1. 1 Addressable Bacterial Conjugation UC Berkeley iGEM 2005 Michael Chen Vlad Goldenberg Stephen Handley Melissa Li Jonathan Sternberg Jay Su Eddie Wang Gabriel Wu Advisors: Professors Adam Arkin and Jay Keasling GSIs: Jonathan Goler and Justyn Jaworski

2. 2 Overview I.Project Goal II.Overview of Existing Technologies II.Initial Design Considerations III.The Construct and its Implementation IV.Current Status V.Future Directions

3. 3 Project Goal To establish specific cell-to-cell communication between two populations of bacteria

4. 4 Project Goal

5. 5 Implementation NEED: To transfer genetic information from one bacteria to another MEANS: Bacterial Conjugation NEED: To specifically control who can read the message MEANS: Riboregulation

6. 6 Bacterial Conjugation Certain bacterial plasmids are classified as having a “fertility factor” i.e. F + Cells that have a F + plasmid can conjugate and transfer their DNA to other bacteria F+F+ F-F- F Pilus Formation F FF F+F+

7. 7 Choosing Conjugative Plasmids There are many plasmids that are classified as conjugative.. For our project, we used F and RP4 plasmids for the following reasons: F and RP4 exhibit differing pili lengths, biasing the order in which F and RP4 will conjugate F and RP4 do no conjugate with themselves F and RP4 are among the most studied and well-characterized conjugative plasmids F and RP4 plasmids are readily available

8. 8 Important Facts about Conjugative plasmids Conjugative plasmids are very large, from 60k – 100k basepairs long The TraJ protein is a regulatory protein responsible for initiating the DNA transfer cascade DNA transfer during conjugation always begins at a specific sequence on the plasmid, OriT, the Origin of Transfer.

9. 9 Modification of conjugative plasmids TraJ was cloned and placed into biobrick plasmids under the control of promoters of our choosing The OriT region was also cloned and placed into biobrick plasmids thus creating small, mobilizable plasmids The OriT region and TraJ gene were knocked out with Lambda- Red mediated recombination to prevent unwanted transfer of the F/R plasmid

10. 10 Conjugation Results An R-plasmid bearing cell can conjugate with an F-plasmid bearing cell The F plasmid and R-plasmid knockouts fail to conjugate The biobricked OriT-R plasmid is mobilizable by the R-plasmid knockout

11. 11 The Riboregulator Isaacs et al., Nature Biotechnology, 2004 Method of postranscriptional control of gene expression cis-repressive sequence (“lock”) upstream of a gene’s coding region forms a hairpin, sequestering the ribosome binding site trans-activating (“key”) mRNA strand binds and opens the hairpin thus allowing access to the RBS. Highly specific activation occurs. Very similar lock and key pair sequences do not exhibit crosstalk

12. 12 Biobricked Riboregulator Lock from Isaacs Paper Tacking biobrick ends onto the end of the lock sequence would be ineffective due to the distance restrictions between a ribosome binding site and a gene’s start codon The mixed site was thus incorporated directly downstream of the ribosome binding site The five base pair region between the hairpin loop and ribosome binding site was used as our address space to create two new lock sequences RBS region Biobrick Mixed Site Predicted mRNA structure of one of our Locks Address RegionHairpin loopStart of locked gene

13. 13 Biobricked Riboregulator RBS region Biobrick Mixed SiteAddress RegionHairpin loopStart of locked gene crR12 lock taR12 key Lock 1 Key 1

14. 14 Biobricked Riboregulator Activation by the key sequences was highest when transcribed five nucleotides from the transcription start site (Isaacs, et al.) We created a biobricked derivative of the E. Coli rrnb P1 promoter to provide constitutive production of our keys Three nucleotides of the biobrick suffix were nested into the 5’ end of the wildtype sequence in order to transcribe the keys at the desired five nucelotide distance.

15. 15 Unlocking the Riboregulator RBS region Biobrick Mixed SiteAddress RegionHairpin loopStart of locked gene Key 1 Key 2 Lock 2 Lock 1 RBS now accessible

16. 16 Biobricked Riboregulator Constituitely On RFP Lock 1 Lock 2

17. 17 Riboregulator Construction Locks and keys are separated at hairpins into pairs of easily ordered oligos ~ 30 bp. One of each pair is ordered phosphorylated for easy ligation of annealed products Anneal pairs in separate tubes (heat to 95°C, unplug heatblock), combine, ligate. L11 5’- ctagag.aactagaatcacctcttggatttgggt L12 3’- tc.ttgatcttagtggagaaccta - p L13 5’- p - attaaagaggaga.tactagtagcggccgctgca L14 3’- aacccataatttctcctct.atgatcatcgccggcg When annealed and ligated, result already has XbaI and PstI sticky ends…ready for assembly Keys require extra pair due to inclusion of key terminator (hairpin) within the part.

18. 18 Construction

19. 19 Parts Used J01004 J01005 J01000 J01001 J01002 J01003 J01006 J01008 J01009 J01011 J01010 E0420 i12351 E0840 E0420 I0500 R0040 I12007 C0051 B0034 B0015

20. 20 Construction Path

21. 21 R-Cell Plasmids

22. 22 Sequence of Events arabinose TraJF F-Cell R-Cell

23. 23 Sequence of Events arabinose TraJF F-Cell R-Cell cI TraJR

24. 24 Sequence of Events F-Cell R-Cell spoOA

25. 25 Modular Design Why didn’t we just lock the fluorescent proteins? Modularity and flexibility of design (send out inquiry for message verification!) with the addition of spoOA, cI signal

26. 26 Progress thus far… CFPkey2OriT F lock1cIGFP lock2spo0AYFP pRM pspoIIEON ONNON RFP key1 ONONN pRM TraJ R OriT R Non-mobilized plasmid Mobilizable plasmid Moblizable plasmid Non-mobilized plasmid F-bearing cell R-bearing cell pBAD TraJ F ara RBS Unforseen Eco site site-mutagenesis time

27. 27 Implementation Issues Fluorescence is sometimes inhibited after conjugation Slight leakiness of the lock we designed Need to add or knockout an antibiotic resistance to one of the plasmids for selection purposes Time-scale of conjugation is slow Cloning is a subtle art

28. 28 Next Steps in Implementation Finish constructs Test our keys with our locks to observe activation and cross-talk Prove that our biobricked OriT-F plasmids is mobilizable Create a ‘stop’ message to end communication after the programs are received or a reset method to return the cells to their original state Determine the effect of copy number and mobilization

29. 29 A Look into… The Future!

30. 30 Application: The Bacterial Network Because “channels of communication” are limited only by the large number of unique riboswitches, multiple pairs of cell-to- cell communication can occur in a single culture One can envision a network of celluar strains in the same culture, specializing in different tasks and communicating specifically to necessary related strains Coordinator Cell Intermediate A Reactant A Reactant B

31. 31 Addressable Conjugation vs. Chemically Based Communication At its heart, our construct creates two one-way channels of communication Quorum sensing exhibits cell-cell communication by using a chemical signal as its message One can easily imagine a construct similar to ours with two different chemical carriers. For example, AHL and DHL.

32. 32 Addressable Conjugation vs. Chemical Communication: Disadvantages Slower Conjugation ~ 8-18 hours Chemical Means ~ 2-8 hours Conjugation occurs in clumps Heterogeneity Limited multiple usage

33. 33 Future Projects Post-conjugal disengagement Multiple “call-and-response” Library and characterization of multiple key-lock pairs Extending address space

34. 34 Addressable Conjugation vs. Chemical Communication: Advantages Rational design of separate specific communications channels Ability to transfer complex genetic information, instead of a single chemical signal

35. 35 Addressable Conjugation Paradigm The Bacterial Internet Cells that transfer information are web servers Cells that selectively express the key can be thought as accessing the web page. The key becomes the URL. Cellular “Server” User Cell A User Cell B Selective key expression Requesting File Download User Cell C

36. 36 Addressable Conjugation Paradigm Function Modularity We can envision cells with modular components Different common genetic programs with similar responses can be loaded on-the-fly depending on differing environmental circumstances Common High-range Module Lock A Common Low-range Module Lock B If (High Conc.) Load A Else Load B

37. 37 Berkeley iGem would like to thank the following people

38. 38 Plasmid and Gene Providers Dr. Virginia Waters: RP4/RK2 plasmid Dr. Laura Frost: F-Plasmid Philip Silverman: pox38 F-Plasmid Dr. Farren Isaacs: Lock and Key Sequences Mike Cantor: SpoOA and pspoIIE plasmid

39. 39 All the members of the Keasling Lab

40. 40 Mario Can clone anything Jon Dueber Can clone anything

41. 41 Connie Lambda Red Doug Lambda Red

42. 42 Professor Adam Arkin & Professor Jay Keasling

43. 43 Questions? One more animation