Directed Evolution
Try to direct evolution (on shorter time scales) to produce new biologically based
materials.
Take advantage of the power of mutagenesis and screen/selection to engineer new
biomolecules.
Rational design of new proteins, etc are limited by our incomplete knowledge of
relations between sequence, structure and function.
Directed Evolution is an alternative and also a complementary approach.
Diverse applications:
Industrial
Biomedical
Basic Research
Examples of goals:
Improve existing enzymes to function under different environmental conditions
(pH, temperature, etc.)
Create new enzymes to catalyze specific reactions
Industrial enzymes used in over 500 products from detergents to beer making
Advantages of enzymes over conventional catalysts:
Renewable resource
Biodegradable
High selectivity
Work under mild conditions
Examples include replacement of phosphates in laundry detergents with
proteases and cellulases, replacement of emulsifiers with lipases in bread
making, replace sodium hydroxide with amylases and pectinases.
Can also create proteins that specifically bind target ligands (diagnostic
applications, biosensors, templates for self-assembly, …)
To implement directed evolution, must have a means to link
genotype and phenotype. Some methods for linking genotype and
phenotype:
SELEX with DNA or RNA
Phage Display
Ribosome Display
Direct screening or selecting of cells
When some or all of the steps are carried out in vitro then often called
in vitro evolution
Earliest example (that I am aware of)
was from 1967, PNAS 58:217-224
Sol Spiegelman and colaborators:
“An extracellular darwinian experiment with a self-duplicating nucleic
acid molecule”
Evolved Q replicase – an RNA dependent RNA polymerase
“What will happen to the RNA molecules if the only demand made on them is the
Biliblical injuction, multiply, with the biological proviso that they do so as rapidly
as possible?”
Start with the RNA genome of phage Q.
Note, random RNA fragments cannot replicate by Q.
Add replicase, incubate 20min
Remove small amount of RNA product and add to new tube, add replicase and
incubate 20min
repeat.
After 74th serial transfer 83% of original genome was eliminated. Replicates 15
times faster than original RNA.
SELEX (Systematic Evolution of Ligands by Exponential Enrichment)
– a particularly simple means to link phenotype and genotype
e.g. select for functional DNA molecules  genotype and phenotype
are directly linked to the DNA
start with random library of DNA sequences and select for those that bind substrate
(“aptamers” - “apt” - fit)
can be used to find the best DNA sequence that binds a DNA binding protein
can also be used to create diagnostic tools.
Similar procedures can be performed with RNA - in fact for applications RNA is often
more popular. The extra hydroxyl on RNA can make it easier to bind, also provides
greater chemical reactivity.
Note - with RNA or DNA libraries can start with an enormous number of distinct molecules
(32-mer  432  1019 - in fact it would be difficult to really have quite this level of diversity
in your library but even something several orders of magnitude smaller is still enormous.
Can also screen or select for enzymatic activity: ribozymes
Now need to be a little more clever in designing the selection or screen
If want a riboszyme that cleaves a particular chemical bond in response to a stimulus
(e.g. pH) then can construct a library that is immobilized via this bond, apply stimulus and
collect whatever comes out of column.
Example of Selex application:
find DNA sequences that will bind to spores of B. anthracis (anthrax)
www.arches.uga.edu/~mfield/selex.htm
Conjugate one set of
aptamers to microscopic
magnetic beads and another
set to a reporter enzyme.
If spores are present then
magnet will pull out reporter
enzyme (e.g. chemiluminescent signal)
Phage Display
Display peptide or protein on surface of bacterial virus
(in principle can use other viruses but phage viruses easiest to prepare etc.)
Some proteins on viral coats can accommodate peptides or proteins and
will present them on the surface.
The phage genome (or alternatively phagemid) contains the sequence for the
protein or peptide so isolation of the phage with desired phenotype will also
provide the genotype.
Most popular is filamentous phage f1 or M13.
pIII on the end or pVIII along the length of the rod-like virion
for pVIII ~10% can be loaded with alternate peptide
Advantage of phage display: easy to screen over 109 sequences
Can either clone library directly into phage genome
or use a phagemid (plasmid that contains f1 ori) with replication
deficient helper phage
5 copies of pIII and pVI 2800 copies pVIII - all can accommodate
peptides
http://www.biochem.unizh.ch/plueckthun/teaching/Teaching_slide_shows/filamentous_phages/index.htm
www.neb.com/nebecomm/products/productE8100.asp
Phage library selected for binding anti -endorphin antibody
with NEB phage display library
www.neb.com/nebecomm/products/productE8100.asp
http://www.unizh.ch/~pluckth/slide_shows/Slides/ribo/
Can also express libraries directly in bacteria or yeast and
screen or select for desired phenotype - all depends on having a good screen
or selection.
e.g. yeast surface display. Boder and Wittrup used expressed antibody library
against fluorescein on the surface of yeast. Screened for fluorescein binding by
flow cytometry. Achieved (femptomolar binding affinity!)
Can screen for improved green fluorescent proteins by simply
monitoring fluorescence of cells expressing libraries of mutants and
selecting the brightest cells.
http://cheme.che.caltech.edu/groups/fha/
Yeast surface display (available from Invitrogen)
fusion to a-aglutinin
http://www.invitrogen.com/content.cfm?pageid=3458
How to generate libraries?
can use any mutagenesis technique - e.g. chemical mutagens, mutator strains,
UV light, site directed mutagenesis with randomized oligonucleotides,
mutagenic pcr, DNA shuffling, …
Error Prone PCR
Type of mutation Number times observed
e.g. substitute Mn for Mg
in reaction buffer, 5-fold
excess of dTTP and dCTP.
AT and TA
34
GA and CT
26
AG and TC
24
AC and TG
6
GC and CG
5
GT and CA
2
Template length
EP-PCR
doublings
Mutations per nucleotide
position
5
100
bp
200
bp
400
bp
800
bp
1600
bp
0.0033
0.33
0.66
1.3
2.6
5.3
10
0.0066
0.66
1.3
2.6
5.3
11
20
0.013
1.3
2.6
5.3
11
21
30
0.020
2.0
4.0
7.9
16
32
50
0.033
3.3
6.6
13
26
53
Current Protocols in Molecular Biology, 2004
DNA Shuffling of family of genes
to evolve class C cephalosporinase to moxalactam resistance
Crameri, et al. 1998
With directed evolution you get what you select for!
Very often the system will find a way to solve the challenge you pose in a way
that you have not anticipated.
Example:
Saggio and Laufer, 1993
Used phage displaying a library of peptides to look for binding to a
rat mono-clonal antibody against nicotinic acetylcholine receptor.
Antibody was conjugated to biotin which was then immobilized on a surface
bound with avidin. They ended up selecting for peptides that bind to biotin instead
of to the antibody.
Note, with all of the techniques we have discussed there is an amplification step
between rounds of selection
(Selex – pcr; phage display – growth of phage; direct selection of cells – cell
growth) so you may end up selecting for what is most easily amplified, not what
best satisfies your selection. Selections are never white or black!