High throughput gene synthesis and cloning of polyketide synthase modules

1
High throughput gene synthesis and cloning of
polyketide synthase modules

• Kosan Biosciences
• Sarah Reisinger

2
Kosan Business

• High value
• pharmaceuticals

Technology platform
polyketide alteration production
3
What Are Polyketides?
Product
Company
Therapeutic Area
Azithromycin
Pfizer
Antibacterial
Clarithromycin
Abbott
Erythromycin
Abbott, others
Josamycin
Yamanouchi
Minocycline (Dynacil)
Wyeth-Ayerst
Miokamycin
Meiji Seika
Mycinamicin
Asahi
Oleandomycin
Pfizer
Pseudomonic acid
SmithKline Beecham
Rifamycins (Rifampin)
Novartis, Lepetit
Rokitamycin (Ricamycin)
Asahi
Tetracyclines
Pfizer, Wyeth-Ayerst
Aclarubicin (aclacinomycin)
Bristol-Myers Squibb
Anticancer
Adriamycin (Doxorubicin)
Pharmacia-Upjohn
Chromomycin
Takeda
Daunorubicin
Astra, Chiron
Enediynes
Wyeth-Ayerst
Idarubicin (Idamycin)
Pharmacia-Upjohn
Amphotericin B
Bristol-Myers Squibb
Antifungal
Candicidin
Hoechst Marion Roussel
Griseofulvin
Schering, Wyeth-Ayerst, Ortho
Nystatin/Mycostatin
Bristol-Myers Squibb, others
Spiramycin
Rhône-Poulenc
Mevacor (Lovastatin)
Merck
Cholesterol-lowering
Mevastatin (Compactin)
Sankyo
Pravastatin
Sankyo, Bristol-Myers Squibb
Zocor
Merck
Schering
Zearalenone
Ascomycin (Immunomycin)
Merck
Immunosuppressant
FK506
Fujisawa
Sirolimus (Rapamycin)
Wyeth-Ayerst
Insecticide
Spinosad
Dow Elanco
Avermectin
Merck
Veterinary Med
Lasalocid A
Hoffman LaRoche
Milbemycin
Sankyo
Monensin
Lilly
Tylosin
Lilly
4
Polyketides Defined

• 10,000 known polyketides
• Produced by soil micro-organisms
• (actinomycetes myxobacterial)
• Diverse, complex structures
• Produced by modular enzymes
• Similar precursors, similar mechanisms
• Each 2 carbon atoms encoded by DNA sequence

5
Polypeptide - Polyketide Analogy
DNA sequence (5,000 bp module)
enzyme module
PK 2-carbon unit
Change DNA sequence ? Change PK structure
6
Polyketide Synthesis
module 3
module 4
module 1
module 2
PKS Gene Cluster
Assembly-line blueprint
PolyKetide Synthase (PKS)
The assembly-line
The raw materials
The polyketide product
7
Change Module to Change Structure
module 3
module 4
module 1
module 2
module 3
PKS Gene Cluster
PKS
Polyketide
2-carbon building blocks
8
Change Module to Change Structure
module 3
module 4
module 1
module 2
module 3
PKS Gene Cluster
PKS
Novel Polyketide
Polyketide
2-carbon building blocks
9
Morphing

• In theory, could sew PKS modules together to make
  any or many polyketides
• In practice, difficult to obtain functional PKS
  module interactions

10
Morphing Objectives

• Learn how to connect PKS modules from different
  PKS gene clusters to make any or many polyketides

11
Morphing Toolbox

• Objectives
• Develop a library of modules to express in
  genetic host
• Connect modules in all permutations
• Determine which module sets produce products
• Learn how to correct inefficient module sets

12
Develop a Library of Modules

• Possibilities
• Natural modules
• Pros
• Already exist
• Cons
• Requires isolated genes
• High GC content possible expression problems
• No convenient restriction sites

• Synthetic genes
• Pros
• Control of GC content fewer expression
• problems
• Designer restriction sites simple to
• mobilize module/domains
• Cons
• Huge effort to create synthetic genes
• (100 modules 500 kbp)

13
High Throughput Gene Synthesis
14
Objective

• To develop a fully automated process to
  quickly and efficiently synthesize and engineer
  large PKS.

Output Synthetic Gene of Interest
Input Gene Sequence
Gene Design
Synthesis
15
Module Gene Design

• Develop a system for generating synthetic PKS
  modules that allows for
• Codon optimization for expression in E. coli
• Common restriction sites at module and domain
  edges
• Additional restriction sites within modules to
  facilitate partial domain or module
  swaps/replacements

16
Module Gene Design
Generic design for 200 known modules
?identified conserved regions for engineering
restriction sites between domains within
modules
17
Software Automation

• Developed suite of tools for gene synthesis
  design and analysis
• Synthetic gene design
• Split gene into smaller parts, codon optimize,
  restriction sites
• Oligo design/specificity testing/order
• Automation input information
• Sequence analysis
• Database

18
Gene Morphing System (GeMS)
User selected Restriction enzymes, Distance
between sites, Fragment size
Protein/DNA sequence
Input

1. Codon optimization
2. Restriction site insertion/deletion
3. Oligo design and testing

Design validation
Output
Oligo ordering file Automation files for oligo
mixing and cloning
  http//software.kosan.com/GeMS
19
Gene Synthesis Fragment Generation
Input Oligo components of 500 bp synthons

• Distribution of individual oligos to gene
  synthesis wells
• Gene synthesis
• Clone into vector
• Transformation into E. coli
• Isolation of colonies
• DNA sequencing

Output 500 bp synthons in plasmids with correct
sequence
20
Flow Chart of Synthesis
21
Gene Synthesis
40mer oligos
22
Generation of Synthetic Fragment
U-U-U
23
HTP Cloning

• Criteria
• Purification of PCR products unnecessary
• High efficiency
• Amenable to HTP automation

24
HTP Cloning UDG Cloning
5-UXUXUX
UXUXUX-5
PCR
AXAXAX
5-UXUXUX
UXUXUX-5
AXAXAX
UDG
AXAXAX
AXAXAX
No purification necessary!
transform
Synthon in vector
Vector with long 5 ends
Annealed insert-vector
25
Generation of Synthetic DNA

• gt 500 synthetic DNA fragments generated
• 100 success rate
• GC content from 44-69
• Size between 129 and 1400 bp
• Over 250,000 bp synthesized
• Average error rate around 1.5 errors/kb
• Fully automated most steps in process

26
Gene Synthesis Module Assembly
Input 500 bp synthons in plasmids with correct
sequence

• Digestion
• Ligation
• Transformation
• Isolation of colonies
• Verification of correct clone
• Repeat until full-length gene assembled

Output Complete module (gt5kb) in plasmid with
correct sequence
27
Gene Assembly (Synthon Stitching)
Criteria Accurate Amenable to HT
5,000 bp module
28
Parallel Ligations to Assemble Modules
29
Synthon Stitching Method

• Utilize Type IIs restriction enzymes
• Cut DNA outside of recognition site
• Use different Type IIs enzymes to create
  compatible overhangs
• Same enzymes can be used for all synthon pairs to
  facilitate automation

Bsa I 5 ... G G T C T C (N)1 ... 3 3 ...
C C A G A G (N)5 ... 5
30
Stitching Method Use of Type IIs RE
31
Synthon Stitching Method

• Unique selectable markers on two sister plasmids
  eliminates need for purification of fragments

32

• Alternation of vector pairings allows for unique
  selection at each round of stitching

33
Results of Synthon Stitching

• 26 complete modules constructed
• gt 250 successful ligations
• Selection scheme works extremely well
• Majority of ligations performed gave only correct
  product
• Use of Type IIs enzymes makes method amenable to
  automation

34
Improvements of Gene Synthesis Designer Vectors

• 3-plasmid system for synthon stitching
• Counter-selectable markers
• Allows 4-piece ligations of unpurified digests

35
Synthetic Vector FamilyMultiple-synthon
Ligations
Use of counter-selection allows for stitching of
multiple fragments without purification
36
Second Round Stitching
Can combine 8 fragments in 2 steps with no
fragment purification!
37
Testing of Modules
38
Proof of Concept

• Expressed synthetic 6-module DEBS gene cluster in
  E. coli
• Protein subunits observed on SDS-PAGE in the
  soluble fraction
• Product (6-dEB) identified by LC-MS

39
Results of Module Testing

• Tested 14 synthetic modules in 154 bimodular
  combinations
• 72 of the 154 combinations tested produced
  measurable triketide lactone
• All modules tested worked

40
Summary

• Successfully developed method for high throughput
  gene synthesis
• High-throughput method for assembly of DNA
  fragments into larger genes (modules) developed
• Populated module library and tested in bimodular
  cases

41
Acknowledgements

• Kosan Biosciences Morphing Group
• Dan Santi
• Ralph Reid
• Kedar Patel
• Sebastian Jayaraj
• Hugo Menzella
• Sunil Chandran

42
Summary of Major Synthesis Efforts
aEach experiment represents the parallel
processed synthesis of the DNA indicated.
bAssuming Poisson distribution of errors cAny
specific error was counted only once