J. Mol.
Biol.
(1982)
155,
517-531
Conformations
II.
Influence
I).
A.
(Received
of Pectins
of Residue Sequence on Chain Association
Calcium Pectate Gels
E.
POWELL,
Pnilever
and Interactions
R.
MORRIS.
M.
J.
GINEY
AXI)
I).
A.
in
REES
Research. Colworth Laboratory,
Sharnbrook
Bedford MK44 ILQ, England
19 May
1981, and in revised form
16 November
1981)
In the accompanying
paper
(Morris
et al., 1982) we present
evidence
of Ca’+induced
association
of poly-o-galacturonate
sequences
from pectin
into dimers
of
2, chain symmetry,
with co-operative
(“egg-box”)
binding
of Ca2+ on specific
sites
along
the interior
faces of each chain.
We now investigate
the role in calcium
pectate
gel networks
of other structural
features,
in particular
methyl
esterification
and 1,2-linked
L-rhamnosyl
residues
in the polymer
backbone.
Acid hydrolysis
of
citrus,
apple and sunflower
pectins
gave polygalacturonate
blocks with a relatively
narrow
molecular
weight
distribution,
and average
chainlength
of -25 residues
in
each case. Since the known
relative
stabilities
of glycosidic
linkages
would
lead to
chain cleavage
predominantly
at L-rhamnose,
this result indicates
that the length
of polygalacturonate
sequences
between
rhamnose
interruptions
is approximately
constant
within
and between
the pectins
studied.
Calcium
pectate
gel strength
is
reduced
dramatically
by the incorporation
of these chain segments
when they are
de-esterified,
but
not
when
they
are esterified.
This
interference
with
the
development
of a network
structure
that resists
applied
stress, provides
further
support
for our model of junction
zone formation
from sequences
of contiguous
deesterified
residues,
with
Cazf -mediated
chain
dimers
providing
the primary
associations
that can offer resistance
to deformation.
Samples
with
different
levels
and patterns
of esterification
were prepared
by
enzymic
(blockwise)
and chemical
(random)
de-esterification
of almost
fully methyl
esterified
pectin.
In the former
series, the extent
of Ca2+ binding
(as monitored
by
circular
dichroism)
increased
almost
linearly
with
the fraction
of free carboxyl
groups,
whereas
the latter
showed
a nonlinear
relationship
of a form consistent
with the requirement
of this binding
for blocks of contiguous
non-esterified
residues
and, in the presence
of excess univalent
cations,
binding
was negligible
when more
-40%
of the carboxyl
groups
were esterified.
Statistical
calculations
of
than
sequence
length
distribution
at different
degrees
of random
de-esterification
show
the best fit with experimental
data when binding
is assumed
to require
sequences
with seven or more consecutive
free carboxyl
groups
along the participating
face of
the chain.
For 2, chain
symmetry,
this corresponds
to a sequence
length
of 14
residues,
in excellent
agreement
with previous
independent
studies
of Ca’+ binding
to oligogalacturonates.
In the absence
of competing
univalent
counterions,
circular
dichroism
changes
o
are similar
in form but so large in magnitude
that site-binding
of Ca*+ mustriow
Ca f+
beyond
the half-stoichiometry
at which
it is arrested
in their
presence.
517
(M,Z2-2836/82/080517-15
$02.00/O
C 1982
Academic
Press Inc.
(London)
Ltd
518
I). A PO\VEl,l.
ET AL.
binding monitored by circular dichroism. and gel strength (yield stress) measured
mechanically. both show a similar dependence on the I)attern as well as the Ipvel of
esterification.
as expected for network formation by co-operative binding of (*a*+
within interchain junction zones.
To fit this binding data cluant,itatively,
it is necessary to postulate a two-stage
process. (1) Initial
dimerization,
probably
corresponding
to the “stjrong
associations”
indicated by evidence from rompetitive
inhibition
(see above). for
which a critical minimum sequence of seven residues is again required but rsterified
residues can now be accommodated
within individual sites provided that they are
paired with a free carboxylate on the complementary
chain. (2) Subsequent (‘a’+induced aggregation of these preformed dimrrs, which ran occur irrespective of the
pattern of esterification
on
the external faces; the evidence from mechanical
measurements shows that these contribute little to gel strength a.t high stress.
1. Introduction
Studies of the gelation
behariour
of pol~vsacc+arides
i~r vitro. after extraction
and
purification,
have proved
a valuable
and experimentally
convenient
route to
molecular
understanding
of their structural
role ir, viw (see. for example.
Rees,
1977: Rees & Welsh, 1977: Morris et nl., 1977: Stockton
et al., 1980). A unifying
principle
that has now emerged
from investigation
of many differerlt
gelling
polysaccharide
systems is that interchain
association
occurs by the formation
of
extended,
conformationally
regular “junction
zones”.
We have shown (Morris
et nl., 1978) that the primary
event in the calciurninduced gelation
of alginate,
t.he principal
polysaccharide
component.
of marine
brown algae (Phaeophyeeae).
is dimerization
of poly-L-guluronate
chain sequences.
These pack as buckled, 2-fold ribbons (Mackie.
1971) with polar cavities or pockets.
which provide an array of regularly
spaced, specific sites for cation chelation
(Grant
et al.. 1973). In view of the obvious analogies,
we have adopted the term “egg-box”
binding to denote cation-mediated
interchain
associations
of this type. Pectin is an
important
constituent
of the cell wall of higher plants, and is based, like alginate.
on a linear
1,4-linked
polyuronate
backbone,
in this case of n-u-galacturonate.
The
poly-u-galacturonate
sequences
of pectin and the poly-L-guluronate
sequences
from alginate are almost exact mirror images, and we have presented
evidence in
the accompanying
paper (Morris et al.. 1982) that Ohis structural
parallel extends to
their association
with calcium, including
closely analogous
2, egg-box geometry.
Although
the formation
of stable
int.ermolecular
junctions
is a critical
requirement
for gelation,
some limitation
of t,he extent of interchain
association
is
also necessary to give a hydrated
network
rat)her than an insoluble
precipitate.
In
the case of alginate,
for example.
dimeric
poly-L-guluronate
junction
zones are
terminated
by the occurrence
in the primary
sequence of n-mannuronate,
either in
homopolymeric
blocks, or in mixed sequences with L-guluronate
(Smidsrod.
1974:
Morris
et al., 1978). Our principal
objective
in the present
work has been to
investigate
covalent
features
that have a similar
solubilizing
function
in pertic
polysaccharides.
Pectin? unlike alginate,
is based on a single uranic acid residue. and this ma)
occur either as the free salt, or as the methyl ester. We have therefore explored
the
effect, on gelation
behaviour
of the extent. and pattern
of esterification.
Other
(‘.dLCIl’M
I’ECTXTE
GELS
519
st,ruct,ttrat
feat)ures
with
clear implications
for the restriction
of interchain
association
are side-chains
of neutral sugar residues, and also the occurrettce of 1.2.
linked
L-rhamnosyl
residues
as covalent.
insertions
in the polymer
backbone
1965: Gould
rt al.. 1965; AAspitiall
et ~1..
(Xspinatt,
1970: Zitko
6 Bishop,
1967.1968).
It is not known whether
these rhamttosyl
insertions
are 1 or p-linked.
Ilut computer
model-building
calrulat,iotts
(Rees $ Wight. 1971 ) show that’ both art
incompatible
wit,h regular
conformations
of Ftol?r-D-galacturottate,
and wouttl
therefore act as junction
delimiting
“kinks”.
analogous
to those identified
in t-he
agar and carrageenatt
series (Rees, 1969.1972).
It. should be emphasized
thatt 1 :‘tlinked rhamnosyl
residues do not. of course, have a universal
kinking
futtction.
Thus. in one of the Klebsieh polpsaccharides
(KS serotype),
where such residues
has
form part of t,he regular repeating
sequence. an ordered chain conformation
lieen characterized
in the solid state (Atkins rt crl., 1979), while an analogous
1.2.
linked
I)-mattttosgl
residue
that forms part of t,he petttasaccharide
repeating
structure
of xattthatt
is essential for adopt,iott
of the ordered
conformation,
b\
allowing
side-chains
to fold compactly
alottg the polymer
backbone
(Moorhouse
rl (11.. 1977). Ln the specific case of pectin, however. incorporation
of such residues
within
ordered
conformations
of the fl(‘lSgala~turotlate
chain
is stericatt)
impossible.
not because of any general
or ittherent propert,y of 1 .&linked
rhamttost
(cf. Xt.kitts et txl.. 1979). but. because t.tte linkage is di#~~nt
from that in the rest of
the Ilolymer
backbone.
In t,his work we have investigated
the distribution
of
rhatnnose kinks along the polymer backbone.
and have compared
our results kvitlt
the known critical sequence length (Kahn, 1975) for the formation
of stable calcium
E”lt~galacturottat,e
interchain
junctions.
,4 prelimittar~
account, of part of this work
has been published
elsewhere
(Gidley st al., 1979).
2. Experimental
The
followitq
samples were used: I. 11 from Bulmers Ltd.. Hereford. U.K.. wvert
unstandardized
citrus pect.itts and had degrees of esterification
(fraction
of urotta.tr
residues that are methyl esterified) of 36’?,, and 7 l?lh, respectively ; 111 (“Brown ribbon“). at1
apple ptvtitt from Obipectitt, contained 72 oh ester: IV, V were sunflower pectins (kindly
donated
sutttlower
by Dr B. de Vries),
and had ester contents
of 440/; and 53%,
respectively.
Thra
pectins
were supplied
pure as freeze-dried
powders.
Commercial
pectin
samples
w(arp precipitated
from aqueous solution with ethanol,
washed
with
acidified
aqueous
ethanol
(5(?;, (v/v)
HCI in 60%
(v/v)
aqueous
ethanol)
and dialysed
extensively
against
deionized
water
before
being freeze-dried.
E&r and urottate values were determined either
by ittfra-red
spectroscopy
(Bociek
& Welti,
1975) or by saponification
and titration.
(II)
Prepardion
qf pectin
hlockn
Pectin
(1 p) was dissolved
in deionized
wat,er (80 ml) and diluted
with hydrochloric
acid to
give a final concentration
of about
I(+<, (u/v)
polysaccharide
in 05 M-acid.
The solution
was
heated
on a boiling
waterbath
for 3 h : a precipitate
normally
appeared
after about
1 h. The
hydrolvsate
was cooled, centrifuged
and. after neutralization.
soluble
and insoluble
fractions
bvere dtalysed
separately
and freeze-dried.
Blocks were fractionated
on a column
of Sephadex
(XX) elut,ed with deionized water, and were typically
obtained
in art overall
yield of -(iP,,
520
Il.
A.
POWELL
(see Table
1). For the competitive
inhibition
were
used;
after
de-esterification
these
chromatography
columns
was monitored
phenol/sulphuric
acid assay (Dubois
ft al.,
(c) Preparation
ET
AI,
experiment,,
both
samples
behaved
by optical
rotation
1956).
and
soluble
and insoluble
identically.
Elution
measurement,s
or
blocks
from
by the
USC of pectinestrrasc
Pectinesterase
was extracted
from
oranges,
essentially
according
to the method
of
McDonnell
et al. (1945),
and purified
by chromatography
on DEAE-cellulose
(Datunashvili
et al., 1976). Enzyme
activity
was detected
in column
fractions
by the following
assay:
1 ml
samples
were added to a solution
(1 ml : 1 sb, w/v) of pectin of high ester content
(samples
II
or III).
The solution
was then
adjusted
to pH 7.0, and allowed
to stand
at ambient
temperature
(-25°C).
Release
of free carboxyl
groups
was detected
by a drop
in pH,
typically
within
5 to 10 min, and monitored
by titration
with
NaOH
(01 M). The purified
enzyme
was found
to be low in pectin chain splitting
activity,
as monitored
by measurement
of the intrinsic
viscosity
of the de-esterified
samples
(capillary
viscometer:
@2 M-NapI).
Samples
having
different
levels of ester subst,itution
were prepared
enzymicallg
from
pectin with a degree of esterification
of 920’ ,0 (prepared
from pectin 11 as below).
Pectin
(1.5 g)
in 0.1 M-NaCl
was adjusted
to pH 7.0 and mixed
with
100 ml of a solution
containing
pectinesterase
(40 mg) in @I M-NaCl
at pH 7.0, and digested
at 25°C. The release
of free
carboxyl
groups
was monit,ored
by continuous
Ctration
with @l iv-NaOH.
At, suitable
times,
samples
were removed
and quenched
by precipitation
of the enzyme
with an equal volume
of
10%
(w/v)
trichloroacetic
acid.
The
samples
were
filtered.
dialysed
and freeze-dried.
Although
de-est,erification
occurred
more rapidly
at higher
pH, the reaction
was carried
out
at neutrality
to avoid
concurrent
base-catalpsed
saponification.
(d) Alkalinr
dr-astrr<Jicatiorr
of p&in
This method
was used to prepare
ester-free
pectin
blocks
for t,he competitive
inhibition
experiment,
and polymeric
samples
having
different
levels
of ester arranged
in random
distribution.
The conditions
were chosen
to minimize
chain
cleavage
by fl-elimination.
An
@59;
(w/v)
aqueous
solution
of pectin
was maintained
at 1°C’ on an ice/salt
bath,
and
adjusted
to pH 12.0 with 2 M-NaOH.
The pH was maintained
by the periodic
addition
of
alkali
and. after
suitable
times,
samples
were removed.
quenched
in acid,
dialvsed
and
freeze-dried.
Intrinsic
viscosity
measurements
showed
that chain cleavage
was mmimal.
Gel strengths
were measured
on cylindrical
plugs of gel (height
12 mm, diam.
12.5 mm)
using an Instron
Universal
Materials
Tester.
The method
used for preparation
of gels by slog
release of (:a’+
from an insoluble
salt. was typically
bv addition
of a solut,ion
of citric
acid
(0.446 g in 10 ml water)
with
rapid
stirring
to a solution
(40 ml) containing
pectin
(0.5 g).
sodium
citrate
(@44 g) and calcium
orthophosphate
(0.143 g). The mixture
was poured
immediately
into moulds,
and aged for at least 12 h before
measurement.
An alternative
method
of gelation
was by mixing
a solution
of pectin
(300 mg in 20 m1) at. 8O”f’ with
a
solution
of calcium
chloride
(132 mg of the dihydrate
in 10 ml) at the same temperature
: gels
were again formed
in cylindrical
moulds
and aged for at least 12 h. The latter
method
was
used only at pH 3 and below:
at more
alkaline
pH values,
a gel formed
as soon
as the
solutions
were mixed.
When gels were made for rompetitive
inhibition
studies,
pectin blocks
were added
with the pectin.
(‘ircular
dichroism
(Morris
et al., 1982).
measurements
The neutral
sugar
of solution
content,
and gels were made
as described
of pectin
samples
was determined
before
by gas-
(‘ALCIVM
PECT.4TE
GELS
*521
liquid chromatography
analysis of alditol acetates, using an internal standard (Bjiirndal
of nl., 1970), after hydrolysis with sulphuric acid (0.25 M, lOO”(‘, 16 h). The chain length of
pectin blocks was estimated from the molar ratio of reducing end groups, determined by t.he
method of Park & ,Johnson (1949), to the total uronate content determined
by infra-red
spectroscopy (Bociek Sr Welti. 1975). All samples were de-esterified before end group analysis,
to avoid possible p-elimination
under the alkaline conditions used in the assay procedure.
Highly esterified (tvpically 84-950;,) pectins were prepared by treatment of pectin II (12.7 g)
\vith 1 M-methanoiir
sulphuric acid for 70 h at 2°C”. Samples were recovered by washing
ethanol
and ether.
soquent,ially u ith NF;, ( r / L-) a q ueous ethanol until neutral, absolute
3. Acid Hydrolysis
of Pectin
When
pectins
(samples I to V) were exposed to strong mineral
acid (@25 MH,SO,) at lOO”(’ they were initially
soluble, but material
was precipitated
after one
to t)wo hours. After three hours. the mixtures
were cooled, and separated
by
c+ent,t=ifugation into soluble and insoluble
fractions,
both of which had high uronate
cont)ent. low levels of neutral
sugars, and were almost fully de-esterified.
the
residual degree of esterification
being in the range 0 to 5% in all cases.
The soluble fractions
from all samples studied
showed
essentially
identical
profiles on column chromatography
(Sephadex
G-SO), and eluted as a single, sharp
included peak (Fig. l), indicating
a narrow range of molecular
weight. A sharp peak
in the same position was observed as t,he major hydrolysis
product in the insoluble
fractions.
but some material
of higher molecular
weight was also present (Fig. 1).
\I:r at)tribute
the separation
into soluble and insoluble
fractions to precipitation
of
this higher molecular
weight material,
with some incorporation
of shorter chains
within t’he aggregate structure,
while t)he major hydrolysis
product is of sufficiently
low molrcular
weight to remain soluble even when fully de-esterified.
Results fol
Froctlon
number
PI<:. 1. Sephadex
G-50 fractionation
of (0) neutralized
acid insoluble
and (a) soluble blocks from
pectin I. The column (400 mm x 26 mm) was held at 20°C. and eluted with distilled water at a flow-rate
of 9 ml/h. Column loading was -0.4 g of dialysed and freeze-dried
hydrolysate.
Fractions
of 4-5 ml each
were collected, and monitored
by optical rotation
(578 nm: 10 cm pathlength).
V, z 80 ml: V, z 330 ml:
elution
volume at band centre ~175 ml. Similar
elution
profiles were obtained
for the soluble and
,nsolnble fractions
of all pectins studied.
522
D.
A. POWELL
TABLE
Percentage
Pectin
yields
ET
B.C.
1
of polygalacturolbate
blocks on hydrolysis
Soluble
Insoluble
blocks
of pectins
blocks
both yield and degree of polymerization
were consistent
from run to run, and a
similar fragmentation
pattern was observed for all the samples studied
(Table 1).
Glycosidic
linkages of uranic acid residues are known to be largely resistant
to
acid hydrolysis
at pH values sufficiently
low to induce protonation
of the carboxy
groups
(Lindberg
et al., 1975). Polyuronate
chain sequences
of pectin
would
therefore
be expected to remain essentially
intact under such conditions.
Neutral
sugar linkages
and, in particular.
those of furanose
sugars or 6-deoxy
hexoses,
however,
are considerably
more labile. The observed hydrolytic
scission of pectin
would therefore suggest removal of neutral sugar side-chains,
and cleavage at the Lrhamnosyl
insertions
that are known
to occur in the polymer
backbone
(Gould
et al., 1965; Aspinall
et al., 1967,196s). Although
acidic hydrolysis
cannot.
of
course, be absolutely
specific, and the presence of some residual material
of higher
molecular
weight indicates
incomplete
scission of rhamnosyl
linkages. nonetheless
the narrow range of chainlengths
in the major hydrolysis
product,
and its isolation
in comparatively
high yield (typically
-700/;, of the insoluble
fraction,
and -85’j/,
of the total
hydrolysis
product),
would
suggest
that
the uninterrupted
polygalacturonate
chain sequences have a relatively
narrow distribution
of lengths
in the native polymer.
Estimation
of the chainlength
of this major product, by assay for reducing
endgroups (Park & Johnson,
1949) gave an average degree of polymerization
of -35
Since t,he latter
using galacturonic
acid as standard.
and -25 using L-rhamnose.
seems the more appropriate
standard,
we conclude that the regular, uninterrupted
sequence length in all the pectin samples studied is approximately
25 residues.
4. Competitive
Inhibition
of Pectin Gelation
Competitive
inhibition
is widely
used to study the specificity
of biological
recognition
events, for example,
enzyme-substrate
or lectirr-hapten
binding.
This
approach
has been extended
to characterization
of chain-chain
interactions
in
regular
chain
gelling
polysaccharides
(Morris
et al., 1980). using structurally
segments to occupy binding
sites on intact polymer
chains, and thus inhibit
the
formation
of a crosslinked
network.
Thus, for example,
when poly-r-guluronate
blocks,
isolated
by partial
hydrolysis
of alginate,
are added
to gelling
concentrations
of the parent molecule,
calcium-induced
gelation is inhibited,
while
poly-n-mannuronate
and heteropolymeric
mixed
blocks
have little
effect, in
C’ALC’TUM
Concentratm
PIG. 2. Effect
of added
pectate
blocks
(slow release
of Ca* + ) : (0)
(yelation
from heated
solution).
blocks
84%
PECTATE
of added
523
GELS
blocks
(X,
w/v)
on calcium
pectate
(1 O/o (w/v) gel strength).
esterified
blocks
(slow release
of C’a* + ) : (A)
(0)
De-esterified
de-esterified
blocks
agreement
with previous evidence that cation binding
and inter-chain
association
are largely restricted
to polyguluronate
chain sequences. We have therefore
used
this approach
to probe the primary
structural
requirements
for junction
zone
formation
in calcium pectate gels.
I’oly-r-galacturonate
blocks, prepared
by controlled
hydrolysis
of pectin, were
de-esterified
by mild saponification,
and purified by gel filtration.
Two alternative
gelation
procedures
were used. The first of these involved
addition
of an insoluble
calcium salt to the polysaccharide
solution,
and subsequent
controlled
release of
calcium ions by addition
of acid. In the second method, calcium chloride was added
to the polysaccharide
solution at 8O”C, and gelation occurred on cooling to ambient
temperature.
Gels prepared
by the first technique
were consistently
weaker,
but,
this is probably
due to inclusion
of air, or mechanical
disruption
of the incipient
network.
during the final mixing step when acid is added. In both experiments,
the
presence of polygalacturonate
blocks had an inhibitory
effect on gel strength.
measured as the stress required
to rupture
the sample (yield stress). Similar
large
reductions
were
observed
in the maximum
deformation
that
could
be
accommodated
by the network
before failure
(breaking
strain).
As shown
in
Pigure 2, the reduction
in gel strength
increases progressively
with increasing
concentration
of polygalacturonate
chain segments.
until network
structure
is
almost completely
lost at a block concentration
approximately
half that of the
intact polymer.
The presence of esterified
blocks (84% of uronate residues methyl
esterified),
by contrast,
was found (Fig. 2) to have no significant
effect on gel
strength.
5. Effect of Degree and Pattern of Esterification
Studies of Ca2+ activity
degrees of polymerization
in the presence of polygalacturonate
chains of different
(Kohn. 1975: Kohn & Luknar,
1977) indicate a critical
524
I>.
A.
PO\VEI,I,
E7’
.4/,.
chainlength
requirement
of about 11 residues for co-operative
binding.
For the
longer chains studied here. this need for a critical sequence lengt,h should be seen for
junction
zone formation
in calcium-induced
gelation.
What
is not obvious.
however.
is whether
esterified
residues can be accommodated
within the junctions.
or whether
they will act as jllrlction-termirlatillg
structural
features. in the same
way as kinking
residues. To explore this further,
UY have examined
the gelation
behaviour
of pectin samples n-it,h both different
levels and different
patSterns of
methyl esterification.
and under different
conditions
of ionic euvironmr~tlt
Since thr distribut,ion
of esterified
residues in nat,ural pectins will bc det.ermined
both by partial
saponification
during
the rxtraction
procedure
(whicah xvi11 bcb
essent,ially random)
and by the pre-existing
distribution
in tlirv (which
is probAt\.
syst,ematir).
we first. prepared
material
t.ha.t was almost
c~omplctely
methyl
rsterified
( -9Os,,) and then partially
de-&erified
using both chemical and enzymic,
methods. Alkaline
hydrolysis
is known t,o remove ester groups in a random manner.
and will thus gilye pectins wit.h an a.pproxima.tely
stat.istical
distribution
of free
carboxyl
groups. 13~ contrast,
pectinesterase
initiates
its a&ion adjacent
to a free
rarboxyl
group, and proceeds along the chain in a stepwise
fashion.
producing
blocks of free carboxyls
(Kohn rf cl/.. 196%: Itexova-Bcnkova
& Markovic,.
1978).
Starting
from highly esterified
material,
therefore.
the enzq’tnic
method
should
yield pectins in which blocks of est.erified residues alt.rrnat~e Lvith blocks of fret
galact,uronate.
The characterist.ic
circular
dichroism
chmgcs
((irant
et (1.1.. 1973: Morris et c/l..
1973 : Bryce rt ul.. 1974) that. accompany
calcium pectate gelation arv seen onIF for
co-operative
binding
of Ca* + . and not, for simple electrostatic
int,eractions
with
polygalaeturonate
chain sequences that are too short for co-operatirity
(Kohn KT
Sticzay,
1977; Bytrick?
rt al... 1979). We have therefore
used this technique
to
quantify
the extent of junction
zone formation
in our various pectin samples. a,nd
structural
requirements
for stable
interchain
hence to identify
the primary
association.
Ln t,he ac~companying
paper
(Morris
vt (I/.. 19X2) lye argued
that calciurnmediated
association
of unesterified
polygalacturonatt
chains in thp presence of
swamping
levels of competing
univalent
ca.tions is limited
to t,hc formation
of 2,
chain dimers.
Under
these ionic conditions.
the extent
of (‘a’+
binding.
as
monitored
by circular
dichroism
change
(Fig. 3). decreases
dramat.icalty
with
increasing
content of esterified
residues randomly
distributed
aloug the polymer
chain. The form oft he observed dependence
of (‘a* + binding capacit)y WI degree of
random
esterification
suggest,s that the format’ion
of stable interchain
junctiotl
zones requires participation
of uninterrupted
sequences of free carboxyl
groups. To
test and quantify
this hypothesis.
we have compared
our experimental
results with
st,atistical
calculations
of sequencr length distribution.
If the fraction
of esterified
galacturonat~e
residues isJ’. then the rjrobability
of u
norl-esterified
residues occurring
consecutively
along one face of the polymer chaiu
is (t -f)“.
As shown in Figure 4, a value of IL = 7 closely matches the obse~~~~etl
dependence.
This corresponds
to a minimum
vhaiulength
of II residues in a chaiu of
2, symmetry,
in excellent
agreement
with t,he t,hreshold value of 12 to 16 residues
determined
for co-operat,ive
binding
of (‘a’ + by galacturonate
oligomers
in free
(‘ALCIITM
PE(‘TATE
Degree
of esterhotlon
GELS
525
(%I
FIN:.
3. Effect
of degree
and pattern
of esterification
on the calcium
binding
capacity
of pectin.
as
monitored
by circular
dichroism.
(0)
Enzymically
de-esterified
pectin:
free availability
of Ca’+ : (0)
alkaline
de-esterified
pectin;
free availability
of Ca’+ : (A) alkaline
de-esterifed
pectin
; Ca* + binding
in
competition
with 0.5 M-Xa+,
In the latter
2 cases. the continuous
lines show the best fit obtained
from
comparison
of the primary
sequence
requirements
for co-operative
binding
of Ca*+
postulated
in the
text. and statistical
calculations
of the distribution
of esterified
residues.
cd.. circular
dichroism.
solut,ion (Kohn.
1975). consistent
with dimerization
of 2, helices as previously
proposed
(Morris et al.. 1982). These results indicate
that co-operative
binding
of
(‘a*+ in competition
with excess univalent
counterion
requires a minimum
critical
sequence length of seven unesterified
galacturonate
residues, and is terminated
b>
the occurrence
in the primary
structure
of an esterified
residue on either of the
participating
chain faces.
l’nder
more normal gelation
conditions,
where Cazf is present as the sole OI
principal
counterion,
the extent of Ca*+ binding
to unesterified
polygalacturonate,
as monit’ored
by circular
dichroism
(Fig. 3). is approximately
twice as great as in
the competitive
situation
discussed above. We attribute
this to extensive
dimerr
dimer
aggregation,
with carboxyl
groups
along both sides of each 2, chain
participating
in cation chelation,
rather than only those on the interior
surfaces of
isolated dimers. thus doubling
overall Pa’+ binding
capacity.
In the absence of
excess amounts of competing
urlivalent
cat.ions, gels prepared
from pectin samples
of low ester content are highly turbid.
This is reflected in the increased scatter of
our circular
dichroism
data for these systems (Fig. 3). and provides
further
evidence of extensive
aggregation.
I’nder
these conditions
of ionic environment.
Ca2+ binding
capacity
shows
(Fig. 3) a sigmoidal
dependence
on the degree
of random esterification,
with little
reduction
in the magnitude
of circular dichroism
change until approximately
onethird of t,he galacturonate
residues are esterified.
From this we conclude (1) that
there is some limited tolerance for esterified residues within
the ordered, interchain
,junctions.
a.nd (2) that ester groups incorporated
in this way experience
the same
526
1).
A.
POWELL
ET
A/,
200
0
Fractm
of non-esterifled
residues
( f)
Fit:. 4. Variation
in cat,ion
binding
to t)ol~galacturorlatr
with degree
of random
esterification.
bound
calcium
that
is resistant
to displacement
by swamping
levels
of monovalent
counterion
Sa’),
as monitored
by circular
dichroism
(r,d.)
change
(a)
between
solution
and gel (see Morris
1982) is compared
with
the statistical
probability
of occurlp~~cc
of 6 (- - -), 7()or8(---,-)
consecutive
nowesterified
residues
along one chain
face. for a wage
of methyl
polygalacturonatcs.
The
(@5 Met crl..
electronic
perturbations
from the r~lf’orcw~
proximit,.;
of site-bound
calcium ions. to
make the same molar contribut,ion
as free carboxyl
groups to the overall magnitude
of circular
dichroism
change.
To explore further the primary
seyuence
requirements
for the formation
of stable
interchain
junctions
in the absence of excess univalent
couuterions,
we have again
attempted
to match
t’he observed
dependence
of the magnitude
of circular
dichroism
change on the degree of random
esterification
by calculation
of the
probability
of occurrence
of specific dist,ributions
of esteritied
and unesterified
residues. This is clearly a more complex analysis than the previous calculations
of
the distribution
of consecutive
unesterified
residues. and a number
of seemingly
attract,ive
models were tested, but failed to reproduce
the observed dependence.
.A
very good fit (Fig. 3) was obtained.
howrvrr.
on the basis of the following
simple
postulates.
(1)
(2)
A minimum
critical sequence length is again required
for co-operativity.
In the initial dimerization
process.
calcium ions may occupy egg-box nest)s in
which one or both of the participating
carboxyl
groups are unesterified.
but
,jurlctiorl-zone
formation
is t#erminated
when rsterified
residues
on both
chains coincide.
(‘AL(‘II:M
(3)
PE(‘TATE
Subsequent
extensive
dimer-dimer
residue sequence along the outer
527
GELS
aggregation
faces.
occurs
irrespective
of the
If we again denote the fraction of esterified galacturonate
asf, and consider the
growth of a pre-existing
dimer, then the probability
that the next residue on both
participating
chains is esterified
(thus terminating
the junction)
is f2. Hence the
probability
of any other combination,
to extend the junction,
is (1 -f2).
Since the
initial residue pairing is a function
of the relative positioning
of the chains, rather
than their primary
sequence, the probability
of occurrence
of a dimeric junction
zone of IL residues along each chain face and with no disallowed
coincidence
of ester,
groups
is (1 -f’)“-‘.
Subsequent
association
of all such dimers
into large
aggregates
will then increase the fraction of residues participating
in (:a’+ binding
t.0 2(1 -f’)“V
with the proviso that the maximum
value can never, of course.
exceed unity.
(This proviso
takes account
of the fact that at low levels of
esterification
the outer faces of dimeric
junctions
may also satisfy the sequence
criteria
for inner faces.) In terms of this model. the best fit (Fig. 3) to the
experimental
data is obtained
for TL = 7, as previously
found for (:a’+ binding
in
competition
with excess univalent
cations. Although
success in matching
observed
results in terms of a particular
model does not, of course, establish unequivocally
the validity of the model, we regard the excellent agreement
of the critical sequence
length
calculated
for both
sets of ionic conditions
studied
with
previous
independent
investigations
(Kohn.
1975) as additional
compelling
evidence
in
support of t,he proposed interpretation.
For pectin
samples
prepared
by enzymic
(blockwise)
de-esteritication.
the
magnitude
of circular
dichroism
change
011
addition
of Cazf
(Fig. 3) was
approximately
proportional
to the fraction of free carboxyl
groups, confirming
the
need for consecutive
un-esterified
galacturonate
residues. We further conclude that
the polygalacturonate
chain sequences form egg-box junctions
in the same way as
isolated polygalacturonate
blocks. and the fully esterified sequences take no part in
Degree
and
FIG. 5. Gel strength
as a function
(0)
alkaline
de-ester&d
pectin.
of degree
of esterification
of esterification
b&o)
for
(e)
enzymically
de-esterified
pectin.
528
D.
A.
POWELL
ET -4L.
the binding mechanism.
By this circular dichroism
criterion,
the extent of junction
zone formation
in block (enzymically)
de-esterified
pectin samples is greater t,han in
randomly
de-esterified
materials
when the degree of esterification
exceeds - 154,,
To test this conclusion.
we have compared
the mechanical
strength
(yield stress)
of calcium pectate gels prepared
from materials
of both types. The threshold
degree
of est,erilication
for the formation
of gels with a measurable
yield stress (Fig. 5) is
for “block
de-esterified”
pect,ins. and - 50% for randomly
-SD/b
de-esterified
samples. For the former series, yield stress increases approximately
linearly
wit’11
decreasing
ester level below this threshold
value. while the latter show much
greater sensitivity
to reduct,ion of ester content, below 5OS,. Xs shown in Figure 5.
yield stress measurements
indicate greater interchain
association
for the block deesterified samples at degrees of esterification
above -15?;,,
in good agreement
wit’h
the circular dichroism
evidence.
6. General Discussion
In the accompanying
paper (Morris
et al.. 1982) we present evidence that’ the
primary
mechanism
of calcium-induced
association
of poly-o-galacturonate
chain
sequences is by dimerization,
closely analogous
to that previously
demonstrated
(Morris
et al.. 1978) for poly-L-guluronate
sequences of alginat,e.
In this respect.
pectin and alginate
may therefore
be regarded
as courlterparts
in certain land and
sea plants, respectively.
III the present work, we have shown that this analogy
extends to competitive
inhibition
of network
formation
by chain segments identical
in chemical structure
to the principal
binding regions of the parent molecules (polyu-galacturonate
and poly-r-guluronate.
respectively).
In both cases. the major
fur&on
of other chain sequences present in t,he molecule (totally or predominantly
esterified
regions
in pectin.
and homopolymeric
or heteropolymeric
regions
involving
o-mannuronate
in alginate)
appears to be in solubilization
and hydration
of the network,
and these sequences
show no such competitive
inhibition
of
interchain
association.
It has been shown (see Cook & Stoddart,
1973) that at an early stage in the
biosynthet,ic
pathway
pectin exists in an essentially
fully esterified
form. and the
pectin t,hat is most readily extracted
from mature tissue also typically
shows high
degrees of esterification
( -709/,).
The presence in plant tissue (Gould et 01.. 1965:
Rees & Wight.
1969: Cook & Stoddart.
1973) of pectic material
that can be
or by chemical
degradation.
extracted
only by the use of Ca2+ sequestrants,
however,
indicates
that. some chains have undergone
substantial
de-esterification.
which
is believed
to occur at a later stage of biosynthesis.
by the action of
pectinesterase.
and would give rise to a blockwise
arrangement
of free carboxy
groups. Our present results explain
why this product
would
indeed form tight.
(‘a2+-mediated
associations
in the plant
tissue.
Biosynthesis
of structural
polysaccharides
as soluble
precursors
that, are subsequently
converted
to the
structure-forming
species by enzymic
modification
at the polymer
level has also
been demonstrated
for the agar (Rees, 1961), carrageenan
(Lawson
& Rees, 1970)
and alginate
(Larsen
& Haug,
1971; Haug & Larsen,
1971) families,
and may
represent
a common mechanism
for biological
regulation
of tissue structure.
(‘AL(‘JI’M
PEC’TATE
GELS
529
In both alginate
and pectin. selective binding
of calcium in competition
with
much higher concentrations
of univalent
cations is terminated
by the occurrence
in the primary
sequence of residues typical
of the sotubilizing
stretches
(methyl
gatacturonate.
or D-matttturonate.
respectively).
While such conditions
are relevant
to the external
ionic environment
in oi~o for alginate,
the same is not true for
pectin. Therefore.
it is perhaps significant
t,hat, white mattnuronate
is incompatible
with interchain
association
even in the absence of competing
counterions.
it
appears from our present results that some degree of methyl esterification
can then
be accommodated
within calcium pectate juttctiott
zones. This may be as high as
about, a third of the participating
residues. provided
t,hat they are distribut,rd
randomly
along the chain. rather than grouped
in homopolymeric
blocks.
The covalent
feature
in the primary
structure
of pectin that exerts the same
absolute delimiting
effect’ on junctiott-zone
length in pectin as I)-mattnuronatt
does
in alpinat,e is the insertion
in the polymer
backbone
of 1 .&linked
L-rhamttosyt
residues. which have been shown by computer
model-building
(Rees 8 Wight.
197 1 ) to
be incompatible
with
inclusion
in
conformationally
regular
F”)t?lgata~turortate
junctions.
In this work. we have shown that the spacing of these
is such
that the tbngth
of
kinking
residues
along
the polymer
backbone
uttinterrupt.ed
sequences of n-galacturottate
(free salt or ester) does not vary widely
bvithin or between t,he samples we have examined.
and is around 25 residues. (‘CC
operat.ive binding of (“a’+, irrespective
of ionic erivirotitnetit.
appears to require the
ittvolvemettt
of at least seven residues along each of t’he participating
chain faces.
bvhirh for Z-fold symmetry
corresponds
to a critical
sequence length of about 11
residues. Thus. any appreciable
reduct~ion
in the spacing of rhamttose
insertions
would he likely to comprotnise
the stability
of ittt,erchain
association.
white wider
spacing would decrease the tot,al number of junctions.
The natural dist.ributiott
of
rhamttose kinks itt pectin thus represent’s an efficient (possible optimum)
balanc~c
brtwern
the number
of ittterchaitt
associations
and their stability.
Such periodic
strrtc~turat irregularities
in the polymer backbone might also offer art explanation
of
t)hr pronounced
structural
periodieity
visualized
directly in electron micrographs
of
pect,itr stained \\ith rut,hrttium
red (Hanke B Northcote.
1975).
For l)oth algitta.te
(Morris
rt 01.. 1!)7X) and pectin (Morris
rf (rl.. 198:!), the
concentration
of competing
mortovalrttt
countcriotts
required
to rest,rict ittterchaitr
associa,tion to egg-box dimers is 0.5 %I. Under these conditions.
t,he critical primar?
strucatural
requirement
for the formation
of’ stable calcium pectate junction
zones
apl)ears to be seven consecutive
free earboxyt
groups along the interior
faces of
each of the participating
chains. The same crit,ical srquenre
length appears to
persist in the presence of (ia’+ as the sole or principal
counterion.
but in t,his case
t~strrified residues can be accommodated
within t’he dimer structure.
provided
that
thr corresponding
residue on the other chain is not also esterified.
Our results
suggest that under these ionic conditions
extensive ditner-dimer
aggregation
ttta?
o(~~tr. irrespecbtive of the residue sequence along the exterior
faces. This lack of
specificity.
and ease of disruption
by competing
cations,
indicates
that the
associa,tiotts btxtween dimers are far weaker and more t,ettuous than those bet,ween
participating
chains within dimers. Further support for this conclusion
comes from
the abilit’y
of short chain segments
to sigttificant,ly
reduce the strength
of gel
530
D.
A. POWELL
ET
AL.
networks,
even under
ionic conditions
where there is evidence
of substantial
aggregation,
since only interactions
of the latter type will be inhibited
(Morris et al.,
1980). We may therefore
regard the calcium
pectate gel structure
as a strong,
cohesive network
cross-linked
by dimeric egg-box junctions
between
unesterified
chain faces which, under appropriate
ionic conditions,
may be reinforced
by weaker
associations
of partially
esterified
chain sequences to form additional,
less stable
dimers, and by dimer-dimer
aggregation.
Under the experimental
conditions
used
in this work, fully or predominantly
esterified
chain sequences appear to have no
structural
role other than as interconnecting,
solubilizing
stretches
between
junctions,
although
the gelation
of pectins of high ester content at low pH values
and reduced water activity
is well-known.
and utilized
commercially.
In summary,
the pectin chains in our samples contain blocks of n-galacturonate
approximately
25 residues in length, joined together by sequences that contain 1.2.
linked L-rhamnosyl
insertions.
Since the geometry
of these anomalous
residues is
incompatible
with
the regular
2-fold
conformation
required
for interchain
association
in egg-box junctions,
they act as an invariant
restriction
on
chain
packing,
to prevent
total
insolubility.
The extent
and pattern
of methyl
esterification
exerts
a more subtle,
regulatory
function,
which
may be of
importance
(Rees. 1969) for differentiation,
growth
and consolidation
of tissue
structure.
We thank
providing
the
Mr E. J. Murray
samples
for experimental
of sunflower
assistance,
and
Dr
B. de Vries
for
kindly
pectin.
REFERENCES
Aspinall,
G. 0. (1970).
Polysaccharides,
Pergamon
Press, Oxford.
Aspinall,
G. O., Cottrell,
1. W., Egan,
S. V., Morrison,
I. M. & Whyte,
J. N. C. (1967).
J.
Chem.
Sot. ser. C, 1071-1080.
Aspinall,
G. O., Craig,
J. W. T. & Whyte,
J. L. (1968).
Carbohdr.
Res. 7, 442-452.
Atkins,
E. D. T., Isaac,
D. H. & Elloway,
H. F. (1979).
In Microbial
Polysuccharides
and
Polysacchurases
(Berkeley,
R. C. W., Gooday.
G. W. & Ellwood,
D. C., eds), pp. 161~
189, Academic
Press, London.
Bjiirndal,
H., Hellerqvist,
C. G., Lindberg,
B. & Svenson!
S. (1970).
Angew.
Chem. Znt. Ed.
Engl. 9, 610-619.
Bociek,
S. M. & Welti,
D. (1975).
Carbohydr.
Res. 42, 217-226.
Bryce,
T. .4., McKinnon,
A. A., Morris,
E. R., Rees, D. A. & Thorn,
D. (1974).
Faraday
Lhcuss.
Chem. Sot. 57, 221-229.
Bystricki,
S., Kohn, R. & Sticzay, T. (1979). Collect. Czechoslovac.
Chem. Commun.
44, 167173.
Cook.
G. M. W. & Stoddart,
R. W. (1973).
Surface
Carbohydrates
of the Eukaryotic
Cell,
Academic
Press, London.
Datunashvili,
E. K’., Tyurina,
S. S. & Kardash,
N. LJ. (1976).
Appl.
Biochem.
Microbial.
12,
36-39.
Dubois,
M., Gillies,
K. A., Hamilton,
J. K., Rebers,
I’. A. & Smith,
F. (1956).
ilnal.
Chem. 28.
350-356.
Gidley,
M. J., Morris,
E. R., Murray,
E. *J., Powell.
D. A. B Rees, D. A. (1979). J. Chem. Sot.
Chem
Commun.
990-992.
Gould,
S. E. B., Rees. D. A., Richardson,
N. G. & Steele. 1. W. (1965). LVature (London),
208,
876-878.
C*AL(‘IUM
(:rant,
G. T.,
Morris.
E. R.,
Rees,
PECTATE
D. A., Smith,
GELS
P. J. C. & Thorn,
531
D. (1973).
FEBS
Letters.
32. 195198.
Hanke.
D. E. & Northcote,
D. H. (1975).
Biopolymers,
14, 1-17.
Haug.
A. B Larsen.
B. (1971).
Biochim.
Biophys.
Acta. 192, 557-559.
Kohn.
R. (1975).
Pure Appt.
Chem. 42. 371-397.
Kohn.
R. & Luknar,
0. (1977).
Collect. Czechostovac.
Chem. Commun.
42, 731-744.
Kohn,
R. 8r Sticzay,
T. (1977).
Collect. Czechoslovac.
Chem. Commun.
42, 2372-2378.
Kohn.
R., Furda,
I. XL Kopec,
Z. (1968).
Collect.
Czechoslovac.
Chem. Commun.
33, 264-269.
Larsen.
B. & Haug,
A. (197 1). Carbohydr.
Res. 20, 225-232.
Lawson.
C’. J. & Rees, D. A. (1970).
,Vat~re
(London).
227. 392393.
Lindbcrp.
B.. Lonngren,
*J. & \,‘vesson,
S. (1975).
=Idvan.
Carhohydr.
Chem. 31, lSeF-235.
JlcDonnell.
L. It.. Jansen,
E. F. & Lineweaver,
H. (1945).
rlrch.
Bioch,em.
6, 3899401.
hfackie.
W’. (1971).
Biochem.
J. 125, 89Y.
Moorhouse,
It..
Walkinshaw.
X. D. & Arnott,
S. (1977).
In E:ctracellular
j!Iicrobiu./
Polysaccharidrs,
llCS Symp.
ser (Sandford,
P. A. & Laskin,
A., eds), vol. 45, pp. 90-102.
AC‘S, W’ashington.
Morris.
E. R., Rees, D. A. 8r Thorn,
D. (1973).
J. Chem. Sot. Chem. Gommun.,
245-246.
Morris.
E. R., Rers, D. A.. Thorn,
D. & Welsh,
E. ,J. (1977).
J. Supramol.
Struct.
6, 259-274.
Morris,
E. R.. Rers, D. A., Thorn,
D. & Boyd.
J. (1978).
Carbohydr.
Res. 66, 145-154.
Morris.
E. R.. Revs. D. A., Robinson:
G. & Young,
G. i\. (1980).
J. Mol. Biol. 138, 363-374.
JIorris,
E. R., Powell,
D. A.. Gidley,
M. J. & Rees, D. A. (1982). J. Mol. Biol. 153, 507-516.
Park.
.J. T. 8~ Johnson,
M. J. (1949). J. Biol. Chem. 181, 149-151.
Rres. D. A (1961).
Biochem.
J. 81, 347-352.
Rem I). A., (1969). Advan,. Carhohydr.
Chem. Biochem.
24, 267-332.
Reea. I). A. (1972). Biochem.
J. 126, 2577273.
kes.
D. A. (1977).
Polysaccharide
ShapcCs, Chapman
& Hall, London.
Rres. D. A. 8: Welsh.
E. J. (1977).
=Inqcrc, Chem. Int. Ed. Engl. 16, 214-224.
Rcrs. 1). -4. & Wight.
A. W. (1971).
J. Chem. Sot. ser. B, 1366-1371.
Rexova-Benkova.
L. & Markovic,
0. (1976).
Advan.
Carhohydr.
Chrm.
33, 323-385.
Smidsrod.
0. (1974).
Faraday
Discuss.
Chem. Sot. 57, 263-286.
St,ocbkt,on,
B., Evans,
L. V., Morris.
E. R.: Powell,
D. A. bt Rees, D. A. (1960).
Botanica
Marina
Zitko.
23 563-567.
c. T. (1965).
V. & ‘Bishop,
Canad.
J. Chem.
Edited
by
43, 32063214.
V Luzznti