Download Designing & Testing Flippable Hix Sites for Bio Computers: Study by Bruce Henschen - Prof. and more Papers Biology in PDF only on Docsity!
Page 1 Design and testing of one-time Flippable Hix Sites for use in Biological Computers Bruce Henschen Biology Department, Davidson College, Davidson, NC 28035 BIO 372, Spring 2007
Abstract
Davidson College is currently working on biological applications of the Pancake Problem, in which Salmonella Hin recombinase is used to flip DNA ‘pancakes’ that are flanked by symmetrical 26 bp Hix sites in the DNA. I propose multiple mechanisms for mutating these Hix sites to engineer pancakes that can only be flipped once by Hin, enabling better control over the Hin flipping mechanism. These mutations included insertions, deletions, and point mutations in the Hix site. In addition, a restriction enzyme and phenotype-based screen was proposed to test the functionality of the Hix mutants. All mutants were cloned into vectors and assembled as Hix mutant:Hix wt paired constructs to test functionality independent of other mutant. Data from wt constructs was compared with data from a pilot Hix mutant construct and showed that restriction enzymes provide a more reliable, although imperfect, screen for flipping activity. Data also revealed that an insertion mutation caused Hin to alter the properties of the Hix plasmid, causing damage to the plasmid and altered function. Further study should analyze the effect of point mutations on Hix site functionality and should develop a more reliable screen for determining the orientation of a segment of DNA between Hix sites.
Introduction
At the iGEM Synthetic Biology conference in 2006, Davidson College presented work on using E. coli to solve mathematical problems by using a system of DNA flipping mechanisms (Campbell et al., 2007). According to the mathematical problem, called the burnt pancake problem, one must determine the least number of pancake flips it takes to have randomly-arranged pancakes oriented in the right order and direction. In E. coli , genes and promoters in random combinations were used as DNA pancakes; the bacteria
Page 2 were able to flip the DNA elements using a combination of the Salmonella Hin recombinase protein and the HixC sites in DNA that Hin recognizes. The Hin recombinase system in Salmonella sp. allows the bacterium to undergo flagellar phase variation, or changes in flagellum surface protein types to evade a host’s immune system (Adams et al., 1997; Glasgow, et al., 1989; Sanders and Johnson, 2004). In vivo, Hin recombinase flips a promoter flanked by 26 bp symmetrical DNA sites called Hix sites, changing which antigen is produced by the cell at any given time. The two Hix sites on either side of the promoter vary slightly from one another and are labeled HixL and HixR. Symmetric HixC sites have been synthesized that work in the same fashion as the other binding sites (Lim, 1992). Outside its normal function, Hin recombinase can be used to flip any DNA sequence that is flanked by HixC sites and complexed with a Recombinational Enhancer DNA site (RE) and the Fis and HU proteins. The Hin protein consists of a homodimer in which each monomer binds to one half of the roughly symmetrical Hix site. In vivo, these dimers complex with the Fis and HU proteins to form homotetramers (diagrammed in Figure 1), which induces a conformational change in all four Hin proteins to expose DNA- cleaving regions. Cleavage of each DNA strand occurs via a two-bp overlap cut in the middle of each Hix site, followed by rotation of two of the Hin proteins in the tetramer bound to DNA segments, which results in DNA strand exchange (Dhar et al., 2004). The process is complete when the two new strands are ligated together and the complex, or the invertasome, decomposes (Heichman et al., 1991). In both E. coli and S. typhimurium , DNA flipping has been observed (Dr. Haynes, unpublished), showing that E. coli possess the required protein elements for flipping when supplied with two HixC sites and RE. It is unknown how many times Hin inverts the two segments of DNA before the DNA Figure 1. Hin/Hix mechanism, adapted from Hughes, et al., 1992.
Page 4
Materials and Methods
(some adapted from Simpson, unpublished) Bacteria used : To begin, I used E. coli JM109 cells grown in previous semesters from the Davidson College Freezer Stocks. After growing stocks of previously transformed E. coli , I used pSB1A2 plasmids in these cells to make my Biobrick constructs. When transforming ligated parts, I used either Promega JM109 competent cells (Part # 44-0301) or Zymogen JM109 Z-competent cells (Catalog # T3003) for ligated plasmids. After following the protocols included with each cell type, cells were plated on LB plates with Ampicillin (100 μg/mL) and grown at 37 °C for 10-16 hours. To isolate plasmid DNA from grown cells, I picked colonies using sterile toothpicks, dropped them into LB Amp+^ (100 μg/mL) broth, and grew the clonal populations for 10- 16 hours. After growing the cultures, I followed the protocols included in two types of kits: Promega Wizard SV Minipreps Kit (Promega part # A1460), and Zymogen Zyppy Plasmid Miniprep kit (Zymogen part # D4036, D4019, and D4020), to isolate plasmid DNA. Validation of part identity : To determine whether ligations were successful, I used agarose gel electrophoresis to compare the size of the Biobrick part with the known length of the part. Gels were poured using “How to Pour an Agarose Gel,” written by Dr. A. Malcolm Campbell, and found here: http://www.bio.davidson.edu/courses/Molbio/Protocols/pourgel.html. Once gels were poured using Promega Agarose (Promega part # V3121), digested samples were loaded next to the Invitrogen 1 kb ladder (catalog no. 15615-016; ladder information found at http://www.bio.davidson.edu/courses/Molbio/Protocols/gels2002/1kbladder.pdf) to quantify the size of the part. Restriction Digestions :
Page 5 DNA was digested using “How to Digest DNA with Restriction Enzymes,” written by Dr. A. Malcolm Campbell and found here: http://www.bio.davidson.edu/courses/Molbio/Protocols/digestion.html. The following restriction enzymes were used to digest DNA: EcoRI; XbaI; SpeI; PstI; PvuII, all ordered from Promega. Double digestion protocols were found here: http://parts.mit.edu/wiki/index.php/Double_Digest_Guide. DNA Purification : After running DNA on an agarose gel, the DNA must be removed from the gel before being ligated into another part. To perform DNA purification, I used Qiagen QIAQuick Gel Extraction Kit (catalog #28704). I stored collection tubes for future use according to a protocol developed by Dr. Haynes. DNA Quantification : To determine the amount of DNA in a sample, I used a NanoDrop Optical Density Reader. Ligations : To ligate parts together, I used “How to Ligate DNA,” written by Dr. A. Malcolm Campbell, and found here: http://www.bio.davidson.edu/courses/Molbio/Protocols/ligation.html. Annealing of ssDNA Oligonucleotides : Annealing of single-stranded oligos was performed according to “Building dsDNA with Oligos,” a protocol written by Dr. A. Malcolm Campbell and found here: http://www.bio.davidson.edu/courses/Molbio/Protocols/anneal_oligos.html. FACS Analysis of Flourescence : To measure the amount of GFP fluorescence in a particular culture, I grew cultures in 2mL LB Amp+^ (100 μg/mL) broth for 10-16 hours prior to the analysis, and
Page 7 shown to be involved in Hin binding and do not have a known function. From the data, I decided that the best places in the HixC site to modify were the bases between the cleavage site and the first DNA-binding site (between bases 2 and 4), base 7, and base 12 of the HixC site, where the DNA either does not play a role in Hin binding or has been shown to have altered Hin activity when altered (Hughes et al. , 1992). Two manipulations to the DNA might result in one-time flippable Hix sites (Figure 4). First, bases could be inserted in one Hix half-site and deleted in the other paired Hix half-site (Figure 4). Upon recombination, the inserted and deleted half-sites would combine with each other, resulting in a Hix site that had a cutting region 36º altered from the Hin binding domain (Lim et al., 1992). Hin is a flexible protein and could theoretically bind to a site with one insertion or deletion (Glasgow et al., 1989), but when recombined, it would not be able to cut rotated DNA in the proper location, resulting in one non-functional Hix site and one fully- functional HixC site. Other studies have examined the activity of Hin when certain base pairs in HixC are altered (Hughes et al., 1992). In particular, mutations in bases 3, 7, and 11 reduced, but did not eliminate, Hin flipping activity. If one Hix half-site with one mutation is paired with a Hix site with another, unique mutation, recombination of the two sites would create one wt HixC site and one Hix site with two mutations in it (Figure 4). While both mutants cause low Hin flipping activity independently, a double mutant could produce very little or no Hix recombination. In addition, sites could be designed with symmetrical mutations in each, such that when recombined, each Hix site would contain two different mutations and would be nonfunctional. Figure 4. Method of using mutations to create one wt hix site and one mutated site upon recombination by Hin.
Page 8 I used both insertion and deletion mutations, as well as point mutations, when designing Hix sites for use in the experiment. The full sequences of the Hix site designs are in Appendix B. I first designed a set of oligonucleotides with truncated biobrick ends, to protect against the possibility of mutation during oligo synthesis. However, efforts to clone these into vectors were unsuccessful (see below), so new oligos were designed with full biobrick ends as replacements. The next step, after these sites were designed, was to determine how to test whether Hin was able to flip a segment of DNA once, more than once, or not at all. The first assay developed is diagrammed and described in Figure 5. To build these parts, I used the ligation plan in Figure 6. Condition pLac Orientation Phenotypic Effect No Flipping 100% Forward Green Cells under UV light, No survival on Tetracycline-coated plates Flipping only once 100% Backward (^) No Green Flourescence, Survival on Tetracycline-coated plates Flipping Multiple times (wt) 50% Forward, 50% Backward Green Cells under UV light, No survival on Tetracycline-coated plates Figure 5. Diagram of the Hix functionality assay and Table explaining theoretical expected phenotypes for each possibility of Hix functionality
Figure 8a. 0.7% agarose gel, run for 35 min @ 150 V, to verify TetB-RBS-HixC-pLac- HixC ligation. All lanes show successful ligation; plasmid from lane 7 was preserved. Page 10 The next round of ligations included ligation of TetB-RBSrev-HixC into pLac-HixC to make a second intermediate, TetB-RBSrev-HixC-pLac- HixC (confirmation gel shown in Figure 8a). A final intermediate was made – pLac-HixC-RBS- Figure 7a. Construction of intermediates.^ GFP-RE, confirmation gel on Figure 8b. Positive ligations are seen in pLac-HixC 1-3 and TetB-RBS-HixC 1-2. Figure 7b. Construction of intermediates. Positive ligation seen in lanes B3 and B for RBS-GFP-RE. b Figure 8b. 0.8% gel, run at 120 V for 35 min. Confirmed ligations of pLac-HixC- RBS-GFP-RE reside in lanes 1, 4-6; plasmid used in lane 1 was preserved.
Page 11 After completing the assembly of intermediates, I was able to complete the full wt construct – TetB-RBS-HixC-pLac-HixC-RBS-GFP-RE (Figure 9). While completing the ligation of the wt construct, I attempted to clone the received oligonucleotides into vectors. I chose the oligo Hix1 (i+4TL), which has a “T” i nsertion at base pair +4, to be a pilot for development of a protocol for all mutant oligos. My first set of oligonucleotides were designed with XbaI/ SpeI sticky ends to save cost and reduce mutation risk, but these would not successfully clone into the vector pSB1A7. I originally had wanted to use this different vector since it included Transcriptional Terminators beyond the Biobrick ends that prevented read-through transcription. However, cloning Hix1 into pSB1A7 was not accomplished for two possible reasons. First, cutting plasmid with XbaI and SpeI produces complementary sticky ends; these ends of the plasmid are located near each other relative to oligonucleotides floating in solution around them, so plasmid cut with these two restriction enzymes has a greater chance of ligating back together without the oligo insert without incorporating a Hix. Second, the TT sites in pSB1A7 are inverted repeats, which makes it extremely difficult to ligate another inverted repeat sequence of HixM (mutant) or HixC into the plasmid due to DNA geometry and the tendancy of complementary DNA to fold in on itself. To solve the problems, I redesigned the oligonucleotides to include full Biobrick ends (with EcoRI and PstI sticky ends) and ligated them into pSB1A2, the vector that the wild-type part was constructed with. After ligation of this part was complete, I performed the same procedure on Hix2-7. Ligation of Hix1 into this plasmid is shown in Figure 10a, while ligation of Hix2-7 into this plasmid is shown in Figures 10b, 10c, and 10d. Figure 9. 0.4% gel, run at 150 V for 35 min. Confirmed ligations of TetB-RBS-HixC-pLac- HixC-RBS-GFP-RE reside in lanes 1-5; plasmid used in lane 1 was preserved.
Page 13 In the previous step, Hix2 plasmid was lost and was not used in the final ligation step, but the part was ligated with TetB-RBS in the next step along with the other Hix mutants. Figure 12 shows the completed ligations of Hix1 and Hix3-7, as well as the confirmation of TetB-RBS-Hix2. These parts were stored, freezer stocks were made, and full constructs with Hix2-7 paired with HixC will be used in further analysis. The Hix1:HixC full construct was used in phenotype and restriction enzyme analysis by comparing data to the HixC:HixC wild-type construct. Results from Phenotype Analysis The assay in Figure 5 shows that, theoretically, pLac should transcribe either tetracycline resistance or GFP, depending on which way the promoter is pointing. Because the promoter begins in the direction of GFP, cells that do not have flipping- capable plasmids would not survive on tetracycline coated plates. If the Hix sites allow only one DNA flipping event, pLac in all plasmids in a cell would point toward TetA(c), a b Figure 12a-b. 0.5% gels confirming plasmid identity: Hix1 full construct: Gel a 1- TetB-RBS-Hix2: Gel b 1- Full Hix3-7: exp. Lanes in Gel b Plasmid preserved from arrowed lanes a (^) b
Page 14 enabling the cells to survive on Tet-coated plates but without green flourescence. However, if the promoter could be flipped multiple times, then approximately half of the promoters would be pointing toward TetA(c) and half toward GFP at any given time, causing cells to fluoresce green on Tet plates. During the course of the study, other research showed that pLac not only promotes genes in the forward direction but also has a slight non-inducible backwards- promoting capacity, causing TetB to be produced when pLac is pointing toward GFP (Dr. Karmella Haynes, personal communication). Since tetracycline resistance is conferred through transcription and translation of a small amount of the TetA(c) gene, any transcription of TetA(c) from the backward-promoting activity of pLac means that all cells are tetracycline resistant. One option would be to use a different promoter, such as pBad, used previously in experiments at Davidson College. However, construction of parts had already begun, and pBad is a weaker promoter than pLac, so I chose to include pLac in the final design. Therefore, GFP expression, quantified using FACS analysis, could be used to determine pLac orientation. The predicted effects of flipping on GFP expression are tabulated in Figure 13. TetA(c) is toxic in high concentrations; it is possible that cells that have pLac reversed produce too much Tetracycline resistance and die as a result, providing another screen for testing the direction of pLac. Condition pLac Orientation Effect No Flipping 100% Forward 100% expression of GFP Flipping only once 100% Backward Minimal Green Flourescence Possible death due to overproduction of TetA(c) Flipping Multiple times (wt) 50% Forward 50% Backward Reduced expression of GFP Possible death due to overproduction of TetA(c) I used Flourescence-Activated Cell Sorting (FACS) to determine GFP levels in each construct. In preparation for analysis, I cotransformed the HixC:HixC plasmid with the Hin expression cassette (obtained from Dr. Haynes) and plated the cotransformed cells on Amp+^ Kan+^ plates to determine whether it was possible to observe flipping in the wild-type. Once these cells were grown and the cotransformation was confirmed to be Figure 13. Table for determining pLac orientation based on FACS and cell survival.
Page 16 g (^) h e f c d a b Figure 14. FACS Data gathered on multiple days. Graphs display number of cells counted at variable levels of GFP Fluorescence. Graphs collected on the following days: A-D: Gathered on 4/4/ E-H: Gathered on 4/22/07 ( controls appear very similar to A and B )
Page 17 Results of Restriction Digest Analysis: In addition to using phenotypic data to determine pLac orientation, a DNA-based method of determining flips that involved using restriction enzymes was designed with the aid of Dr. Haynes. In this method, I cotransformed the Hin and Hix plasmids into cells cells using Amp+^ Kan+^ plates to select for cells that had incorporated both types of plasmids. Colonies were grown in overnight cultures containing Amp, Kan, and IPTG to induce Hin expression and flipping, if possible. Third, I mini-prepped the overnight cultures to isolate all of the plasmid DNA present in the sample, which would consist of unflipped plasmids, possibly flipped plasmids, and plasmids containing only Hin. Samples from the miniprep were transformed into new cells onto Amp+; since Amp plasmids are more numerous than Kan plasmids, because only Amp selection was in place, and because cells normally only incorporate one plasmid each, it was assumed that surviving cells contained a BioBrick plasmid but not a Hin plasmid, thereby “freezing” the pLac promoter in one orientation while multiplying that particular plasmid in a colony. These colonies were mini-prepped and digested using PvuII, a restriction enzyme that cuts according to the map in Figure
- Because PvuII produces fragments of different sizes based on the orientation of pLac, agarose gel electrophoresis could be used to separate these two possible bands that can be Figure 15. Restriction digsst map from NEBcutter (http://tools.neb.com/NEBcutter2/index.php), as well as a graphical interpretation of the effects of cutting with PvuII.
Page 19 shown in Figure 17). In addition, I digested the 12 plasmid samples with EcoRI and PstI and ran an agarose gel to compare sizes (Figure 18). Finally, overnight cultures from these samples were made after the digestions were run. Toothpicks from these cultures were used to streak Amp+, Amp+ Kan+, and Amp+ Tet+ plates. Growth on Amp Kan is indicative of the presence of the Hin plasmid, while growth on Amp Tet is indicative of an intact BioBrick site, since TetA(c) is transcribed. Growth on these plates was monitored and recorded in Figure 19. Figure 17. 1% agarose gel, ran at 120 V for 40 min, of PvuII digestions described above. Hin = sample ran without BioBrick constructs HixM and HixC (-c) = samples without Hin Figure 18. 0.8% agarose gel, ran at 80 V for 1 hour, of EcoRI and PstI digestions described above. No DNA available for HixC G+ 2 sample.
Page 20
Discussion
Discussion of Experimental Design While developing the assay to determine the effectiveness of each HixM:HixM pair, I determined that neither the DNA nor the fluorescence-based methods of screening can differentiate between plasmids that flip multiple times and plasmids that have weak Hin activity and flip at such a slow rate that only 50% of the plasmids flip. Some Hix sites could have weak functionality, such that half of the promoters flip once while the others do not flip at all. This result would show the same phenotype and the same gel bands as cells that were allowed to flip readily. Methods must be determined to differentiate between constructs with weak activity vs. constructs with full DNA flipping capability. Despite this initial uncertainty, I was able to gather data using phenotypic analysis and DNA-based analysis. Discussion of FACS Data These data confirm that pLac is a highly leaky promoter, producing variable levels of transcription with equal amounts of input. Because these results only observed data for HixC:HixC constructs, Hin should be able to functionally flip the pLac promoter. Theoretically, pLac flipping would cause a relative decrease in GFP production, since Colony Growth on Amp Growth on Amp/Kan Growth on Amp/Tet HixC G-1 + - + HixC G-2 + + + HixC G-3 + - + HixC G+1 + - + HixC G+2 + - + HixC G+3 + - + HixM G+1 + - +, but less HixM G+2 + - - HixM G+3 + - - HixM G-1 + - - HixM G-2 + - - HixM G-3 + - - Figure 19. Colony growth on multiple types of plates.
