sea
Res. 40:2043-2060.
Carter, K.R., Tsai, A. and Palmer, G. (1981). FEBS Lett. 132:243-246.
Chisholm, S.W., R.J. Olson, E.R. Zettler, R. Goericke, J. Waterbury, and N.
Welschmeyer. (1988). A novel free-living prochlorophyte abundant in the
oceanic euphotic zone. Nature, 334(6180):340-343.
Chisholm, S.W., Frankel, S.L., Goericke, R., Olson, R.J., Palenik, B., Waterbury, J.B.,
West-Johnsrud, L. and Zettler, E.R. (1992). Prochlorococcusmarinus nov. gen
nov. sp.: a marine prokaryote containing divinyl chlorophyll a and b. Arch.
Microbiol. 157:297-300.
DeLong, E.F., Franks, D.G. and Alldredge, A.L. (1993) Phylogenetic diversity of
aggregate-attatched vs. free-living marine bacterial assemblages. Limnol.
Oceanogr. 38:924-934.
Geiskes, W.W.C., G.W. Kraay, A. Nontji, D. Setiapermana, and Sutomo. 1988.
Monsoonal alternation of a mixed and a layered structure in the phytoplankton of
the euphotic zone of the Banda Sea (Indonesia): A mathematical analysis of algal
pigment fingerprints. Neth. J. Sea Res. 22:(2):123-137.
Giovannoni, S.J., Britschgi, T.B., Moyer, C.L., and K.G. Field. (1990). Genetic
diversity in Sargasso Sea bacterioplankton. Nature 345:60-62.
Goericke, R. and Repeta, D.J. (1993). Chromatographic analysis of divinyl-chlorophylls
a and b in samples from the subtropical north Atlantic Ocean. Mar. Ecol. Prog.
Ser.
Goericke, R. and Welschmeyer, N.A. (1993). The marine prochlorophyte
Prochlorococcuscontributes significantly to phytoplankton biomass and primary
production in the Sargasso Sea. Deep-Sea Res. 40:2283-2294.
Greer, K.L. and Golden, S.S. (1992). Conserved relationship between psbH and petBD
genes: presence of a shared upstream element in Prochlorothrixhollandica.
Plant Mol. Biol. 19:355-365.
Hope, A.B. (1993). The chloroplast cytochrome bfcomplex: a critical focus on function.
Biochim. Biophys. Acta 1143:1-22.
Hurlbert, S.M (1971). The non-concept of species diversity: a critique and alternative
parameters. Ecology 52:577-586.
113
Kallas, T., Spiller, S. and Malkin, R.C. (1988). Characterization of two operons
encoding the cytochrome b6-f complex of the cyanobacterium Nostoc PCC7906
and highly conserved sequences but different gene organization than in
chloroplasts. J. Biol. Chem. 263:14334-14342.
Kimura, M. (1980). A simple method for estimating evolutionary rate of base
substitutions through comparative studies of nucleotide sequences. J. Mol. Evol.
16:111-120.
Knoth, K., Roberds, S., Poteet, C. and Tamkun, M. (1988). Highly degenerate, inosinecontaining primers specifically amplify rare cDNA using the polymerase chain
reaction. Nuleeic Acids Res. 16:10371.
Lewin, R.A. (1981). Prochlorophytes. In The Prokaryotes. Vol 1. (eds. M.P. Starr, H.
Stolp, H.G. Truper, A. Balows, and H. G. Schlegel). 256-266. Springer, Berlin.
Li, H., Cui, X. and Arnheim, A. (1991). Analysis of DNA sequence variation in single
cells. Methods 2:49-59.
Li, W.K.W., Dickie, P.M., Irwin, B.D., Wood, A.M. (1992). Biomass of bacteria,
cyanobacteria, prochlorophytes and photosynthetic eukaryotes in the Sargasso
Sea. Deep-Sea Res. 39:501-519.
Li, W.K.W., and M. Wood. 1988. Vertical distribution of North Atlantic
ultraphytoplankton: analysis by flow cytometry and epifluorescence microscopy.
Deep Sea Res. 35:1615-1638.
Moore, L.R., Goericke, R., and Chisholm, S.W. (1994). The comparative physiology of
Synechococcus and Prochlorococcus: influence of light and temperature on
growth, pigments, flourescence and absorptive properties. Mar. Ecol. Prog. Ser.,
min press.
Morel, A., Ahn, Y.-H., Partensky, F., Vaulot, D. and Claustre, H. (1993).
Prochlorococcusand Synechococcus: a comparative study of their optical
properties in relation to their size and pigmentation. J. Mar. Res. 51:617-649.
Ohyama, K., Fukuzawa, H., Kohchi, T., Shirai, H., Sano, T., Sano, S., Umesono, K.,
Shiki, Y., Takeuchi, M., Chang, Z., Aota, S., Inokuchi, H. and Ozeki, H. (1986).
Chloroplast gene organization deduced from complete sequence of liverwort
Marchantiapolymorpha chloroplast DNA. Nature 322:572-574.
Olson, R.J., Chisholm, S.W., Zettler, E.R., Altabet, M.A., and Dusenberry, J.A. (1990)
Spatial and temporal distributions of prochlorophyte picoplankton in the North
Atlantic Ocean. Deep Sea Res. 37, 1033-1051.
Partensky, R., Hoepffner, N., Li, W.K.W., Ulloa, O. and Vaulot, D. (1993).
Photoacclimation of Prochlorococcussp. (Prochlorophyta) strains isolated from
the north Atlantic and the Mediterranean Sea. Plant Physiol. 101:285-296.
Reimann, A. and Kueck, U. (1989). Nucleotide sequence of the plastid genes for
apocytochrome b6 (petB) and subunit IV of the cytochrome b6f complex (petD)
from the green alga Chlorellaprotothecoides: lack of introns. Plant Mol. Biol.
13:255-256.
114
Rock, C.D., Barkan, A. and Taylor, W.C. (1987). The maize plastid psbB-psbF-petBpetD RNAs encode alternative products. Curr. Genet 12:69-77.
Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: a laboratory
manual. Cold Spring Harbor Laboratory Press, New York.
Selander, R.K., Caugant, D.A. and Whittam, T.S. (1987). Genetic structure and variation
in natural populations of Escherichiacoli. In Escherichia coli and Salmonella
typhimurium: cellularand molecular biology (F.C. Neidhart, ed.). Am. Soc.
Microbiol. pp. 1625-1648.
Swift, H. and Palenik, B. (1992). Prochlorophyte evolution and the origin of
chloroplasts: morphological and molecular evidence. In Originsof Plastids
(R.A. Lewin, ed.) Chapman and Hall, pp. 123-139.
Vaulot, D., Marie, D., Olson, R.J. and Chisholm, S.W. (1994). Prochlorococcusis
highly synchronized to the diel cycle in the equatorial Pacific and divides up to
once a day. submitted.
Vaulot, D., and Partensky, F. (1992). Cell cycle distributions of prochlorophytes in the
northwestern Mediterranean Sea. Deep-Sea Res. 39:727-742.
Vaulot, D., Partensky, F., Neveux, J., Mantoura, R.F.C. and C. Llewellyn. 1990.
Wintertime presence of prochlorophytes in surface waters of the North-Western
Mediterranean Sea. Limnol. Oceanogr. 35:1156-1164.
Vermaas, W.F.J., and Ikeuchi, M. (1991). Photosystem II. In The Photosynthetic
Apparatus: molecular biology andoperation (L. Bogorad and I.K. Vasil, eds.).
Academic Press pp. 26-111.
Widger, W.R. and Cramer, W.A. (1991). The b6/f complex. in The Photosynthetic
Apparatus: molecular biology andoperation (L. Bogorad and I.K. Vasil, eds.).
Academic Press pp. 149-224.
Wood, A.M. (1988). Molecular biology, single cell analysis and quantitative genetics:
new evolutionary genetic approaches in phytoplankton ecology. In
ImmunochemicalApproaches to Coastal,Extuarine and Oceanographic
Questions (C.M. Yentsch, F.C. Mague and P.K. Horan, eds.) Springer-Verlag pp
41-73.
Wood, A.M. and Leatham, T. (1992). The species concept in phytoplankton ecology. J.
Phycol. 28:723-729.
115
Chapter 4
EPILOGUE: RESPONSES IN THE LITERATURE
TO THE PUBLICATION OF CHAPTER ONE:
MULTIPLE ORIGINS OF PROCHLOROPHYTES
WITHIN THE CYANOBACTERIAL RADIATION
116
Since its publication, "Multiple Origins of Prochlorophytes within the
Cyanobacterial Radiation" (Urbach et al. 1992) has generated numerous citations and has
contributed to the formulation of new theories about the evolution of pigments in
cyanobacteria and chloroplasts (Bryant 1992, Bullerjahn and Post 1992). This Epilogue
is a brief review of published responses to this paper. The conclusions of Chapter One
relevant to the discussion are (1) according to 16S ribosomal RNA (rRNA) phylogenetic
analysis the three known prochlorophytes and chloroplasts each fall into separate
lineages dispersed among the cyanobacteria, and (2) this polyphyletic distribution
suggests chl b photosynthesis was "invented" several times during the evolution of
cyanobacteria and chloroplasts (Urbach et al. 1992).
Most authors citing this work have no argument with the first of these conclusions
(e.g. Bryant 1992, Bullerjahn and Post 1992, Cavalier-Smith 1992, Palenik and
Haselkorn 1992, Swift and Palenik 1992, Cretiennot-Dinet et al. 1993) which echoes an
inference drawn by Turner et al. (1989) and by Morden and Golden (1989a, b) from
analyses of the evolution of Prochlorothrixhollandica,and which agrees with the
analyses of Palenik and Haselkorn (1992) and Lockhart et al. (1992c). However, several
workers have pointed out that the independent invention of chlorophyll b in each of
several lineages is not the only evolutionary scenario to fit the data, and that the
distribution of chl b among cyanobacterial and chloroplast lineages could also result from
lateral gene transfer (Palenik and Haselkorn 1992) or by descent of the entire
cyanobacterial lineage from an ancestor containing both chlorophyll b and
phycobilisomes, with subsequent loss of phycobilisomes in some lineages
(prochlorophytes and green chloroplasts) and chl b in others (most cyanobacteria)
(Bryant 1992, Cavalier-Smith 1992). I wholly concur with these observations, and with
the suggestion that information on prochlorophyte chlorophyll biosynthesis and
chlorophyll a/b binding proteins will help to clarify this issue. To this end, Bullerjahn
117
and Post (1992) cite findings that the major chl a/b binding proteins of Prochloronsp.
and P. hollandicaare of similar molecular weight and are immunologically cross
reactive. They combine these observations with sequence data indicating that P.
hollandicaantenna proteins are different from chloroplast light harvesting complexes,
and form the hypothesis that photosynthetic antennae in P. hollandicaand Prochloronsp.
share a common origin separate from that of green chloroplasts. Future work in this
direction, including analysis of the P. marinus photosynthetic system, will surely reveal a
fascinating evolutionary story.
One research group, including P. Lockhart, A.W.D Larkum, D. Penny and
coworkers does, however, take issue with the phylogenetic branching pattern inferred
from 16S rRNA sequences in Chapter One (Lockhart and Penny 1992, Lockhart et al.
1992b, Larkum 1992, Lockhart et al. 1993). Their skepticism comes as an extension of
their work showing that branching patterns inferred from protein-encoding genes are
confounded by nucleotide substitution biases (detected as differences in G+C
composition at third codon positions, see Chapter Two). Lockhart et al. (1992b) argue
that phylogenetic trees inferred from cyanelle, prochlorophyte, cyanobacterial and
chloroplast protein-encoding sequences are invalid because patterns of similarity in third
codon position G+C content are congruent with the inferred phylogenetic branching
patterns. They argue, in addition, that 16S rRNA trees, even those inferred from
sequences without detectable G+C bias, are also invalid, based on the following premise:
Although the effects of substitutional bias upon inference are best
demonstrated by analysis of protein-coding nucleotide sequences, bias (a
genomic effect in eubacterial lineages) must also affect variable sites
within rRNA genes and may also have an effect on amino acid
substitution (Jukes and Bhushan 1986; but see Prager and Wilson 1988).
Inference from such datasets (particularly where long edges link taxa:
e.g., Giovannoni et al. 1988; Morden and Golden 1989; Evrard et al. 1990;
Kishino et al. 1990) may therefore also be misled.
(Lockhart et al. 1992a)
118
Larkum (1992) proposes that, since phylogenetic trees linking the Cyanophoraparadoxa
cyanelle to chloroplasts and showing prochlorophytes dispersed among the cyanobacteria
are invalid, it is reasonable to propose that prochlorophytes are monophyletic and and
share an ancestry with green chloroplasts.
Lockhart and coworkers make a valid point in saying that phylogenetic trees,
inferred by methods which assume constant substitution biases across taxa, may be in
error when calculated from data having varying third codon position G+C content. Their
extension to 16S rRNA sequences quoted above is a subtle argument, but also may hold
true. Resolution of these phylogenetic problems must await the development of methods
demonstrated to be insensitive to nucleotide substitution biases. However, the failure of
P. marinus strains to group with chloroplasts having similar third codon position G+C
content in trees inferred with psbB and petB/D sequences (Urbach and Chisholm in
preparation, this thesis Chapter Two) argues against an evolutionary link between P.
marinus and chloroplasts.
I'll close with a quote from Bullerjahn and Post (1992) which is appropos:
If one reviews much of the literature on the prochlorophytes, it
becomes clear that taxonomic classification and the inferring of
relationships between organisms can be a risky enterprise. Some
microbiologists have been eager to show a direct relationship between
prochlorophytes and the chloroplast, whereas others were determined to
place them among the cyanobacteria. The underlying problem in such
exercises lies both in the danger of overemphasizing one aspect while
belittling others and in the import of externally imposed bias into the
judgement of a certain property.
119
REFERENCES
Bryant, D.A. (1992). Puzzles of chloroplast ancestry. Curr. Biol. 5:240-242.
Bullerjahn, G.S. and Post, A.F. (1992). The prochlorophytes: are they more than just
chlorophyll a/b-containing cyanobacteria? Crit. Rev. Microbiol.
Cavalier-Smith, T. (1992). The number of symbiotic origins of organelles. Biosystems
28:91-106.
Cretiennot-Dinet, M.-J., Sournia, A., Ricard, M. and Billard, C. (1993). A classification
of the marine phytoplankton of the world from class to genus. Phycologia
32:159-179.
Everard, J.-L., Kuntz, M. and Weil, J.-H. (1990). The nucleotide sequence of five
ribosomal protein genes from the cyanelles of Cyanophoraparadoxa:
implications concerning the phylogenetic relationship between cyanelles and
chloroplasts. J. Mol. Evol. 30:16-25.
Giovannoni, S.J., Turner, S., Olsen, G.J., Barnes, S., Lane, D.J. and Pace, N.R. (1988).
Evolutionary relationships among cyanobacteria and green chloroplasts. J.
Bacteriol. 170:3584-3592.
Jukes, T.H. and Bhushan, V. (1986). Silent nucleotide substitutions and G+C content of
some mitochondrial and bacterial genes. J. Mol. Evol. 24:39-44.
Kishino, H., Takashi, T. and Hasegawa, M. (1990). Maximum likelihood inference of
protein phylogeny and the origin of chloroplasts. J. Mol. Evol. 31:151-160.
Larkum, A.W.D. (1992). Evolution of chlorophylls, light harvesting systems and
photoreaction centres. In Research in Photosynthesis, Vol. III (N. Murata, ed.).
Kluwer Academic, pp. 475-482.
Lockhart, P.J., Penny, D. (1992). The problem of GC content, evolutionary trees and the
origins of chl-a/b photosynthetic organelles: are the prochlorophytes a
eubacterial model for higher plant photosynthesis? in Research in Photosynthesis,
VoL III (N. Murata, ed.). Kluwer Academic, pp. 499-505.
Lockhart, P.J., Howe, C.J., Bryant, D.A., Beanland, T.J. and Larkum, A.W.D. (1992a).
Substitutional bias confounds inference of cyanelle origins from sequence data. J.
Mol. Evol 34:153-162.
Lockhart, P.J., Penny, D., Hendy, M.D., Howe, C.J., Beanland, T.J. and Larkum, A.W.D.
(1992b). Controversy on chloroplast origins. FEBS Lett. 301:127-131.
Lockhart, P.J., Beanland, Howe, C.J. and Larkum, A.W.D. (1992c). Sequence of
Prochloron didemni atpBE and the inference of chloroplast origins. Proc. Natl.
Acad. Sci. USA 89:2742-2746.
Lockhart, P.J., Penny, D., Hendy, M.D., and Larkum, A.D. (1993). Is Prochlorothrix
hollandicathe best choice as a prokaryotic model for higher plant chl a/b
photosynthesis? Photosynthesis Res. 37:61-68.
120
Morden, C.W. and S.S. Golden. 1989a. psbA genes indicate common ancestry of
prochlorophytes and chloroplasts. Nature 337:382-385.
Morden, C.W. and S.S. Golden. 1989b. Corrigendum: psbA genes indicate common
ancestry of prochlorophytes and chloroplasts. Nature 339:400.
Palenik, B. and Haselkorn, R. (1992). Multiple evolutionary origins of prochlorophytes,
the chlorophyll b-containing prokaryotes. Nature 355:265-267.
Prager, E.M. and Wilson, A.C. (1988). Ancient origin of lactalbumin from lysozyme:
analysis of DNA and amino acid sequences. J. Mol. Evol. 27:326-335.
Swift, H., and Palenik, B. (1992) Prochlorophyte evolution and the origin of chloroplasts:
morphological and molecular evidence. In Origins ofPlastids(R.A. Lewin, ed.)
Chapman and Hall pp. 123-138.
Turner, S., Burger.Wiersma, T., Giovannoni, S.J., Mur, L.R., and N.R. Pace. (1989). The
relationship of a prochlorophyte Prochlorothrixhollandica to green chloroplasts.
Nature 337:380-382.
Urbach, E., Robertson, D. and Chisholm, S.W. (1992). Multiple evolutionary origins of
prochlorophytes within the cyanobacterial radiation. Nature 355:267-269.
121
Appendix I
PRELIMINARY ANALYSIS OF GENETIC DIVERSITY IN GULF STREAM
POPULATIONS OF PROCHLOROCOCCUSMARINUS
BY DIRECT SEQUENCING OF PCR PRODICTS AMPLIFIED
FROM FLOW-CYTOMETRICALLY SORTED CELLS
122
Chapter Three details an investigation into the genetic diversity of natural
populations of Prochlorococcusmarinus from the the Gulf Stream and the Sargasso Sea
in which P. marinuswere sorted from field samples by flow cytometry and the diversity
of petB/D sequences in the sorted samples assessed by cloning and sequencing individual
molecules from PCR amplification products. This Appendix describes a preliminary
investigation which preceded the cloning experiment in which PCR products sorted from
Gulf Stream field samples were directly sequenced. Autoradiograms of the resulting
sequencing gels revealed comigrating bands at numerous nucleotide positions, especially
in the intergenic region between petB and petD, indicating that multiple petB/D
sequences were likely to be present in the amplification products.
METHODS
Sample collection, flow cytometric sorting and DNA isolation. The DNA
preparations used for this preliminary experiment were the same as those used for the
analysis of Gulf Stream populations in Chapter Three, and originated in samples
collected from 50 m, 85 m and 135 m at 370 30.68'N, 680 13.69'W during July of 1993.
Samples collected from 85 m and 135 m each contained a pair of flow cytometrically
defined "double populations," subpopulations of which were sorted into separate
samples, resulting in a total of five sorted samples (Figure 2 in Chapter Three).
PCR amplification. Sequences at the petB/D locus were amplified as described
for the analysis of petB/D sequences from cultured P. marinus and Synechococcus, using
biotinylated forward primer PPETBD314 and unbiotinylated reverse primer
PPETBD1160R (this thesis Chapter Two). PPETBD1160R also served as a primer for
sequencing. Conditions for PCR amplification were less stringent than those used for the
cloning experiment (this thesis Chapter Three) and produced sufficient amounts of PCR
123
products to serve directly as sequencing templates. Templates were prepared using
streptavidin-coated magnetic beads (Dynal) and sequenced directly with Sequenase
(United States Biochemical), both according to their manufacturers' instructions.
RESULTS
Autoradiograms from direct sequencing of petB/D amplified from sorted samples
revealed multiple comigrating bands which differed in numbers and intensities among the
different samples (Figure 1). Most dramatic were the results of direct sequencing of
petB/D from the 85 m Bright and 135 m Dim samples in which multiple comigrating
bands were frequent in the intergenic region above the petB/D start codon and were also
visibile at some third codon positions in petD (Figure 1, lanes b and e). Multiple
comigrating bands were less prominent, but visible, in sequencing reactions from the 50
m amplification (lane a) and could not be detected in the 85 m Dim and 135 Bright (lanes
c and d, compare to reactions for a P. marinus culture in lane e).
"Major" sequences could be read above the relatively low or absent background
of comigrating bands for the 50 m, 85 m Dim and 135 m Bright samples. All of these
major sequences were similar to Allele 1, the most frequently recovered allele in the
cloning experiment (49 out of 191 clones in the combined dataset for Gulf Stream and
Sargasso Sea sorted samples, 22 out of 139 clones for Gulf Stream samples alone, Table
1 in Chapter 3) (Figure 1, lanes a, c and d).
DISCUSSION
The presence of multiple comigrating bands in sequencing autoradiograms can
result either from sequence heterogeneity in the template DNA or from suboptimal
124
reaction conditions which allow premature termination of nascent DNA chains. The
latter cause would not, however, be expected to produce a disproportionate number of
comigrating bands at third codon position sites and intergenic regions, a result which is
dramatically evident in autoradiogram lanes from the 85 m Bright and 135 m Dim sorted
samples (Figure 1, lanes b and e) and, less dramatically, in lanes from the 50 m sorted
sample (lane a). The results of this analysis were therefore consistent with the presence
of multiple petB/D alleles in at least some of the flow cytometrically sorted field samples,
and prompted the cloning and sequencing investigation in Chapter Three.
The investigation in Chapter Three revealed that, in fact, all of the PCR products
from the Gulf Stream and similar products from the Sargasso Sea contained multiple
petB/D sequences. Consistent with the results of direct sequencing, the 85 m Bright and
135 m Dim samples contained the largest variety of petB/D alleles: when corrected to a
constant sample size of 19 clones these samples contained 12.1 and 13.7 alleles,
respectively. The 50 m, 85 m Dim and 135 m Bright sorted samples yielded 5.3, 5.2 and
5.4 alleles for the same constant sample size (Table 1 in Chapter Three).
There was, however, one type of comparison in which the results of the direct
sequencing and cloning experiments did not consistently agree. The major sequence and
the most frequently recovered allele were not consistently identical for each of the sorted
samples: while the 50 m sample yielded a majority of Allele 1 clones (17 out of 23
clones) consistent with its major sequence, the sets of clones recovered from the 85 m
Dim and 135 m Bright samples were mostly Allele 5 (17 out of 24 and 16 out of 23
clones, respectively), with only a minority of clones belonging to Allele 1 (three out of
24 and two out of 23 clones, respectively) (Table 1 in Chapter Three). This discrepency
may be attributable to two potential causes: 1) Allele 1 may have been preferentially
amplified under the less stringent conditions used to produce PCR products for direct
125
sequencing as compared to the cloning experiment, or 2) Allele 1 may have been
preferentially chosen as template during direct sequencing reactions. Discrepencies in
the identities of major sequences and most frequently recovered alleles for the different
sorted samples reinforce the necessity of mentioning the caveat in Chapter Three:
genotype frequencies among PCR clones (and major sequences of PCR products) derived
from natural populations using degenerate, inosine-containing primers reflect both gene
frequencies in the sample population and PCR amplification bias (as well as sequencing
reaction bias) for or against specific alleles. It is therefore speculative to draw inferences
from gene frequencies (or major sequences) in datasets such as the one at hand.
126
Figure 1. Autoradiograms from direct sequencing of petB/D PCR products
amplified from flow cytometrically sorted P. marinus from natural populations in the
Gulf Stream (a-e) and from unsorted, cultured P. marinus MIT9313, isolated from the
135 m Gulf Stream water sample (L. Moore pers. comm.) (f). The arrow marks the
initial petD ATG codon, above which is intergenic region sequence. Dashes mark the
positions of third codon positions in petD. a) 50 m, b) 85 m Bright, c) 85 m Dim, d) 135
m Bright, e) 135 m Dim sorted samples (Chapter Three). f) unsorted, cultured MIT9313.
127
L~YP
mAJ
REFERENCE
Hurlbert, S.M. (1971). The non-concept of species diversity: a critique and alternative
parameters. Ecology 52:577-586.
129
Appendix II
JUSTIFICATION FOR THE 53.5% SEQUENCE DIFFERENCE
CRITERION FOR SEQUENCES ASSIGNED TO A SINGLE ALLELE
130
Introduction. In Chapter Three, genetic diversity in natural populations of
Prochlorococcusmarinus was investigated by comparing DNA sequences of cloned PCR
products amplified from flow cytometrically sorted field samples. The genetic locus
examined included the 3' end of the petB gene, an intergenic "nonsense" region and the 5'
end of the petD gene (the amplified locus, including both gene fragments and the intergenic
region, being referred to as "petB/D"). Clone sequences (which were inferred from single
gel readings, and not confirmed by sequencing complimentary strands) were categorized into
Alleles, or sets of clones within which sequence differences were indistinguishable from
PCR and sequence determination (PCR+SD) error, corrresponding to genotypes present in
the field samples. For ease of analysis, each Allele was characterized by a single clone
designated as its "prototype." Among the 191 clones analyzed there was a great deal of
sequence diversity, with 59 sets of clones exhibiting intergenic region sequences that were so
different in length and sequence that it was impossible to identify homologous nucleotides
for use in sequence alignment. These sequence differences were easily identified, despite the
undoubted presence of PCR+SD error, as ones which originated in genetically distinct
individuals in the sorted populations. Three of these sets of clones, however, contained
sequences exhibiting sufficient mismatch to suggest that they were amplified from
individuals containing different, though related sequences. These sets were subdivided into
Alleles according to a •3.5% total sequence mismatch with prototype criterion for clones
included within an Allele. The total number of Alleles arrived at by this process, including
the sets of clones with unalignable intergenic sequences, was 71.
This Appendix presents arguments in support of the •3.5% total sequence mismatch
criterion used to group clones into Alleles in Chapter Three. Initially, bounds are set on
possible values for the criterion using a theoretical calculation based on published values for
Taq polymerase error rates in PCR and an ad hoc estimate of the range of sequence
determination error rates. Next, evidence is presented from the distribution of closely related
131
clones among field samples suggesting that some sequences within these bounds do differ as
a result of natural variation, and not simply due to PCR+SD error. Finally, a practical
method is described for optimizing the choice of criterion using the fraction of sequence
mismatch occuring at third codon positions, and the method is applied to the dataset.
Theoretical confidence limits for the frequency of sequence mismatches between
cloned PCR products attributable to PCR+SD error. During the course of PCR
amplification, replication errors give rise to PCR products containing one or more
mismatches with the original template sequence. The frequency of errors within individual
PCR product molecules can be seen to follow a Poisson distribution, with variance equal to
the mean, despite the fact that the frequency of errors in whole PCR reactions conforms to a
Luria-Delbruck distribution (Eckert and Kunkel 199 la,b), and thus has a very high
variance'.
In the final, cloned PCR product the mean error frequency (f) is equal to the
probability that polymerase error has caused a nucleotide to become mismatched with its
progenitor in the original template (Eckert and Kunkel 1991a,b),
1The Luria-Delbruck
distribution was first proposed to describe the widely varying frequency of mutant cells
in replicate bacterial cultures grown under non-selective conditions in the classic fluctuation test (Luria and
Delbruck 1943), but the distribution is equally applicable to the frequency of errors in replicate PCR reactions
(Eckert and Kunkel 1991a,b). In a Luria-Delbruck process the rare occurrence of a mutation (PCR error)
during an early generation (cycle) causes the mutant (error) frequency in a small number of cultures
(reactions) to be substantially greater than the mean, giving rise to a frequency distribution having a high
variance. This is because the final number of mutants (PCR products) resulting from an early mutation (error)
is amplified through the process by a factor of 2x, where X equals the number of generations (cycles) between
the mutation (error) and the analysis. Thus, in a Luria-Delbruck process the number of the round in which a
error occurs influences the final error frequency attributable to that error. In the case of the distribution of
errors within a single PCR product molecule, however, the situation is different. In this case, each nucleotide
is descended from its original template nucleotide through a number of rounds of PCR replication, each with a
probability of error. Every nucleotide in the chain is subject to this process independently, and the final error
frequency attributable to any error is equal to 1/L, where L is the number of nucleotides in the chain, and is
independent of the number of the round in which the error occurred. If the probability of error is assumed to
be equal for every nucleotide position, errors within the chain are seen to be rare, independent events with
equal probabilities of occurrence. Thus the frequency of these errors conforms to a Poisson distribution.
132
nm
f=_
(1)
2
where n is the number of PCR cycles and m the probability of error per nucleotide per
cycle2 . Since these errors are distributed as a Poisson distribution, the variance for the
frequency of errors in individual PCR product molecules is equal to the mean, and
confidence limits can be calculated for the frequency of PCR errors in individual product
molecules using known values for the polymerase error rate. The expected frequency of
differences between two cloned PCR products that can be attributed to PCR error is equal to
the expected frequency of errors in a cloned molecule twice the length (1)of the region
compared between the two products (neglecting coincidence of errors at homologous
nucleotide sites and assuming the two cloned molecules are not amplification products of the
same template molecule, which would decrease the expected frequency of their sequence
differences). Accordingly, the 95% confidence upper limit (L95) for the frequency of
mismatch between two cloned sequences attributable to PCR error is
(2)
L95 = f + to.1[21-1 'f
where t0.1121-1] is the critical value for student's t distribution for a one-tailed test with a =
0.05 and degrees of freedom equal to 21 - 1.
A term may be added to expression (1), the mean frequency of polymerase errors in
cloned PCR products, in order to account for sequence determination errors. This gives
nm
f =_-+g
2
(3)
where g is the rate of sequence determination errors in errors nucleotide- 1. If sequence
determination errors are assumed to follow a Poisson distribution, this expression for f may
be used to calculate 95% confidence intervals for PCR+SD error using equation (2).
2The probability
that a molecule has inherited a DNA strand containing an error introduced during the
previous round is m/2, hence the 2 in the denominator. This factor applies to the product of the final PCR
round as well, although in this case it reflects the probability that the progeny of a mutant strand in a cloned
heteroduplex will predominate in the final plasmid preparation. This formulation neglects the effects of
superimposed errors.
133
Sequence determination errors are estimated ad hoc at between 0.0 and 5.0 x 10-3 errors
nucleotide-1.
Application of the theoretical confidence limits to analysis of the data in
Chapter Three. Mean, standard deviation and upper 95% confidence limits for the
frequency of sequence differences expected in cloned PCR products compared over 71
basepairs (the length of the shortest sequence in the dataset) were calculated using the lowest
and highest reported Taq polymerase error rates (Eckert and Kunkel 199 la) (corrected to
reflect nucleotide substitutions only3), over the estimated range of sequence determination
errors (Table 1). These upper confidence limits provide theoretical bounds for the value of
the criterion to be used to declare sequences different or indistinguishable. If the lowest
combined error rate were to apply, sequences in Chapter Three having alignable intergenic
regions and differing by up to 3.1% of nucleotide positions would be considered
indistinguishable from PCR products amplified from identical template sequences, at p =
0.05. However, applying the highest reported error rate raises this criterion value to 18.5%.
Thus, the 95% upper confidence confidence limit for the amount of sequence difference
between cloned PCR products attributable to error is shown to be sensitive to estimates of the
polymerase and sequence determination error rates, the true values of which are unknown for
this experiment (Table 1).
3 Values
for the range of Taq polymerase error rates inPCR are available from the literature (Eckert and
Kunkel 1991) but require correction in order to be applicable to data analysis for Chapter 3. In Chapter 3,
apparent insertions and deletions are not counted as sequence mismatches (i.e., gap weights were set to zero)
on the assumption that these mismatches are likely to have been caused by PCR+SD error. Accordingly, Taq
polymerase error rates used to calculate theoretical upper 95% confidence limits for PCR+SD error in Chapter
3 have been corrected to omit insertion and deletion errors using the original data, where this is aplicable.
Specifically, the highest reported Taq polymerase PCR error rate has been corrected from 2 x 10 to 1.6 x 104 errors nucleotide - 1 cycle-1, using the original data (Dunning et al. 1988), while the lowest reported error
rate is unchanged because insertions and deletions were not observed in the original experiments (Goodenow
et al. 1989).
134
0
a)
1o4
A
*1
4-
0d
8
oi
.00
a
0i0 0
coou
0
4.1
CUD
0
4-000.%n25
.c
O)
f~o0
00
a.)
a)
ao.
c7
t:
LAu
.40. -
a.).
44
0
.. 00
0
U) 0
4-
0- 0 W
Cl)E-
20.
(.1)
UC
78
-
)
00
-L4-
po cn
Cd
0
-~t
Q,,
4
04 *
0
~Cn.
cdu
0
4El
Ev
LV
I...#
135
C.)
t
Applying the calculated extreme bounds for the criterion to the task of lumping
indistinguishable sequences into Alleles using the set of clones recovered from the Gulf
Stream and Sargasso Sea yields a total of 74 Alleles in the combined dataset for the lowest
error estimate and 59 Alleles for the highest one4. While the impulse to err on the side of
conservatism would dictate the choice of •18.5% mismatch as criterion for inclusion within
an allele, independent evidence suggests that this criterion would lump together a number of
cloned sequences which do, in fact, derive their differences from genetic heterogeneity in P.
marinus populations. A lower criterion, while still above the 95% upper confidence limit
calculated from the lowest reported Taq polymerase error rate, can be estimated by
considering the pattern of distribution of closely related sequences among field samples and
the ratio of their mismatches at third codon positions to mismatches at all codon positions in
order to distinguish sets of related sequences giving useful information about genetic
diversity from sets in which members differ mostly due to PCR+SD error.
An upper limit for the sequence mismatch criterion suggested by the
distribution of closely related sequences among field samples. In the dataset of Chapter
Three four 100.0% identical sequences belonging to Allele 71 were recovered from four
different field samples: the 85m Dim and 135m Dim Gulf Stream samples and the 40m and
70m Sargasso Sea samples, and therefore were cloned out of separate PCR reactions. Allele
71 is highly similar to Allele 1 (8.1% mismatch), which includes numerous identical clones
distributed among several field samples. If the 18.5% mismatch criterion were to be applied,
Allele 1 and Allele 71 (as well as several others) would be lumped together as members of a
single Allele. While it is not unlikely that a single Allele 71 clone could have been produced
from an original Allele 1 template, it is highly unlikely that the identical set of errors would
4
Application of the upper bound criterion calculated using the highest combined PCR+SD error rate causes all
sequences having alignable intergenic regions to be lumped with each other. There are 59 sets of intergenic
regions which cannot be aligned with each other, therefore application of this higherst possible criterion value
results ina dataset containing 59 Types.
136
occur in four separate PCR reactions and four separate sequence determinations. Therefore,
at least some sequences in the dataset that differ by 8.1% represent true genetic variants and
not the results of PCR+SD error.
Optimization of the criterion for distinguishing sequence mismatch attributable
to PCR+SD error from mismatch due to genetic diversity by considering the fraction of
sequence mismatch at third codon positions: a practical method. Evolutionary change
at third codon positions is generally faster than at the first two positions because degeneracy
in the genetic code allows substitution at the third position of most codons without changing
the encoded amino acid. PCR+SD errors, on the other hand, should be random with respect
to codon postion. A high fraction of sequence mismatch at third codon positions (fraction of
mismatch at third codon positions = fractional mismatch at third codon positions/3 x
fractional mismatch at all codon positions) is therefore evidence that a pair of cloned
sequences differ mostly due to true evolutionary divergence under selection for a conserved
amino acid sequence, rather than due to PCR+SD error. Conversely, a low ratio is evidence
of the activity of random processes like PCR error and inaccurate sequence determination.
In an ideal dataset, a pair of cloned sequences which differ due to PCR+SD error
alone should have a fraction of sequence mismatch at third codon positions equal to
approximately 0.33. In an actual dataset like the one in Chapter Three, however, there is
likely to be a significant amount of scatter in values for this ratio at the low end of the total
sequence mismatch scale due to sampling error associated with the small number of
mismatches in highly similar cloned sequences of finite length. Additionally, in a real
dataset some sequences differing by small increments of total mismatch may, in fact, differ
due to evolutionary processes, but these individual comparisons cannot be distinguished from
PCR+SD error because sampling error obscures differences in their fractions of mismatch at
third codon positions. The effect of the presence of these small evolutionary differences will
137
be to increase the average fraction of mismatch at third codon positions above the 0.33 value
for comparisons at the low end of the total sequence mismatch scale. In the face of this
complexity, the task of setting a total mismatch criterion below which sequence differences
are considered indistinguishable from PCR+SD error (for experiments in which the error rate
has not been experimentally determined) becomes a process of examining the distribution of
fractional mismatch at third codon positions along the scale of total mismatch and
identifying a point of discontinuity at which the sphere of influence of random errors gives
way to one of evolutionary divergence. At this point there will be a maximum difference
between the means for fractional sequence differences at third codon positions for sequence
comparisons above and below the criterion. This transition point should fall between (or
perhaps slightly below) the theoretical upper 95% confidence limits for mismatch
attributable to PCR+SD error, determined according to the range of probable polymerase and
sequence determination error rates.
Optimization of the criterion for sequences in Chapter Three. For comparisons
of clones having alignable intergenic regions in the dataset of Chapter Three, a plot of the
fractional sequence mismatch at third codon positions (in coding regions only) versus total
mismatch (in coding regions plus intergenic regions) exhibits a marked biphasic appearance
(Figure 1). Consistent with the expectations of random error due to PCR+SD superimposed
over a smaller amount of difference due to evolutionary processes, comparisons at the low
end of the total mismatch scale have a lower average fraction of mismatches at third codon
positions than do comparisons at the high end, but this average is greater than 0.33. A
criterion chosen to distinguish domains of total sequence mismatch in which the dominant
cause of sequence mismatch is PCR+SD error or true genetic diversity should divide the
dataset into two subsets having the greatest possible difference in their average fraction of
sequence mismatch at third codon positions. A graph plotting this difference for
138
Figure 1. Fraction of sequence mismatch at third codon positions (in coding regions
only), plotted against total mismatch (including coding and intergenic regions) for
comparisons of clones with alignable intergenic regions in the dataset of Chapter Three.
Dotted vertical lines are theoretical upper 95% confidence limits for percent sequence
mismatch (L9 5 x 100) attributable to PCR+SD error for the highest and lowest combined
Taq polymerase and sequence determination error rates (Table 1). Datapoints below 1%
total mismatch are omitted, and some symbols represent superimposed data points. Solid
curve: fourth order polynomial fitted to all datapoints. Dashed curves: upper and lower 95%
confidence intervals for the solid curve.
139
.
12
1.0
0.8
0.6
0.4
0.2
0.0
-I0
0
10
Total mismatch (%)
140
20
hypothetical criteria placed along the total mismatch frequency scale at intervals of 0.1%
mismatch shows a local maximum at 3.5% total mismatch, between the theoretical bounds
imposed by published values for Taq polymerase error rates combined with estimates for
sequence determination error rates (Figure 2). The value 3.5% total mismatch was therefore
chosen as the best possible criterion to distinguish sequences differing due to PCR+SD error
from those differing mostly due to evolutionary divergence, for the dataset of Chapter Three.
This value is consistent with a Taq polymerase error rate near the low end of the range of
reported values5 .
Reexamination of the graph plotting the fraction of sequence mismatch at third codon
positions versus total mismatch frequency (Figure 1) using the 3.5% criterion to divide the
data into two subsets and analyzing each subset independently for the mean and its 95%
confidence intervals, illustrates the contrast between sequence differences attributed to
PCR+SD error at the low end of the total sequence mismatch scale, and differences due to
natural variation at the high end (Figure 3). The mean fraction of sequence mismatch at
third codon positions for comparisons to the left of the criterion equals 0.48; for comparisons
to the right of the criterion the mean is 0.81. 95% confidence intervals for these means do
not overlap, indicating that the criterion chosen by this method separates the sequence
comparisons in the dataset of Chapter Three into groups which are significantly different in
their fraction of sequence difference at third codon positions, and therefore in their ratio of
random to evolutionary sequence differences.
5It should be pointed out that this analysis would be improved by an additional sequencing effort, providing
full-length sequences confirmed by sequencing in both directions for all clones having alignable intergenic
regions. It is predicted that the number of sequence comparisons yielding ones and zeros for third codon
position/all codon position mismatch ratios would decrease with increased sequence lengths and accuracy,
reducing scatter at low total mismatch values in Figures Al-1 and AI-3 and perhaps altering Mean R - mean L
values at low total mismatch values in Figure AI-2 as well. An additional sequencing effort before
publication of these results is being considered.
141
Figure 2. Optimization of the criterion identifying the upper bound for total
sequence mismatch considered indistinguishable from PCR+SD error. Datapoints are
differences between mean values for the fraction of sequence mismatch at third codon
positions for datapoints in Figure 1, grouped to the right and left of hypothetical criteria
placed at 0.1% intervals along the total mismatch axis. Dotted vertical lines are theoretical
upper 95% confidence limits for percent sequence mismatch (L9 5 x 100) attributable to
PCR+SD error for the highest and lowest combined Taq polymerase and sequence
determination error rates (Table 1), and therefore represent theoretical bounds for the
criterion. The curve is a second order polynomial fitted to points between 2.0% and 5.9%
total mismatch to aid in identifying the local maximum, which is marked with a solid vertical
line at 3.5% total mismatch. The criterion is therefore set at the value of total mismatch
dividing the sequence comparisons into the two groups having the most different mean
fraction of sequence mismatch at third codon positions possible within the bounds of the
theoretical limits.
142
0.4
S0.3
Z"
9
0.2
0.1
0.0
0
10
Total mismatch (%)
143
20
Figure 3. Data from Figure 1 divided into two groups by the criterion (solid vertical
line) drawn at 3.5% total mismatch. Data to the left of the criterion are comparisons between
closely related sequences which, on average, exhibit low fractions of sequence mismatch at
third codon positions emblematic of sequences which differ mostly due to PCR+SD error.
Data to the right of the criterion are comparisons between sequences differing by greater than
3.5% total mismatch which, on average, exhibit higher fractions of sequence mismatch at
third codon positions, characteristic of sequences differing due to evolutionary processes.
Error bars enclose 95% confidence intervals for the mean of all datapoints to the left or right
of the criterion.
144
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
0
10
Total mismatch (%)
145
20
Conclusion. This Appendix presents both theoretical and practical methods for
discriminating between true natural variation and PCR+SD error in the analysis of cloned
sequences amplified from populations of mixed genetic composition. To my knowledge,
both are currently absent from the literature. The practical method is applied to the data of
Chapter Three for sorting clone sequences having alignable intergenic regions into Alleles
corresponding to genotypes present in natural populations. Even without this analysis, the
dataset of Chapter Three shows abundant genetic diversity, with 59 sets of petB/D clones
having unalignable intergenic regions; differences among these sets must be attributed to
natural variation and not to PCR+SD error. Application of the practical method to the
analysis of clones having alignable intergenic regions adds 12 additional Alleles to the
Chapter Three data used for population genetics analysis. The Alleles produced by
subdivision of sets of clones having alignable intergenic regions exhibit an average fraction
of sequence mismatch at third codon positions of 0.82 when compared to related Alleles,
significantly different from the expectations of random error. Therefore, even if their
prototype sequences contain errors, these Alleles serve to identify genotypes present in the
natural populations and are legitimately included in the population genetics analysis.
146
REFERENCES
Dunning, A.M., Talmud, P. and Humphries, S.E. (1988). Errors in the polymerase chain
reaction. Nucleic Acids Res. 16:10393.
Eckert, K.A. and Kunkel, T.A. (1991a). DNA Polymerase Fidelity and the Polymerase
Chain Reaction. PCR Methods Applic. 1:17-24.
Eckert, K.A. and Kunkel, T.A. (1991b). The fidelity of DNA polymerases used in the PCR.
In Polymerase chain reaction: a practicalapproach(eds. M.J. McPherson, P.
Quirke and G.R. Taylor). IRL Press, Oxford, pp..
Ennis, P.D., Zemmour, J., Salter, R.D. and Parham, P. (1990) Rapid cloning of HLA-A,B
cDNA by using the polymerase chain reaction: frequency and nature of errors
produced in amplification. Proc. Natl. Acad. Sci. USA 87:2833-2837.
Goodenow, M., Huet, T., Saurin, W., Kwok, S., Sninsky, J. and Wain-Hobson, S. (1989).
HIV-1 isolates are rapidly evovling quasispecies: evidence for viral mixtures and
preferred nucleotide substitutions. J. Acquir. Imm. Defic. Syn. 2:344-352.
Keohavong, P. and Thilly, W.G. (1989). Fidelity of DNA polymerases in DNA
amplification. Proc. Natl. Acad. Sci. USA 86:9253-9257.
Luria, S.E. and Delbruck, M (1943). Mutations of bacteria from virus sensitivity to virus
resistance. Genetics 28:491-511.
Saiki, R.K., Gelfand, D.J., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B. and
Erlich, H.A. (1988). Primer-directed enzymatic amplification of DNA with a
thermostable DNA polymerase. Science 239:487-491.
Sarkar, S. (1991). Haldane's solution of the Luria-Delbruck Distribution. Genetics 127:257261.
Tindall, K.R., and Kunkel, T.A. (1988). Fidelity of DNA synthesis by the Thermus
aquaticusDNA Polymerase. Biochemistry 27:6008-6013.
147
Appendix III
COMPARISON OF NUCLEOTIDE DIFFERENCES AMONG CLONES
ASSIGNED TO ALLELE 1 TO KNOWN PATTERNS OF NUCLEOTIDE
SUBSTITUTION ERROR BY TAQ POLYMERASE
FOR DATA OF CHAPTER THREE
148
This Appendix provides an analysis of the pattern of sequence mismatch among
clones assigned to one "Allele," or set of clones within which sequence differences are
considered indistinguishable from PCR and sequence determination (PCR+SD) error
according to the 3.5% total sequence mismatch criterion (Appendix I), in order to
investigate the hypothesis that their differences arose through Taq polymerase error
during PCR. Results suggest that different sources of sequence variation, possibly
sequence determination error or natural genetic variation, contribute to sequence
mismatch within Types, in addition to Taq polymerase error.
Sequence mismatches among clones assigned to Allele 1, the allele containing the
greatest number of clones in the dataset of Chapter Three, were compared to nucleotide
substitutions characteristic of Taq polymerase error. Allele 1 was spawned from the
subdivision of a larger set of clones having alignable intergenic regions (Appendix I) and
contains a total of 44 clones, 11 of which are identical to the prototype sequence. 11
additional clones, 10 derived from the 70m Sargasso Sea sorted sample and one from the
50m Gulf Stream sample, contained the identical G->A substitution as their only
mismatch with the prototype. This nucleotide substitution was present in the remaining
S70 clone sequenced across the site in question, as well, and may have originated in a
natural variant of Allele 1 P. marinus present at 70m in the Sargasso Seal. If this
nucleotide substitution is not attributed to PCR+SD error, then at least 22 out of 44 (50%)
of the clone sequences classified as Allele 1 are error-free. Among other mismatches, T>C transitions were the most frequently encountered substitutions (14 out of 44 total
substitutions), with C->T almost equally abundant (12 out of 44), followed by A->G (5
out of 44) and G->A (5 out of 44)
1 Although
it is possible that a significant fraction of pet B/D sequences cloned out of a pair of pooled PCR
reactions, each beginning with about 105 sorted cells could contain the same PCR error, especially if this
error occurred at a hotspot for PCR misincorporation (Keohavong and Thilly 1989), it seems unlikely that
all of them would have the same error.
149
(Table 1). Detailed studies examining Taq polymerase errors have proven that most Taq
polymerase errors are A->G or T->C transitions (Dunning et al. 1988, Saiki et al. 1988
Tindall and Kunkel 1988, Keohavong and Thilly 1989, Ennis et al 1990, Eckert and
Kunkel 1991). Therefore, the type of sequence mismatches observed between clones
assigned to Allele 1 and the Allele 1 prototype do not conform to expectations of
variation due solely to Taq polymerase error. It therefore seems likely that natural
genetic variation and/or sequence determination error contribute significantly to sequence
variation within Alleles.
150
Table 1. Nucleotide mismatches between clones assigned to Allelel and the Allele 1
prototype sequence, excluding G->A at position 1243.
Nucleotide Identities:
Prototype->clone
Number of
Occurrences
Number
of Sites
T->C
T->A
T->G
C->T
C->A
G->A
G->T
A->G
A->T
151
REFERENCES
Dunning, A.M., Talmud, P. and Humphries, S.E. (1988). Errors in the polymerase chain
reaction. Nucleic Acids Res. 16:10393.
Eckert, K.A. and Kunkel, T.A. (1991). The fidelity ofDNA polymerases used in the
PCR. In Polymerase chain reaction: a practicalapproach(eds. M.J.
McPherson, P. Quirke and G.R. Taylor). IRL Press, Oxford, pp..
Ennis, P.D., Zemmour, J., Salter, R.D. and Parham, P. (1990) Rapid cloning of HLA-A,B
cDNA by using the polymerase chain reaction: frequency and nature of errors
produced in amplification. Proc. Natl. Acad. Sci. USA 87:2833-2837.
Keohavong, P. and Thilly, W.G. (1989). Fidelity of DNA polymerases in DNA
amplification. Proc. Natl. Acad. Sci. USA 86:9253-9257.
Luria, S.E. and Delbruck, M (1943). Mutations of bacteria from virus sensitivity to virus
resistance. Genetics 28:491-511.
Saiki, R.K., Gelfand, D.J., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B.
and Erlich, H.A. (1988). Primer-directed enzymatic amplification of DNA with a
thermostable DNA polymerase. Science 239:487-491.
Tindall, K.R., and Kunkel, T.A. (1988). Fidelity of DNA synthesis by the Thermus
aquaticusDNA Polymerase. Biochemistry 27:6008-6013.
152
Appendix IV
ALLIGNED SEQUENCE DATA
153
w El
qi)
•
0
C
N1
1-4
O4
0
oa
0
m
mu
C
m
0moo
Om
,-U
mm N>1
C)DO
ý4
-H•cco -
=
n :D
o04t
.- ,•m
J:
ci)t-l
0i
*H
CO U)
0-,-cy)-c
UU)
-
:D n
C
00
I I 0
I
KC4
F:Cg4 4C
4 000
l<44
40
O:
0000000
<00000000000000~I ; z c
f< f<
I t < 0(90
c OI Ohct<
-
I
00
0c00
I
I
I
I
I
I
I
l
II
I
co
00000•0
UUUuuUu
0
c)
=) =)O
0000000
uuuu
(9(90(90(9(
m i 3
Wor;l
oc4
,1-.
4 -,4.
010000000000000
mci•
kD 0>
w- CU
H 4*1
,-U)
0,W E-1
SZ
UUUUUUU=UUUUOUOD
000000000000000-
U . Ou
tj*-H 04
,-. - 4-o1-4
m
0'-4
<<
oc <
lI
00
l
l
I
001<
ou•
0000000
UUUUUUU
l<X I
(9
$
o00000>000(9(ouuli
UUUUUUUUUUUUGUUui
•
UUUUUUUUUUUUDDO•
000000>OOOOOOOO
ou
IIIIUI
I
UUUUUUUUUUUUUOUU
0000000
00000
0000000
0
UUOUUUUU
0000000
D00000000000000
uuouuuo-uuuououuu
S-U U 04
cUN
0 .-o
E
. 00
00
0000
(9000000000000(5-
0 ý0 Uo
la)
0: O
- 14 ECtý 4
00
0•0
00
0000
#£,.
(1) ."0-0
mn Mc
00000000~0000i
Q-)a4o
000000
00U000
-H
0C
ciJcci
' 0 O0
U)ý4OD
o-~•
0Q4
ci'4 00.:
C0
ci
)
ýcN
co)
44cc) cur5ý4
44-) o MN
1.-- 0
i) -Q
'A
r->
c>
-4
ci)
NC'
4-4
0w0
00
00
0000
0000
0
00
000
o u
u uu
oUUu
ou
u
<U
OctPi
UUU
0 . N-
-
00
04 04
0Ocm
0
00000000000000
00000
00000-4
UUUUU
UUOUUU
00009
U0000
' Z0mVam 4)4
4-3)
rM 1.4 -H r.
1-) (acc
ý4-
D•OD
.OD000
ai)n
-H
a
04
00
4
40
N0
0
0 *0Ea
o0)
0 .-
000000000~
UUUUUUUUUOO
000000000
K
l<<<
c1:X
MD•
:D
D D $4<<
nD
0000000000(000
000000000
<4Otu
000
(99
(((
OcC
o>C
0mm
m
co w :2f EIXo
0)a)H4 n
XX
X
m
U in
EA
N:R:
E
•
Oc0-40
N
(n
O
1)4 '
C
.4
0
U
(a0
aA I
MC
o
o-,-i
OiJ
cU'
o
C
0 .4oT
n
00000000000000
4-
. 4 U
<')Xm a4
-4I-4-4-4-4-40-4
00
U~ca0
~2
0000ou000
00000000
0
c>1
0
0
(9000
U00000
u
t4<<
0
0
1)
r,0
E'Q *110
U)
)
U) -H - 0 ci)
M U~4-4
-
-4>,
0000
0-
UUU
UUUUUUOUUO
UUUUUUUUU•
00000000
000000000
0000
u :
4444
E
>
00000000000000U'
--- 4-4-4 -4 -4 -4-4-4 -4 -4 -4-4-4 -1
(9
0wHa
(9 X9(949
r-I"
M
Cmm
-
uO
mo
% (9(9( 9(9(9(9(9(9(
00000000u
(9
c(9mj-99%9-c9omc
CD)
D(
C DC
0- - Ný
- qr4 ml
- - -
)
-
9. 9(9M94(9aý
DC
9
T )
DC)C
004
r-iCN)-40
S
r-4 CN r4I
m
0 S40
0o
C4'0E- M~a3O -4W
oo4m
0•
•ZHmbM
uO Om
E
N-
-
C,
S0 C
A-) -4
ý4~
~
Sa4
CMI
(l0 C)
~rE10EHm =
mo
ON0>-4UO
M Mr>1 4
C
r
44 W =ý
W
:3:M
.
4U
44u
MM
Mf
0O~r
4
XC1
O
m04m
> 0
AUNM
UUUUUUUUUUu
UUU
UUUUUuuu4
UU U U U U U U U U V U U U U r-4
U UU
U U U U r4
U UUUU U U U
UUUUUUUUUUUUUUUH-
88888888
80
u. I-
00000000000000UUUUUUUUUUUUUUU00r4
U UU
UU
BBC
U U U U U
4 U4 U4 UU
fBC
4 U4 r4
r
8888•
8888•
0
00000000000
UUUUUDDUUUUU
U -I
-1
ý4
u
u
uuuuuuuuuuu
UUUUUUUUUUU
00000800
08
u
UUUUUUUUUUUUUUU-4
u
1-4
C,--'
uH
BB
OU'r~-4
0U-4
U
UUUUUUUUOUUU
UUUUUUUUUUU
U
u UIUUU
8UUUUUUUUUU
U
Uu
U ()U
UUUUUUOfr.UUUUUUU
UUUUUUUUUUUUUUUM
:3
UDDDDDDDDDUUODU
3 :DU
DDZ)
ououuuouou u
UH
<o•
O i
UUUUUUU4UU1
88888888888
uuooooooouuu
Mooooooooo
uUUUUUUUUUUU
OOOUUOOO
u H
0
uu
I
8•
ZDDr4
UUUUUUUUUUUU-4
•666•66666660
~8r4
oooooooo
ooooos
•ULUUUUUUUuuuuuurH
UUUUUUUUUUUUUUUH
UUUUUUUUUUUUUUUoo
2888888888:
04
U-
U-i
o'-4
UU
4UUUUFUUC
-411
un
iý uvUnu
2!.
$114r.vUnU9-4
u :: u :ulil
666 Dn
r4
uuuoououuouuuouus
888888888888880
vooouou
ou o U 7o
,
UUUUU
UUUUU
r-4
OH
U44
00000 000
UUUUUUU2UUUO
UiUUUUUUUUU
80
UUUUUUUUUU
UUUUUUU•uuu
OUUUUUUUUUU
UUUUUUUUUUU
uuuuuuuuuuu
880080088008
UUUUUUDDUUDUUD=H
UU
888
U8U88U8888U
00000000000
oooooouuouu
UUUU<<<<<<<
UUUUU
8888U8QOQ
iI
~i
~
s
TI
4
0000000000
00
I
0 oO
4,4
~uUUUUUUUUUUUUUUUU
UU U U
U OO
UUL U
O~U
U UC) UU
U ) UU
UU -4-4
88 ULUUU
8
8
88
8
8
U
8
I
I l
l
UH
UN
0
ý
vouuoovuuooovuuvU UUUUUU
OH
i
8~40
U<<U<U4444U
UooUuouoUuU
UUgggUUgggg
U-.
88888888888
uuuuuuuuuuu
UUUUUUUUOUUU
Oc< 4 4= i 4 4
U-.
<on
U'-.
UOIOJIOI
UH
U.-.
U.-.
UUUUUUUUUUU
UUUUUUUUUUU
4:-.
U.-.
UUUUUUUUUUUUU~UH-
('-
4:-.
4.3
U
U'-.
LI
m
4.
.1111.1
HI
o
A4 HOEX XO4
0('
w rc
"
C
r-i ('4 r-I
-.
w4 L)
0
m.n
o
U
u
O.0 U
o0-4 U 0
U oL)
fu 0.0
''ko
.Q -4 O
C14
0)
0
tn wH
Oi m =
r-i r'4 mI
r Ln %o0r-
w
cm 0>
w.
4 "' r'1 q Ln
r-4 T-. ,-4 v-4 v-4 r-4
vA
S"
m v
L %or-
wo
00000000000000004
0
0
-4
aO4-
r-4 ,-4 r r --I -IA
-fr-4
ý-4 4
v-4
44ý§" 4 r4 -4
CO4('41 C). C>4(D4 C>4
(40C
(4 C'4 (' C>4('C>
C>1
> 0'
.......
.......
0
"
m -
L
-
E
N
0D
a)
m
C my U
.- 4
Fh4O
O
moo
O-4uo
mmr->1
o
o7 0 r-4 U 0
.
,-4 ýo
0
cc,
E x M:
<
-Q
-4 r-4
llýu
Kc a cz
ý: 0
I4 -a
M
PO
a
J0
u
000000000000-
UUUUUUUUUUUUUUUM
UUUUUUUUUUUUUUU•U0
88
0UUUUUUUUUUU•U0
88888
8888
Fc PC Pi 0:4
oz
4
UUUUUUUUUUUUUUU-4
UUUUUUUUUUUUUUU•
000
0000000:4000
888888888888~88i
•UUUU•UUUUU•
UUUUUUUUUUUUUUU000000000000000
000000000000000
UUUUUUUUUUUUUUU,-l
UUUUUUUUUUUUUU
UUUUUUUUUUUUUU
•U UU
C)U
C C)U
U = D UU
U0
UCDU
uUUUu-4
UUUUUU~UUUU~~UK-i
UUUUUUUUUUUUUUU
UUUUUUUUUUUUUUUU
UUUUUUUUUUUUUUU
00000000000
00
000
UUUUUUUUUUUU
0000
UUUUUUUUUUUUUUU,-
U- 'A
UUUUUUUUUUUUUUU•
OOOOQUOQ©U0004U
0000C000ut00
xc
F<<P: x<4
4
UUUUUUUUUUU888888888888Uuu
0
-
UUUUUUUUUUU~UUUU
8888888888888U88
UUUUUUUUUUUUUUU-
000D0
SU
UC)Uu 0U U 000
UUU)U 000UUC-4
JUUUUUUUU
UUUUUUl
ououuouoCu
uom
000000000000000
DUUUU~UU
c0UU'-4
U
MUUUUUuUUUUU
0:4
nn n
U
n
M
0
-4
-4
n
0 0
o
01UU0 u 0 UI 00
>
u uu
00U0 1UU>4zuUC
10I 0000
nII
UUUUUUUUUUUUUDUU-
DýD
ý:3= = ):)= u u n-4
00000000000
00000000000
000
0000000000'
DUuUU
UU UUUUUUUU-4
O
O
000000004
0UUUUU4U4UUU0,-i
UUUUUUUUUUUUU~UU4
DODOOD DO•ODUDD
0
00000
000000
OUUUUUUUUUUUUUU-
0
000000000
UUUUUUUUUUUUUUU-4
oououuouuuuouuuu-
cci
UUUUUUUUUUUDDDD
ii
cci
eW)
E
r-
M
.0
r-4
OmCl
moo
0
ID
U)J-
wOHa4
U)'t
-i c4 m " Ln
00000l0000
oa
C:)
I"
ONCD mm r-
mee
"w=xFCf<
D
a
C)
AZ
0
>,l-
r,"HH
D -0 " m
w
0) a) H 14L
o
cJ
O
0000000000000000
w r-Ivw
w r- w
- w
- lr w
U4 Zc
<X 0.4
xc MxEIn
HA
'-4
fA
o0
0--
0000
.I (N -1
0D-4
U c:
0
m
>
N
0
4
m tN
u'L
U'o N-
m
i
(Nm)
r-4 -4-4-4 -4
mc) 0
LA-,o
-4 -4
-44--4ý4 -4 -4-ý4 -4-4-ý 4-ý4-4-4-4 -4
0000000000000000
LAL L L L L L L L L L L L L L L
'V
~ m~
z
~
o-uc
a
in
4E
in
a4
U
-L Cq
0 zr0)D
~000000000000040H
I
I It
I
I
I
I
I
I I
I
I
nCD C
Cm 1
CD(r
Uio
D r
000U000000000000-
I I 1
00 0 0
0 00
80000880808000000-
o<
4 4
OtOtl
t c
0 0 00H
~ :cx
:D:
: :))nD
4 FC 0: DC4
40: i
000000000000000H-
l<0:
lA H
UC)uu
0: < :4
f<o<
40
:: MDH
-)u
K c f
uu
4 :4
:3 C Cd C) UD U C) D CD UD CD UD C =D CD
u CD CD CD CD CD CD CD CD CD z u'
0
00
C C
CDD8
CDD
D-I
'
u
(-)
0000000(
00
000H:
OOOOOO:D
OOOD
OOOOOOOOOOOOOOOHUU U
D D
M :D :D000
H
o<<oqtc
CDCCD:CDCCDnCDnCDDCDCMD
000000000000000-1
CDCDCCDCDDCDCCDCUDCDC0-4D
CDCDCDCC:CD:DCCDCDCDCCDCDCD
u u U
u U
DCDCDCDCDCuDCDUDCD
u
~000D000D00C000000000CDDCDC0CDC0CDCCDCD
r-4I
D00C-)
U U
u u u
u u uu
u U
u
0:4
u U-4
U UU u uUCDmD:0
0
00U000H
000000000
000
0000000000
00-
< 04CCD<DCcDcD<CDCDC4D<K4D<C<CD
CDCDCDCDCDCDuCDCDCDuCDCD=CD:CDCDCDH)
r-)u uU
u
uU uU u(U
0
00
0--
~CDCCDCDDCDCCDCDDCDCCDCD
CD' DCDCCDCDDCDCCDCDD~CDI
MMMMMCDDCD
CDCD~~
u0C
CD D-
0 .)C 0
0
0 0
0 0
U () 0
0 0
0000000000000<0
u uU
0)>~
U~-)
0)~~I
40)
-
oi
J-i
u
-i
m0
CO(nLn
01%0-A
0
m .- i wx .o r-
-i %.o,Q
-iH
E
CJ)4
H
~ ~n~ ~ E~
H
HHHH
0000000000000000
a
u
uu
ý4'0 E- in :
v.0)
(TI H
uu
H
(
0
rH( "
H
H H
C. 0 0 0 0 0 0 0 0 0 0 U 0c < <
< < DDCDCDD<DC<D<DC<DC<C<
CDCDCDCDC DCDCDuD0DCDDCDCUCUCU
CDCDC) )D D
::)DCDnD DH
:)DCD D
0)
EN
:::
3-)
0
H-
ý4
in
E-4ulC:3
0'o'(3-
0C0
0c
o 04
m I-l in I'
r-4
0 r-4 -40
0'0
r-
H HH H
r 0'0
r
C) C) C) in CD N C) C> C0 H
H)
H (D C)
H C)
H in
H C.
HE
HomoH
HH
HH H H H HHH
H H HHHH
0000000000000000
4)a
r=
J
-4
MoC>
m
n
enoo
C>D
C
0
f
N
.
4 UO
MM
o
C,4N-4 C
o 0C-4CUOO
m m r-enr->U-4a
U
omm
mCC
A
OH7
ý4
-4
UU4U4 4 oc4 4 4
()UUUUUUUULUU()U
C"
en
V
0
>
UNDaQU-W.04
O
)
E
0
Q
X XO
mUm4
CooJUUUUUUUUUU
U4-
.)U LUCUUUUI
r4UUUUULUUUUUUULU
UUUUIUU
guuuuuuuuuovuuuu-4uu
ooo
oo
oo
4
o
oo
U
~~~~~~U
LUL) U
IL U U )U )ri
a
U U
aI
4
0
U
U
UUUUUU)UUUUU
U UU
UUUULUUUU
00000000000
U
:> U
4000000):
000000o
UUU UU
UU
22222
moom
n=6
2
o
' -4
22
r-I
888888888886
o
en
u
Z0)
-1 V-'
u
UU00000000
4 U
r
0
0-4f
UUUUUUUUUUUUUUU
U U
U
UUUUUUUUUUUUUUU
4
vr4
>1 0
ý4
41 r4 :D cn
mC <<<q
> m r r- r-H
mm24
4) TI
r-C"
z3
mo
MM
m04D4
M U O W
N
- LU o
>4UC
~MaVLO b rW hMM ýe
4 rIT4 r0 rEEý23
o o
vo 4
vc o r- oo-o
-4o -4o ro o- v4w4
m D
ý >C>C C
4
e C e0 ýCl(
C> 40 C>C
DO D0 DC 0C 0 DO 0O DO
em
[a0OO
1
LU
~~UUU-4
O
U
-i
O
:n=4U
M:
m
m
<4444
4
U
U
UU
U
U
8888u88u888888880
uoouoooouuuouuuu,4
2
424
0
,,,,,,,,,,,,,-
I
u
UUUUUUUUUUUUUUUi
a
m
N
e
Z-)=)=
0000000000000000
4
4
mn
)Z
)D
44
4C'D.-n
m
4
44i4
0000000
0 00 MM0000-4
r= rý
,->1ý
mnr-C)C)
4J
v
a
e
-4N n
a
L
m
t ra
o
K
M
m
>
M~
D h D ic,4m cn%
r- 1-4ýý 14 -4 r-4
o o
,
ric4o o- ov - o o-i-4rf
C
C
) )
> )C>C
>C C
mONO Y mO N0
%4
mý(
-t
00
88888888
UUUUUUUUUUUUUU-i
4J
4*
U
=n=nDnUD
-4
)UU
UUU
u
u uu L
V u
uUU
UUUUUUUUUUUUUUUr-4
0IoUUUUUUUUUUUUuu
UU0
4
0
MI
U:4440
0UUU00UUUUUUUUUUUUi
UUUUUUUUUUUUUUU.-4
rU-i UU
888888888888
..-
0I-4 I
UUUUUUUUUUUUUUU
UUUUUUUUouuUuUUU-4
UU
I
L)
DU
U
U
UU
UUUUUUUUUUUUUU.UUUUUUUUUUUUUUUi
:3D n
UUUUUUUIq
0I
~
~
~
zr-Iu
00
0
00 0
Z:)
U U
I
u U
uuu u ::u=
u
uu
CUUUUUUUUUUUUUUU
000000l0
UU
U
- 000000
000,00-4
CO000o
000004
UUUUUUUUUUUUUUU0UUUUUUUUUUUUUUUZ-4
UUUUUUUUUUUU=UMi
:D Z'3
UUUUUUUC
I
0
UU
-i
UUOUUUUUUUUUUUUUUU
~~C")
UUUUUUUu
UUUUUU
Z:)-4rC,
0000000,-A
n :D
ý:8888o
o4r4
o
>C
%0 Y
)
W-
)
40
m C )
o0mm
m oo
N.: o
0
o
MMr
mo
mmr
0
1ý
UOO
CN,4
0
:
N ) 0
0mm
o
U
E-A
Onvc-
moo
mao-iw
OtIEmZE-co
0o
F:
L
UUUUUUUUUUUUUUUM
OBOOO000000O000m
UUUUUUUUUUUUUUUi
DOODOOOODODOOODO
0)wHx
04<<
C:C4
<<<<<<444444444
4
OZ4
< <O 9
O<o
UUUUUUUUUUUUUUUM
UUUUUUUUUUUUUUU-f
<<<<<<<<<<<<<<<
UUUUUUUUUUUUUUU,
UUUUUUUUUUUUUUU.H
000000000000000
u
U L
U
ouuououuouooouu
000000000000000,A
c <
:
o<
8UUUUUUUUUUUUDUU~
UUUUUUUUUUUUUU,-
U
UUUUUUUUUUUUUUU
l<oc uU
:404 IUU404
UUU 4
I)I)uU
U Oct
U
U u u <u.-4 s
UUUUUUUUUUUUUUU
UUUUUUUUUUUUUU
U-I
UUUUUUUUuuuuuuur-4
U UU U u C C)uC
U
U UU U
U u u 1-4
UUUUUUUUUUUUUUU-4
UUUUUUUUUUUUUUU.
UUUUUUUUUUUUUU
00000000000000-
000000000000000
000000000000000
uouououuuououua
ouuouououuuouou
UUUUUUUUuoUuu-i
UU UUCC U U UUC)U U U UU UU
0OO
0O'H00000000
<I
,:Cxc
0000000000000
F:4F
4 04 04 xt 4
IcL
:C5
I
I4
r<
I I
I
I1
I
FC 00
888888888888888~
0r-i
UUUUUUUUUUUUUUUrH
Oct<0000000c
UOOUOUOUOUOUUUUU
000000000000000UU
UU DU
nU nU
U UU
UUUUU)UUUUUUUUUU
OOOOOOOOOOOOUUUUU
UUUUUUUUUUUUUUUM
DDODODODOODDDDH
COOOOOOOOOOOOOCH
DOOD•OODDDODDDD
•
UUUUUUUUUUUUUUUI-4
000000000000000<
MMIý
1444ý
UUUUUU•UUUUUUUU•
OUUUUUUDUUUUUUUU
00000
4:4:4:40004:
4444
0CD00
t0 0-I
-4
00000000000000
UUUUUUOUUUUUUUUU'OZCU
000000000000000
UUUUUUUUUUUUUUU
UU UU UUU UUU UU U U
9 0: c <<OtOc
000U000000000000
U0000000000U0000
UUUUUUUUUUUUUUU•
UUUUUUUUUUUUUUU
000000000000000<
UUUUUUUUUUUUU
DODOODDOODODODOO
OOOOOOOOOOOOOOOU
OCC
X
0000KC
oo<
n444
44
U 4 4 0044
0
04444
4
00000
UU UU U UUUU
4~~~~2
Oc
U
0:
UUU-c
0)0
0
moo
0
0r)
0)
mm (In0a0
oo IU
O)
o.
ý
%-H
t3
a)
-H-oCI)
01)
E
41
4
O
No
.mt
am
0mm
moo
ma.-wN
)C
oo1o
ooor
4P4P
rooE-oooooE-
)Ha
ao'o -lUO
mmN
4
z
0000-40
omm
o1
0
n
xý
~
o o
4oc
C-I C>) mD(D L3
000000000000000
0
U.)
4
D
N C) C) C0 C- C') m> D CD C)
l
4J
r0
-
mm
2ý
O
E
0 N4
Cm
oI)O-mm
r- >,I 4
Dmm L
moo
0
-1
a4w
I
moo
.
Wr a
a4wHH
cq -4
o0-1
0
m
mmr- ,4
M M OlA
Me
00
0
0
O
000000400
0uouuouu 000000000
000000
U
0
U
0
U00000000
0
UU000000000U
u0 U
0
U
ý:
::)44
4 :: 44
00000
0
1
0. xD UUt<
n=
U
A
40
U
~
l
I
000000000000
i4
0
11
CI
00000000000000
0UUUU00
Unun
uuu
Z.4
oUU0
004
00 o
U~
I<<<
00000000
00400u0Uu0u0u
UUUUUUUU
0000000
1000000000
Uo4U
-
ouu000000uuu
I100000000
I'00000000
I :)= n :)n
= 0 =
0
0
'0000000
UUUUUUUUU
00
0z
000000000
000000000
000000000
<-
0
0
ID--
uu 0
D
4i
E 4
-N
4
u
000000000
u
uu a'U
uoO
Ub
omHmEr4
UO M
m<
oocPI
mmc-><..
04
Aoo: O
mw
.- 4c'4
~
E~-
u-3
-W4
OVIO
~
)
(N
-4~.N~~. rcE'rm
\
1-44-D4)
000000000
ID
(DD 1
M
-4H4-4-4-4--4-4-(-4-4-440-rAk
ý4
OO0
0
0Mr 0 ,1ý 0
-4-14-4-4
0
-44r-I4
Q 0M
Ft
mo<<U<
0
ID>.
Usd
C
0
-44
-cfýO4
Zi
000000000000000SI
FC
000000
l
0
U
0000000
000000440000000
UUUUUUUUUUUUUU0
UUU404 0
uC)UU
000
0000000 0000
0auuuooi
0
U
0
0
C C t
C)U0 U0 UUU
0
0
U Uu
0
0
0
<
0
I
C U U
U
0000
0000000
U u0 UUU0
U
000000000000000I-
0
0
#:CZ
#C ,
00 0cýP
0004<
U
000
00000000
0
UUUUUUUUUUUUUUUl
1110
1Il
HM 'II
MI
00
U0UU
U
i
l
00000000000DU
000000000000000uu-I
UO
U
UUU4
0000000000U0000U-4
00000000000004004
Uu
0
M =)UUUDUUU
II
8888
U
00
uU
0
OD0000000
D
08
=0
00
0000
U UUUU
U
#Uc9U :=n
4<
n:8
0000
0
:
000000000000000-4
0:40:
=
uuu
00
88 8888
U
00
000000
U00000000000000U
0000000000
0000
u
000000000
Xa
N
m14
-4mC)C
=
n0 00
30
mam~
uuuuUO
m0
0
0
mwa
Uuo
Ou
UO0m M
mmN>1u
D
H
)C
C
m4
r-)-4M4--4-4-4
0-4-4-4-4-4--4-4w -4-i4-4-4-4r-4
-4-4
)4r
00000-00000
00ý
ýu a0000
-
ca
(1)
o,-i
0ý4
lz
0U)
3mn
.1.1
CU00
LV
EMlEU
01CC
u 0
(a0
0. N
in
a4 W
rAýE
lFCUC aO
C
0
000
0o0
000
0
U
DOU
000
000
U
0U
0
000
0U
u
UU
U
U
000
UUU
0 U()C
ou
U
00
88
000
U U
000
00 u
000
D
M =)
Soo
00
OZ
C):EU:
00
U
U
U
C,
o,
0
00 0
0
U
U
00,
0
uO
UU
UU
0 00
UUU
UUU
U00
E<<0
0UUU
0
UUU
U
Do
f< 0
UUU
8
U
o,
U
U
u•
o
ou
OO
UZ
OO
EE
U
olo
UUUU
UUC)
u~
000
U
U
Eu M:C)C
U UC)U
0010
0:
C,
0E
<
o
< X4C
1)
EU:
u CU
10 0
=)CnFc.4C
)=
'U
<<
DI
I
IUU
•oao
Cl)
0m0m
CDnC
0
4-)
0
0000
a~ E-cO
0000,
0)000
C')
1-1LE
rnm
r-Ns,
P0ý
000
oU
0-4
) 0.
0
UU
UUU
UUU
UU
UU
UU
UU
00
00
UOU
000
000
EUEDDE
DU:Dl
00
000
4U)
JJ
o
U)--
S0
0 W .Ov*o
0mm
moo
me.-m
UOE-4 11 3: E-4
oo E2
DOHamu<
9uam
(x
zz
c3
l
ý4
0 -,
0)0U
U)c
4
o
C
o ,-I UO0
m m N >, 14
-4 .0 Q r-4
0
EUU
00E
ilý
u M
ý4
)
a
0000000000000000
") IlE "U)IrI lr11
C)
I
V
r-rITImIV
q
c
ONC-0
moo
mmr
rO mýu) 3: E- N r-
eO
-,O
w0
004
omin
ini
in
>
)
.
m
cý o4 w
>0
u
m
o in
mn
0000000000000000
inO r-U
I'Di
n in in
in
mn in
ý rIn r-I
in
" i
-A
r-i
r4I C
r-i
r-i
HAC
m C
C
C CDCDCCDCDC>Cm
T- oi - o o o- rA
4r- rI
- ri
-4r4o- oý
tot H
mmo --
C0
V
, 4.H
-
00
- r
00i-i.L
co
(1)H wC)
(C)
f-4 F-4Un an
EL.EL0
x mxHHA
00
C14NCNC
-9 W
i-
04J
Um
U
Co
00
r-1m..HXX8-1-4,-44
-0
) 4J
-44
rOoE-4
F4~Lfln .-000CC
4 rH m
>,
>,xýc
1
U U t a)N0ra-H -H
x
0rE-'C
(NI
Pt
aýa
L
it,
It
=xU
Cl)
u :
'UU c
U'U
UU1U1 UE
U I-4
1U1dU
00
ID~ICID'CUI
()IU-L
U
U r-O
I:
U
- FUI,
(- <U
EU
a4U aU
U a4U
iI
iu
IýU
u a'
BIblululu,
sI•I•I•I•I
IJ4
in
c
P
U ) IUU
rU U OCU
U
U
I
U
U -I-
I
U
1E4
E
E D
E-
I
I
E4P L3
,U4404ItU
i
iE-
I
IU
Iu
I
E U
U
u
E--'C
0- o
Eý
018,9,9,9,
14 4
E-U
'U
U 04iU U
U
C>
i-U
I
'U
U
'F-'
t,
,-4UH
'-I
10
U
E0
Q0
>
U
EIU
E- isE
i
iUU
U
u
UIUIUIUIUI
ý4-
U
*0
U<
ON
a) U H
U
1
I
i
U
U
iU
I
Q E-
4
U
I
U
UIUIUiUIUI
I~
IF--'I
u
U
U
U
0202020203
U
F-'
CoD
-- UEn UU13
UL
F- IF-I
F-'
F-'
C)
QOO
.0
~
U )
I
ui
EI)a)U
H.0
C)
I() oI
C)i
E-
E
iE-
U
U
rug*
I'U
IU
r-
i:U
E14
F4
U- DI~C)l
EU F.
IU
E-. I
Ell
1
" Ci E- i
E-. (At Uri -iE
UIUUluUIU4
ý)CV)-4
x U i E
U
-A UC)0 C
)4)41 0-I0IONCY
r
- 3)a
rNA-f
L
ý1 r r
Q1 N - O
>0 > ý ýýU
M M -I-
000000000000000000
U E-'
I-I
a4 UUUM
aUa
IZ
U
oUlUIUIUIU
1 E- 13.
E--4'-
E-'
P > UE
U
toU
H 'HE-.I
IHIU
u
U
itE-' i
EU IF-
o"4a
CýM
0
4.) r U Q)
,
00 00 0
00 00 0
0
CD
rc r5EýE
L
Hý -)rI
,In( XXU0 4JL) -)U r r0M
'Xý X24I E :
E n1 ý 4 W' C:
1 '
r, -rLnk
r D h(D- EU "()
nUum rU 0Li 0U- C
r- - 4 1U 4! 1ui
-4 -1 4 1-4U14
r- --iri
rIrU I Ar
UAI4ý1
4 ý, ,!- I U ýI,
C)C)C)(
C C ) ) )C
CU(DCUC
D U))CDC)C
rýr r4 ,!ý!
-
- -
-I
-
Iq
l
-U
I- U
Ari
U
Iý
iE
a
4-1
U
U aU
U
Utoo
a)
IiEi
I
UUUU
00080oupy
10
1
1
1
1a
0144C
'U) a xU a 'U<)a
'U
LOU) u
N< c
E- - Em4 a)-I
E4
I
tN
0 t
*Ha)I'o)
i )a)
a) -UU
a)a~~i~
H
'u Lf
U-to 004
oo
I>
u 4
-4NI
C
oC
40 (14
Ln in r4ELI)W
z-
u
E O
)
-"v-P(-4
L9
0I
0
O1C
*HC)
tOo
0OH
ELU L)
~
SN
*H0'(i
r-i
N.0X
"
T
-a) a
N
N
UU C)
CM
C)a) a
M
r
M
M0 (D
)~
U
c UJJ : U
,>X
co)u--(-n
C) a)
; FL0 N
0N a
mn.~.~cw
U
En
a) -HAI
00
.- I r-4
-wOq ciai
VO V E- E-'
H1 H
N
CI
CO00ý
I
00
-4-,4x
IN
10
C) I ) ..i
-4 %t - r- w w
IN N -4rWMW
I
4f- N
4OQ N N
m LA ýi-4
w >, >4.0.
U-u oU m id-4
14
1
a4 a4MCM =
U U 4 -) -U U Z ZtutM
Q
~-w
ta
LAnLA -4
CNCV N
U I
iE-4
uIU I
UI
L) U LJ
En C)
Ena
'U
IttBtGI
I
C E-"C
Ut 44
E ý'U
MIUIUIUI
>4
L'
10 I-)
u
C.to
10
U
104
F E-BU
U
>04
to
a
EI 4-11
1111
WL 0:1
W
EII
i
II 12
.t U
II 1 0111IIII
IUIII
tC.
I
IIL
I~
>U
04I
I
III
)
I
E- >
010E4-,,
i
E-i I
II
U
,,IE-
L IE U a4 I E- U E-UI
U
U
EI
u
lUil11llI I
I I I11
tI
cIN 14
IN C4
N C%4r4
IN4IN I4NU
4t IN I4
"tQC14
C14 Na
bI4
B>>BP>H
u•14
0
UE-'
IIUIIE-i
0I
t~
I
I
,,.
I
I E-,
UUII
Nt
N
1
E-'
I4
I
1
U
I
It,
E-4
III
.lll
I
I
P
040
'0'
0' UE-4UiU-U
11 IIOIII
U tI
tU Ut tU
ul
I
5
I
b
DI
I
10
EU
U
UIUIUIUIOIU
E-
E-
I
I
I
11
It,
I
I
U
E4
>4
lEIE-
1
I
'E0
tI
U-
UtU
CUC,
I
H
H >E-4>E-4
U- 'U
3
U.1E
00000000000000
U
NN
00
4) >4
**
U -)
UG0
0-4•
IVtC4l H
H
A a
UE-4
S00
~
E-
M
Ia
N N ININ
X0tu00
000
-44~-4000L
O
-W 0 ýNN
d1
En En =
0VU -4
v-4 v-4 r-( 4
1LA
v-4 r-
U
N CD
141 a
-4 .t v-4 v-(C'4
N N
0Co
-4 I
IRV"IV
Iol
ON 0%
IOU)~C
mm
C) C>C
C14C414
o
DN~
CDNII
>1, :>,..C:
U U to (a -,H -,
U U4J.J4-UU 0 z aU
En
.- t~~~
O0
-4
X X
tu r-I -A
4- 4J -H -,1 0% CN
N) )
O WO1t 14 N- N- .0 .
N N
n
14a4a4a4fr4
m
mNm~L
at 0
-i N m v LA wD Ný wD m0
r-4-4-4 -4-ý4-4-r4-4 -4 -4-N
En-40
V-4 r-4 W-4r-4 V-4 r-4 V-4 r-4 -4 9-4 -4 -4 -4 r-4 W-4V-4 -4 -4 r-4 rCA
OH
-4-4rl-4 -4-4i-4-4-4 -4 -l4-T- 4-T-14-4-4V-4V-4 -4
D
C
(Da
000)a D00C>000000000000000
(n m m ml m mnm m en m m m m m m m m mým mn
(J)C)
1J)U)
oN
VVHH
I C1)
"\ (N
.H .- 1
i
U1U)0oo
rO CI)
<csE-4Ea4 a4 Ejx
M
(DCo
El) -H
E)IJX
-H
AJ
4-) -4
UHI
(1
I U
I
UW
UE U
U
-H.H
C
:CNNC
11
moo
HqE- "I Ul 000C40
i Om
m 00 >1 -140-HCYC
>1
ýa U U (a R -H-)a-Ha4r4
n
U U
JJ U U r- 1- r0 M
EI
HIt)
U
i U
U.
a:
X X -I,-1
l ý -1--H
m
ocu
HH-I
>1 a:
U 0u
J
Lna
0)
>1
U < -,<
He
c
H
II C)
I U
IaL
I
E
U
-• I
I
IHI
H I
i 0
i
4U
U
UU 14U
10
0
U
U
i4 E
I
H
01010000000
0U 10 10
'H
IUlUiIPIBI
IUFUI
:I
EI u.. i
p
UUI~iIal~il
it
I
i
I
i
ZI ZI • 4I F ,iZ ,
P
H
-H
U
IIbE-U
E-• Ip0
CH
I
H
0000000000
010
01000101
EE- 1010101
E-.
0Ocal 'H 0 04H
0 O
U'a
I ý
I
E10
I
I - I
101010101
u
-
S
01011U
I
Pa E-~
P 1
U)
lU CE-)
01011
010
0uIH
IHuau'
a'
H>PHH>H>H>
01a:10I018
'a 101
U
>'•10
H
10
H 101
H 10 H 'H
H101
00808T8T8(
U
E-J i
a0000000000
101010
'a:
U>HZU>U>U>
-
i
it)U
(.
Ia:0IUIPi
U IUI
U
HUIHulIPi
E,
F-4o
4101819191
a:
UIUIHIOI
i
a41IUDUIOd
<: I <D:i
010101014
<:
Hi~li~ilit
S1010
i
IU
',U• IU
HIUIUIa<laI
c U10
w<DFl
I P I '0<0
I 01 I
0000000000
0101010
'H'
U
I
0
F9
010
(E1
UIUIUIHIHI
U
H 10
1010. 1
IU
1
p
I
1010
P
I
4HP
H
1U0
U U) U
(-)
U
Oct
I
HIHIHIHIHI
urbiVOUiPlull
U
0108H U
<t5
lpt IP7
0
0
0
I
H- I
I
I
t9
ulbI0I
UIUIIUiUI
0U i 0) 0. U U
0i - U0it
i 0
L.9
U
EIU
U
C E.n
p
-I UJ Im U
U ( Hp
U
E-OHH
pH
010101011
0000000000
H
:i
''Hi
IU101
0'
U
0 1
1 U0
1
10
U a U a4 U a4Ua: r,)a
U128121
H1UIUrIU
U
I U
0004000000
01010101
121 F4 1:
4
P
1010
IU
lU
1 F)4 11
410
E 'Cu
W,4
010101
F:
I UU
I
'H
I
I
W
,
I
1:It
010101010I
a:
U I
I
U
101
U
IU
H>H><>H>H>
SH
10101H4
Ia
UIU:IUIHIHI
0o000000000
1 0 E14
1
CEU
l Eo
.u(9 Eu,
Uc:ý
B 10
1II<i
l<uE <Euu- C
U00000U000
0'0
U U
mU8
UM
00
4
T40
c
=
WU
En
)mm-1 -HXX7.CUU - r-4
NN"
00%m
ml)
aO "4UE
.- l000 MX
VHmm-H-Hw
o-H
EJl)Ell
U U 4-3
EnM)L
CD C>
00000000000000000000
4
0
-H
CC
m> D
00000000000000000000
mm
m
NýC>1CD C)
.)
C) CD Cm
C
m m0mmmmmmmmmmm
N)
C)
OD 0>(
0o
W
0
w En
N N
00
Co Co
I-I4m--Y
INCl
L9 W
-H X
CD IN C) C )4-) 4-)-
~J0J
m-H
0000
o"
u-)
AAn
II
I
CN
0) CoN
N-r
u OMf
00
U '
00
W4.r
(14CN
UIUlUIUIUI
H
-Tll 'ITININI
)-,
00000000000
HU I U
Oc
E- <
UI
~ ~
Do
u
'U '
10
'0
0 I
I
u I2
E101
Hi0 l ui
H l
0AIUIUI•U
<I~
1
Hct
If
0101010101 0:
I
U-
Ol•l•l•l•l
E-4L
i U
I4
E,
.
F-
•
OI
I
- E,-iI
01 lu
H I>,,,F l(-)IU
,
U
01t,1010101
lU U'
O
I
E
Ex
U (o ro -i-
U
EU IU
I
f<
0302070?03
0101010
H
I
)
S010101
8101010l1l
ý
,c
.0
I
-IJUOI
10
0
10~~~
E- LnL)rirO E,
J
U
010101
19
I
0110101101
01E
H I,
U>U>U>U>0>
00
H
1l010
0
10
000000000000000
i.E
Uý
HIOIbUIUI
P-• E-•
0c,80000
0I
I0
H
I
00101010
- HOI'Oi
I
E E4
P
34EH141
P
P-
F -IH
-,,
UI I
OllIOi0l
UIU~Ibl0
I
I
- I
01
HI
H I 010100101
I ,1 i H I
H:Ii
H
U
0
0H,:: H:•
H>H>H>HiH>
,<u u'
<
UHIUIUIUIHI
I 01U1
I
u 01010141
P u0000
E
IE4
P
U
4
I •
•
UI
i
H
iI.
8
4
E
>I E,
_
I
Eý >
E'- >
p
H- I
Hc
ý> H->
H>U
010101101
H>
I-. iI< p . I<
01010101
u o P E '
I0U0
~ r
010101010
0 U
It,
HI I
iF EiE'
u01C1
u9 I
0
10
E0110101101
E,
HI
H
E,
UIUIUIUIUI
I
E.. >
I
010101014Hi
Blblblblbl
0010 'o'o'o'0
1.
E,ý
1
101014
0101
I
•: I I
Hi
H
I
010
I
I
H
UI
0000000000
01110181
H
U
0101010101
UI~1~IUIbI
00000000
00
0
H-
OUIUIIUIHI
010101
H
H0
0101010101
H~IUUiI>I
xt101
HHiHHIH
I
E-X
c0
0
0
uU
0
0
0
X~11I
< )w
U$:I
OHcUNC)I c UU CA
0 SI 0 0 0 0 1 0 0 0 1C0
1-)1
0
00
0
01
( >
: - z
1 1 l z 14-1z14E4>1
C>0 Cl
(1) wL9
01% "
i i w 4 - - 4N 1
'-4~~~~
>N oC E-'10
rOEýL4)~~~nr4~ rdo0. ~P Nm
11,c ~~
-ENC Iu
AU)x~ Hx
U4)
81) Hl 41 >ooa1,aý.:(3r
;H,-l .H .1(,
1
0I01
EýIU
I
0HC101
0 10
f101
10101
H 101010101
E-4
-
Ur-IrI-A A -1r4-10A q
T- - L
lT1rIý
A
A U- - Ulr
C.~U CoC CC)CC)C
CC>0C
I CDC>CC)(C><
U
koko ko%1 o Dýok 10%110k %0I'D I'kDko
r-4U
I lu
N N
'UC. Co
'GHHm
'U
i
4-O
00
IN I
Lf),-i 0t
a4 U)U
m U
00o0
N, Ný N> ND ND N, N
L9
01
0 .0-,'•
000o0o0
N N N
000000oo
N ND ND N N N> N
ND ND ND
toil
N N
r-4 .-l
'Mom
N0000
NH
-. -4
-1.MM4
A
123Hi4
- §O X
3
9.nn
8
V U
E En
Li
00
C)4.-
$4 t-
'ar&&LA
0
C'N
It
x x
w
Li
4
Lii
1
8
-4
00 -
0
0
000M0J
rU O E E~- L
4a
'V 4 ý-
4
-4 w
>
U
w
##M~mm
z5Uwwm
,
I" ,I~
I 1Li
b4 -4O 0
13 A a
m
1i Li'Li
Li1
Li
LL'4E-'tLi 44-r4
~ cl440
iEwUU
p , 40x Li0m
I
I
I
'Li
L Li
I
I I4
1>1>
4 >I I >000
I
1 U>1 >
14 1
X X. 9M
S aq
Um
>Ih.c 4
i~
U U 9
U 4Ji-i-i
V
4 -j
4=
'UM'U
i • 1i
i
iE-~i
i
ifU
IDID
&E-
1
11U
i
IU
I
-I I
Li
3I
IU "
I
- I4
U
E
I
EI
I
II
I
UtII
Li
t
) I E4I
U I
1 01
0 1
I
U
IU
Er-
I
I
E-'
Iu
'
E-'
I
i
I
IE
I
I
L
Iu
I
I
i
E-I
I
ru~c uZ
u
1
Li 1
E- U)
I ul)u
1
1-'4l
(i~
~E 1(LU
~~ ~zuEL ~1UC
i
IE,IUE
L
I I 1141
I
i L 1 81 i E.
~Li~Li~L
a 8
Li 51LL
8I
E-'IU 1 116
r
'l.
I E-
A4 u 34
I u
I1Icx IL
i i
I LI UE
E4I
IIU
E1 IciI
: A
4
1~
If
E
~
I
IL
4ua
LiE-
III
Li
b
U
-4I E-4 U I
H04
01-i
$3-.4
4l)4J
0
M-,4
Co
OM--
U''U'
dova
4
001
m IW
14
IU
i I
E-I
u I
Li
IN
•
L i
L
i
E-'
Ei i
UE
I UI
91 9 : E41
I
U I LUI U
4< I 4-IU 4 11I IU I
W
to) to
mN
m
-H-
>WIiI E
wO 0m 0> v-4 N' m~ v
,-I r- .i
r-IV-4ri
LA w0 rý -. I 4.-
IUiF
i
U
_G= i
iv
101010ý0:iolI
I i Ii
LiU
SE-4
E
Li
N -oi
U)i
E-i 1 Li1
-4
iU
E-U
1 E-4
I
i
E
u
'-
'iI
B
C11NC'4 C
W-
['
, , ice
iiu 4
-i
0I1u ý,
Z u9
f<4
01
<u
ý-,9
ý,
9Ur ý,
,9
>1
0
f10L)
10 ,E,,
0
1(9U
()c101
:1,:
,I
: 0I
r
~
,
I'I U4N'1 U
A
A
U14
U1LIU04
04Ua4
404
a4 Ua
0 1l I U (
4r
0 0 00O
13
4J4J -.4 -- 4 M~ M t- t( "4N r
-.
4 Co Co I' 14 (- tN Q £
Li LiU 10 9
LA v-4 w-4 OD OD >I >1.9
CU )AM
UU4J4 U iLi
r r
v-4 ('4m' qw LA 'o t-
i'
I
I
IM: F:, a:F•
:F, F,
Li
10E- 1 01
P BUf<
4 14 4
> E- >
I U I U I L) E4
I1 U t1 U
iU
--l ll u 0E P
0 0nL)
L
0E
E
U
ni El
EIUEn0
E-14
1ý :41 E- E- E-I K <
I E- I F 41>
0>1>E->
>
< 'IE4 'E I
10 1 'U
1 E'<I'
UOOOOOOOOO
HUH0WUE
E-U4UPUE
4 1UIU1
1414tUIUI1I
1
IC
iU
>E-
E-1iE-
U1 4 1E-1
Uu:Au
IUIUIU
4 1F
4 1
i
4 1U
F,
0I0i
UI I I
u I u uI
I UI U I Q I U I L)I
u
N
E4
I Li4
uI
UI
Li u U Iw-L
E-'
14 4-L
4u~L
LL)x-I0
E-010
41I
iI4IciI
E
IU
I
lU
U Li
I
Iiu
ILILLIiLIL'LiNi
IU
E-
13 u 04
io L) 13 U
4 'I
IuI U I u
I U IUL I u IL0 I
i
Ii
)
Li
LiI LI I'LI
Oc
l4t
co
f iI
LI
LiI u Of4~
I I
ti
El
-I i
IL
I i U IE
i u ) I EI ' E-- II
I ELLI'LIQILi
E-
LiU
E4I
E4-1
>I.E4
I
IE-
1f<o
E L41I
1 '
i01I1 iL-1
u
EU 'EU- >E-
>E-'~
'1'I
I
w m 0
r4 -.4 "'
-4 r-4 -4 v-4 r-4 ý4 v-4 v-4 r-4 v-.4 -- v-4 r-4 r-4 .-4 r-4 v-1 -f T-4 v-4
40 C> C> CO CO C> CO CO CO CO CO CO CO CO CO CD CO Co CD C>
cO DODO
oo DCODO
o
o DOODOODOCOO00OOD0
o
oo
o
o
'0'
EN- N
W
0ý-4H
C-N
CoW
L
-4
MOW
a4 11 M M X
U Li U
Li.
>1>X.t
U01-U 4J 4J U U 9
mm M ELALl. w
mOOc 0: m- m 4e0
c'JN14rINI
L
'0
aN
-.4 .- i T-4 T-i ,-4 V-4 T-4 .-4
~0%
0% 0%
al
ON ON ON
0m
-,A -1
'U
Oc o~
0
4 r-I rq
ON 01 ON ON ON 0h 0) Ch Om0% 0
UW
ON ,
.tv
w0000
t
V
C> C>
ti
.J
Al
o4
Colo co .-
m m00
0000-4
(>
v-I
ai
.-
E'u) Ln
rtI
t E-'
H4ELA
12 4) H HO
04 a
MU
1
0
-1 E4
10101051
141414
EEI-IM
181
E-4
E-
00 Di
I~~
01
1
1
1E04I E-
010I
E-I1
1
-W
md md -.-i -i
Id It
Z
IUE-4I P
IU
IU
S1414
411
EAU
I
441P4>
0010101
UI
IE- E-
010iF
E- lE-4
1
IU
0"IU
0
0 I1 0
E-4 U0101010101010'
U U) 001
1
1010141
I
r-I
010
10 1E
409 10 01 U00
4 w1 E409
I IU U I
I fU II
I)
001010101010101011I
U
101
101
010 01
O0
4-H
a
4-r-iCo oD5,5,>.z .q 0 0
rL)UJJ1 'o 10
(nEn
x
101
101
101
C0
U1 '
101
EIOI
0
0 0
1I
U0
IE
Ix
I U
i0
E101
101
E-E-401
10
UIEU
U 0I I U
010I110101010lI
101
U I
101
E- I
1> >
UE-.
E-4
O E-4
E4001010101010
O-E>
>1010101
101
101
0i
z
1
Oh'>-IIU'101 E'U I U
ý0IW0' '
01UU010
z
z
I L4 I
Z,4
14tb40
0t-i i1
lt
U
IE-'
1
1010
I
I
10
>
44g
4E-E I
1
10A
U
l4
~
I
U IU
I
I
IEE-'
0
010
IU
IU
101
I
1 0
I10
iE1U0
E-4
Ii
IB
S
10
I
I W
0101
E-I
1
0
0W
U
101014114
I8'8I
0 E-4
0
IU
1
1L0
0I8I8l8I0I
E- E4
04Q
I
0 w
U4
UI
U
I
UC
l0I0I1l1l00I
OiUI
H
U0 IE-4 IE-
U
U
U
I
101
40H
l~l~l
U
tU
>1
1011
E >
0101010101a101
IIU I
E-4
U rlIu
I
8
ol
ol
I
p
a
1
4
IU
1
14
1
8
41 U 6 1
I
SU
101
101
101
U 6i8E-86
484 1
1 E-- 1 ý1 1
1,
U 14
ILr 0011
UEE-UE 0 UE-4
0
ý ý:
101
101
U IU U I~
U
1
10101
II
U
101
E-4
l1
0l101C
01
O
E-' E-
E-' -E'
SU
14
0
01
0
0
o
i00,
0101011101010
HP
Im
C~
-41
a4
101
UI
101
Ia
~
10
S10101
11~
10
ID
10101
00.
101
101
W,
101
10
aU
1 pU
0 HUlHL)l
1010
1i
E-4
101
I
O
1
II-I
IO~
Z
E2
E
I
I
I
I
I
-1300
I FW14II r 1N I I N
0.1U
1
20.02
4
M
E-1I
I
I
DB
1 -- E
P'
WE- W E-4 [IEHE-'I
0110101010101
>U
- I U- >E-4
>I F-'
-4
> IE01010101010
E- -U
I C u- I
101101
001010101010
IF'
U10 U1010
U
U1010 UPP
101
M
E- U
I
'-
> > >U4 > 24 4 > 0
I-) I
I
r-
Id
4J
I
T1rI
UV
4)4
1r-14
a t)
40C Da41 "U-4
U I
E, 1
4 14
q
M
1
ýr
BL) U
I 0
0
4
E-
D
U
129
I
U
E-L, ý I4I ý
rh 0
IL4E,,
E4
uI0E'
I
1I
1
1i
400C
FO
0-
r
0U
101
-r-i
-v-Ix
04 Id
12O>
I
0
Z
E•H
LA LA ,-4,oi wO w
MlE94
0
m- wN r-
3:
0
a4
-. H•
10 -vi
War
T-i IN
m~ Id LA k
r-0
N CO aI
r-4 ý-4 r- r-ir-4-4
-4
v-Iv-ir-4v- v4ir4iv-ivI vi T-4vi v-Iv-iv-Iv-Ir-4 v-4 v-ir-4v-4
0000000000000000000
00000000000000000000
r- v-I
r-4 vivv- v-I
v4Iv v-4v-iv-Iv-Iv-iT-4 v- v-4 v-iv-ir-iý
Ef E 0
E14El. En
4a4 C4Or44 m
- wOO 0)
md
0 W-4
(In -W L%0
wd mA 4DNC
r4 r4Iw-4v-4v- v-4 v- i v-4 vi I4
0
IN
SCmO
LA LA v-IC>
) C
u u41
a4a4
En
-
0
T-i
v-i
x q -s'
.4
N N
J-l0
U
4
4
UZ
4
t
t
m m
i IN m -wiLA w. N- CO m 0
v-I ,4
r-i
v-i vi v-i v-i
r-i
IN
ý
r-4 v-i
v-4
v-i
r-4
vi
v-Ivi v-4 vi v q-1v-iv4 vi v-Iv-4v-iv -I
r-4 r4 r4 -4 r-4
vi T-4 v-i V4 vi r4 vi -4 vi vi v-I vi v-ivi vi
n
X
-H
00-<-H
.LJ
4.J-,A
nU) ý
o
oUUMJ
H
IHUH
x x
IP
1(
C- I0101 UI101 rl
(9(IUIUIUI(9l(9I(
10
I U I CI(9I(9
1E
I
I ui
1(
I
o
o
-
F-.
>1£
C9( 1919(
uaw-H
m m
0
0 0
a
0- 01OD00
MN
(.H
a:E.
mcoo
H
(TIU1m
i
U)i1> U)H.1 1 U
u NU
NQ
z
m
OIOIE-I8IIOilIBKIOiK
EtE-
P
I
i
i
EB
i E-4
(9
HIHIE-
U4
P(9
(9
I
Uipi
101
0I0
U E191E
E0
(9
14
10
i E
I
1
0
1
0000000000
UlglbibrBI
Ip I p
0
01010101010
I H149IH
H
..
H> H> E- H
HIHIHi(9I
I
I
0I
10101
H
U
IP
1(9101
~
IQ
IHIHI
1
U
I-
pI
ElHi
1,E
0404040404
U
u
u I U
0000000000
H0'010101101
'
'E- iE-
I
C1
Q.1
UBUPUBUBUB
(9191(9191(91
E-I
EI(9
(9 I I
-
I1
(9
D
U
0IUI•IUIBI
E-
I
-IHIHI
UIUIUIUIBI
(E(E~I{
>1
UH(H0
04 1
149(H(E.9E.
1
14
I
1
i
l
I(
1
jI
1
H>QQQ0UP
> I 910101
> P>
( I(9I(
H 1(91(
0100010'0
O
0
I
1
1
1
14
18
UIUIUIUiPi
(91(91C40 wE-0
U
0
U
MU
10O
01911011(91(91(91(91
4 l4
CA~~~o(
8
m
4
U
uE-.L
(
H II
HIHIUIHIHI
E1
P~
101 L4 E
01010101
Ul~lOlolII
000000000000000
UO:
(11)
F
(D
>
'(91ol
1
1
1(
E-.IHPI>E-4
4
I~
Eý
Eý iO
(91(1
-1(914
i
H
1(90
1(
I
'H
(9HU9H9H(9lUI
(91
1
1 0 KS
:0
F : (F:~
KS 0:S(10 1((91
:21(919F
HP>#<Q
KQ(18 1
1
:
1 1 1 C)41
10 (() 1 1) :9.
(9
9 91(91(91(
1(
10
1(91(91
I
it IU
I Pl If<
I<
I
I
101
ot
E41
:4.J
U
IU
14 1
OtC FtC
E- I #CtI (-)1
1
H 1(1U
H
xE U
b
'10 10
10
1
0;
8313
11"k p
E-
0 1
i
E-
iF-4
iE-
(91
0101
1010
10101
4141
00000000
'OE-.
1010
0
0
0000
1
I 1I4 I
Esý
Oi IU IU
O4 14 1 I
AR
100
00 00000000000000
U1 E-1 E-,I>
91 Iý : ~ -IUU EUU UU- U EU
4
14
1
14~~~1041
1
lH
Ol4l0l0l4l
8 u
U
U
Eý 1,
BU
00~
1 -)m
I E .'F-I
(9 (91
t
I
1010
101
1(91(9
lolw
C
,ocW1 <W
1 uU
U
I up
f,I E,
P f3I U
p 14E
KS
10
010 mo10
U
0
KSIK
'<
oi'<0 o:
114 10<r14
4 1w 0<4
I
U
Iý
s
E,> E->ulu
i
UP
O0o
ES
B
10
Ii
U
I8
E-4
1(
I
b
U I1
OlGlOlolol
H-1
1
1
,
I
I
2 4
19 14
U
u
0 E< KEU2
KUS
W
4
0 0
(N ( (,4
N
N
*H-H
X
4J 4J r-4
C)
,H (N m
I'm
m.
r-4 r-I .- I rH
oe .,H
0(00
12O1.
a4
4-1
U)>.
0C1N
N
N-
>1
>1.
U U
UD
LiU)
U to
049I
0-1:~
Ct)
010
rH -4
.H
r-I .H .H .-4 .H
00000000000000000000
C14 <.l
m
I'D N
r- COz01
m.
NH
(N
.HI .4
1H 1-1-4 -A ,-H rH 1H (N
-4 ,HI r-
rA
H
v-A r.HI
C) C)0
1
mH--XX
•'0.-H.H
C)
0
(NN.1.
U) U)- I-ix
-N;T
1-4N'
N
) CD C-01)
Ci
C >.1A>.1 -H (9 (9r
r r-l() H -
0H(
) a
"D
N 0.1 01
'cH-1 -.
(N
000
in
,-I
U0 E)-4F4 in
13.4L.En
tul
U-,.0
00
o0
0000000000000000
.H
H
l
-4
)
I 0
,H
(N
l U)
r- .r-
N
cooooooooo.o-oooooo
ZH XH
LHLH.H
14 MH MH
mý ml mv flm
.H .H.Hr-H -4
m 1 m m
.H .H
r-4
.
MH MH 4
aH a4
N101
$.H <H XH XH,
v-A.
(N (N (Nq
(N (N (N (N (N (N (N (N N ( (N ( (N (N (N (N (N
.- .-1-4 -. 4 . l -A
. ,-1 -4 r-.4 .
.--4 4 -1 -4 ,r- A . 4 H -4 .- 1
M m m mn m m m (ý ml m ml m
.-4 .H r-H ,-1 .HA.H -1 .HH
rA r-4 -1-
co
00
-V -:
'0
'0
mm
000300H
.-4E-4
mON
-14
W r-i "-4
r coOD
E-i E-4LA
o 0
X~w
1
CooC>
4O.NNGJW
w]
ý4NrNr
N
i>,w)
>i.CIi £.
( U U Q
ra M -N *H
E
U
U
U -
0
AE
"
A
Pl)
E
w
E-. i
Fu
UEUEUEUE
E
E40 0
U
0
U
0
E0
U
F0
-<
0
0
[3
0
uU
U
u
UC
U
E
E-'
EU
E-
O l 1010:1
O
E-4
'
41
0
u
F-
E-
U
SU
E-~ f
U
U E&
1 011
E-~
0
000
c
09
0t9F
U
u
eC uU0
0
~E00
F3F-UF-F<
80
I4
Q
t.. ,..,N
'cl, C
-T -4
4- _1
J -4x -1x 0'O\N
N '1 El) E)w
C 0mC) -EH
00
)
El)ý4
N4 r-N.0 N N
C14CN ,- A En
-i r-i
Zw>>
U U MI Ea H -H
n4IQ0-CU
WW
U
U.J UU
C
liEl
N
(01%
Cl
rO ro
00
(AEll
N N
900 -4--
0
Hm 0
OCU
E
,-4
EU
4
U
U E-.
>E-
I1
E-L1
D
-4
U
U
U
EU
[.
EU
'-
00
E
U
0
E-
0
U
U
E 0 0
Q<
0 9 -'
uFU
0
00
00
F-
U
<
0
9U
U
<
L9
UE
-
0
F
IU
I
m m
-4 H -) s
-4 H4-a)
- 1 -T
> C C)C.
~
-I'r-r
0)
0
8
u
rlEU
U-
EU4
U
EU
,
u9
0)aH
U
U
EUU
E-
UF-U,,U,,U,,UFEl
C a4
4 C1
0
U
0 0:4
1•0
• O
C)$ý9O
0
Of
u<<< U< WUo
0 EU
Er
89
E0E4E
010E-E3
8
0
E U
E
E
U
PIE0IýE-4
OEUEFIOEOEG
OEOjE{MO
pU
14
IU
01010l
14PI [1
EU
IEU4
E
- l E
U-IUOUOEOE
EU
gF-- F-
cy8
EU
00EU
c1
EU
EU
FF-'
01EUE10E
-'
018
EUuE
~
U-
0
E-4E
EU
I
F(-)iE-E f<
EU
U
pER
F-I
U
U-
U
U
~ 000
IU1)UI1U
0
E,08
01
E
U1) IF
4
CU I
UE-- H
EF-
10
-
U9
:
0
r-1ri
4r- q - rIr1ý 4 14ý
r ýATA
4
1
(D
)CCoCC>C
DC)CC>C
)CCDDC.CC)C
U
U)UEL
-. 1 r'IV0
I~ Uh
.'rI
Z.ý
0
0Vq
0 ;0
0TE--4
r-riv4 - -I rf - i I ý
I El iri 0 ýl 0zE-i
el -
~
u EU C
UU
4
N q
0N N
U U
EUF 44 U- -14 ý EU
rOE-Lnm
OE
v ýioo- 1o C>
>,i.aur)mn
HU 4E-U
XlfMM3
~
~CWWa
i
La4a
I~
i
E U E U E-uo
r-4~8 n% -m m >
n-Ti" m-,n%,
0 -1 - -10 --mci>
r-4P
r- lr
~ Ar48
b
U
o
E-> 0
Oc )0: U :E
4
EU
iEID :FI U
U U) U9U
O
L4 -E 2 (D L4EE-I EU
EU
EU
-c N
-4
03-
F[3
C <
S
C 0
E•
[9
UE
U U L9
8 ,< 0EE0 0 UO ,<
,<
E-E
u
uU
EEEO -4© UE
UE
UE
EE
O•
8
L9 i
9
I[9
U
xC
-
EU'
I0
UE0UE- EU1010
-
F
F
U
aEU
<
1:gcl<12
S(
U
U
c
EUEU
EU0914
F
0000000000
I
OE>EOEOE0>
00000000
EHX-4
i
004
0000
'0' F-LALA~-4,E
(
E
IC)
I
0U U
~ ~
~~~
U
l--4LL
*HHEO'NE-4
U
JJ
UU 9
~UE->
ED
El)
Eli-H-
1-)iEl
N
ED>'
aX
N
H
En
DoJ -H
rsr E-4 E- Ur4
E/)-4
Ou
'0't--A
El) l)
H-4Q
U UsU)U04
r-$ (
(A 0
0 -H
04 4-
CE) CE]
m ýr LA 'XE
-44444----
N- M (h
0D -4
-4
"-
M
-A -I
,T
LA
r-4-4-
%,EN- M) M
-4 -I
-I
0D
-A
LA LA LA LA LA LA LA LA LA LA) LA LA) LA LA LA LA LA LA LA LA
-4 -4r-4
-4-4
-4 -4 -4 -4-I-4 -4 -4 -i
r-4 -4 -4 -I 4--4
jO
A-iA"
o
90
-1:1 -IT 0)
q(
00
00
oo
H
I-JO
NN
w) > > .-H .-H
0 H .-H
(O
u
"N C'
U n - H
"
0"UA CA
4-4 4-4 (A U)
rz 0 OH
.,-I I-i
O4
o
0
C0
mmNX
m
0000.-H
"e a wo m
o
44 4M M
OOHHU)~mI-JI-UU
m
oo
H00
m>>-•.
o 04
304aUU
00- N N
I-i0
U)
>w0- > H-HA
0 -H
(uA
'
C-i
0
0
" 0
u10
Hý
I u
U
'U
1
14
'U
'
E
I
4HHu U
I l 1
I E-'
'H-
I
iI
14
1
1
'E
I
UIUIUIHIt)'
H 'E-.
UIUIUIHIUI
'H E
EEE_
E- I rH
E-H
H
I U..
H'IH'IHU
U
-'
U-,I
UIUICDIIHlI
UHUHUHU1U1
. Eu _1
u
I E' I -- I
EI CU0 U)
I En ) En
'
I U
IU
UI.
o*
0c0
C>U)
P-
E-I i
E-i
E-i
E-I i
-i
I4 ,F
F- ©
(:
,E:.I H
F rH
P
- E.I (,: 4.I[.)
H E<
U-.
'0
U
U
FCC
I
Oct
r) F + IE-1
U
U IE
U..E- C)
C-o<
iý U1<0
o+ ' o
',c ',
o-[41,
E-14 14
4ý
z',
FE4
~;,,)
o
EN
'i<
U
u
HE
p-'
EH
C
U
'1
i U'
o0
'C
'H
H
H,-.I U)
o)
0 SO
-1U--U)%
,ii
o X
uI
CD 'U
*H
-H 0.01,...
-H
>0
H
It EUE
p
00
-(4- H A0
m It
A-i 0
H~
~c I
i-i
P
t
E-+ I •. I E-+ i
I U-+
FG,
UI
E
1
H-4UIUEI
14 l4'U 134UP D
0
U I
U
UIU10
'C' 0 ' I~iU
MA4
4-i
PIUIU)
EBH
::1: I
uU)
CD
'U
PI
H 'CD
'H uP
CDU
>p'H UH E-4
H
'HE-
0
Q
10
0
1
1O
BlUIUII~I
ON
'
UIUIU
0
U
N
144U
I1
1
1
1
H >C)C/U
'H'E- >
~
ig 'I
IC I
1
t
0
C
91
8 P8 ACD ' 0
uH
.0
-H Ord
C),r,
rU
I
'i
UJ i'Ui
Eý
CDU
E.,I
'U• UI'HJ
E'
I
F U H•F I
U
FUF- Ei -iE,
E
F11
E,
P > -1> > E >
ý 1-1) 0) :
: E,
U<'11
'U .. 'H H..U
1
=U
xt
<UGf<=
H
r..
'H
E-8:81:IF
: '
uH>HPHPH>>>
: H:.
u 'U
: u :u
U u U f8t
1
44I
C100i
'i
ui Hý
E-
44W
1
i
F0 ,
F
U u
'U
0
'H
'U
'H
'F
H'H'H'H'H'
H
'H'
' U '
'•
UH>UHUHUHUH
U'U'HIHIH'
U 'U
'-U E-HU E
UHUHUHUHUH
E-,
F... FP.->-- P-- ->U'UIUIU'UI
U
H
H- 'H
'H
HHHHHHHHHH
EEU-H'HUsIHU'U
EH
14 F
-1
04E'H F
14E
U'H'UIHIH'
EF
F
E 1<
u CIUU4U
u 1u
FCC
UI8'
cYU H
UHUHW44W
4 '-co
OO
(1U t-.I
I
' EE-'H
11
EUO
H'H'CDO
u
0(1A
-
EE-• 'E
E-9 E-+
EEE -'P i
I I
I It) . II
' E-4
'H t-4 E-'::
UHUHUIUIUH
E-,
-.
uE " P-.
i E-,-i
iFCi
i
'U (.-E-,
< bU2U2UEU(U(
I -U
U IUE
Et
i
1
U 1
U
u
I
u
-)14U
u
F
u ýE-
u
q'i-
U
aU r-)U2U12
: U :
;
H
'u
E-1
EH
'H
0
04
4-10.
, 0
So
000
E
41
v 1..r:A
'(a
4i
0 'C
Cl
oB
0
m
E4- O
'0
z
0a
-O
H
U0
4J >
0 \D
-HAU
a) 4
C)-
HH
H m
f
C)
I
Xx
-H
C>
ý4 ý
A-i AIN
a4
00rcsoUa
S1H
HC
A En
Ar-4 N- Ci
r--tý4
oo
,-E
CNI
-
000)
OHHQ
0
I-U
Lo -H
m_
NoN
00
0>
IDJ
0
(.,4
C-N
C ý mmXX-i
DC- -q-C>
N NI
ý
oo
0000- ooH
c
O
SOMNN
0
-H-
OOHHQ.Q.U)mI-I-UUU
- N1 C
0
X
R
:
"0
0H
0 ..
0
CI
O ,-
HD
00)
D
>0
N0
00000000000000000000
HHH
H H
H H
C
HH
)
H0 C)
0Z
0
H
H
H
C)
N
H
C
C00
H
00
A.;r -;I.
0ý4
(N N
m-XX
CC
0
0H
C-H
O""r-4C4
C
-:
>
-H.,-H
H-H-Iwx2
C fu
m
UU
r A 4 0 124
a
0NC
00
OOxxc
mmXX00UU~
D
)0000
q)
-I
r
-J
WWNN
mo ýCý
U U
u En
jU) > > -H- -,-i
li J u)UU 0 0
-H -H 1a 0)
CýC)
0I
-.
UU
)UCH
" 4- M O(
w MlHH
Xi
ZHH
4-
04
C)C
'0'
,
00
-4-4
0
E-oF
H
"n in r-4 r-i
Q,) 0i
") (n U
En
N
4:
U
i4
F4
m
F > E4> meHHQ
F> F>
<1<14:4414:1
U U IU
4:4:1414:14:1
'H
H
'HHU
(H'410
U iU
Ui(iUi4:
iU
4:14:141
U'
E-4ý0E-4
E 111E
U
C40
U
:14
i
F
ti
HHH>
-4> F>
E4>
11
0141
:14:
F
(
1: 0:'
14 K
4
0
1
H
i-
FI
U
4 00
Fw
(U)
S1
H, U
r1 FH U'Hi
Fb
4:
'H
'UI
'H
I1 <
0I
0
4:'(
'4
01
~
(5)
()i
-
E-4
P(5'(P'(5E-4
o<
I1
HIH
l4:iH
HP ý'HPLý
14 4 H
8~
'HPEý
14 U
I
I
KC
0 I
PUC'PU P U P-
PU
UUiUIUIU'U'U'U'UIU'UI
U
I
E(5
I E<5 I4
O
I(
1)
1
H iH
U H
14
>mouU> (_
Hc ICiH<ib
HiU
01
0I)
i
14
H
U
ýý
0000H
C
IU
1 fI:l I
P U
H
I
E-4(DI
E
I9
u94'Hu
9B
'U4
14E
U
0
iE14PI 0H
14P4aH141la
U
I
u
UH
C:
PH
'HH
(5ý
l(
I uI 'HF L3
'(5 H'
p
11410 FUU H
UiU:
IUiUI 'H 0
H 'EH 'H
H
UUUUsU
~H
'U'U
IFIUI
FI
H Hi0ig:
( C E-4
181
IU
000
E
(
'(5i5l(5
UHUi
1 EE>lU
U
u
U
U
(u(F(
U
1P
5P
0
HI2
IH'
'44
U4:U4:U4:U4:U
(5'
(5)
i
u
>
'H'C
C
'(5'ro
l
Fi i~ E, 4 EU'P
UIUIHIHIUIUIUIUiUIUI
E-4
H 'H 'U
ELL.P1
HIHIU'H'UIUIUIUIHI -E14
U
U
i'U
14Ff4
H
P
4'HP
HI
'H
14E
'
:f :IUI
UI:C
H>)UHUHUHUH
H
'UI
(5io
UFUIU'UIH'U
HIHI
5 '(5'H'U'H
(0)i
14
PUIUUUI
(54:
'4:
1
(5
I(<
4: 4
u(5
U
U4
H
P 000000000000000000
Fu
0
CEN
x0
40-0
0
O0-H-H000000W
UIUI-ibmUIm
Ui
b ib i l~
u
lu
l
0
C)X4
C)
n
xC)c)
,!
x$:4
c
Fil
C4:Kt:4:4
r-I -1
(D C) ) ) - - (D C: 0
ý
-V
0)ON
r4
A -Aý4 4 - r
AJAJW
0'H
mU
0B
0
ý4 (1
W N
ror>PF >nP>r> uu w> > ., a " ri
m =wrlj-) uu 00
--I a r
4-44-0 M M 3::'12W '
4 a Z
O- n -z n;o -wm
c>-i"
)
:rLnkDt-oD
r-4i-dl
r- r4 4 4 - - ý a- cC
4
ý4rI
ý
l
UIr4rI
r
12
r
ýU
iU
IU
iý
UI
1-IX
4ý
Aý
1
I4rfý
<D(D C>C) C>P
(DCDC
DC D< DC C DC >C
m r) r l ) r) n l)rl)(1 m m
m i)m l)m
E--4-4E-44E
0
e
00
TH
J
UZ -
0000
rO rO F FH
'0VHH)U)-w
W) 4a)
F4
H
~ ~ ()r
~ C)~
mmxxoo
C>
m
Q4
-H -H 0
4
4Cn
wUU
En (A X X
4J U U
A
ra
XEE
Xm X444EM R Z
-i
004 -A
O
~
"O m~ -,z n
'--Il0 0
a)C
D
0
)
W >
)NN
> -H -H
0 -H -H 0) '
4a4a
) N- mOCO 0D - " in qu)
%o N- CO CO%
0
-4 -4 q-4-4-4
-4 -4 -q T-4 rACO
-4 -4 i-1 -4-4-I
-1-4
q-i 4-4 -1 -4 -4 -4 -I -i -i -i -4
CD CD C> CO CO CO CO CO CO CO CO CO CO CO CO CO CO Co CO CO
.o.
o
o
o
oo
o
o
o
o
0
0 000000000000000
P 4 P
-4
E44444----H -4
P-i
-4-
4C
P
H E-4 H
1-1OO4
'
~
mmXX0oUU
00
J0 10
C> C>CW> >,-I -41
0000
-4 F-A
11 H1-4
E-F -
4
>-
p>
tO
tO tO
0-4 tO0101010
O
1 <tI<O
0
-L J1 u
"- ". m. m) x) x
E-4
>p
1
I
10
1
N C-
0
e
>4 $4 (1) - )
0C)
00
0 -H- -H4Im
11
r-H -4
-IT NP m 0)
'O 10 F- F-rn
tt
I
I ,
I
tdi
O
40
~
101010
~
I
1
F
a, u
rF
c
0II1
I
0
pI
>4 >
4
>F>>
u0
0
>
>
FEF IIF
UU
0
s
8 0
>4>
U
BUO
O
E- I
F, > E,> E4>
E-U E-u
1 #0
: ý : r-) I0
< I1 4
u
EH8
E->
>HE-H -HE4
u8 U
I u
>
U) E-4
I u - II u E- 10 UIF,IU -UF- I
IiUU101
Ii UE 1F-i >1 E-i I EE-4 Ii
I
U U iFE-4
U to
U U,
Et
E
Er, ,<ý
10
E IF-IF-I
<
t 0 toýU 1 101w01C401WC
tut
wu
IFtO
IF- - 101
IF-r)P
tlý
x4
I <
I
14:
t
itýi
00000000000000000
I
l
Ip
>
>>E
Ix
8
>
u IL U8F L E-UE
Pc
1 : p I o41 :1 < I
L)
0
4
--:r ID' m0)0
"
"
mmXXooUU
=-H -
('-C'
)
00
NN-. Li
WJJ))N
N
,'El
,o F-4 F- LA LA r-i'-4 C) co c0.
0 w)in) > > -H
'0'0
ItI)
0.0E
M x) X) LJ 4-1 0 0- 0 0 H"- a0.
0
4-4 - WDLNH
3:4"1114ZZ
O
0 -A
QD
U)A-)
H
OH
COO
'I
"N C) qv' LA %DtN- m m)
010
,-Ir-
(n
"H
r -I
.
I
IF-
I
1
Eý> Eý> Eý E
-4 E-
H
00000
- -4'-tL
I
FC4 L,
4-F4-I<--~
-F-I
IF>-F>
4,4
,L, '- i EL,'
01U
a4
u(9 U
iI U
E4
iI PuU P)
:F- EO
(9t4t:t
0
0U
(90
UItU(9
0
U)
0-) iI 0
EI E,U
UU
U
UE
1 14
Q O4 J1 r) 4
Fý
U ( U E
818U
10 EU0U<UUUU
4
E4
0
E
UmUGUGU
EO,
0
FC 14
tutu1E
414
1
E-
IE
18
tO
UIF IE
10
tO01 IF-<
IF
01
01
~
UiFI Ui
i0
> E, >E, > E
0A000
Coooo
IýUEU
E-4F-
F-i
NN
OO
- '
I
F-
E-
4U
I
OF-tF-0tOItot0IF-IF-l01
ED
F-IF-I
>
ooc4
I
iE-
tO I
1
1
F-
KC14
tF-
to
tF1-1
ol
0
>
.LIO<F
>F
t, ,- t-
1
1
u
0
101
IF-Ft
U.F
E-
E
I
tu Fý> - u
P B
F 41
)
E
1
I u
uI
F>
g , -'>F1 E :I
IuI
I
0
~
L) ; j 1 r) I~U
IF
0 f< C
1 411t FIF-t
E,
F- to F-U Ft-IUIOL
I FIFto
F -0 10101w0
U~E
14
0
u
o s
F- IF UtF- Q14
IF ctQ
u
L IF uF- 14
F tI I c~o I
tU
>>
z
tc
101
tutu
u
u u0
>
>
'F>F->t
tiib
Iu
to -
I'E- tot
tF
E-t
04
t uI Uto1to
I <F-I
a4Il ut
U)
a,I
tO
I<
I
I EF.
U FCFuEF-I
u
- SUI U
F10
to O
tF-
o
1
0 F-UM0wF-IF-w
0
to
010>- E F>1
u
0
ct FC u
Oc
0101
) ' t (j
E4
1
Et
1
I
iI~
t~tI
C)
14U
I
l<
OI--U
u
E
0 0
IIFtF1010
I
4) 4t)>>
u
U
~ ~ ~~ ~ ~
00r
>
p
iFtotutu1
tSF--Et
u)
t_ U14E110
a,I u) a4 t uut01aUuu
oEý
u~~~
()IL) I
-)Iu
-)
00 u 00
u
0 0 1-. I4
UUO
I
,cc
INC)
00
00
Kv LA %.o r- wX 010
-A ,-4 A -1 -4,-4
C"
ýi4 .-4 ýl-I-H
.r-I
i .-ý-4 I r4-I .H .-4 ýq r-I v-1 ,-I .- I .-i r-4 .H
00000000000000000000
w
I' D'w
D'ID I' D'ID I' D'ID I' D'
D'IDw ID
41E-'>
000000HOJH
N
10 r-4
L
L
L
-i r
L
L
~I
0U
I
E
0
OUttE
OO
r14 (N
x Co
00 o
C m
C) m .x .1
0
C')
00
-H
000.-H
-4
N
N
LA) LA) 1-4 "N 4H -1 14
,-4t2X0000Ja11.
U)9Zmm
= 4-1 4-3
44 4-4 U)
00000000000000000000
LAL L L L L L L
IF-ý
F-
1
IF19
0
0
Oi w
1) N N
w) > > -H1 -H0 -H-H-- 4 MM
ID LA '-1) N(0
0
0
L
L
L
L
L
L
L
00 r40co
-W-O0%
N)"
v-
rO'0EE-4" Un
LnaH
EnVEn
U
S10
1
,C
1,0C==U
A-)U U CO CO >
U0
01
I
U
I
E
FCI
4
oq
~~~~
I
I U
5E-1Fi
It
U
I
IL
I 1I
a
14I
10 t0
i
1
l
04 u
U~
~
4I
U
E-
I~
L)
E-I
1
a4 u
a4 u
ua
Iu
11II
L)I
I I
U
I
10 I
14
uI u I
owl U.-410
1
IBu
b10b 1
3U CL
I
0 2ItoI0 1 0 E- 10
I 1U IC. I U I U I U I U I
I U I1u
L
I1
I
I
~ 4 ~14 c 1U
U
4Ut 1Ut
~ ~
E- EI
oo
14 "
I
mm X o o
NN
C'4
OO4
C4 a~()
En M) X 4J 4J
-OtNi4C
)
C- -
w844
0H
4J
0N
N
Q2l
~ C ~ m~ UUr14NtC1
2.0e00~~
0J
vv If aiý1-14C>
m
H H
1
E-H0U E-U E-
14
L) E- 1i oi
0000 -11 *'1 In In
n"
1-1V-17-1
0
0-,.-
I
10'
U10 0 U0'
U 3 E-
0)0
0
I
E-,0
-I
0I
l l lb II I1
10
4 1
W0)
0)I-4
U
IU1IUI
ru EI
U 1
U0U
IIOCI I IL
I IIU1 I 1 I1 1
4J
5-.4
0
I
4101f< udI u
6
0
E- I 0 u-10IU .10
I0
4
It
ua
I-I
E-,1U totI
1
E0
14
IU I u I
4- H
l
-O
-4 1-4 T-4 H
NC
14-4
10
w
0000CC
C'4"
xoo
C -HI-,4 00
C
l- 1.4orr-
x
HD ww
'0 E-4E-' mn n r -i
WHH
M XM4
aiommoo
M444uMuM
L
M- %D4r-
W CD
O0
M.
-i ('
-4 T-4-4 -4 -4 -4 -4-4 -4 r-4V-4V-4r-4 -4 -4 V-4 -4 -4 r-4
000000000)C>C)C>ClCD000000C>C))00000o
C
w- wO%D w. '0 w. %D0w. %D%0 %-0%0 %D0%P0%D%D0w. %D %D
w
.4
41 4
EII
1-
1
U•:
Ii E
I
I
1t
IP IB
UIl
L- U
I II I-
UI I
, E-4
4 41 E8t8H8E-48H0.
10 4 1 4 U1
8
ýx
1 E-4 11 IL
E4 I
'U U U I, I
I
1810
=
0
II
I
I E- i E- E
E-t
4
04
#
in
U0
I
U0
10 0
uIt
U
O I
,
410 U0
tU
E0
1 114
uB;u
IL)
E-
I U
E--•u 1 0
Ei E-l
E
u
I
'U
1I
U IU
oc
Mx
w
00
0
)
)NN
U U L1 Lq > > ._ .,q
UU
0 0 -j-H4 (a fa
4W1
C
qr-4 w wD ._
L
aa
C
I
0I
I 1 1 E4g
0
I0 L-4
UI~i~i-~i
IU I
I
I
ua4
I
S0H
C
I
1 14rU
IL I4U0 UI11 2
'ti
> -4 -4
In
-4
E-4 E-' L
H
BI
Irlu !MMI
I I
~~I
0-
0 -4-
01
I
U
o JCo
0 0
m
N CV
cm
X coo
c 0:t•.•..-,
0 0 >4 W
J4J 0.)
N N
m
0
-0):t W
>)N
)-J)-UU
-4 N' m' qz in %D
w mD 4o~
r-4 -4 r-4-4 T-4-4 -4 -4 -4 -4 "-
-4 r-4 -4 -4 r-4-4 -4 -4 -I -I -4 r-4 -4 r-4-4 -4 -4 v-4 -I v-4
000000000000000
C Cl00
0
o oN - rN ro
N
N tN N N N N N N N
r-
mmXX00UU
~0N'-00
H 0000
00
'00 E-4E-~Ll um
1)WH H
"~ am
-i -i CD cc) .. c:
m) x) x 41 -1
M
GHHc
0N'00
00
r-I1-4
0C>ooo
C>
--
-1
U U w] Ln > > -Hi-H
UU 0 0 -H--H )u r
M
4
XX00oUU~
C)C>C>(D-AH-H0C0C)0
NN
''0E-ý E- L Ln-fl
b 4444M
00)H Xo
4m
am
0
ý
() ()
JMM
JWWGNN
A- oo oo c.0.0 U U w) u) > > -H--H
m~ x) 3
x) 4-ýaýa
m£
41 u uU
0) 0)j
:1 0 0 -H -HU
>
~
~
'U
E'U
I C
E-'
iEU
IU
'_9<
'ci
1U
I E4
'U
E-
14E
U
'U FCC
E'lCt7ý
o<
1-
U
10
14
U
:8
'C,
'c
10
U_9
C)
CU
pc
'c
'F--
C)
'Uab(9
U
'I 'U0
'U'U
'U
'U
'U
'I 'U
'U CU
'U
'U
'U
rE-'
'U
I
'c
'i
'C,
'I
'U
'ci
'U
'ci
'U
Eý
'c
'U
'(9
'U
'U
1-'
'U
8
U0
'U
lU
IU
'U
'U
'a:
'ci
'a:
'F-'F-
)>.
U41
00
-4
0)-)al
-VH
)
mm~Xo0Ouu
(Co. 0
000C-H0C
0
- ý4r
("O--qr4
OCCU()E
NNH
-1N-1O
LnN414C41J
4-4
:r 04 a'UU
0U4
0 0
0
ý4 4
a)J a) N N
>-a-4
C~oC)
9'0-o
NC'-
E0
-' F-' Lu Ll
H
.- 4 .- 1 M
a)
H a) U3
n
M~ w']m
Lfl ý) r-H 4r-
0-O0
0-H
0-40-40-4T-A4--4-40
ODO000
D0
D00
00
CDC
DC
DC
DC
00
oDC
-4
H -4
-4 -4-4
r- 4
,H
r-I
r-4ý4r--44-4 0-4r-4-4
O00
0 0 DCD
oo
DC
0 oo
0 CD
0D 00CDoCoC
00OOD00
CD
0 0
(cq
n r
c)(D
u u>ýý
as0000H-00000-4-~00
CD .0.0
U U
J u)uU
x~ x.
WDC 0)C
.4
JNJJ 0)JGNN
U) W] > > -,A -,A
0 0 -H--H- M) 0)
H C'-i M
'
UlC)
-4 .H .HI -4
-44--44--44--44--4-4 -44-
-4
C'
- OD (3) 0
H.H, VH. '
r--4-
a-. CD
o C>
C C)
a oC
g C-.%
) ol (C> C) C3 C)
.)C>
nMC-. <D C) C> C> C> C>
(7)
coooooco••••ooo
ooo
0 0
INN
NI~
00
mmXXoo0uU>
4J~~LI
W 1W N
Oi) ID
V Kv Ch ON LALA CN CSI
-4-4 1-i1,4r- rcoooCd.ooU
4141
4 v-1C4
4141414
0 -4
E4
Ln
4~.U
> -H
>
IDa ) P0
U U
HH o~
C4
" -it
I U)Mw4X w X 4J 4~J
0 w
0 *-l(0 -H4
MI
U m)
I
I
I
i lE
lII
I
I
i
IIEIt
I II
I
I
Ii
I
I
II
I I I I I I I=
I
l I I
iI
It
i l l
I I I I I I I I
I
I
I
I
I
I
I
I
I
l l
I
I
I
I
I
I
l
I
I
I
I
iI
I
I
,
UU
U a4U
EI-llI
I
I
IC,
lE-
1
I
I
IC
U
I
ii
I
11i
II
i
i
I lIII
111111liE
lit
I I
I
I t
I
Il
I
I
I
I
I
I
l
IttI
I
I
Il
it
i I
I
III
I I
I
i
till
al
I
I U
l
ilIII
111:11:tillt il
I I I I I
I I III
I It
I
t
I
I
I
I
U
I I I
llIII
I
I
I
ii
iIll
I I I
I
t
t
t
II t
t
t
I
I
I
I
II
f
11III1
I I
I
I
11111llill
II t
I
I
I
II
-l
I
E-4
U tU
tE
>4~lE'
U I-
I
I
I
II
I
U t
IUI
f
I
-I P' ýl
l
It
IUI
I
I
I U
I
I
I
I
I
I
t
I III
l
i
tI I
II
tI
ll
I
I
t
I
tt
I
It
1111
I
I
I
I
1111
I
E-4
11111lil
tI I I I
I
I
It
tIll
I
I
UE-U4
tI
I I
I I
It I
I II
. I I
c i I I
I I
~
I
II
I U
t
Ii
I I
Ii l l
I I I
I
i
t
II
I
l
I
I
I
t
I
I
C
i
I
II
I
-1C1C11
I
I
~
I
III
Iii
I
L I U Ia U4 QUI1
OE-C-'
E4
EI0'c
IU I
IC-'
t
E-MIC-< I
1C
I
U11aIUI
it
I
ill
I
E-R E-18
111)4
p
VI
I
1
19
U
U- UA'c UM 1
UE-EIUE.'C
U U
U E4U
u
IUI
1
1 u~
c-Ic-i
I
U
U0
1
U0 1014U114U114 U
11
V
V
EE
I
10
1 U11o LI)I LO I14
U 0ICE-4 EI 'c EnI c'
-C
UU MtCE4
' EA
10
U2
)I
U
U, U
a4
IU
1 U
0 4000t u04
84 I
%> 0000-c4cI
IL
E-EX10)1 E')I.' 0'r.
1
5
I3
0%I
18
I
I
I
I
I
I
IIII
1
1111111
ill
111111
I 1111111
I I I I I I
I I I I I I
I I .~
I I
rco N I-j414
4a
ýf -ir-4
ý v-4 ý4 v-
cooooooooooooooooooo
C)
C 0CoMo4o
C
o
-4
4
-4--4-I - r4 M
-I44-44M
-I r-f
l
1
M
r- r4-4-4--4-44-114
I1iý0 4ýIý
1
r- -IIrfý 41 I-qý IU 4- IU 4V
qrI
T-4r- - - r4 4 4 4 4 -fý4
U
LooOO.z
-4-4-4
Ir1
r- r4r-I r I
ýA
U4-4-1-4
II
I
I
I
I
E-E
U
' LPL)XU
UPE')
ID I)H 0.0 ) U) 60
r1I41-4
44
1U
IE-'IU z 4 I 4E-I E- C19U0I E
#CCU<U
FlIC D 110F10
0 :F :810
U1414UMU1014
U
00h
ee
00.c4.40014-0I
0*HD
I
i
11111lItti
m -w,Ln a r-aDchcov10
I
I
U
4A)rIl
I
C~'zIP
U'
IIIIIII
11II
4j
E
MUMMME
CU
lit
ill
t
Ililt
I
I
E
'c~lc~l'~l~lcE-4
.1E
E- .IUIC
i
t
It
II
I I
t
l1Ii~t
I
U
itt~ll
lit
-V 0010c'
I-
it I
I t
I
I
t
t
Ull
1111111
1111111111
1111111111
I I I
I I I
, IU
I
blii
lII I tII
I
I I
I I
I I I I
I
II I
tt
11I111II1ItUl
11I
U
I EU)14 C
1
E-4
i Ei 1E
iE-4
i'
E1
i
i
ti I
i
I cC i t I
I I Ill i i
i l
l lll
I
II
I I
U I I I I I
I
II
it
tIIIl
itt
I
Istt
lilit
I
I t I
I It I I I I I4 I I I I t I I l
lIt
till
ill I
ill
1111
l
It
'cU UI 1
i
C
IIIIII
Ut
I
Itlililt
I II
111111l
ti
tit
l
i
ICE-' i E-i
I
lil
I
I
I
I
I
IIII11l111I'c11
I
I I
i
II
I I
I I I I
l
I I
l
I I iI III t l
II
I I i
l l l
lI
lI
I I
I tI I I I
I
I I I
I It I I I
I
II
isi
I1111111IlUI
I I
l
tII
I I i I U I I I I I I I II
tili
UtiI
ill
It
II I II I
Ut
I I I I t I II
t
III
Ill
it
I
I
l
lI
I I I
I ItI
I
I
I
I I
II
It I I I
t
I
I
I I
II
I t tI
tI I
t
II I I I I
I I I I I I I I
I
I
II
I
111111ill
l i
I I
I
I
III
ltilt
I
I
till,
Ill
lit
Iiii
I
IU
U
IU
C- U1 IE
ii
li t I
1111111111
i
l~l
illIll
till
:
iU
II
U
U
E-' >i4
.
I
11
I
II
II
I I II
I I I
~4I
liii
I
lilt
IS,
1U U
4U a4U 14U 11
Ua U
I11111
IIt
tilt
it IlUt l
II
tilitti
tltIR11t
lillt 1111
1i
llil
Ut
I I
II
I
Il
it
l tl
i
1 1 1 1 1 1
I
i IE
I
N
U U 0 0 .-l- -14 (a Im
aD aD En (AI x.0 =) 4-1 4
E
U AEl
i u
u
0
U)E-
1 F-
I
Ocn
4 x
0 c-n
U c
U< U1
'
0
Enu
' LA'0 N
> Ec->c*H
1c- I
1
Ol
-44--4 -4-4--4--4
Q)-4)-H-4"-4-4n-U-4X4X4-4-44U-4)-0-0 -4-4-([4c
m*
n,0r-a
OD40
-l900000000000000
r-I r --I
-1
cI-4i T-IT-4Iý r-4
T
C C O C>ooooC)C)
) D
)
0000
0000
0D 0) 0) 0)
D
00
0
000
04
N
N
" CN
00 -4 M
-1 (N
r-4
,--4
m a% "
4(
n
11) 21 H- H- 0444 (n2 U)
-ý. q.
-w
00
-400
-1:rI'
N U- 1-4
- U
4 a 0 0
:
ý4
NW
X X 00 c
(1 (1
00C0
10
U 'Fý -. H
F-. i
U a'U a4U a4U aU
U
EF,
F
o E-
10
mm
29285
" a
E E-~
H n
1)1)4-4 4-4
X Xo00
E
N
-4,--4
2
2
14-4
r-
t
0n •0
r-J.)4- (3)2
N) Nq
u) r-i r-4 Co co c0 .0 u u w Lo > > -q .- I
LOQ~U
(n2
a
J
04a Z Z0l . H1 1 12
i
I
1 010(
ý
,00
U<0U<U.1U
I U
U I
U
I
I
81UI•I•I
UIUIUIUIUI
I
U 'U
U a4U a4U 4 U
IU
a4 U a)
E-4
U1 1 g
U
U04010U
0
'UU
'F-UC
0
101
p10101
H
I
I
F-U
II
401
U0 Ii U
IiU E-U
E
-
iI
UIUiUIUI
I'I iUI
U
-
U
H>(IEU H a' U a E-'
U a-' U
I
l
000U~o0000
04 I 0 101
oUiUi~I~i•
800
I
0u
E- i < i P
0 U xt I -
: ý
: ý i
'HiýFP
Ei EE01(DIl
( ,I E- I
E- ' U
i <~
i<
U -u
SE ooE.i888
~~
-iE
)
E- F-i
U
U I
0 1
iE-
<
Hu
U
I
4H 11K
,
HE--
'H
010
I10
U
U
C
Pi i E- i E-
U E
U
U
U
U
U
iI'U
U
I
I
P
i P
E FiI
•01010101
'UU I'U
E- I'U
i E, Ii
H iH
'HPIPI
I
H
U4 10
iF-
u-
'F-. I
U
u
I.
.o
'F-
U a'U QU
1I
,
iiUUE-U ui
Fi
U U Eý
i
F- i
U
Po -' E->EI
E-4U E-4U-
'UU
U
U
•oooo..
U
E4 -
U
I
0I
EH -I
U
UE-o-••
U
U U E-
UH)OýUHaU
uI u1
Ul
u
I
~
U
'U
'
U
EI
F-4U Ha
I
I
El rj C-
41J
(d
FE--
.N(N
N) N3
-4-
NC4ý
44
U)
0
0-,(N
12).J
0~ -H
12)4.1
0
(1
000C
( :
Mx
xoC
0: -H -H- 00 C
a414
a, 4NaN
r-Nn COw0m
M~ C>
(Nr~
"
r-4-14-4 -
ý4
00
F24 E14
'E-E. L4
3
E00
E
(N(
-ý-
Nul> > -H -H
0 -H1-- 112 112
,o -0N CO 0)>
-4r-4-4 -4-4-4
(N
0-4-4-4000--04-4
4-4-4-4-4-4-4-4-4-4-4-4
(N (N (N (N (N (N (N (N (N (N (N (N (N (N (N (N (N (N (N (N
T-4 -4-r-4-4 -4 -4 -4-4l -4-ý4v-4-ý4-4A-4-4 -4 -4 -4 -4
.H Ll
( II~1~'N C
~
-4
.-T-4-)-mH
0-4
-
"-"-4
-4 -4-
H1 1
"1 1" (12 m
-44444-----
0
~C -1 ~
(1
N
C1
-C4'C
-A -4 CD4
C) 0
t
0 C
0
4
-4 -4 -4)(
30
ý4- -4 -4-4-141- a)a)4-4-4
4- 412(1
u1 (1 (1
-4-
1.-
N
(12
-4 112(
ý4- -4r-4 -4 -4 -4 rA-N
r- r-
m m
00C
mmXXoo
00
X X
C04CN 4 "n "In
En x )
4-4 4-4 U) Un :ý
-24ai
4-1
0000
ý>>-ý.
0 0
0 ý) a)
-H4- 00
tr IV
U
U
I
U
I
> > -,I0 .-H1H rU
(A
I
H
E-
F
HF
I
E-
F
H
< : 0:4
9U
U
uH
U
u F
0
H~
u
U
Si EU
H
U
E-,
U
U
IU
uu
I
IUiU
0000000000000O0000
i E-
E-4 I4 F-
U
U
80
I
F
Eu
I34
iU0
U
-
H
LL14
H C-)
H'H'
UH'U
Bi
'H
u
F
4 u a4 U a4 U a4
uH )U
UlUIUIUI
uU
UH
H H ;F-E-'
U 004U u
u4 4 uH
u
ui u
U U
1U
I
'U IF
U
El ~~hEli1 1
U
IU
u
o
U
coI
~: D
HF
'U
'HI
I
U
H 'U
iH,
F Ei F E- •
EUp
UU4Uc
-.
,
F
u
a,
12
u HUE
I1
FU
u
U
i
H H0
H
$< <
I
UHUIUIHI~l
I U I~
Fl',
>U
'H
:
~ 1212
,:4
UJI000000)
U I:F U F
C2:x4
U
I E- i
I E-4>I
U1 E-U FU
IE-4
EUU
U
101
F
>U
U- I F I
.HH
'4
UHUHUHU
~ r-4 ~
C)
N
U)> '4
N
o
-4-4m
)U
00
U)
C4
A
E4 0
" -,4
in %-D)
L
m
-Iri0
oUUML
'r,[-
>>.,
)i
'0H
'0
H
C'1
uu:?
Un 'HC'c
•
a) ) HmH 0"4
ro
0~ 0
CD
N O0 0: HHq H H
00000000000000000000
a)'
C.'Co' >a a)*)
C' C
HV
Hw HV HV Hz HV HV H
H-
ý in
n %.-0 w aN c)
H- H- H- H4 Hi H
-
Co'C' C)'
a) (
HV H H H- H
) C)'
a)ca)'a)'aD
HT H Hr H HT HqH
v)
ui n
4 -4
~0-4
4
-1 -4 -H r
in n
in
m
m
r-A~ A ir4- A -
~
0-.
n i0 t-i W al C)
n40N
~
i
Un UT n "D r-
rH C'] m
-4 -4 - -4
-H -H-1
in
m
in in
I
in
H
in
w -Hm
- rul 4-0
A N U)i
-H _
in in
in
i 0N
--IH
in
m
-j
mm
n
H- H- H- H- Hj H- H4
_ H
Hj
rO
00
to~S~
t
CD
0C
4
nD
0)
Le
ON
0 0
C>r- C>ý4C -
C
0
4
4
E4>
r=
fir.
ý44)
4.-4
0E.0- I
44
W
)
o<
I
1 SD-'EE>E-
04
4I
ý4El-,1
I
I
I
0I-4
1
I 0
H
EV00
00
<
U)
4J
S0
14 101
r_
*.
U) 0.-I
4.~)
E
ý4 00
t0 .s:
I10
I
>
I
I
>
U4-10
10
0101
0
1
1
1 0
0
0
E-
E-4
I
01
I
0
1
I
f1 El
1
1
-
IIbI
1-
oou
U U)UK
E40
iE
lu
E,
E- ýw8
E- D
E-1 E-
a
au~
I
I
E-,> E->
w
94
L):
a4
m
m
aU1
00 01 1I
0
0101
E-4
01
U 4iIE-4
U 4m
IL
>E
1
1
I
0101 EU1
-4I -4
I
E
B
0 Pmem
0 0ý
C4
m
1
01
::'
<
41
0E >
a1.0moem-IC
m 1
u
E0em
m
0
L(
U)
B
E-4>
~
01I
sw010
000000000000000000000000000000000000000000
oooooooooooooooooooooooooooooocooooooo
o
>E
1I
0G--0E.
Oz
E
I0
E- 4
E'
I
1
IOO
I-I L10
0
'E
II
- >1
-4
0LA0 Dr
0ý
L)
E-4
4
0E 1
IBI
EE->
e01
1
I)
$41
4
-4Nr4
o
r-- X
0E-4
I
o'
I1
0
E-4U>',44
UI
u4 04 u am
U II)Iu em
0UI0
0
0011
G
0 1
1IHLA'0C-
010
lA
k
I-4%OC40MC.)C,
UUE
E-4
IM
0p0
01
0E4
>-
814
w 4 (N
CIO 94
d
00U3.
S4I)
I)U)
E-4
1
1 0 1
41041
'1
01
4
LA'
wOowowow
0I
I
1 I
1
I
1 41
oE
S141F
L)04 I01ua
4 uE-iw u
U Ia L)wowm
Io L)so
Uo
L)
U--
E4E
u
E
010101
10101I
I 0 >1 4
>1 -i 0
U)-4
O1 0
CHI
0U0
1-0-I
L 11
1 441100
10
~
E-
4
>E->E>
-
1
11
8
oc
, >,
z
1
0-0-0-0-0-0-
m
I
oo..
•
1
C*4
r
-IMf4Mr4M -
ii
1 EH>
>-0E
-4
u<uma
•
O
0)0
C14
-I
-
u
I E II
>
>
14 1
IU4 IU ra
O
U
O40
I
1 4 W U •u
oEnUE4U
1
c14
-
r-I
i -
Mr4M - M18
I ý1M -
4 41
~ a~ omo
mu
r
0>
%4D
0o
r-4
T-4
D
q q
i
01,
-4 -4-4
r-I
V-4
C4M m
r-
Or4M I M -I Mr4M-I
14JI
[ý1 f.I
e-
100
-4
* 0
0
Iý
40to)
ra IVI
Ený41
S(a
0
-4
e
0
t-U
)a
0E
0 8
e
0
U0
00
l
mM a2
141 r-44
L
s
-
0
-4 141t-41U)-4
Lfl -41 U
-414
U'-4
LA ,-A
o
C
11
r-4
141 -4
a
e\4(N
Lj1 -4 LII -4 LII -44Ln
-4
0
C
0>
,-q
*- C
0
SE-
C
-4
-4
0 r
C
iUE1- i•
I
00000000
0000000000
uu~~~~l
000000
0
<x I,p ,,°
Ei
l
0O0
u
U
CDEC• 'CD
u i OtI I,•
<
DI
Ui•C
Z
EOI
E--XEE'
UEIUIUEIUE
UE-
E-4
U'
E- > E4
El
a
a
E-
UElUE
E01
UIUI
E
D
E1
E
1U
UEI
CD
UI 4UD
1
1
E-'
Im
U- -IUE)I'I
HE-
EUii
E-E- a
UCD
Ký~~~aa
-E-
IC
0
I,
EýIU P
EI i EI
ý
Cu,
<C
<
I
I
E,
CDi
0: E ~ b0 Eý-IEP
0
Is
EýIUD
Ký
-UIUI
E->E>
E-
'0
01
EI
18
0
E
Oj.
E1 lIu
I E, 01
El
E-E
E-'
rý f-4
v4
Wa~~a~ata
U ED I
UEi
CDIICbiUI
C):p1
u 1
u
m
I
Eulu>
<l
OI
K
11ýEý
Ii
E
u
uE- u0I U1
01
-) 0 I :ý I
Eý,UU E
a4
•,LI,
<
I, :
>
E-
u
CD'C '
C'
UEU'
u-i o
<UFl
CD'
1 >1
: E-:
u x Eu >1104
u >x u
E-
'
i
F i
O
usuu
z(z
UI 4
1
0DýAC
ý4
1
U,
0: ,
SE-4
0D
4
-
I
CDICDI
Eý
ý I>E,iI
E-II CD¢
E
E
E-•
F I
ý uI EI
CDI
E-
C)
0 r
-4
-'IU
E
I
E-i
C
i
-I
U i
Eý E-2~
00
m
-00
,o0 (CN
-A
o0 ,:io ,L N- L-I
on
aI)
,-i
II)
I)
0- -40
r- -1 0 En ,4 n -4 Nu)-
0 4
~
veC
orn %De m ommmamm~f
)O
-It
,N
0 U,
Ii
(N -4
-4 -4
N
N
N
F-' I
I
_N i
Eý i Eý I
N-'
NI
N
N-
E-.
<
u<
i
E-E
>E
E- i
EI
>1
E-N I N•
p
10
E
I
a:'a<:
I
EIE-
~U
uIuI E 1
EE- C I
0
O
>
u2x
NH
EE->
00
8<1<1ii•
01
oi0101 •
.1I
1<I
N 'F-
0101
UAuN
U E E
i
>
E- I E-4
U'
11
>1
E-
E-•1
U>
I
•u
(
Nrq( -
12t
I
I
a)
N
C
-4
UoU
I
•
N
I
)
01<1
Ei
14
El;
I4 r_)
Q) C'
C
Lfl
Ni
_N>
E-uE-'
Eý
E-•
i~
~~~l~l
-
(N a) (N - (N)C (N)
C
(N)
C)
(N a (ND (N)C 4- (N a) (Nl -43(:) w~ ml a) Lfl
C
.H r-4 .- i
04 r-I
-qr
r-A
r-4 C4 r-I "- r-I "- r-i ,-I
" rA U)
C -I m
E,
N
r-i
o% o
u QýuND
: o
UE-4
E
>11 -i11
i'
I
N
'u
E-4101ýE3
EEI
u- u-
I
N1N1
0U
: N( I
EuE-
0101
uD
p rD E
UuN
I
~~
E-
EN>N>
> E
E
EG-'
Eu
-
u<gu 'u
U
NI
ru)OWiU
P
UNE
rI
CLi
U0uN
E-4'
I.4E-
a:NE,< I
-
IE, >
I
E-
N-DE--.
I E- I
< :II~
I
U
0 E-'
0
I:
E- >
ulu' E-'
E-'
i 81<
E
u0
o
uo-u
Ug
Elug)
4)U°
D a
Ir
-
m 4
-urN n i
m w to (D
u
U
N
I
m 0p ,-q
Co m)0o
iN
i
o
o
o
C
"•m
~
'.D '-)
N
LI)
Ln
L '.o
o o0
r ((A
"m1-4 -4 -4 -4 En4 4 .4 Oct En -4 En4 4
En 4 El) 4
N-
N
u
Fl.
ko I'D o N CoI'
(N r- (q ,-
r
m
U)UUaUa:S:))claL
4.-m
--4
-n -4
-n
ko k0 4
-
j
-4 EA -4 En 4 En- 94
"N C') -o Ln '0 N- Co 07%0D r-i "N C) - Ln %D0N- Co m 0) ,-i (N m
Ln %.oN- Co m' c -4 (N
)0m'. 4m0a )0
a%
0'Na
.
6
6<
0o Dc 0
-c0
o)0 A i
-i4 .- 4i- .-.-i
1 4.1 "-4((N (N
-4 --- i 44 -Il l A - 4 -4 -4-44---4 4-rl-4-v4 -4-4-4l-4-4-4l
oooooooooooooooooooooooooooooo~o~ooo~o~ooooooooo
000000000
Co 00000000
CDC0DC000oC000
000000000
DC)ClC
>Cl DC
0 J-)
a-40
Ni
-
-
>1
'U
ugui
0C04
(N,--i
U
~i
u E-'
N
r--U r
0 10
NC-.
I
V
U - 4Dr4(
E
EE-l
a)
U
I
N>N>
P00
<go
oko-0
(1
N
uI
U
E_,
0P0
E-
0000
F40
IL
E-i i P
P
Pg
0 Q
E
i
N0I•i
N ii NEU E-
N>NE>
0101U
0101ý
I
DC-CoCoClC
O0
r
U)
0i
CD
:a:laUa
-T
3 LI) '.o N-m
(N (N (N (N (N (N
~lC
-4 -4-l
DC
(D0 4 0
r-4
.4
$C~ i ft4 0I 0< En Oc En
Co
N m
4-4-4 -4 -4-4
)(DC
)C
1-4 (N mn -ti
m' ' ('1 m(
-4 -i
-4
>0000
C
M
(m Cl) m I
m
Nr
0Ln0C)Wm r-
mmm
r
0 -0i-c c0
-4
c -4 V V " " LI)
U) r-4LA,_I
1)M En a
040
F:4mt
oV-I Lfl,M
oi _4d
XXMMD4Li4
Fi
E
I E-4
U
o0
0 w0
F- F-'F-4E- cn a U U [ ". U)
LIC)
4M
1 M4XXH H
F
0
oooo.--
"-H
U U4 J
=4X
ý:ýa
>>
4J
a4aZ0
-H 1-
S
iF-4i
Ui
i
E-.
E-4:>E
C,
,-4 ,--,- 0.-i i,--0
(N "
00"-4--4m (mX00X
kN r- -
E-,
E-'iI E-i IF' H I
U• 4•U
141m
umummumumumu
UlUIUlbiblBIUI
101010
iC, i
EýC,
> C,
E- >
' P >'C,
<- >I
Ip
I f IC,
-'O>-Po>-PE>
0E-H>P
0
IU (IUIU
IU
I
U
I
U Eý>
EI
4
E4>
01010
>~~HH
UHiUHiU
IC, I
C,
IC,
>
E-4:
U404:
04
10
UO
I
OE-'Ep
p>Fp >E- >
> iHI4:IC
HO
F-'
I
I
E-
>1
4:14:1
>'
C,i
1
S
I H
II
C,i
P E-i
u < uu•
u
4
I
I
uou0
1I
P E->~~
F-
zUIF4
U
U
E
S E0 I 1 0
4:UHI
gaIHIH4 II-I
gy I
U
I
14:
I
IF-
u
Cl, 1C
ICx4 I1
000.U<.00<.
H01010101
I FEE4
lU
01U41
a4
1004
U
4
UIUIUIUIUIUIUI
U 1 0 1
I0
a
U1
U
01
E,
U a4U
•
4E1
L
14
lO
I4U
IC,
U
I
I u
Ula PC a4UaE-
IC,u
I
I
4
C4Ua
l 00i
iD
I
IF-P14
4:'UIE
U
ER I
0
i
IC,u
C, I
U
UIU uU
U
U
,
PCICI
C P ,I
IH
HI 114:14:14:
0l
() I4
u
01
CI
F-'
,
CI
U
4U I
U E-4OFPiE-io-
U1 EF
U
U
E-
U
H
uu
i
lU
4
UC
UIHIOIHIHIHIHI
F-il4
4:0'
F- ý 10 ,4
.
F
cS o
I
oo-o
101
E-4
u
E
I-
I
U:I4p
I
C
C, 10
0104P
I
0
Eu U
,CU'H
>'PC, U4:P
C,-
C
P
EI3HIP>EIHIHIHI
CIC
IC
I>
UC I CI,
C
8 :8U
S010
CE-iCI
F-
~Ln
•4-)
U4)
N-o
m
c>%o0.0)(co
LANNN-4NN
41) la) 41) w r1'mNO
0> -i 0> -i 0 --i 0) i-4 V V
Ln r-4LA ,-4 -iq' - LA r-4 4L) 4)
M/
N Ln u) U
u) a4ia~4
H
P
H-
:
(N (1m
Dmm -4m m
r-4 rn l00C
C
cimmmoooo
m m .- l-HrrM
u Hý Hý H H, co'
ITu -u
-
u
u
-
mA %o N- m m 0) .- i N m -4, LA %.oN- mo) cX.
0
H "- m ,v LA LAo N- mX m 0D r-, " m .r LA Lo r- m
meIm
m m
-rq
-4
4-v -4' 4-e
LA
LALA LA
LA)LA LA
u-ALALA La) LA LA Lo LA LA LAoLA LA
r-4
4-4l-4 -4-4-4-4-4l-4
-4 -4 -A4-1 -4 -4A-4
-4 -4 -4
-4 -4-4 -14-4 -4A-4
-4 -I
me-4-4-4-VI
mmo
mo~
oomem
oewm
0 01
0-O
U) 0*-
P
I
00
0000
00
0000
0
0
0
0
0 0
0 0
ooooo
0
0 0
0 0 0 0 0 0 0 0
S00000000000000000000000000000000
o
.D
OD m
(N
.-
NC)
C>
"C -1 (v)-1
Qo - ) o a o Q)
<D rco C> rcI
irn
MM
,R
n rq
-iL
(
0
.l
.-4
(N
%D
a
o
n4i - u
ri
%o
m
%D
o 0
Q) 4 4)
-Co
c)
r
4 .-d
: 1 4 :
iu)
C) , r-I
r-A
H.-41-4
m
-iL
-4
r-
(N
N
C)
in)I'
(N -i (N
I (N
C
i ý1
- in .- 4'
-4 N ý4 W)
-1
iM
.-4
4
-
)
ý4~
(
A r
M
q
)
in
r-i in
r-1 in
-1 in
ýi in
(N<
-Hin
L ýin
4
04:
EIF<
4 c l're i : i
OlE-4
0)
in
1 '0(3
(1 ) C)
N( a)~41-1(N
CNC)(
(Ncqr
-i
r-41 -4
I E-' I E-' I
1
U
-H
ma)1i
in) -1
r~-i,-..,-i
in"
,~--t,
U
E-4E U ri E- E-.
4:F
H>
ui
E-I
0i
0
E0
H iE-,
H>-'HHHH
0
I
a4U a4 U a4 U> a
oi
01
EH
Oi
CHi a
U
a:4U: a~4U
0a4
u40:I4
I U
U 4:
04IU:
U
Ei i Eý
U I U
u
8
l<.I r
00
•HF-
Oi
ý 4:I
0
r4,
01
E:-:
H>H
UU 13
I U
bH
E-H>
81
Oi
E-HI
>
u
40
' I <E I ,: I
r...
4:04:04E-
>P
E-
E.,-I
>
Oi
00
01
I
. 04
4U
u 01>1
i
I
',
,I
U a4 :1U
8
S U a.4'
U..a'.U 13 U a4
ul
0000-I E-•
04:
E0-
P• I P.. I r.
c
8o4H>:
801
uOi :
Hi
0i~
F-''
E-i
F-'
E- >
Ei
rH0
E-, i E-• i ..
E-'
i
E-> P
Eý
0U)>
E-'
Ol~lul~lol
0I-'
-
uot
'OE-
H>
i
r,..
.• I
- E,.p'
lI
U i
U
OUIO)
1>
0-4
> Ei
8?
UI
FI F-''
E-'i
H
ouu
u0000KHI
041U13
E-'
F->
014
04:
0i1
oý4I
E1 '
H4 E-
F-'
0
I
.
E-IO
I u''
E. I Iu- I
E-' i U,. i E-.' i
F-iý
-0X-'
C-
E-' i -
I
0
E-
F- '4:
Ia U
a4
CL4
4E-'F-' I
E-t
E-I
IO
HI
01
1
E">F-'OHO-'OE'
>
> Eý>P>
E-'
>
>
04U04 U 4
a
a
'4
OE-'OE-4
10)
E- i E-, i •
C
UP0101
P
1)
U)E
4u<:'HI
K>•: . >:
UE1 UEý U P•
0000
P > > E- > P >
0
'0
U-4
U) -
C)
0) r-q
aI)
oT -i C)
U) OC in
.-in
i
N-
en
m
(N
-4
Am
,-4I
(I
1)
O
vin in in
"
a0
C)
r-I
.-I
kD
C)
.-4
.-4
(N
m
'0
1-I.-4 -4(N.-A
in)
%D
NO
0)
-1 i .-H
ý4W ý
4W ý
4 W ý
4()ý
r-40U)r4w C)
-i
I r-I
-4 L)
U)
U)
Lr-4
U)
U) r-4 U)
-i
) r-4
U).-iU
in -4 E'- 0:
n
F40-4
in
in o4 in ,4 in -C4in g4 in - in
CD
-;
4:
in
in
O
'0
N- in
w o)
in
in
in
-•
••
oooo~ •
o:4:c
x
1
•o
• •o
4o
a)
UE- I
0)
r-A
(N
N
)
in)
(N
ml
C)
(N
(
N
n
'0
in)
in
in)
1-4
-4N,
i--4C
ra)4)%.D(N 0
NN C4 ý4
4NiNý
ý4Q)W4Q)W4Q)134)ý4W
(I)a)
C)OU)OQa) m () Mla
M) r- -i
M) r-i M) r-i CO -i Q
1:3Z
r-ic4(N -A in .- i un -i n
in .--I in ,-4 in ý iLf rinI
r-4 in .--4 in -1 .l -i
C')
Mr4) .-- mn 4
•o.•
Noooooo
n u
o• ~'ln,I LIn'0
ooooooooooooooooo
c) .m inr in
in in
'0o rC) in)
,-4 in
(N in
C'T in in
Ln in
-,DrN
CD in
C)-i
(N inM' in
- in in
%.0inN- inM7%q)
C) in-A-4in
in
ini in"N in
in inmc)
in in
in in a)in
in in
in in
(Nr) in
T-1 .-4 1-4 r-4 .- I .4 .-4 .-- 4 'A .
(N (N9
(N (N (N N4 (N4 (N (N C') m' C') m mC')
C)
m m m
IV) IV) V' qz
•
'I
C) C) C) C) C) CD C) C) C) CD C) C> C> C) CD Co CD C)- C) C) C) C)l (D C) C) CDIC) CD-CD CD C) C) C) C> C) C> C) CD C) C) C) C) C) C) C) CD
C> C) C) C) C> C> C> C> C. C) CD C) C) C) C) C> C> CD C) C) C) C) C) C) C) C) C) CD)C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C)
0-4
mDN
m r-
(N
ua r-i L
r -44
mC>~~~
m
(N
r-4 LAr ,-
r-i AC
~~m
m
r-i
m(
m
mm
mCe
m
m
LA
T-4
m
-1 Nc
m
W- C)M)-r-
m
mN
Ln -i LA r-i LA r-
0
,A
LA r-i LA r-i LA- 4 LA r-i LA
u- r-4 c)
.lF:-4 o< r-4 4- r- < r-4 04 r-i 0:4 r- om Eni
m -4 i-n KC
-4
-4
N-
N C
"IT"4, LrI
C>'A C>r- C. LA
.-zr rLo Oct
U)f U)
U)
m
k0m r-()mX 11
C
iACLA
Lm LAOm : k
LT,
o) U)L a)
(1)
C) r-4 C' ,-i C'
1- C' r
N ~N-1
NU) ~i)
(A
HIi E'
uCDD I
ct
0 a1
FCC
H>I
FCC
0 E,>
0CDE>
>>
00
H0
E
H >
H,
,u
I
E-
<
u
I
p
I
1
0
1
. i
U
Ei I
u'p
u
uo
uc
U
Hi
H-
UI
PH
p
u
U
PH
H
UI
f<
uH
Hý>
Hl>
H'
E>>
E-
E 1
1
0
U'
I
'('
E-
0
10
0 1
>
UI
Hi
0H'it1t9L
r•IE
H i -i I1I
i
S
U
xI X4Ol
U
Hit1
0.
.,
II
U'
,
I
0
I
Hi110
U
BI
H'
CC
H>
H
U
01
I
HI
i
<>4E
CE,
Hi
E :
'
I
u
UC) .lUEn. um CA.
E-
E-
IC)
U UU.
'H E'H
'H I I
H
'>H'H
H
E,
'H
i
CDCDCCDCDDCDC
1
H -4
H91
U IH
1
C1D
CDDCCDDC
E-
Hi~E-ii
HIU
10
HiIl
Hi
U'
HBiB
i-c'
CDI
0019
' u
H
H''--E~
1I
ilU
1
I
E-
I
IYB
U
U
<
EOct
u4l
BYzY
IH
P
IH -4HIE- iUH
P:
E4
CD
U' I
CDC
CD-
C
C< DC
E-4
P C D
CDCDCDCDCDCDCDC
CD'D
EC'CD
d-4iP
UHUHUHUH
CD 'CD 'C
wo
(N
u -- I
vo
o
LA
C'
.- I
a
N-
"T
m N •00 00 ODO• 0
(N
U)Q
(N
) Q)
(N
Q
0
.- I
(N4
in
C) ON -4 ON N (7.(
in
)aU)QC)
(Q
mn
m
r-4
-~i-4
Ný
U) -H-
,4 0
aO
O) .-
w)
m)
C'
T-i4
mN
en
T
-i
in
LA Lo N- m) mC c'
~-i4
LA
i-n
m
r-i
LA
i-n
4<T-
,-I
.- N-
(N
r-A
C) CD
O (N
N-
CD1-1 rin (N ,i mv LA tT c> LA () LA MC N- (n MC %,0C'
a)
a)
)
)
LAI
CLA-1
UDCD
UI(
- )
U)qLA r- LA U) AC)
r- LA r-i LA ,- c'
-i C' -i C'o U)
r--4 ,
C'
C'
-4
C r--4-IC' -- C' ,- -4C -4 C' r r-l
-4 in
m -4 m r-4 qý .--I .:I -I q~
C'
A
<E
om<E
N -4 N X4 N F: N .- 4 LN
En4
En
NC
N
m
l N.
(n<
m
4) Q 4)UCI4)UQ a U
LA r-i LA r-i LA r-i LA r-i LA) ,-i LA r-4 LA rimn
.-4 c m
ri-n-m r-4 m- .- I mn
-nivM
,-4r<-444_
o'
rI4
•
LA
N-
ONLA, 4L
m
CD
m
mN .:n LAv
m
r-ND C)
m t-n mn-4m '
m,-I
C'i
pC
-4(N
n qr LA LAo N,
Dc) (=)C'
-i
"N e) ý
LA LAoN- m )
C'-i
c
(N
CD CD CD C' CD C' C> C' C' C' <D C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C'
C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C'
CD C' CDC' C' C' C' C' C' C' C' C' C' CD CD C' C' C' C' C' C' C' C' C' C' C' C' C'
C' C'l C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C' C'
LA LA LA LA
LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA) LA) LA LA LA LA LA LA LA LA LA LA LA LA LA LA
OD
C)
Ui
I'D
("N 'IT( m
(l r- -4 OI --:
N- iA
LA
LA)
LA
LA
LA)
ci) C> ) ciCo ai) C) (2) C) (1) C) (1) C)
C)
Nr- ,-4 (N .-4 (N .- 4(N -i 04 r-4 (N .- I (N
O)
0N -,
LA
ai) C>
r-4 (N
rlo
'-4,
C)
a
-4
m
qzc
N
OD
0)
(1)
c> LA .- i LA (N u) r') ui -,:rLA Ln Ln ýo Ln r- t~o CX
(NLA C')
ýC)
k.0
kC'
kC)
' .0 ý.
'D
I'D
-10
o ) NLn
i1 OD~
i-i 11) r
CP
ai) C) ai) CD (1) C) II) C> ) c o) a)c) C) (1) C) ai) CD (1)
r-4 (N
A
C-4(
-I
(N
N.-4
(N
-1
(N
4
r- C(4 r-4 L) -1 LA( ,-A C)oc-4 Co c-A C)
m
m
m
c-q N- c-I NOctm tm
r-4o
rr- N
A- m
- sm
- A04WFCE
- =-1ý lýo 4rqr4rirAEnw
fZ
nF4(
EnF:
C
n<E
w re
:CE
4- nKMFC(
- ~ r~ 4"- m
- "nPCU
1"1ýc A4
vm
U
Hi'
S i
i
E
4- I
E4'
I
4:'
PH
0'
010
IE-'H
UJ
I
8
8
0000
0101
0'
0
Oi
8:
S>
r-I Cl --i CD -4
c-IA U)
LA -4
f<
< o
< i
cci:,
'UE-
U
-Uci 0 LA C)o w
N- w N- N- r-
PH
0101
u i
oU
u i
0
01
H'
E-u 4
P >E->
U
U
00
E
20000 8
H-E-c
U i U i
x
ui
U'
I
0
i
PH>
4
o~u
,< I
E
'
U 04 U ccc
<
I
4:'
a.
E-H
0i
04
E-1
H>
u1u1
UI~i
0000a4ia4EU I
0101
U101
H-i> P
u1u
Icc
>
I
EH
U'I
H-BIi
HEý
b>
0
UH->
E-
H'
0'H>
0i:04i
E-
H1H1U
0'1
4:'
E1-
0H
0000
1I
UUU
UI
0 0
0'
H>
Ee>E>
H>H>
UHOl H lI
00
Ei
•141
H101
0000
ci
I
H
c0c 1
T
I
I
0 1
0101
0101
Ei
4:>
U
I U
0i01
0
U
1 0xt
0101
8 ;
40
U'
000 :
0
i U
4U'
0C1
0101
O4 U 4
0101
H1H>
0101
U
IU I
0000
0101
a4
ai C-)
P0
E-4
u0I01
00Z
U'UM
UH
0iH'
0101
U
IE-
Hi
I
I
ccc'cc
cci
0Oi
Oi
100 :
U U] UU
0101'o<H
900
H
Hý
cci:
IE-
U Ul) Um
x
H
U
I
-
I
u A
Hl
u
l< : 'ccIi
c-c cc Y
UHEH
cci: '
I
UHUE-c
H1
0'
cJJ
0 0
C)
Ln
'.0
0
-- ml
(N -(N (
r-.
N ,
0r
Mi0- %D
c) LA 'D
,-iLAN'.DL"mm
L I , Ln
LA M WL M LA
LA 41 LA%O
4) 4.-J NI-U -,1
o Ciw CDwCo) C), i)C w
D a)
D
-)
iDw
cwiCD
iQ)
4-1
N -
c--I c-4 c-1 c-I c-I c-I
w
En ]4:U4U4U4:):):U)
4(
<
c--I c-i r-4 c-I c-I r-I -- 4 c-i
ns
E:<Ul
<CAgE
c-i c--I -I c-4 c-i
<E
:c LI)04:
0)
m
m
OD1w-c'
0.--"
ONvNNLi
%oo , ý4%,
-kor- a-) Nra,-)L nr1O)'Ln
Dr
-r.o
C)• a) CD-) mai)C,. i )
IO
0
'.--o00
k
(N c--I (N4 -I LA)c-I LAcI C) c-I C) cI C) c-i C)I c-I C) c-A
r-i r-i ci r-i m~ c-i mc r-i LA c-IN
r-I
Nq
r-I
LA c- LA c-i
En < C f<-
o
-
0
M < E f<0
ememoooooooooooooooooooooooooooooomm
0
a4
0.O
< 0
mc'-z wA %o Ný wX m C) c-i" (NvC)-c LA %.0 Ný wX m C) c-i (N m ýr LA %D0N, mX mC c-- (N m ccc LA ko N- mX m) C) i (N C) -T LA '0o
M)0MM 0) 0M 0M0M C> C> C) C) C) C) C) Co C) C) c-4Ic-I ýl c-1 r-I c-4 c-I ci c-I cI (N (N (N (N (N (N (N (N .N (N crn m' m' m' C) m m
c-I c-1 c-I c-1 -l c-i c -ý c14 c I --c -Icrc 4 - c 4-I
r--I rc-IA -I r-4 c--I c-I c-I c-I c-I r--i cI cI c-I r-4 r--i c-I c-4 -I c-i c-i c-I
C)
C) C) <D CD C) Co C) C) C) Co C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C)
moo
O000
0 Co Co
00 C) C) C)
0 C) C)
00 Co C) C)
0 C> C>
00 C) C) C)
00 C) C) Co)(
0 C) C) C) 00000000000000000000000
C) C) C) C) C) C) C) C)l C) C) C) C) C) C) C) C) C) C) C) C) C)
LA LA L LA LA LA LA LA LA LA LA LA La) LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA) LA LA LA
r-
mX
C' M
c-i c-I
C) C)
C)l C)
LA LA
rn
r)
1
,-4 r-- r-r-m-)
Fo E-4
F'
0, -i 0D r-4 a o0 ( (N In LA u U
E0 -4 A -4 2) X) En U) 4. CL.U
a4
m l)
1
-L4
:E
-4
cq-4
0
D>o)11
O
F o
H
Co u
u
UI
1
-
0
c
U4
P4 al
0-
0
uiLo-
->
0
U
C
Z0
C
-I
o: 1 4 1 E--i
1
1014
-4
0
i
I
I r-) I
u
4.
10
i FFF
41'010101
UIUIUE
I
u<
U
U
U
U i
rU UC
u iHE4 U..>
FF-H
U Ii.:
i~ U
U C U
E4 '.HU
E, F
0>>,
1
Fuummu i
00u
F
' F F
F
F
F
0
U I U
IC
uUIUIUIU
-, FA
- Fo E-AH
C
'UE-4> U
IF-_. - I
E-4>IIUUIEI
U
E-4>
4
muo
I I
o
0
1
18<
, >,
I
00800080
U EU U
E C)M U
F
u
01010101010101
U~u•<U
U a4 U a
U 04 U a4
U
aU
134
u
Ig
101
uIF-.
i U
UIUIUIUIUIUIUI
I
UISUIUIU
IUI
U-
U
E
I
F-
I
i
0 a
G
0
U
IO
0
:
F',
UA
ui
U i
UI
UI
I
FHI F
I
F
I
F-.
I
u I
U a4
U
IFu
UI
I
iEH'
F
00
1
)
U
U04
a4 U
Uf
01010101010101
E- H-4U
-- U-.H
-H
>1i•
I 9
10IIE-.•
I IUII
0 OIE. 0E-.
I iE-iýý
oIE-.
II
UF-'iF-'
U
U
i-'I
F U
-UI F FU-. 'U
IE'FF iF-- #FIC
'F-'UE-,
I <
U)U)U
U 'U
U
)
U
uU U
()
F-'U IF- F
U
-
I]
I
'F--Ei
MU
A U UM U'un
UlIUI~
01010101
01010101
U'I
P
IF u
8E
g
I UiIF10i1
p ,,
OF
4
U
I
IF-'
UI
UI
U I(.
F-' F-U-U '
U
E, I
FI
-
UF'
F
U E-' U F-, u F- UgE-L
UI
o:V
8C?
r
U --4
'0
14I O
O
O
I4 1
0010101
m
-1-40
0m
m
r-cC
U1)
U))4
"TMM
)
0D -40c> -4i'
qlA,--4 U
EAK0
-4
U) -40
U
m010
4E X
1
NLA-LA
134
~u) u)
U
:I m 2 m.0
-4 (N 01 'UWLA I'D N- w1 1
U E-E-PPO
Ua(0HH HH
m42 E :zxý::E
(N (N
00
- 1 -N1 r-4NrUU
X-X
Z
oo or4 (N
o 01mo.qo LA
o %.
UUJ )
,a A
-
0XXU -4
- -4 1-4
00 14 4 4
raE
x
O-)
0
.a A
0 0U
--
w
>>
uU)-
w01O) <D-4 (N 01 -,' LA 'D r-
w
mLA
omo) oC)4-C)
mU)'
q)-o--o -'-T'
U) "r vu)' LA
o uA
oo-T
LA
LALALA
LAmLA
m '
%D
)o
4-C
moammeme
m -4 -4 -4 -4 -4-4-4lý-4-4l -4-4-4
o -4 -4-4l -4 -4
m -4 -4-4
r-4
-I 4-4-4-4-4 -4r-4-4l -4
000000000000000000000000000000
000000000000000000000000000000
LALA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA LA
D
0N
t
a
a
ac,-i
Co
O-4
0
-'00
m
0
oI- -
LA
r
ý4q
4
En
kD
0
ko
am
C
r-4
r-
N
N
T-1
0ý
N
N
%0
O
OW
-4
q T-1
4
r-I
LA
m
0
0
r-4
r C r-imD q4Q
i-Q) v-4N
o> -4 0> r4 0> r-4 0D H4 mi r4 IMr- m~ r-4 IM H4 m H- mD H- IM v-i mD r-i ca r- mD-i
,q' v-i Ln r-I LA r-4 qc r-i LA H- LA r-i LA r-I LA r-i LA r-4 LA r-I LA r-i LA v-4 LA r--i LA v-i LA r-4 LA v-I LA r-i LA r-i LA r-i
O E 4
0 I- ) EIL.
11
0U
01010
I
IIU41410
I
I
I
I
S101
0I01
I L)I1 0
010101
0101@
1U U
I U
1018
21814
0' II
IUUIUI
U
I
I
I
814
I
1
1 f< I,1
It41
I
I
114
I
11 4~
1 y
U
tUI
I
IIUI
I1
0
U PC
i U
1
E
E-4
i
.4:
vI
L
I u
U
,,
0•
U
UU ii,
IU2I¶
,
01,t E-4aE
t Eliu t
uI
E-I
Eu1LI
mtU I)
I
1 u0 mIU
1UIEUI
U,•0- s I-u
1 ~~~
4 I4
1
-, "-u
I
I
UI
1)41 10
U
,
I0i01I1010
E
1I0
U
4u
E-4i:
I 4.
U U
U UU
1
0)0 (AEn0
u
Fa
,
,
I• ,
,
N
u
IU
LA
LA r-4LA v-i Ln r-q
m v-I m r-4 in r--i
101
,
I
t4
I
E'-,
I
Ln
40N4
,
I,
,,•,
1 0
:
I
00
10101010 4J~Ev
01
:
U I
4Ui
01EU
E-0
E-4 i
OXU ,UI
I
II
I
UIU,IUU
1 ,10
0
1. 1
01010101
. 1.E-4I I04,[-
1 E1- m1
A
v--i
Om
Il
Cm4
E4E 3 1 E-
l .
1,E-4 ,o 4ýE
S
LA
004
C44
N
r-4%D
r-4
m
N
%
( 0
iN0
I
IU
1
:4
I0
I
I
1E1
IU
IU
I
1
,Ii u '
U 1IU E4 IE4 &1
U 1 E4 UI iUE-4
t
CA2 A0
E - IE 1 9
1
1
16)91 1.6
E- 91 9 9 E@
I II
u I u ,1LI
E-K
W 1 ,4E I
E- • 44
Ei
,
10
I ~
E- 14 U1 i
1
uI
uI u
L~i
I
b
-1
.:1
01.4
4 uluE-14
E-4tuE-491
>) E-4
E4-IC4
1EiE-4
SEE<
4
4 4EU
-4
u-E"-UU E-4,,,
#ccIE Ii <E4
4
10
E-i
01011
En 00E
I
01
141
E-L E
u
U
4
.3II
n 0X,
z
- 0.<
4-
0181
I ~01101
Li
01EE
4 E0
XE.
(0UE-t
E
cn10
0 t1
UE-Ln 0E-4
IEI001
,1
U EUEI
U
U
E-
0101
I 1 I
10181
1
E- U...,IU
01
010
t
1110
U
19
I
L)
I
I
I
I
P-I
I
UI
E
u ,Iu
dig eo
I
l V9i81014
I u
I )
Iu
4u4I
4
1
1
1
I
1u
a
S
uJ-L
t
D
ij
0 -H
OW
0
NT40
4)- En
134-i
4S
Ez4
1
I
t
UJU
4
4EU
a, ý ,I
4 0
>111 U -
I
E1
1U I
14
W1
1I
I~tlI0
~~M11
)
UI
--
I
,1
EI
i4U0
E-1
L
4E-4i0
E
41
I UI
1F
1
1
I
Iu
4I u 14
Ex
1I Eg
1
101
1
o
IU
-
I
L)1I
E-0' ý 0 0'8 0
OD
N
N .imT4q
M
T-1
0
0
LA
-
ko
r-4
NS
CIDPCOD 4 00IO
V-I
m
t-
0
DO
%D
%D
C
kD
C
'
h
D4O
0
i r
Eh
-I
v
3
v-4
OD
v-I
q
r-iL
m
v-I-4 v-ý
D4O
r-4
CN
N
r-4
04O
m
N
rN
LA
M
r-i%o-i0
K
T-4
T-I
D4O
r-
D40
D
NI
1
m
4
kD
D0 0 (D4N4
-i v-IDN CS
O1
4
LA
m
1-4 C4
01
LA
LA
C4
LA
f
4 UoIDCDI 4- cDIc -4
0oooooooooooooooooooooooooooooooooooooo0ooooo°
0ooooooooooooooooooooooooo00
0000O00000000000000
C•o
mW.n
0) kL al rCN
Cl
) ) a)
in r4 wn ,-1
M r- M r-4
•'
N
4-4 N
• ••
0
OD 00 MD ON OD 0)
(N
(N
m
(: ( ) Q 4)f)w
n r-4 wn -4 In r-i
M r-I M ,-i M -4
•
•
Ch r-4 0) (N C
a) U)
m
m
Mmea
C
)
1z) a) rQ04) C ()
In r-I U) -4 I) r-4 Lfl -i
Cv r-q M' r-A M r-I (1 r4
N
••00O•
-1
,D
rIn
()
f) a )
In -4 In
fn r- m
~H
(N
r-l
N-i
OD wD
m
w C04)
-4 In -i 0)
r-A m r-4 .4.
o
-o
m' N
01)
-i 0>
r-4 01
N
o
0)
-i r-A M
(N
0
Wi -In 0C, I
M
T.
0a
10)
010a0
)
-40c
r-4 C, ,-I 0> r-4 0:
r--4
-4N - r-l
N - -4
N
E-
mD
)
Mn r-
OD I
M
0)
0)
0 10)
0)
-4 c0 -4 0) -4 0D
c0 -ic
-4
-4N - N- -I N- -4 N- -4 N- -A
~-i
E-
E- C.-4Hý D. E, U.H)
E-H i H
i
F'
LýE
I
Ez
i
E-FE,
F>H
U
U
IU
U
IP
SF
I E-4 ,
' 1
4:I1 P,.
>10
U
0
-I E-
I,,:10
-I
P.-i
E-4
E-4':
H- i
L) a
1
U
r-i
CE, X4 E-,
HH
'H3i'0'
0I•- I
r,>
E-
E- >
E-4
E- >
SE E u E•oc E
L)E0404 U E-
F
E-4EF
OH--I
00
E-
E-
U4
u) I u.
E-'i
p F II p.
F-' P-
Eýej
Uz
0
4
UX
EIII
E-
Ii
-I
CL
i
E-i i ,,:: 1
E-.
F F
UM
H.
I UE-'
I
C-)E-4
0F
U
U I IFE-
UXE-oH
O-
E-1l
t-
1F-l
-14:UE- > Ecr >
OXOF-0uou)
0u
4:
4:4:
>i E0 >
4:1Eý
E
4:
OU
EOl•
E'
E-.iI
E, <i
U (.)
E-4
01o<01
0101
U Et
4>4:
Ei
HE-
xt: <Iý
--'I EU
u
lg
-
101
0104
U X
IEý
0 I
I
F-' iF,
E1
01010I.0Ui
E-I-
u 1
HIH
U..
I C) 'I
En
'F
HC-.
Li0
.
HU'
Um
E-4
-
I
-
E- i u
E-HoIo
Ei i E0-10
E-' I
U414
E- U) wE
t14 E-) r4E-
3
U
0I0I
U:E0,
.
P -,
U
i4
) E-
I C
I
E-
E-
0. I .J
F-
i
EF-H
En
0 10
C) I1
E-Hi
0
H
0~
01
,.
w.
H 'Hi
E-H
>
H
0>
N-4
(
m)
N
'-4
CD mD ON wC> m' 1-1 m' "N m
m C' n
INC
(N
m
m
m
m
m)
r -q
~0.
NCD C'
0 4-1-
0C> C
0C>In
0D 000
a4J
mmammmo
r-4 m -4m.a
C-4
0)
C> C) C1
~
I
CD C'l CN
CoIn In In In In In In In
0D 00%
0
0'
0o 0D 0D 0'
ch
-" C--(n•0•'KT
IM LIn0 C
O
InM ýI'C'N
m CmD
CnlO
01-- 4 -01T-r0
0' -IT
- 'IT 01 In
0'-- n ,n
-IT01
% -V %--I
Lfl
C )
-0iC)
-AC) 0
) 0l D -0- )
0- )
)C0 014C )
01' -101T
r-4
Nq
Nl-4
Nr
N
1- N -I N-r- rN
l -4N
U)fc4 $cc
Cfm 4: (A : uI4
1m4:m) 4:c mi4PU)
m <:
-44:OUn -1
-4 m man
r-4 m F-A
M i
ar-aI~~
0:
r-l4:I
C'
0o
Cn CD
In CD C,'C
ND C.4
0
CnID Co Cn C>I>
0o 0D 0o
00
(DNC
C') 1DIn In N> CDC'C0 C.
C> Co)CDC> > D N N
ID%0
0 0'
0o 0o 0o 0l 0.
0
DN N N ND
k0
0
C) C') 01' In
Cn
N>CCCD C>
0ý
(C-
C0 CD C) CDCD CD C0 C> C>CDC0 C0 C'
k0
0D k0 k0 k0 k0 ý0
10k0
%0
0o
0D0
ý0 10
0D
0.
-1
N
N
m
*U
m
So()CaCako
a)
C)
0)
(N -4 (N -4
-4 .- r-4 -4
N
C,
C>]q C'
4 E
E4
41
3
SU (U
U
U
i
-I
U
U
E-
E
IU
(U
4
u
< u)
Iu
HI
U
I
EH >
H
- >H
I
I
L<L
)
E- I
Ua U 04U
-U
i
P
U UI U
U
U
E-
L14
L4
'
U U) U
I
E4i
'0
u- U U
U
4 E- 'E-' ECl
E-4
HUUH•
U H( E-.
I
I
U HHI
HUIUI U)
0
U
i
i
iID
o
U C. ' EH
E-4
U H U
, -.E-- Li
U'
H'
OH
Ll
4
•U E
'I
4
U E
IIU i
UEH
u
(- u
u
U
44
a
UEHUHUEHU
Ei- u
U
u0 I u
f3.
UtU
U
U UIU UQ.AU
UI uU
U IUu
IU I
H
(
H
iU
FCC
U
ui i
X4
U
E
E4
E4
ý
E-
'-
U
(<IU(I~liHIUIH
0'
010'0
E,'
o E-
U E-U E-
D)-4
-q 0) -A 0> r-4C
0
- 0> r-:4NA :4 En :4 U) -il ) -<
LI
004
.,2:UI(Uol LI
U
(-1 x
l -I
E
'H'
H U
E- .
I
F4
-
<U <
u
U
E-4
i
-
Nmcmeommo'w
rm
r2a) (1) (1)0
i
o4
U a(U a4U 04U 0
U
'u
o )ooa)QJowow
oo
wm1
-(N r-4(N r- (N
-4 - (N A (N -14(N
-4
"- -1 mI rC4 En4-4 M- ,4 (A 4 In OX n pc4En -C4 n f;4C)
-4
i1U
010
m4
m
ammmmmmr-wmeme
-4 1
Ln
U1
I
I
H E-4I
pp
U4u EC
U
2
-C
. U E2
o
u
4 U E
E
Ul
E2
U
1'
U
EPI
U
U PHUi
I
HE-.
UUI
0
0
0CD
0i
< I
1
U
i
U) CXD
c000
4:
4:'
CD0N
I
4'
C
0o0
U'
U'
.I
•JE-•
H
01
0
C
00
U
H:
0 i
P
EH
UH
'UIUHi
I
I1 iH
~00000000
4:1
U
E
Ea
U
U E-
Ui
U H
E
U -1
0
H >
I- >
0i
>
.:
%D
0)%
( OD
"
o0>
m
N LA
)o0)
A-4
U)
U)
LAn U)
LA
U)
U)
%DU
(2)0a
)0 (2 )0C (2)0 ()2)0 (2D
(2))0
(2)0 (2D)0
0ý -4 (N r-4 (N -4 (N -4 (N
(-4N
,-4 (N -i (N r-4 (N r- (N
0LA
Nl
m
Ua-1
-4 En
-4 (A -4 U)
-4 En4,4
En4 4
En -4 EA -4 En
E>
U0
-
-.4 Ln kD - m) 0m 0o -I (N m .* LA %DN- w) 0m 0
(v
m U)
%D
N-
(2)0C
-4 (N
(A4 <
1-4
.
r-m
0Y
Eno0 U
-,T m
mr)
mo
U)L)UC ) >a)C) 4) m a)a)4)
(N r-I (Nr(N .-I (Nr- LI) - LA).4
-4 rA
~-1
(NAC
m
-
w-4 r-I .- I r-4 '-4
En
-4 (N4 m
0)
) N. 0)
m
nC
y.-4
)
M' r-
-4
a)a)NU
N
N
aN)
a)N a
0
4 0 r-4 0D r-I 0D .- 0 r-4
M" r-4
- tA
N
f r-4
N r-4 LA)r- Lfl -4
004
LTU)
ND 0) 0M 0m
r- CD
)
z
LI )
N, 00
ooooOO•O•ooooooooooooooooooooooOoooooooooo
00000000000000000000000000000000000000
0044
OJ
OOOO0o0ooooooooooooooooooooooo0oOOOOOOOOOOOOOO
C)
Cp C) CD C) CD C C) C)) Co C)
U>
C> (D C)
U)
UD U)
C)
U) U)
CD
UD)
C> C> C) C) C
CD
)
UD UD
C)
(D
UD
UD UD UD
C) C) C) C) CD C
CD C) C)
CD
mmrl)m
m0
m
t--C
lU)
o r-i
"Z
OD
a)
.T~]
C
h
U) U)
-t-OO
c
M
M
M
gUUU
a)
••
w
0 0W
r~O~-4-i,-r--r--ý
4
tW) N N
0-4--4 (1) '04) Un](Ci] Q C).4
L MU (U0 " E "E-HP Ho X X U
U
-)_ j 4-J 0n 0.0.0
En
]
-H
tn
U U
C -H >
r-H_
(U M
uA•
U
U
,d
R
U)
U U U
I E-t
E-4 >E-
m
U
u
>
p
A~tibttiD
•••,
U
eU U
tIU
>E
>
>
Xu 4u
Ut
u
UU
UIU
14p
U
>l>
u
ou<uu
U)
IU
>
UI--'
Ut
Ut
,oum o o
ou
ou
E-~~
um~
UUU-U
)uU
U
r)
c o0o
m0E
000000
<
o
,
E-
IU
o
u
p<,
:lul
g=' IC
Ch0U
xiC4
u! L <9
Oci0000CU
x#c
u
(A
IL
X
,u
,F ,
u
UUE-
u
iL
U
IulE-t
u
, ]UP
,I
XE,
u uCD
oum
Oct
mCD
m 0
) (Ao0o
< xI Oc xX I
o
0( )m0 'E cmU(
: 'ý
lull10lul
I
I
Iu
E-UUIUP
, u:>
>',E E -u
u
,EE
E-
eeom
Iu
p =0
om)u
<:
E-
UE-
me
IC
C
I
UI
oe om
u
UU
E-]UU
U
p>
0
IUIUIUo<I
'U
I
IU UUU
IU
oj
uFe
u i
E-4 u : o4u
CUUIUIUIUIUIUU
u
-4 -4 U U
-400
(N 'OO-4
OO
XXUU
~mmmocoo-.H.H0OO
q -I m m co CDu X
oýU u
U U4J
J
0
( c,
C>
U i
Eý
<
•
••
•
•
U
N
rN-
*00o
Oj-
UJ
1])
U-H
0
4
lb
I
E
U
•oooo
I
oOH
mmmmoooo
PIC
00<0u00000<0z00C0000 U
T
Y
U
cmoo
Um
.•
0
U
1 &
XX
iu U
HEO- 0
Fl'
: OC4 I $rI-Ott
u0000000ooooo
-tE-U
U
I
•
0
U
U)
U
U E- C U U) E
X4J I00w Ut N N
<0000
::
C0
-4
4)
4) L)
a)U
1 4q
• AQ
0l - C r4 D .- I C r-Im -q m
-4
4
t~ r-qLl r-lLfl,-IqT r-4Lflý-! I1.-I ul -4
0o 0
I I
I I
I
kD
-
)ý
0
I HH m
in ,-4Ln r- un r-
I
I
I
j
IE- I OI
I
I
I
I~
~
1
U
~
I0"~
~Iio~ 10
~
iE
I
~
i
E- I
E
01
)
()
U4W
I
UI
I O I
I
~
C)o~~P I
P1
4
,-l
4 r-4m
I
I
q O
" r LH r- Ln r-HLfl
1
I
II
UL
a) Co)0wm q) Q
Lflr- Lfl -4 in
E-U
-
,-1
PI
IP
E-
10
P
0
~
I
v
I
01
C I
E-
~
~
I
: I~,F
o41I
~ P
IF41-
F-
IE
I -
IEP
IE
PI
I0<l
E-4
IF
0g
u
Uu IJ-U
PH
U
01
-
U
i
'
~
X
0100001010
I
-IF
FI
~:0-
:
~
l4l-, '
~
P:'p
1
~F~
4:14:4
'I
14):E
p
u
E
i
U
~
-
U
p~
E- 0Q~1
Ox, 1~~
0:
P
PH
P (N
IN
(N
PH
UIU)U
U
(N
E-
uH
~~ ~I ~ ~ o
zcFC
)'(<
1
U
t I
E-4
~ (~iF4
P NP
P II UI
P
ftctOc
14
14
-
o~
0
E-
<F
C
C
H-HP
I
r-1 E-:
01
)101
e Q~
Po
PN
IP I
0 1
ElU-)UIU
o
I
0-I
-I
1010101U10101P1u10101P10
:
0p
-
HIEI
0~- ~
uE
EI
I
'-'EUoo
'
- E-,EI
~ E
I
11001100
xE
UHFHFH'HEHi4
1010101
E-
I
U0
U
1
i
E-1
U)U
-
.-
Lfl
I'D
oo
E-H4
U)
0-H
)W
o
rinin
P~
iEiU i
i
U
000
IE- I
(
mN
1Jý44()ý
M
II
IUI
0 IFU
E-4
4W ý • )ý
m
L'
0
CN
n r-4in r-j n r-Hin r- L
E-II
E I O
H
oo
i o
C'J
-
ý
a)
r-
(
,--4
C)
0 A
a
I
I I
I
C
m
(N-4
C
-
r- OD
U
(
r-J
IN<N
moD10
E
lor
0
El-)
0I
I u-4:
ENN
U
P NN
NN
P
I~..
IF4L NN
10P
0
E-0
E-4Io m~~g~
P~g~
P:tI PUP <)
Pý P NN
Pm
P PNN
Pag
m
a
m e
w
m
a
m
o
a
om
w
o
a
m e
m o
m
o
m e
e
a
o a m E m
P0
00 @ -1NNN.N11
Pm
E
EPww
P
P
U a Ep-4U0 EP I4UP
P aI P
PU ai PU I U I U 10P 1 u<I000
< P
p<I000
u PU 00
IE-P0000
UO
4UEUa 04U1 a
4U010a
WU
4U
0
4U
a
4U14U
a
4U0
4C
1)111
1
Oi
U
1010
U
10110101018
U
IU IU U I 1010ilfi10
P
4
:14
:4:1441414:1414141414144
PI
14141414U14141
r
o
In
0D
4) 01
0
Q
1)
t 10
1
Q
4)
0
('4
('4
LA rLA
I (' LA .-i LA.-
-4
m
CN
0
01
U4 401I
01
SU
LA rr-4
LA .-4 LA
A
01
0
'-4
t
C
(4
.-4
4)14 Q-4
Im
lU
B
UI
I
c
I 0
I
0101
~-'I~~
B
10 II
U'I
1f
I
i
0-
LA
-
~IUIU
t'
-
14 I
IU
1
04
IB
E4
00
uc
E-,
I
0
I
E4
E-4
E4
u
1
1U1IE-
U4 IiI0E-
0
1
IA1LALALLALAALACCOCOCC
00000 UIPU I
0UI
O4 9 4I
Us.
N N
I
NN
0
u01 UBI
d1 I
8
E
N
4u IE
Il I
I
4 U0 a U a4
10
a
4 I
U04 U
134IL
t c0 r4 0o
A
"'
14I
4
C)
0
si
IE
1
J
u
E-4 I NEU
1E
:E:I
0 01
0 U
0
i
E-
L
)
B
Iu
1
0I
~ 0'r
I
U
10
E-
E-4
0
0
1E
14
I
II
I
I
IE
L
L
I
I
E4
u) II
L
l
E-'-4-4uE-4
U
U
IP
I
0 1 P0 IIUE101 0UBI 0)II1
I
U
I
10
1'4
U
I
E-NNNNNNNNNNNNNNNNNNN
iEI LI
u Iu I
uB
I
B
0 8
1 r) I0
IN L) I
1B
u I
P
1'U 23I
'19 0'
I
E-P 0 F -Ig,
a4 U U a4U aU aUU U
aIua
ua
I
04
U IU
L)aU
0
E-4
I
E I
I
r
u
E-l
0 1
1 EU
0
14
4
14
1u
E-4
EL
4 i
~~E-4
I
I E0
-U o
0
0 0
I
E-
-
1 I I'I
L)
000000000000
- E00
4HE-4
IIUI Ir
IU U
1 P I L)
L)1 H I1
P
14
1
UNNNNCDCCDCDCCDCDDCDC01I010
2 I E-IP1
10 I U
E-
IE-UU
E-HN,
E-
u
Pg
L0
oi
FII
1
P
U
E--
9 Bu lE-0I
4 E-
1COOCOCCONNN
0 1
04
L) a4
0I
UU
E-
oUlE4IU
~ ILU
:B~PiiEIUIU IU1E-4EIUE
0p 1 r0
E- E-i ru i9II
I
I
0
0
C
L)
E-4'cE
E4
-
UE4
I
I- U 10I
UIUI
IU I
EI
II1
IE144I
IU
0 0 EUIUIOIU:0
E-, -4EHHU
0UI
0 1
1
4I
0UI
NN4
I
L)a
1
P
4l
~I
4 C)
1
0
u
IP(40
L) 1 u4
I
IIIP
E-I1
I
IEP
I-
E
I P
'UU"U'A
I)
I
00000
t-
I
I
pcii UE- E-4
r- Iii N
N
I
'
4
CI
0
4 0
1
0
IE-4
1
E-~ IPZ
II4 UIU
U
uU
UII I: U
1
s
A 0U4
0ý -0 C)
i
4:
I0
dI
1
u
~-4
E-4 ' -E
P:
i
1
IUE-I4
EI
1i
10
I
1
IEB
1 (
-Iq - t1 - 0V4 r- I r-rIP
14M
r- IPr- rI r- I r- r r I
I nU ) -4 En4I En I En I N
En4 -4E0
En En En NEI-4
En N-4
1 u
I
E4I0IUIE) I
E-I I E.E
0
UIO:
IUIU UI UI 10I
I1JI I I~I UIUI PU: I1 I
UIO
I
UIUIUIUIUIUI
a)
0
ic
E4
I
i
-0
E-4 Iu
U
(' L
01
t
I
LA r-4
0lb LA r-4
LA r4tU
r-i
LA r4
14
M I4 M -IM
4 -I (nrI M r4 fn -4 Mbr-4M rý
r- PCv- s E4 i<rI
r I E-4
q 14 I v-4
1-4
.-o<rIc
- v10
U
U 1E
E-4
Nr
-4
N
a) -4
I'
I
I II1 14
1 Pic L)14u U1I I
44I
14IL
0 I u L)0,81IUI IuIuIr1i0 IUIuIU1,6Iu0I000000I
IL)
IUI L)rI P
I
1 1041
1I II I I INI 14 4 10 I140CI
1040
I 1404
0I
E-14
L aU
1
n
04
1
1 IL
EnU En En
1u UU M0
1)u AI
1 E-,I
1H
.. ..
I
10I
- H E4H
I010
E-,IU)P
E,
E-4
H E-H
1u
IM I
10
UI
0JE5
t,1E4- I
L) I U
E-
1
I0
U
0
IU
II
10 U
L) I
IU
E-413 11I
E EfE-4
IC
PCoo1 ooo
Ol1
O
IU
U
n
Enu U)
U
L)EnU InU)E)
H
14
r I
El0 1 E-4
rI
U1
I
I
IP I
Oc
" EU U
IUE-4
P
r-)1 I
U I
1
u
E u0
U)1)
E10
En
E- E-
E4
-
10
U EnL) En
E1 0
E- E-
L
E-4H
I
U
1
1 E-4
0IU
E- E-
z0I
I U IB U u
0I
I044I
E1 0
I-H -4 E4
I ~
~101
~
1 IU U U
U
IH
U
EU
P 10
44 1
iP
1i
I 1 0 0I
I I~
~~
U
E4 i U4i
I
IJ E-1 'U 'U
4EE-
U
E-4 IU
E-4 EC I1
1
LU
I1
U
IUE-
4 ED.
E-I E-4
E-I
I
P14
1u
4o1UE
0oo4
1o
4 1oooo
4-
T
Iu 1u I
I
I
E4I
I UBI
<0 10
E-U
UiI
ooooo
EIT 1U
U U U
EEýHIP
UIP <
<0
1
4 ,BI
U
IU
I
I
U
E4E3 EEE-14EE-D
E
oooo
v
0>
cD
%D
LA
-4
C>
0
r
Nn-V
E-
uIE'
0
lEu-I
lii
iEIU
I
10
IE-4
I
14
0
M
I
~
E
I:I
EU'
-I
U~IE-I
I
10
II
I
41 I
V I
I
I
LI
Ic
I
E-4
ll
E-
E4
U
E
,
I ,
I ,
IU 1
U
ii
U 'U
10 1
I
I E-I
4 a4I
1u
t
1 0 J1
UU
i
UE
I UI U
u
I
a4
E
IUJ
1
10~~~~
u
1
E'<4I1
U
0
E-: E 4
I
1
UII
I
I
I
I
I
511111111111III
I I I I I
I
II
I
II II
I
II
I
I
I
I
I
I
I I
I
I
i
E-
I
'U
U
Eg4 4 1
Iu
1I
I:
I
E-
I
I
E-4
E-I
U
E-4
, I
u
I
I
I I
I
I
I
IE
I
I E-4II I
E--iu
IP
I
I
,P
I
I
u
lEp-1
I
PI
I
4
4 1
01
I
I
0 10EI i
I
El- i
E-
I
I
I
1
U
I
p0
IP:b10
'L
' IE-' E IEU I
U
I
I
L-'
-I I
iUE I
1:
IU
I
I
I-I
I
B
0 l
04UaU04IU
I u
a4U 04U a4 U
I U
I u
I U
I
E4
1 14E- 1 - I14 -4 p o
P
IU
I
nuE
P
I
04U M
IU
uI
14
14
I
P1
AuU) Iu E
nU EU cn u
14
E 14
'E-
14
u I0
1
14
'<4
1ZI
I U I
Iu
U01 U u 1 u0 U L)
1 H4f-4I
I
I
E-4
I
0
~ I E-I 1
u Iu I U IU IUp
0 0 PE
I
E-E,
m' -cr~ LA
%oN-
E-
4'E<
1
IEJ
.1
oOOOOOOOOOOOOOOOOOOOOOooooOOoOooooooooOO0oooo0
cooooooooooooooooooOoooooooooooooocooooooooo000
N N> C>
N N N N N N N N N N N N
N>
N N
N
N N
OD Ch
14
0 .-4 C')
M' V
-4-4-4
-4r-4r-4-4
N
N N
N
I
.1
1
B13
U Ei
1
U
E4I
I
N N
4
1
-4 #c 04
LA
LA %D N-
OD cm
N N
N
P
t EUI-
I
4E"-
1
r4 LA0r-40o
-I -4 -4v-4v-1
N N
0
.13
10 1
00
00
00
0O
0 00
0)0
-4 C4') C') .4 C4) e4 C4) rA C') -4 N' r4 "' r- LA
E
)4E
f --nOtE 4-E4-4
,-P4n4t
n4--4
%oN-
0
b
U I
OD ai 0 r-4 c') mf v LA
P
U
I I L) 10 I L)
L) u L I U I U L) L
.11UP
.1 JlLU)
911.
LU
6
E--I
1u
fU
IU I
u
uIUu IU I U I L)
r-4-4 -4 -I4 -4 -4-4-4
-1 -I 4-ý4-4A"I4-1-4-4
-4
O
O-i
I
II
LE I1 E1EE- IBi EUI
Ie IE
E
U
I u
I
L) IU
1 u
WE4 E-4
li.
0
>000 C)o00
C
0
00)
00
00
C'r-4
N
IN ' .- 4 C4) ,- N' ,4 C4)r-4 C') r
'-4 rC 0')
to -4444
En
E4
44--444g-44---44--444
fcw4E
0~4
**
I
Iu IU I u IU IU Iu I U Iu I
U
I
1u
UI
0
0 .- 4
I
I
En
m~
I
Ru
U
110
E-4
m~ qv LA% r-0
NOD
I
M 4
4J
0)-
I
I
0
04 U 04 U a4 U14U
1u
U
uI
U I
I4
U I0
0
I
-
I •,
1u
:I
4•J,
I
I
UBI
Il<
En
EU
1
Enu AuI
1u En U
• ut
uUluCI•u
:
UI
I :
1 uU:
IUu 1 guu:
I
UI
Iu I
:10<0: 'E<
E
IE
B U
4~~f
4I
0
P
4
s
Iu
E'
E
I
P 10 14
I uLI
10
U u-
P
I
I '
8aE0 4
§O
I'E4 1'064 IE4I E4
t3)
uI 10 1 uI I u 10I
1
0 14 1
R
01I
U0r4 U ',auauau,uauaua,
I I1 u I 104I 4I
u
I
14
S
I
-4
IE4 u
u
I
uI
0 ul
1 E- lu'ui
1 0 140 ulI1 Lu'u1
10 1
Iu
I uBI
4
E-1 E-1 'E-I
E- UI
E- II
L
1 Eu0 Ii u
U IIi
u I
I
PBIP
i E 1 E-4 L) L)
E- Ui
o< 4 IiUE4 1 EU4
IA
104 >
uII
P E-4 E- E, P EI
10 1 P
E-4
0 10
1 14 U
IB 1
0 I
I
4 14 14 I 14
-1IB10
-1E- -I E- .I I pF.3I 9
I3
U :0
E- Z
14
I
I E- E-4
1E-II
I 0 0I
p
E- E- viP E- L)
E-4 E- 0U
14j 1
0L)I uI 1,u Iu,1u II,
I
Ep I P<Iu8IM
U U014
UI
E-
of
°1oE-I I o,
EIR
1 .o
'P I,
,,
U
0
I Ui4E-4
I
i i UE-
•"
1o :r
14 :~
I •
10
I
4
I
1
I'
' 4: E1
, E-U
I
IE-'''
E-
I
9 14E-4 E-
1
UL1.
I
I EUI
I-
E-
U :IEuiIu
1. 1.
E
iE
IU
1 Ii B
IB 10
'
u
I
10E-4I
I
IP
0:4
I
1
14
E-
III
IL!
I
E-4 '.. 1 . 1 0
I
I
4
uIM
U
I
I0I1
1I
m
%D
k wk
N~N'LAn
- or
4m rr k r r
••
W•• 0)
0
0)
0
N' A c4
C'-4 c'4 .-i LA .-4 LA .4 0) A40 -4 0) -i40 -i c
r-4 r-4r-4 r-4 r-I
-4M -I M
l r-4-4
r4L
N .4i N
4 LA r-4
LA
tU IBE-
I
r-I
ý-
m~
%D
14
I IP I
II
I
EI
P
IEI
E-I I4
S
Ch
CID
i-LA
LA
LA
L LA
LA
LA Ln
%D
k.0
ýo
10
.D
O
0
••
40 ,-4C'N,-IN C') ,4 C%3.-4 C') ,-4 C13.- 4 C14`4 c') 4r-i
.*-4 64),-14
C' -4 c,4.- i
N ý.-I r -4r 4-4 T-4 r -4 T-4 r -4 rA-4 -4r- r-4
4 I r-4
.,-4
ý r-4
-4
.4 ,4
ýl-4 ,4r4-4
,-4 -4 r--4
1
0
0
0
-40 .)I 0> -40C>
En -4 N) ,40 LA
0)
i
4C')
0
-o~ LA %.DN- OD
-4-ý4-4-r4-r4--"4
N N
N
N N
0
40a 4
4 La g
N
N N
r-4-1
N
N N
a) TTr
- a%E-4 aH --iru U
o r-4> -i '
0'0 o4 (N LI) m
') ,-4Lfl'-4 0) l) U)E (n
U)n a1 a4
:
',E91 a
0:O<1'10
10 I
I
00000<
rE-
I
IU
1
E'
1)
El
Z
I
I Ul
o o
00
10E- I
1
E-14
1014
0,
II E
I S E-I4 E-4
H
10lE10~~~~
IU
I0
1 1 I 1
IEI
1
UI1
U
U
U
I
U0 1 U 104U
124UB
U14
U0
UB
am
0iJ
I UmImEUd Io
u
I-Hr)
zH I
E-4L2
P
l.
a
u
PC
UOE-1
10
I
0'
1
I
u
4U
-.
F
I
I
I
IUIUIUU
0)
0)oro
4U
b
u
0
-f
UII
4IU
I IP
Fe,0IB
I
E0
- I
1U
H
'~
I1
10
I
z1 0~~'
uP I
1b p
I:
Oct IB U
- 10U
p
U
IU
I
I
oc
iu4 U
m
0:0I
I
, p- ,E
HH
1U
IO4 U
L)
U
U
HEI1
2
IB
IU
U
m
w
0U P
U
U<
Ip
0 4U1 0 4U
U
rU
I KC1 EU)
KC
L0
4 4 W1 u Ll0P
l.
u
O
U-H
m
0 0
4 I
pII
)UUI P4 PU
IU
1IU1 0
pU EE-~:II~ E-E4
0~u pE
E-4
- Io4 ~ 14
E 1
IE4IEIE
~
S
E,1
I
1
UB1
1410
I 0B 9 10 1U0:IBIB
p-IB10
p
U10
I0 1I E-IIBU
Ip I p
IU
H0 1i
iHI
II
I
0-
E-)
4 1
1
10 :E.
10E-4
:
U L74~ 110 00:4
1
Fl.
IB
I- I
0
H:
Ii0
I U
UI
I
I U4
iE
E-'
0I E- p
UH 'E- I 0K0
1I
0IH IE H I1I4E-'
:H I -<
E-I0HU
UH I
U 14
I 4
UH 1 Hp
P4
E-IIE-4
1H-0 I14
U E<-.
1 U
I CU)
E- 440
U
I0Or'U
Uo-
I
U
1 0I
I
I•
'
IC
IUu
U
'OUIUi
c
0
> >
r_ -H -
' Eý'I
'
1 P I'
100
c
r-) :B
I
p
N N
-1-4H
40 tro
En
) E-4 r-4
U 41 4-1 0 0 .c
00000000
U
I
F:4
0Ii
II U0
EI I U0
I1
u
:I
••
ý4) J 000 ) )
0=
lIU
P
1
I~I~~
I
u,
-
i,0
4x I
E') 'P 1 E')
4 1
I P E,1
U
4
•
-
0 o •
10 :
I U
10
0:
1
0I
U
U
u
U0
000
IU
I U
U
H
U
-.
4-!I-Ix
H
o
u4o U,
E-4
I
o' c
i4 to I4
IUUIUH
U)
Up
IB
1I
r
U
u I E- H
4
•1
E E,10E
U
UIUIHI
U
U
'-1 '-l4 1- 4 0 HNOO
0 0
•mmm C
oo--mmooXXUU
•
•mam
oooo.•-
••
oPI
'U
IUiHbIHIlU
014
~IU P
;U
-
I IP4 U0
4 UHP
1U
IU4
4U
(-)
IU IU
I PI
Io4 U 1
UB
I
P
I
40)1
U
u a
U IU Iua IE
m
m
I
Iu
IU
I L) I u
I U I
4
IX PW
44P
4 PW
E0 P. P0I P I 00
F I E0I PL0
E
H1
1U
ZZUU
U
U
FI'
I
-
.CX04X<=F4
X4
<x04xX
<xK
E,4--44--444-P
44-P
444--44--E4-4
E
PH
Flo
:
0 D, 024 0' pc '0
P
-4-4E- OO I E,
-4
E0
0
E-4-4P
-r--
n LAC0
oO
0)rC - Om
010 40
C-4
14-V
0W
~~ ~ ~
LAr-4 LA
C)
CD
0
ý:r
v)riqrriLC
-
U)
WW
w
LA
k
m
r
w
C0
-4 o-4
(N
ý.
OD
0)
O
-A ,-i
~-IWW
())
m
14 .-r-4
.-
m
a r-4 LA
am4c
r-4 LA
co ,-1
,-4 u)
PQ -4
r-4M
-1 Ln
CO,-i
r-I LA
Co
LA-4
n
iLA
r-A
LA r-i
-n-
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
t
N
CN
-4
.-I
WWW
0)
L
I-l
N
(
k
.4
m
-4
r4
W
W
a r-i
. LA
-4
L)
LI
N
(N
I'D LA
m
Ln Lo
Cr"
,-q10 N 0(N
" ý4 " ý4 " ý4 (N
W
C) (1)
a)m
a) m a
-4 a)
C1 ,-1 Ul)A-4
(: .-ý-4
" -I i LA
U) r
i aCO
mA ,-4
a) r-A
- L)
Am
-A
L ,-1
ml
'U
'f
'U
'U
II
IU
li
'ci:
'U
ILI
'U
C0 a
ia)>,
UH4
W9o
Ul) -'-4
-4
Ii)
oý
rrIM
cN
4)4)(1
O(0
I:
0
a)ý
uLA
'o
1 4a
a
40
)Wa
(N
(n
ON
LA
LA
Ln
)
'IVm LA
CN W N ý4CN-4 (N
1) CI~)
() ma)OQa)mI)
ý
(N
0 .4 0
0o r-4
o o4
0D - M r-4 -4 C) or-4 ) ,-IoM ,-IM) r- a -IM -4 M) -4M) r- M -1 M -I M r-4C) r-AC) r-4( r-4m -4 L)A U,o' r-iLA r-4LA r-jwz -4LA r-4LA ,-iLA r-iLA -4 LA r-jLA r--4
LA r-iLA),- Li r-iLA r-4u) ,- L-4LA I)LA -4 LA r-i LA r-4 q -i mw r-I M~ r-ALmw -4
r-i (,4 m~ ý LA ,or
D
Ch
)c0
0 -4 (N mn -- LA , r- co mf0
0)
--
0 -
0
U)
OD
O)C4
rN
m
LA
,-0
-1
C> 0
-1
r-4
(N
(N
m
INCc m
k
4
(N
e
14
-4%.0
IawDC>
NOD
m~
,-i -4-I
.-4
.-4
1-4
r-4
1-1
'A I'D "Ch
4a
)WQ
4a
WV
4Q
(1
W
cW()
ci
a
ODD00 C
0)0
)0
00
0O
0
D0
) )0
)0
)
4r-I
A4
-I
- -4 v -4
OD000
00ODO00
)0
)
)0
)0
00
)0
~-i
(N m z
-4Nr4 C "-4
LA '0Nr-0
"
04(N
c)( 0D
N C
D 0 D000000000DO
D DO
0
)0
0)0)I'a m
)a
4 (N)
m T
MM
LA
C- C-(v
'0o N, (I
r- mm
D0000
0000 0)m
(mo00)
o
(o0m(m
0)IXD00
((00(
o)
Mr
0) ,-i
q
D00 00
0
c
c) -T Ln LA
r-j--V Z14-
00
00000
((00
0
Do aN
)
C
CN
m
N
-4
-4
CN
.r- a) c,4 a) m)c) Ln
.o -i r- .- i c) a) m) N o -i .- i mn (N
r LA
C%
CN
m
M
m
mc
Y)
M
"
0) C
:
=1wQw11wlI
lwQwC
w 9=
)
Q)
LA r-4Au
-i mA -4 mLA.4LA r-4 mA -in-4 LA
-4LA-4 mA r- LA -4 LA ,- L
ul
-4 -40
c) -4 0> -4 c> ,-i C> -1 C>
rn) -4 r' -i m r-.4 Mr'-4 M r-4 mn r-4 M r-4 M .- m .-q m~ r- m V-1 m r-A -.4 -f I4 r-4 qz ,-A N- r-4 N- r N
C0
M)
Ln
%D M
'A4
'K0
OD OD OD a)
(N
N
r-
Nt'A
a) c> a)
~ ~ ~
~
~
~-i
~ ~
~c)
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I-,
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I-,
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
La mnLo ONray)
ook
Cz)
m -r
k
a
c)
01)
a)
(1)
1)
01)
-4 (Z) -4 Cý) ,-4 c> r-4 C) ,-q C) ýi
-4 rN
i rN-4NH
r-4
NrN
-ý3 c)
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
U
E-4
I
i EI
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I,.I
I
Eý
I,.I
I
I
E-I
EI
U
EI
IEI
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
EI
EI
I
EI
U
I
I
I
'U
'U
I
I
I
I
'U
'F-'
IH
'H
F-'
I
I
'U
'F-'
'F-'
14
I
It-)
I
I
I
'U
'U
IO
'U
H
'U
'H
U
H
LN
(N
0Q)
LA r-4 LA r-4
M ý r' .- 1
~)
LA
'U
'U
'U
'F-'
'U
'H
'U
F-'
ltD
'F-'
'U
iH
'U
'U
'F-'
'F-'
ltD
(U
I
'H
'F-'
'F-'
'H
'H
'F-'
'F-'
'U
'U
'U
'U
ltD
'U
W
'U
'U
'H
I
I
I
'U
IU
(24-I
U-H
I
I
I
I
I
I
It-)
'$4:
E-
C)
LA
,-I
) %.0 m) N- a)
I
H
'
) a
(N
N
m)
(N
(N
C> m) .- 1 a)
m
m
m
NrI
-1
(NN
N3 m) m' a) LIA ,-1 %0 r-q N- .-- ao a
m
m
m
(l
m
m) " CD~ i
.-
m (N~m
LA -v c,
m
4)
r- LA r-4 LA -4 LA r- m ,-i LA -4 LA -4 LA r-i LA ,-i LA r-4 c) -4 D r-4 co -4 C)
m
-4 (' ,qm
-4 C') r-4 (n2 r-4r - m
r 4 -4 m m ,H ' m
r 4 rv-i r
.-I -14 qý
,-I4vr-4
N1
a•
•
'F-'
a)
-4 <D 1)
N- -4 N- r-4 N
r-40 C
••aoa~ooao
inHo 2o
'U
'F-'
'H
'U
'U
F-'
'F-'
',ci:
'H
'U
it,
'U
'F-'
'F-'
'U
'U
'F-'
iH
'H
'H
It,
LA m %.oo) N- m' a) LD c)
(1)
r-4
ci)
(1)
C)
-1
D
-4 N- r-4
N-
r-4
0)
C)
r-
(1)
C)
r-I N- rI Nl
-- A
,-
00
C)
0
C)
00 0 D C) C)
0 Co
C)
CD C)
C> 00oCoC
Ca
Co
) C)
C>
C OC)C0
C) C)
Co
) C) CDC)0(
C> a
C00 a)D 0a) 0C)
C)C)( C)O a)
C)(DD 0
CD C) 0
CD C) 0
C)
D r-i4
D (Nm
0O O O
M
OD
Nc) OD
ý-4( 0m 00
ý~ uA koD Na) OD
a)) p)
LA LA aO
N- OD
a) C) OD
r-4
" mw
ýV LA %D N- aC
OD a) C) .-) 00
C4(N0
LtA I'D a)
N OD 00
Ch C. 0
r-I "t LALA LALALALA LA LA)
%D ko %D kD LOHLD
%D
%L0
LA
LA N' N- N- N- N- N- N- N- N- N- a
OD O)
a)
00
OD 00 O) O) O
a
m mO
LO-:rLDonwn
oLALALn LA
a•a
000DC
OU
-'-- 0
Oj..-I
0'a
a
a
C 000> Co C>
00
•
• •a •
oHLDHLD
a
CCo 000
D
C)
C) C)C
C> C>C) C> )
C) C. C) C> C) C) C>) C) C> (
~
oaGo
C) C) CDC) C
C
)
)C)C>)
C
) CD C)0(D(
0
O
r-Ln
Ln
LA
Ln
q
LA
LA
00WO00W00
%D
CA
V-1
0
U LA LA
m
-W
ai
OD
N
%D%0
%D
%D
%DO
T-4
r-I
m
%D
%DD0
4
4
m
-
w0 ko
LAn
r % r
o rI'0r
r
000(300oW00o0
o00o0
0
W0
WoWCDWMQ
W
(
03 0)
(
0)
40 r-I "~ ,-I N' .4 N~ r-I 04 uI N~ r-4 C14r-I C14u-I CN u-I r4~ uI 04J u-1 N' r-4 N~ u C4~ ,4 "~ u-4 "N u-I LA -I wA u-I 0> u-I 0> u-I 0> u-I 0 r-4 0 u-4
( -4 rl-4 u-I r-I u-I r-I u-I u-I u-I r-4 u-I u- -IHn u-i u-i uI u-i
v-I r -Iý u-i u-i
n ui -i u-i u-i
n u-I
n u-i u-I
n u-I u-i
n u-i en r-i en ui LA u-i Ný r-i - uI LA r-i Ln u-I
En4 t p nn4C
4J
CD
r-LA
n
4E P
0D
LA
LA
)C)Wo)om
'0
L
LA
LA
n <CAO
LA
O%
LA
4En4
0
'0
O n
N
r-
-4
%0
)
'10
%D0
0
%D
O
'0
%D%'Dý
C
4 -4 r1 40
m
En4 E 40 4
(n
0'0-~
D
t
C0
m
r-
%D r-
N
r-
0
to-
momeoooooooooooooooNN•N
•Noooamm
mamooooom
r-4 C14
u-Io0
N r-4
N u-i% ýI C-4 r-4 0 r-I N'u-i('4 -i ('lui F4r-iu -I •
( -I 64u-4
c%4.I
c,4u-i 64lI LAr-i LAu-I
0>r-I 0) u-I0 q 0ou-I0 -I
r- -i LA r-ILA r-i
0 Ja4u-
mn v LA %o N- wD m~ 0 u-i m m -w LA '0 N- co (7% 0 u-i c4 m qv LA %oN
rOD
0h u>-I N' mn -w LA %D N- OD Ou 0> u-I ('4 mn q LA %0 N- OD
m~ ON (A O'h (AOu 0
000000000
)CýC>ClC>C
au-I u-I u-I u-I
r-4
v-4 u-4 u-I .- I ('4 ('4 0'(4 ('4
C%] N 4 ('4 C4'4
M' Mn Mn Mn en
M en en
M
v-1 u-I u-I *-I u-I u-i u-I u-I u-I u-I uI uI v-i .I uI r-i r-I -I u-4 u-I uI u-I .-I -4 u-I u-I .- I -1 u-4 -4 u-i u-I .- 1 u-I 1- uI u-I A-u-I
on
O000000000000000000000000000000000000000000000
O
0
O
0oOOOO
000DOOOD00DOOOOOOD00OOO
CD D D D D D D D D D D D D D D D D D D D D D D C
0-,0
44-J
Ný u -Iu-
r-4 r-4
u -4u-I u-
wo
woC
)
Cowo
wma)
mmwmowa
w Cý4
m
-I u-I T-4 u-I uI u-I u-4 uI u-I uI r-I r-I u-4 u-4 r-I r-I u-I u-I u-4 r-I u-I u-I u-I u-I Mnr-4 Mn -I
C C
C C
C C
C
)
000000000000000
C C
C
C C
0)
0)
4)
a)
LAu-I N- r-4
C
C
C C
C C
C
C C
C
r-A r-i4
co m
mammrw<
0)
a)) -V lrr
00V-4c>r-Ar o c" LnLn
cz r-i U, r-4 a) la) En En a4 a4
I
I
I
I
I
I
I
I
I
I
I
I
I
u. u EU)
- ýE E
(0 '1 H0
I
I
I
I
D0
H
)4ý4 4LJ4J
Lqn
,
U U 4-)~
X
ý4
--1moo
0
w
01
C\)OO
v -IT
rmmmmmro)oo-0.H.HOO
1CN Y )Ma
-4-r--r--
0 0
M
)
Q) a)
> >:
C00-H-
0 0
4
r-4
I
I
I
I E-I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
10
1E
I U
I
I
I
I
I
I
I
I
I
I
I
I
I
I
10
'U
SU
lCD
0
10Eý
1 E-4II
'U
I
E,
aI
I I
U'
12
U
I I
'U
'U
II I
u
r
UI
I OrI
u
I EAI
I
E-I
I I
U
C)I
u I
E-4
I I
u
u
'U
PI
'U
'U
'U
1<
12
E,
E-4
I I
I
'4:1
12
'u
E-
E-U
'C
'U
'U4
'F-
'U
8
u
'4:
I
oo-I
•
r
•
0
•
•
O
•
-ooo
•4
0000-
00000 (n 0000
0000r-r004- 00a)a)
-I- 0o 0'
-q
-44L
U
C
cI-.0.0
II0 0
(0 F-.
H E-4
H E-.
H- FH E-ý
H E-.
H X Oa
=
U UU
U_ 4-4
4- En
0 W)0.0-HH - > >
) C> 4)-V-CT0) in
o: r-4
A4
rQ) En c'-4
,-T
U -4 V0) ro
U) U,
a4. Ci.
0 4-)
0a0
0) 4
o) 0> r-i N
m q-w
.
CD~~~~~~~
,o
r-
w
mD c>
0-4
"' m
zr.
,o
r-
WN
CD
Co C)
U> C>C
U, U, U,>UU, U,) U,)CU, U,)C
-4
)CD
"H
In
,, Ln ,o
CD CD
U)
r-
m
U) U)
C, C
OU
'0
C) C) C> CD CD CD (D (D C) CD Co CD C) C> C) CD C) C> C) CD C. C) <D C> <D C) CD C. CDl <D
C>
qP .j LA
-i Ln -4j-z ,-iu
L
o
0
T1
0
0)
P UU
1
1
.I,
I
lu i
-1
1
U
I U
~ ~
IUl
l
PIi
a
u
U
U
U
1
S,
I
~aU
11
1c
P
i
I lucu
I
F
i
u
I
'p 'Ia.1
iu
l
Ii
i
i
I
U
~
0
~
_4~~
l r
q ý
i rIr
4r-4rI
4 ý4r4N
1C
I
U
UaUa
Iluau:
P
u
Ua
11
0
IO IU'U
I C);
Il
~
Um
Um
auua-44-
~
E- .11
a
P
a Papl E-
~
~~~
~
P
P
P
U
9
P >
P
:
U m a~ U m U m
a a
Ea
ua U P P P
0:U auaE-P
)0( P a ,UF4uUmItwU
4UMU(
UE-a P
E-- i
iu
Pu iE~
P
- A 'A U
-4 '
L
1A
04
Ir4r
-
LA
4 - IN
q Nrq 4-4 Ar4IN
IN
-
~
LA
LA
r-ar-IN
iE-4 aEP
UP
a a aD a a)
a a ra a a
a a a a a
aD a: a a a
-
-
-A
-4
00
j
~
E-4
i rI - - - 4VA ri
4ý r-4
m
4 0- v-4 i A r.a
1r|
a a a a a a a a a a a a a a
P
P
P -I P -I
Pal
: IU
a
IU
1aP- apP
u<
P
U m U m atx U m U m a m f< x a~ x 0a x
P <:
a P a
aaI 1P g a
aI 1U6aa iPa #a
4a
auiPaoz
4a
a a a u a ua
a
U
|
l
a a, a2 a a 2a a a a4 2a a a a
4 a a a a a 4 a
(a W a a) 4a a
) -a a a 1a a a a
ai ap aq a) a) a- a> a4 aQ a- aQra aI a--aQ a- aQ a- a a- a ai a a- a a- a a4a a4 a a a
aý aq a- an a- an a- aq a- ana- an a- a) a- ao a- an ai an a4 aI a- a) a- an a- an a- an a- a a
a) a) a a a a La:
a
a4 a at a a: aD aI a a: aI a a0 a aD aC a0 ac aD ac a a aD a4 aDa
a
a a a a a a a a a a a a a a a a
W
1
SaIU
0: 1 0
i
U IE0II
f l i 00
00 II00
~-4
,-H LA
r- m --A
E-
P U U
:CX
luaýI IU
I
LA
INC
a)ma)
o
1
U
IUIUIUIU
CUUa
a
IP
l f f 01l,lUl
00
it
P
wU EmU
au
I U
~~~~
al
a
a aa
a~~~~~~~
a a
U P u P aLP a4
ap
a< a a
ac a~
a a a
a En a (n a w a x aCx a x
P E-a : E-E4 a aP F:4a 1Ua L4I i Ea
a 1 aa
mU
l
a
Ia
a~~
a a
E-i
I U
em
a
m
-I
q
l 0
)
l a4ua
i 4uaf
aI
aa4
Ua4
a4u
uU
4Ua
U
a
a
Ua4U a4UaUaU0
0 UOZC00
a4PE- : U
U la4
U
uE()li
a4
U
Pa 1a4u#r!)1a4uUUP a4UU o41f-uaU a4
1UP
a4Ui
u
1r,
iU auEu a4
HuiI
ua4
|l4 U
i()| a4 UUa al U a4
iI UE)
I Ua I U I
U
U
Ca
a0 U
U
U
a
aaaaaU
a a a a a
a a a a a a a
a a I I I I a
a
aaaaaaaaaaaaaaa
aON
1
U 0:4
4UC14U al P
*
U
UUp4
u a4
Upa 00
: r.p a4L)
: uU,
a aI
U
a
I
u llUlua
I
a a
M C
Uaaai~~
I
F
I
a
Ma
IUIUIUIUIUUIIUIU
1~ l
m
U-
IN
U
UUFIUI
Q~U
I
fl
l
l
)r
r-i
r-4
Ua
3'
I~iUIUaUaU
l
uauf4 .1la C
IN
1
io a
E- U P
4
OD 0)
l
r4"T4cn
~-4
1--l- ,-I
14
i
4
(n
%10
IN
m
Dr-r40)
mA CDaL
CD-a
ýl" I "
m
,-I
r-i
r-lAI'D CAIN
Cq 4 " ý4
W a)
)a
am
a
aa
oD
a)
m
C)
m ý4
'I ml -4 EI r-Ica r-af
1 co ,-4CO r- COA
r- -4)
-4 "N1 -4 A
LA
.-i LA r-ian -i Ln r-ju)
- LnrA an r-aLA r-aan r- LA r-LA
u
-A LA - Ln r-4an .- Ln r-I,-ar-q m~ r-4a)
00
oC-4
0)
0l
ý-4
N
rIN
1
m 4m-4 m 14--4z LA
%0
N
a)
a)
)
()
)
C> ffA0> .4 0o r-I 0> r-lcmQ riC-I .a
r-aco
ko
P
U
P
>
1
m
a U CA
E-- 1P 4
a
iP
a-
al 1
a
a a
ýa a
ac a
L4ýA
P
R
E-iP
a a)
Q
a a
a a
a1 a
ca a
a a
a1 a
a a
x
1
E-
a a
a ra
ra A
ra a
IN
-4 LA
L-A -1LAr L-
Aj
-a04
ac
-A
0>
U) T-A
w wD o Nr-w
1w
m
ev4 c13 c4
0
o
Ce
0 40
~-4
o0>
ON
as
CN
No
~-im C'Vm"~ m~mrm,m m~ mm,T-4.-i%m -ir-It-m C3-i mwO r-wO am r-4
"~ 0 -4 ~-4
m c4 v. m m,
0;
%r
- m~ CO
w
m
ri w ýw w"w wvvv-w L
waU
a
Lmn-vn-v
%ow,40 %D
n>
Q
a
0 0
0
o
04
o 0
00)
Q)
o 0
o
m
o
o-
m
0m
>
.
fn
M -I U, r- U, -4 M, r-4 U, r-4U, r-4 M -4 U, r-4U, r-IU, -4 U, .-I 0w -4 0w r-40q r-q0- .-1 0- ,--4
0 -40ý
-l
0
- r-4
0-40r .-I 0
T-1 T~ -4 T~.-1 r-4 r-I -4
, -1 .I -4 T-44 r-44
, -4 4 L.
./)4
-n4 En 4 En 4- En 4 N) g- En
EN -4 N ) 4n
S I
I
U
I
I
I
I
U
P101
14
101
S0
IB
I
U
L
t
I
I I
I
II I
I
I
H
I
I
H
I
IU
IU
I I UC II i UI
i Et
II
I
1
II I
1
I1 I F
I
UI
I
1E10
IU
IU
t
I
I
1
1
IU
S
I10
II U
I II
III
I
I I:I
I II
I I I IUI
I UI I
Ab
HULUUA
I
I II
C
I
I
I
I-
I
Ii
II
I
I
I
UUO
I
10
1
I
I
U 01
0
10
10E
10
I )
10IU
1 0
IE1 r_ IEIU
'0
I E-4 10Ir_) I I
II u
U II
UI
E
I u
EI
I U
UI
I
i:I E
Ie
I
I1I UAUU
III
I 1 I
I ItI
1
I
I
1
t
iI:
It
I
1
I
I
IdI
I
U
I
10
II U
I U
U
0
CII1
II U
E-
II u I
II
I
I
I UI I OII I
LUAUA
•
O•I
I
I
E4,
4 i
0L)I I 1 IL)I
I
IUU
IU
I
ULUU
iII
E-
II
I
U
10
I 4
E
IU
I
I
10
E
101
I1IU
'0iE'E'IE-14
IE-
I:: i
q
I
r-4
I
I
I
0 I
E
0
I
I
I I I I
rEI
EE lZ4 I'
) I E- I0'q
I I
8 1 1 I I1 I I 141
1I 1
Z 4io E-II0'I 40Ef UIomoo~ooooouooooooooooooo~u~mnuo~mnuo••oso
ono
ooI
I
1 u0 I I
I I U0 10
I
I
I
I
I
I
I
U
I
I
I
IU
IUIU
IU
U
IU
II
IUI
I
IU
IU
IUI
I
u
I u I
II
I u
uI
I
u I
I
I
II
II U
IU
I
IE)
I 241#00
I~
I
I
I
S10 U I
I
I
0>
a)~
U,
14 a
4 U, .-I
iv-I
I-
0 4J
IIlI
I
D
I
1I
PI
I
I
I
I
I
I uI I
I U
I EII U E-
I
I 1I I1 I 1I I I 1I
T-4 IR
II
I
Ot4 I
N-
II
I
r-4
I
III
I
II II
I
I I
I
I >I
I
III
l
I tII
L) I
C1-4
m
r
I
U
~
I
II
U
I I
I I I I
I
II UI E-4
I U
IL %b1b
lIL
N-
IIIU IE
I II
-4
0,
V-4
I
I
I
I
IU
I
I
I
I1I
Il)
E- II
114
Il I
I I
I
I
I
I
CN
C14
6I)
I
IU
II
I
u
P
u
II
xý
B
I
I
II
I
I
IU
I
I
I
I
I
II
L)I
u
I
I
u
I
u
I1 4ICII
I
EII
14
I
I
I
I
II
II
I
U
I
u
U
I
E-4
I
I
U
I
u
I
1 4I
E14 Up IxI
I
I
I
I
I
r-I
)Q
4
)a
4
)Q
4
)Q
4
Q
0
0)a
0)
40l
0)
)
4)
0)
40) 0L)
0)()
U, r-i U, -i U, r-4 U, r-4 U, .i U, r-4U, r-4
U, r-4
U, r-i
U, r-i
U, r-i 0>ricýr4
0) r-I 0>
0) r-4 0 r4 0
I 0. 40
C)4 040 co ,4
.4c>ý
v-4
r1 -I
rA -4r
-Irl-4
T
r1 -4
,-4 V1
T-4 .- rI-4
1
m .4
-n4 in .4 tN 4 N .4 N g- En .4 tN gI N) 4 N -4 En 4
q-4
l
tm
L %r
D l
l
4 m ll
l
r
O 4l
i
I
I
I
I
rI
I
I
II
I
I
I
W -
DI I
II "
r- LA
O"TmC tom
LA
S(1
C), . C-
LA)
4 CN
-
LA
U•ov 0)- ao
r-"Tm
w1DLom . - omn mU
meý0' nV0
eo
LA
LA
LA
L%0
4-CN -4 CN-4-
C.4r-1r-4 r-4 N- -4 ,-N1-1 .-4
,--4N
'A
-
m
o
%'
e om
mr-m'
mm
M•4ry)
'
I'D
ko 1-4 ko
Lr-L
o
m
k
-
I 1-1N 1-
1 1-
I
E-I u
U IU I
I
I
'
I
UU aE-Um:U UE
aE'Ew4wZx
.
oPI xu
,'
E-4:•
IQ
UIU
uo
0
F-<a~a~a~aaa
0 0 E- D ::
IL)
u
lI
l i
I i U l
a~ U I U 'H
CUý
U P I E- E-ll
E-, I,
Ul
'H
a U
U
O
a E- a
0
z
Enu
lE-
i lI
Ul
u
au
E
u
~(nEa
H Ua-
I
E-
a a U
J
I Pa a ýa
o U~~
I ~ ~
a4 U a4 U• a4U14U04Ua
0
a
aD aD
%a
ý4 r4rI
I- rI r-
u
ua
Ua
E
waoo
wao
%a
E•••
.I
a a
ko
u
IaE-4II
PIIt
o:
Ua
a) D
N
aD
w
a
U
o<I
u
k
~
a)
D a)
C aa)o
" -4 " -4
- E-aE-P
E-
a
aD aD aDI'
aD aDI'aD
4 m
a1r4rIa
IE
I
i
Iý
I Oc4U04 U0a
c
0Ua a
Q
a)
a) a)
aq aa4
a-
a
IE
P4
a a Ea aý a
-1a- a- -A
a4 a-
I PIU
P P P E
I
I40"4Ua
f< I
•
a- a4 a- a-'1
a4
a. a4a- a-A
a4a--4
a a
a a
a
a a a a a a a
a a a a a a
a a a a
a a a a
UI I U
IE I
I I
E•
a Ea1
u:zx
140 1Ic
10,1
l<
~u
I'
xt
IU
aF-
-49a--a-9a(9a(IE(I
ao-
ýa
f<o u
E P 1 E,
E E, ,- ,
0:
P
E
4
vao
~
E-aE
'H
'UE--
E-I
S aD II D waD waD a) C>
a
C a) Ca
C) a) aD
a3 a) a
Ca1-1a 04 r- "
"a- aN"a4 aN a--a
-a
a-" -4a4"aq
r- "aq-4
aý a"
a- a- a- al "Iaa- 'Aa- '4 al
a- a- a- -1a4
ao
%D %D %0 ao%D%D
EU
-4
4 U a4 U
ýIAýli
I- I
vA vI l rI 4.3
E,
I•u
Poa
a E-a
lI
u
1
u4 uOctu ft um u
~
E-IE-
F0
c 1 4
EUIUIU
uauIUIU
l l iI U0
Il
U
I EI
E
'E-.
UU ua UU
LA
- , - -l
IE
H
10 ,0,1,0
4
~fc4
U
Iaua u
E-
•
°
U
-
(1) 0 (1CD0
)0
c--lC
C
u
Wuo04u
.
°
se
E-'
lgll
I UaI U
E-IUi E- E-II II
E
K
IU
4U
i U
u
,8 01C°•: I
0
40
IP
o
LLr-L.o r-r-
A-ac-c-C)
E
H IE-Eý
U,u o u
,
uu
,
C)
I U I E-4
IE
I
a)
)
4L-4 4L
I
SE
UU
.1 C1
m MMMqe0)MCo4r
%Lor-
)
Iq
- ri
a) a a) a )
4C - an
o- a-1C
-I
-I
n rA
a
a -4
waDI'al ko wao11 %DaD
a- -a4ý
a-
a- a-
al alrIr
Q
,--A --I ý4
M
-ODC>oD-A--mmoo
11
r- r-O
-
t
c)O~-i
XX
y-1t0,0
Lf
nm U U
)
1
4-4JJ 0 0 a) a)
rUuu
:
UEL-A--1 >
U U LJ 4J 0 0 .C C: -H H1
4a
al ZHHHZ
O,, UUU
.O
-I
-4-Om-
E-4E- E-cE-EH H H
4
&4
tr r-4un 1- (1) (1)M) CA a4 a-4 1) 1
W r0
M
:4X
M L10
4 L404 a
HH H
~Z
I
MER
X,OPX,ERX,
UE- I
U
i
E-
UUOct U OtU
XH U~~I04
E
u<1
.I
IH
1001
UHUHU-1--10101
1
-1 E-1
I E4
E-'
il
E
I
HPHIU
UI
I I:
U
UI II UI
U
U
U
U
I
UI
I
I
I
I
I
UI
I
I
HI-4H0u4;
u~u
1 pH
E-4
I
I
II
I I II
U
I
0II
I
I
I
I
i E- 1
4EUt/
I
I
I
I
I
I
I Q
E
:
II1
I I
I
I
I
I
I
0
I
III
I
I
I
I
I I
I0 I
II
I
II
<I
I I04I I
I
I-A
0
I0Li
I-rI I)
I
~ l
I- I) 1- "C
I) I)
I
U
I
Ii
Ut Q: 91
U Q
II
:
I
I
IU'1
I
I
I
1
I
-
I
UU :4
'H
1
I
lI i UI
~
I
:
lI
E-I i
I
EU~r,- ýr, 1I U E
U
I
II
II
II 1
0II
IU
0I
III
I
I SII
I
:)II IU I
I
p
0I 1010
if
U0110
I I. II
EIH
I
E IiP
I
I
I
I
I
U I IU I I U I'I
I U I UI I
I
U I IUIU
I UI I U IEI UI EI
I
E
I
I
UiI
E-4I
U I I CI
U2 I U I
C I E
E
IPI
I
I
UII 10 I
i
i
i
I
Q g
ý g Q
Q
Q IZ
< n g Q
$04I 1 IS 1 U I fI ::
IZD I, I I,
P
-Ha4HHHHHHHHHHHHHHHHHHHHHHHHHH
I t
I O
IDI I
IaiI
HHHHHHHHHHHH
I
'HF- U
I:
U I
I
I
I0
I
I Q
:
U U
4U OctU
I
I
4I II
I
~
I
H
0
H
<
UU
S0
U
U U
#C U
UIUUIUUIUUIUUIUUIUIUIUIUI
0
0
I
I F0
cH U
uH
1
u
II
I
I
908
1I0101
U
0100I0
101IC
pm
I
ý
U w w
a
U
I
I1
I
I
I I
I0II I 0I I II
E-I
I
I11
I I0II
I
1
r-
I
I
l
FI
I I
U
I
I
> xU
I
I
U
I
I I
I
I
I <
I
L
) a)
I
I
I
3
II-ljI
IDmI> mI> mI,mI m M C)
r- M
M MI4 MI -4
I4 rI IArI rI rI-rI
-4 HUPUBPUP
rI q l VA
H
40-,
m(D.a
C
CDo> %0
CDDCCCCD
10
k0%D
AJ
4- W D 0% l oA
I I0
I I
II I
I
I
I
I
I 0 I
I
10
I
I CI0HHHHHHHHHHHHHHHH
I 1 I
e) m co
I
1
I
,11U1I I
I
I
U
I
-' I 0 0
-A r- r-
I
4I
I4 I- I I
0 a) a)
IA l VA l Il r- ri I-U- rI rI I4 I- HI- rI1>U
I
-4>P
o
o
>DCk I'Dl >> W1 I'D
-v
4tA
r-uiAI'D
-n i-I
40
HPUH