Engineering Success

Synthetic biology is widely considered to be a discipline which involves the engineering and development of new biological systems and devices, or the redesigning and improvement of current systems. Engineering widely refers to the process of designing, constructing, and building something. The engineering process is critical to success. Synthetic biology in particular, widely relies on the engineering cycle: DBTL; Design, Build, Test, and Learn to successfully innovate, develop, and implement novel biological systems.

This past summer we initiated the production of our product, PROS. We laid the basic groundwork for a future therapy for protein S deficiencies. Our goal for the summer was to successfully express human recombinant protein S in SF9 insect cells and E.coli. In order to accomplish this, we applied the DBTL cycle numerous times. As a team of 11 undergraduate students, the engineering cycle was critical to our success. Although we faced various obstacles and challenges, this process helped us persevere and succeed in our project. On this page we detail our iterations of the design cycle, and how it helped us build and refine our project.

---

Overall Design

In our project, we decided to try and express our protein of interest, protein S, in two different organisms: SF9 insect cells, and E.coli. We wanted to compare prokaryotic vs. eukaryotic expression, and to this extent, we conducted experiments in parallel. Although we could not conduct the SF9 experiments to completion, both of these aspects of our project underwent the engineering process numerous times.

---

Engineering Cycle for E.coli

Cycle 1: Obtaining Pure Cloning Substrates for the Ligation Independent Cloning Reaction

Design: Our aim was to insert the PROS gene, which was codon optimized for chassis E. coli and obtained through synthesis from IDT into the pET His6 LIC cloning vector (2Bc-t). In order to prepare our plasmid for the LIC, we went through several cycles of minipreps, linearization using the restriction enzyme Hpa1 and gel extractions in order to get high quality substrates for the cloning reaction. Our Gene of Interest was amplified using the primers we designed, creating specific overhangs which would facilitate the annealing step during LIC. More detailed information on our primer design and LIC process can be found on our Design page.

Build: For preparing our GOI and the plasmid for the LIC reaction, we performed a T4 polymerase 3’ chewback process by treating the insert with dGTP and T4 polymerase and the vector with dCTP and T4 polymerase. The 3’ chewback would generate complementary 5’ overhangs which would facilitate the annealing reaction. GOI concentration = 136.7 ng/uL Vector concentration = 7.2 ng/uL

Test: After generating the substrates for LIC with 5’ overhangs, we performed the following annealing reactions:

We transformed the annealing reaction products in E. coli DH5 alpha using 10 uL of the 20 uL samples and incubated the cells at 37 °C for 1 hour, shaking at 220 rpm. Transformed cells from both samples were plated on ampicillin agar plates and incubated overnight.

Learn: No colonies were observed on the experimental control plate, indicating that the LIC reaction was successful. We thought that our plasmid samples had very low concentrations for the annealing reaction and the ratio of Insert to vector might be low. We decided to perform more minipreps to obtain higher concentrations of the plasmid and reattempt the LIC reaction.

Cycle 2: Obtaining Higher Concentrations of the Plasmid and Reattempting the LIC Reaction

Design: To improve the yield of our plasmid samples, we performed another miniprep. We obtained multiple samples of the mini prepped samples to increase the chances of obtaining high quality LIC substrates after going through the steps of linearizing and gel extraction. Moreover, the concentration of the plasmid decreases significantly after going through the processes of restriction digest and gel extraction. After treating all the samples and analyzing them using gel electrophoresis, we concluded that one of the mini prepped samples was concentrated and pure enough for the Ligation Independent Cloning reaction.

Build: Final concentrations of our GOI and the plasmid after preparing the LIC samples were 68.35 ng/uL and 5.55 ng/uL respectively. We decided to perform two separate annealing reactions using 2 different insert to vector ratios, 1:5 and 1:2.

Test: We transformed the annealing reaction products in E. coli DH5 alpha using 10 uL of the 20 uL samples and incubated the cells at 37 °C for 1 hour, shaking at 220 rpm. Transformed cells from both the reaction samples were plated on ampicillin agar plates and incubated overnight.

Learn: We observed colonies on both the transformation plates, indicating that both the experimental reactions for Ligation Independent Cloning were successful. After the successful cycle of designing, building, learning and testing for optimizing our experimental conditions to create our recombinant plasmid, we decided to conduct confirmatory experiments like restriction digests, PCR amplification of the insert from the recombinant vector and sequencing the plasmid sample. We were able to conclude that our GOI was successfully cloned into the plasmid by analyzing the results of all 3 confirmatory tests. More detailed information on our experiments can be found on our Results page.

Cycle 3: Conducting Expression Tests to Detect Recombinant Protein S Expression

Design: After confirming the presence of our GOI in the plasmid, we transformed the recombinant vector in E. coli strains BL21 and Origami B cells. BL21 is the common lab strain used for expressing recombinant proteins. However, BL21 cells are not very efficient in making disulfide bonds in the expressed protein, making it difficult to express soluble protein. Therefore, we decided to conduct parallel expression tests in both Bl21 and Origami B cells. Origami cell strain contains specific mutations which greatly enhance their ability to make disulfide bonds in the expressed proteins, increasing protein stability and preventing degradation.

Build: For the expression test, we picked colonies containing our recombinant vector, cultured them in 5 mL LB and ampicillin overnight at 37°C, and used 125 mL of the overnight culture to dilute it into a 25 mL culture. The 25 mL cultures of both BL21 and Origami B were grown to OD600 at 37°C and induced using 1 mM IPTG.

Test: We collected a 1 mL pre-induction sample from both the BL21 and Origami B cultures before adding 1mM IPTG. After induction, the cultures were grown for 4 hours at at 37°C and 220 rpm, and a post-induction sample was obtained every hour post-induction for 4 hours. All the expression samples were centrifuged at 3000 rpm for 5 min, to harvest the cells in the pellet. We performed an SDS-PAGE and a Western Blot to detect expression.

Well #	Sample
Well 1	Ladder
Well 2	Origami B - Pre-Induction
Well 3	Origami B - 1 hr Post-Induction
Well 4	Origami B - 2 hr Post-Induction
Well 5	Origami B - 3 hr Post-Induction
Well 6	Origami B - 4 hr Post-Induction
Well 7	BL21 - 1 hr Post-Induction
Well 8	BL21 - 2 hr Post-Induction
Well 9	BL21 - 3 hr Post-Induction
Well 10	BL21 - 4 hr Post-Induction

Well #	Sample
Well 1	Ladder
Well 2	Origami B - Pre-Induction
Well 3	Origami B - 1 hr Post-Induction
Well 4	Origami B - 2 hr Post-Induction
Well 5	Origami B - 3 hr Post-Induction
Well 6	Origami B - 4 hr Post-Induction
Well 7	BL21 - 1 hr Post-Induction
Well 8	BL21 - 2 hr Post-Induction
Well 9	BL21 - 3 hr Post-Induction
Well 10	BL21 - 4 hr Post-Induction
Well 11	Ladder

Learn: The results of SDS-page were inconclusive because we did not observe any obvious dark bands at about 70 kDa (approximate weight of protein S), and all the bands on the gel are roughly of the same intensity, indicating no increase in expression with time. The Western Blot results revealed that BL21 cells are unable to express protein S, as no bands were observed for BL21 samples. Origami cells, however, did show bands at around 25 kDa, indicating that protein S is being degraded after expression. We thought that the conditions we used for inducing these cells were not favorable and were causing the protein to form aggregates in the cells. Therefore, we decided to brainstorm the ways in which we could optimize the expression conditions for Origami B cells.

Cycle 4: Optimizing Expression in E. coli Origami B Cells

Design: In the previous expression test, we observed expression of protein S from Origami B cells, and the fact that BL21 were unable to express protein S indicates that Origami B cells are better suited for expressing protein S because of their ability to make disulfide bonds in the protein. We decided to perform another expression test in Origami B cells by manipulating several experimental conditions. We decided to induce two different samples of Origami cells, each at a different cell concentration (OD 0.4 and 1.1). We also decreased the concentration of IPTG from 1 mM to 0.5 mM for induction and the origami cells were induced at 37°C, gradually reducing it to 23°C. Moreover, we decided to sonicate our cells before conducting the western blot and adding protease inhibitors.

Build: We picked 2 colonies from the Origami B transformants and cultured them overnight in 5 mL of LB. Both samples (Ori 2 and Ori 3) were diluted into 25 mL cultures and sample Ori 2 was induced at OD = 0.4 and sample Ori 3 was induced at OD = 1.1. We collected one pre-induction sample, 4 post-induction samples (one hour increments for 4 hours) and one overnight induction sample.

Test: After sonicating the cells in our expression samples and addition of protease inhibitors, we performed a Western Blot for detecting expression of protein S.

Well #	Sample
Well 1	Pre-induction Sample (Ori 2)
Well 2	One hour post-induction (Ori 2)
Well 3	Two hours post-induction (Ori 2)
Well 4	Three hours post-induction (Ori 2)
Well 5	Four hours post-induction (Ori 2)
Well 6	Overnight induction (Ori 2)
Well 7	Empty well (loading dye)
Well 8	Pre-induction Sample (Ori 3)
Well 9	One hour post-induction (Ori 3)
Well 10	Two hours post-induction (Ori 3)
Well 11	Three hours post-induction (Ori 3)
Well 12	Four hours post-induction (Ori 3)
Well 13	Overnight induction (Ori 3)

Learn: As evident from Figure. we were successfully able to detect protein S expression (full length, about 70 kDa) using the HRP anti-His 6 antibodies. We learned that the ability of Origami B cells to make disulfide bonds greatly enhances protein stability, making it possible for us to express recombinant human protein S in a bacterial cell strain. We also learned that some of the protein was still being degraded, as evident from the multiple bands visible from about 30-50 kD. We thought that the protein is being degraded from the N-terminal because human protein S has multiple N-linked Glycans added during its post-translational modification. Glycans might have protective functions for the EGF domains on the protein. As we successfully expressed protein S in its full length we thought that the conditions for expressing recombinant protein S in bacterial cell strains can be further optimized.

Conclusion

After going through a lot of cycles of designing, building, testing and learning, we were successful in generating the recombinant plasmid containing the PROS gene and expressing protein S in its full length by optimizing expression in Origami B cells. After concluding our wet lab experiments, we started thinking about ways in which future igem teams can further improve recombinant protein expression in bacterial cells. We also partnered with the UCSC igem team and they helped us model the expression of protein S in yeast, which, being an eukaryotic organism, is a promising host choice for protein expression. After conducting a literature search and consulting with our advisor Andrew Sillato, we came to a conclusion that expressing recombinant protein S along with glycosyltransferases in both E. coli Origami B and yeast cells is a promising future avenue for protein S research.

---

Engineering Cycle for SF9

Cycle 1: Determining Appropriate Cloning Technique and Primer Design

Design: Our goal was to clone the PROS gene which was codon-optimized for chassis SF9 and obtained through synthesis from IDT, into the multiple cloning site of the Baculovirus transfer vector - pFastBac YMBac-II. After a lot of literature review and consultation with our advisor, we determined that Ligation Independent Cloning would be the ideal technique for our cloning process. We designed primer sequences that would be used to amplify the synthesized PROS gene and the plasmid pFastBac, in order to generate specific overhangs on the cloning substrates. Detailed information regarding our primer design and the LIC cloning process can be found on our Design page.

Build: The building phase for our first cycle of engineering involved obtaining our synthesized gene of interest, plasmid and the designed primers and amplifying the DNA sequences using PCR. The annealing temperature for our PCR reaction was calculated using the NEB Tm calculator and the thermocycling conditions were set using the NEB PCR protocol for Q5 Polymerase.

Test: We ran the PCR reaction for both the Gene of interest and the plasmid at same thermocycling conditions and analyzed the results using gel electrophoresis on a 1% agarose gel.

Learn: We were able to amplify our GOI (~2031 bp) and the plasmid (~5000 bp). However, we learned that these specific thermocycling conditions resulted in significant DNA laddering for the plasmid sample as evident from Figure 1, which can be a result of nonspecific primer annealing. Moreover, the smearing on the gel can be a result of using too much DNA substrate for PCR, high concentrations of the primers or the enzymes or contamination. We also attempted performing the cloning reaction using these products, however, it was unsuccessful. Therefore, we decided to optimize the conditions for our thermocycling reaction and re-attempt PCR amplification.

Cycle 2: Optimizing PCR Thermocycling Conditions for the Plasmid

Design: We decided to improve our PCR conditions and conduct another thermocycling reaction for amplifying our plasmid using annealing temperatures of 55°C and 62°C. We thought that increasing the annealing temperature for one of the thermocycling reactions might lower the nonspecific annealing of the primers, and using the same initial annealing temperature of 55°C would serve as a control and comparison with the initial PCR reaction. We also decided to perform a gel extraction of the PCR products to obtain more pure samples for attempting the cloning reaction.

Build: We made two different PCR substrate samples for the thermocycling reactions at 55°C and 62°C and increased the extension time from 2 to 2.5 minutes and the final extension time from 2.5 to 5 minutes.

Test: We ran the PCR reaction for both the thermocycling conditions and analyzed the results using gel electrophoresis on a 1% agarose gel.

Learn: By trying a different thermocycling condition, we were able to reduce some of the DNA laddering and gel smearing. However, there were still some non-specific bands for both the samples and annealing temperatures. We performed a gel extraction of the PCR products and attempted another cloning reaction, which was again unsuccessful. We also realized that for all our prior PCR attempts, we used the primer overhang sequences for calculating the annealing temperatures. Therefore, for the next cycle of optimizing PCR conditions, we used the entire primer sequence, i.e. both the overhang sequences and the overlapping sequence.

Cycle 3: Optimizing PCR Thermocycling Conditions for the Plasmid and the GOI

Design: Since the previous cycles of optimization were only able to slightly reduce DNA laddering for the plasmid PCR products and the consequent cloning reactions failed, we decided to use the entire primer sequence for calculating the annealing temperature and attempt another PCR reaction.

Build: The new annealing temperatures were 66°C and 72°C for the GOI and the plasmid respectively. We decided to make four different samples for the thermocycling reaction, trying the annealing temperatures of 66°C and 72°C for both the GOI and the plasmid.

Test: We ran the PCR reaction for all 4 samples and analyzed the results using gel electrophoresis on a 1% agarose gel.

Learn: From the results of this cycle of PCR, we learned that lower annealing temperatures are better suited for the amplification of the GOI and the plasmid. We saw significant DNA laddering and nonspecific bands. We thought that the appearance of nonspecific bands on the PCR gel and the resulting unsuccessful cloning reactions might be due to high GC content of the codon optimized PROS gene and the designed primers as well. Therefore, for our next cycle of DBTL, we decided to perform thermocycling for the GOI and the plasmid at lower annealing temperatures and use GC enhancers for these PCR reactions.

Cycle 4: Testing Multiple Lower Annealing Temperatures Along with GC Enhancer

Design: Since we obtained better results by amplifying the GOI and the plasmid at lower annealing temperatures, we decided to try 3 different annealing temperatures along with GC enhancers for the thermocycling reaction. We hoped that using multiple annealing temperatures along with using GC enhancers will increase our chances of improving the thermocycling conditions and get higher quality PCR products for the subsequent cloning step.

Build: We made a total of 8 PCR substrate samples, trying annealing temperatures of 55°C and 60°C for the GOI and the plasmid. Moreover, we added GC enhancers to only half of the samples and used the rest non-GC enhanced samples as a control.

Test: Each sample was created with identical reagent concentrations and placed into the thermocycler for PCR amplification. The PCR results were analyzed using gel electrophoresis on a 1% agarose gel.

Learn: The results of this cycle of PCR suggest that addition of GC enhancers did not improve the DNA laddering of the PCR products. Referring back to all the cycles of PCR we underwent, we decided that the optimal annealing temperatures for the GOI and the plasmid are 55°C and 62°C respectively as these conditions show bands with the least amount of laddering. We were able to successfully amplify our DNA sequences using these conditions, and obtain pure samples following gel extractions.

Cycle 5: Gel Extraction for More Pure and Higher DNA Yield

Design: Even after deciding the best conditions for PCR, we still had some DNA laddering possibly because of non-specific annealing, however, we did get proper amplification of the DNA sequences as indicated by bright bands of the correct size. Therefore, we decided to perform gel extractions to overcome the issue of non-specific bands on the gel and obtain pure DNA samples for cloning. By loading multiple PCR reactions into one big well on the gel electrophoresis, we hoped to gel extract a significant quantity of pure DNA samples by cutting out just the right sized bands of DNA.

Build: Four 50 ug samples of the gene from the PCR reaction were selected. Each one of these samples had mild laddering and a bright band where the DNA of interest would be. These samples were pooled and mixed with gel stain in a 1:5 ratio (gel stain: PCR product). This sample was loaded into 0.7% gel.

Test: The band of DNA that contained the gene was cut very closely to prevent excess agar from entering the solution. Furthermore, the cutting process had to occur quickly so that the DNA would not be denatured by the UV imaging system. The sample was washed twice rather than three times to prevent excess loss of DNA. The DNA was suspended in 21 uL of elution buffer to keep DNA concentrations high.

Learn: Gel extraction produced a higher yield and concentration when compared to ethanol precipitation. When the gel extraction sample was tested on gel electrophoresis, only one band of DNA appeared brightly. All future DNA samples that were used for the LIC went through gel extraction.

Cycle 6: Ligation Independent Cloning of the GOI and the Plasmid Substrates

Design: After obtaining pure DNA samples by gel extraction of the PCR products, we treated both the GOI and the plasmid for the T4 polymerase 3’ chewback step. We used a ratio of approximately 3:1 (Insert:Plasmid), and about 38 ng of the insert vs 14 ng of the plasmid for the annealing reaction and transformed them into E. coli NovaBlue cells.

Build: Since our first LIC attempt and transformations into NovaBlue cells were unsuccessful, we reattempted LIC at various conditions and DNA concentrations. We consistently failed with an insert:plasmid ratio of 3:1 and increasing their concentrations as well, therefore, we decided to increase the final concentrations of the insert and the plasmid to 20 ng/uL in the 20 uL LIC reaction mix. We used 400 ng of both the insert and the plasmid with a ratio of 1:1 (Insert : Plasmid) for this round of LIC reaction.

Test: We mixed 5 uL of both the insert and the plasmid LIC products, vortexed and let it sit at room temperature for 30 min for the annealing to occur. We transformed the LIC mixture into DH5alpha and plated the cells after 1 hour of incubation.

No colonies were observed on the negative control plate (linearized plasmid). However, we observed some colonies on the experimental plate. We picked 2 colonies (circled using a blue marker) to perform a colony PCR analysis.

We picked what seemed like bacterial colonies from the experimental plate and inoculated them overnight to perform a colony PCR analysis.

Well Set-Up
Well #	Well Setup
Well 1	DNA ladder
Well 2	Colony 1 (Original Primers, Tm = 62°C )
Well 3	Colony 1 (Control Primers, Tm = 55°C )
Well 4	Colony 2 (Original Primers, Tm = 65°C )
Well 5	Colony 2 (Control Primers, Tm = 55°C )
Well 6	Control Plasmid

Learn: According to Figure 12, Colony 1 seemed to have our recombinant vector, as we could see bands of approximately 2 kB size, which matched with the weight of our insert ~ 2kB. However, we were unable to grow colonies from the successful DH5alpha transformation plate. Because of time constraints, we decided to divert all of our resources towards developing our bacterial expression systems.

Results from the colony PCR indicated presence of the recombinant vector in Colony 1, however, we decided that this alone was not a definitive proof of successful cloning, as we had to discontinue our lab experiments for SF9. We were successful in optimizing the PCR conditions for our GC rich insert and plasmid sequences, and possibly also optimized conditions for the Ligation Independent Cloning.

Conclusion

We went through multiple cycles of design and testing to optimize the conditions for PCR amplification of our highly GC rich DNA sequences and possibly also the LIC reaction.

However, we were not able to continue our lab efforts past the colony PCR, which by itself is not sufficient proof of successful cloning. We discontinued the progress for the sake of focusing on performing successful expression in E. coli cell lines, which we achieved a month and a half later. However, by showcasing our work with SF9, we hope that future teams, given adequate time and resources, will be able to further optimize the wet lab experiments in order to successfully achieve expression from SF9 cells. This would be the first step in industrializing the process and bringing another option except for expression for creating an injectable protein S therapy.

Considering we have adequate time to continue our SF9 experiments, our next steps would be to purify enough of our recombinant plasmid by mini-prep in order to be able to sequence it. If confirmed by sequencing, further steps would include transformation into E. coli DH10Bac and following the Baculovirus expression system protocols.

---

Discover