Fig. 1. Steps of the engineering cycle
Our team successfully produced parts(BBa_K5254012, BBa_K5254013, BBa_K5254018)and achieved a groundbreaking milestone by engineering the world's first E. coli producing heparin strain. We also pioneered the development of the first-ever cysH knockout strain of E. coli Nissle 1917, while successfully expressing all five enzymes necessary for the heparin biosynthesis pathway. This was achieved by following the DBTL cycle, which is a fundamental principle of iGEM.
After researching existing methods and narrowing in on potential papers that had successfully expressed these enzymes in E.coli. We split these modification enzymes into two groups and integrated them into bacterial expression plasmid systems. We labeled the two enzyme plasmids Ginkgo A and Ginkgo B. By the end we ordered
We planned to carry out our project using in vitro, in vivo, and ex vivo approaches. The in vitro approach involved purifying our enzymes and performing the reaction outside the cell. The in vivo approach focused on synthesizing heparin within the cell, while the ex vivo approach was a hybrid of the two, where we mixed cell lysates instead of fully purifying the enzymes.
Name: BBa_K5254000
Part type: Generator
Short description: pLac+RBS+his+MBP+NDST2+terminator
This part is composed of a maltose-binding protein (MBP) sequence and the sequence of the NDST2 protein. A lac operator was used to control the expression of the part, such that a lac repressor can be integrated if gene expression regulation is necessary. The MBP has been widely used to enhance the solubility of the recombinant proteins, improving protein folding as well as preventing aggregation. This design was obtained from Choe’s 2017 paper, “Prokaryotic soluble expression and purification of bioactive human fibroblast growth factor 21 using maltose-binding protein”, in which human fibroblast growth factor 21 (hFGF21) was attached to several different tags, including MBP, and was expressed in E. coli. The MBP-tagged hFGF21 far outperformed the other tags in terms of solubilizing the protein. Each enzyme has a His tag attached. The terminator of the design is removed due to the complexity limitations of the IDT system, but it’s still in use with the Amp Golden Gate. Additionally, due to the Dou 2015 paper, NDST II was clarified to be truncated, codon-optimized, and also cut without transmembrane regions.
Did not happen because the order never actually arrived.
Fig. 2. Denied IDT order due to high complexity
Delayed ordering process: Our insistence on perfecting the order delayed progress. We should have embraced the iterative process of experimentation, as iGEM encourages learning from multiple cycles of testing and reordering. Unfortunately, our late submission prevented us from having enough time for a second round of ordering, limiting our ability to test, identify issues, and reorder materials.
Graciously synthesized by Ginkgo Bioworks!
Fig. 3. Diagram showing the order of enzymes and other elements included in our Ginkgo A plasmid
Name: BBa_K5254001
Part type: Generator
Short description: pTac+RBS+his+NDST2+C5+terminator
These enzymes are type II transmembrane proteins because they work at the Golgi apparatus in the mammalian cells. However, in E. coli, the transmembrane region is not required and even harms the expression of the enzymes. The extracellular domains of these enzymes have been successfully expressed and functional in E. coli (refer to PNAS and Jian Liu paper). Therefore, we designed the plasmid coding of the 6x his and MBP conjugated human NDST2 extracellular domain (from aa 64 to aa 883) and C5 epimerase extracellular domain (from aa 53 to aa617) under the tac promotor/lac operator with Lac I integration. We expect MBP conjugation to help the solubilization of the proteins and 6X his for the purification. IPTG will induce these enzymes with minimal background expressions. Two Sap I sites are integrated for further procedures such as golden gate methods.
pGA1 digested by Xho I gives 3.9Kb and 6.7 Kb fragments.
PGB1 digested by HInd III gives 1.0 Kb and 6.6 Kb fragments, However, it was incomplete digestion with low plasmid load, so 1 Kb band is barely visible. Instead, we can see 7.6 Kb (nicked one) and 6.6 Kb fragments are visible.
Fig. 4. Diagram showing the order of enzymes and other elements included in our Ginkgo B plasmid
Name: BBa_K5254013
Part type: Generator
Short description: pTac+RBS+his-MBP-2OST+his-MBP-6OST-1+his-3OST+terminator
This plasmid contains 2-OST, 6-OST1, and 3-OST1. The extracellular domains of these enzymes have been shown to expressed and solubilized well with the conjugation of MBP (3), therefore we exactly follow the design from the previous paper. MBP conjugated Chinese hamster 2-OST extracellular domain (from aa 51 to aa 356) and mouse 6-OST1 extracellular domain (from aa 62 to aa 411) are integrated under the tac promoter/lac operator. 3-OST extracellular domain can be expressed in E. coli without MBP (3), then, we also include 6x his tagged mouse 3-OST1 extracellular domain (from aa 48 to aa 311). Two Sap I sites are inserted at the 5’ and 3’ flanking region of enzyme coding regions for further application to combine pGA1 and pGB1 into a single expression vector.
Coding regions for all enzymes are codon optimized for bacterial expression. We designed whole plasmid sequences and Ginkgo Bioworks kindly synthesized the plasmids.
Because we ordered the plasmids at the same time, they arrived together. Thus the building of A1 and B1 happened with a lot of aspects overlapping.
As we successfully knocked out cysH gene and replaced it to Chloramphnicol (cmR) cassetle, we moved forward to transform A1 plasmid into cysHKO/nissle 1917.
First round: Failed to transform procedures
We tested two types of transformation: Electroporation vs Heatshock
Transformation condition
Electroporation (Bio-Rad Gene Pulser Xcell Electroporation)
Heat shock into chemically competent cells
There was no colony by testing different conditions. However given our prior knowledge that A1 plasmid could be transformed into BL21 without issue, these results suggested:
Tested the efficacies of the electrocompetent cells and chemically competent cells
In order to test the efficacy of our electrocompetent cysH KO cells and chemically competent cells, we tried transformations with puc19 which is a different plasmid(not A1). This step was taken to test if it was a problem with cell competency or something else. This helped to rule out the problem.
1. Transform 50 pg of puc19 to into the cysHKO/Nissle1917
2. Transform 50 pg of puc19 into 100 ul of cysHKO/Nissle1917 (chemically competent cells).
Fig. 5. Electroporation vs heat shock number of competent cells
Surprisingly, the heat shock method had higher efficacy compared to electroporation method. These results suggest that the heat shock method is suitable for transforming plasmids into cycHKO/nissle1917. This signaled to us that from now on, we should stick to the heat shock method.
Transform A1 into cysHKO/nissle 1917 by heat shock method
Based on the previous heat shock result (42C, 30 sec), we changed the condition to 42C 70sec.
Fig. 6. Condition 1: plasmid 30 ng. Condition 2: plasmid 120 ng. Condition 3: plasmid 1500 ng
Plating onto LB/Amp agar (2 plates each condition: one plate is low volume, the other is high volume)
Fig. 7. Transformation plasmid A1 into ctsHKO/ nissle 1917 through heat shock procedure
Transformation efficacies were dependent on the plasmid amount. 42C, 70sec condition works on cysHKO/nissle 1917. Based on the graph, condition three is the most optimal condition for transformation.
Fig. 8. Confirmation of A1/cysHKO Nissle 1917 with restriction enzyme digestions
We screened 16 colonies with restriction enzyme (RE) digestion. All 16 clones grew in LB/Amp/Cm media, but 10/16 clones showed aggregation during culturing. We picked 6 colonies (aggregation+, aggregation-) and performed Xho1 digestion (expected band size 3.9K bp 6.7K bp). Three clones (#7, #12 and #14) which didn’t show aggregation were positive clones.
Fig. 9. CBB staining of NDST2 and C5epi induced by IPTG
A1/cysHKO/Nissle 1917 was grown in LB/Amp/Cm media at 37°C overnight at 250 rpm. The next day, the overnight culture was diluted 1/10th, and IPTG was added after 2 hours of culture.
N: Nissle 1917, #12: A1/cysHKO/Nissle 1917 confirmed by RE digestion.
The bacteria were stimulated with IPTG for 2 hours at 37°C and then cultured for 20 hours at room temperature. Samples were collected 4 hours post-IPTG induction (2 hours at 37°C and 2 hours at room temperature).
The results showed that IPTG induced NDST2 and C5epi, although expression was still detected even without IPTG.
Fig. 10. Detection of IPTG induced NDST2 and C5epi expression by Western blot
NDST2 and C5epi induction by IPTG
A1/cysHKO/Nissle 1917 was grown in LB/Amp/Cm media at 37°C overnight at 250 rpm. The next day, the overnight culture was diluted 1/10th, and IPTG was added after 2 hours of culture.
N: Nissle 1917, #7, #12, and #14: A1/cysHKO/Nissle 1917 confirmed by RE digestion.
Lanes 1-6: The bacteria were stimulated with IPTG for 2 hours at 37°C and then cultured for 20 hours at room temperature (RT). Western blot samples were collected 4 hours post-IPTG induction (2 hours at 37°C and 2 hours at RT).
Lanes 7-10: The bacteria were stimulated with IPTG for 20 hours at RT.
Enzyme expression was detected using an anti-His tagged antibody.
The results showed that IPTG induction at 37°C led to higher expression compared to RT induction, although expression was still detected even without IPTG. Despite adding LacI, we still observed background expression. It’s possible that Nissle 1917 has a different regulation of the lac operon, and we should note that Nissle 1917 has relatively high background expression, even with LacI induction.
Additionally, we observed several bands that may indicate degraded enzyme products.
Based on these results, we proceeded with the functional assays, which include Alcian blue staining and toluidine blue staining.
Fig. 11. Detection of sulfated products in A1/cysHKO/nissle 1917 induced by IPTG
Alcian blue staining: Detection of sulfated products induced by A1/cysHKO nissle 1917 (NDST2 and C5epi)
A1/cysHKO/nissle 1917 was grown in LB/Amp/Cm media at 37C overnight at 250 rpm and next day the overnight culture was diluted 1/10th. IPTG was added after 2 h culture.
N: nissle 1917, #7, #12 and #14: A1/cysHKO/nissle 1917 confirmed by RE digestion.
A) The bacteria was stimulated with IPTG for 2 h at 37C and cultured them for 20 hat RT. The samples were collected at 4 h post IPTG induction (2h at 37C and 2h RT).
B) The bacteria was stimulated with IPTG for 20 h at RT.
We didn’t see significant Alcian blue staining even though IPTG was added. It suggests that the entire heparin enzyme pathway may be needed to see strong smear staining. So we decided to move forward to transform the B1 plasmid into A1/cysHKO/nissle 1917 without trying to get a clear blue stain with only the first half of the pathway (A1).
For the in-Vitro approach we transformed our plasmids into BL21. Our plan was to express and purify proteins to do an in-vitro assay that would allow us to characterize the individual functions of our protein.
Fig. 12. Enzyme induction in BL21 (CBB staining)
Enzymes were induced with IPTG (0.5 mM) for 4 hours and the expressions were confirmed by CBB staining.
Fig. 13. His western blotting in BL21
Negative = Pre-induced sample(without iptg)
Expected MW for A1
NDST2 (NDST2 +MBP+6XHis), 138.65 KDa, 1223aa
C5a epimerase (C5a epimerase +MBP+6XHis), 81.07 KDa, 714aa
Expected MW for B1
2OST (2OST +MBP+6XHis), 81.07 KDa, 714aa
6OST (6OST1+MBP+6XHis), 86.85 KDa, 755aa
3OST (3OST + 6X His), 32.77 KDa, 278aa
The bands are the rest sizes for their respective proteins. A1 has NDST2 and C5 Epimerase while B1 has 2OST, 3OST, and 6OSt.
After performing a histag purification we attempted to run it on a gel.
Fig. 14. Purification of enzymes(CBB)
Purification of his tag A1 and B1 in BL21
Lane 1) LADDER → 15 uL Lane 2) A1 Lysate → 20 uL Lane 3) A1 Pellet → 20 uL Lane 4)
A1 Flow Through → 20 uL Lane 5) A1 Final → 50 uL Lane 6) LADDER → 15 uL Lane 7) B1
Lysate → 20 uL Lane 8) B1 Pellet → 20 uL Lane 9) B1 Flow Through → 20 uL Lane 10) B1
Final → 50 uL
We observed a significant amount of protein in the residual pellets after cell lysis, suggesting that our lysis conditions were not strong enough. Unfortunately, the gel showed excessive protein in the pellet lane, resulting in a large smear, making it difficult to confirm our enzymes from this data. We plan to redo the gel to obtain clearer results. However, we do have complementary data that suggests correct enzyme expression in the Nissle knockout, so we expect the proper enzymes are being expressed in BL21. Further optimization of the lysis conditions may be necessary.
Lesson learned: The cells were insufficiently lysed, as evidenced by the gel results. We should have included a sonication step, but we lacked access to a sonicator, preventing proper cell lysis. Next time, we should use different conditions and a sonicator to ensure more effective lysis.
We tested both plasmids A1 and B1 together because the enzymes in our system are split between two different plasmids but need to work cooperatively. Specifically, the enzymes on the B1 plasmid require the end products of the enzymes on the A1 plasmid, making it impossible to test the B1 enzymes separately. The only practical way to assess enzyme functionality is by examining the final product, heparin or sulfated heparosan, as a result of the entire enzyme set.
While our in vitro process should have run more smoothly and provided more data on our enzymes, design issues—such as using a His-tag for all enzymes—made it impossible to assess them individually. Additionally, our attempts to detect enzyme activity using the universal sulfotransferase kit faced significant challenges due to high background interference. We were unable to gather data for the PAPS assay of enzyme activities, but we unexpectedly obtained a groundbreaking toluidine blue-stained gel that suggested the presence of heparin.
Toluidine Blue Staining
After successfully knocking our entire pathway into the cysH KO Nissle 1917 cell (reference data in Assembly Ginkgo A1/cysHKO/Nissle 1917(In-vivo)) we attempted to run the supernatant of our cells on the gel.
Fig. 15. Introduction of pGA1 and pGB1 in CysH knock out cells produce acidic materials which show smear in Toluidine blue staining
Toluidine blue is a metachromatic dye used for staining acidic molecules. Sulfated molecules are stained purple to red, while other acidic molecules appear blue. CysH KO or CysH KO cells transformed with pGA1 and pGB1 were cultured with the indicated concentrations of IPTG at 37°C or 25°C for 12 hours, followed by an additional 12 hours at ambient temperature. Supernatants were collected, and acidic heparosan was semi-purified using DEAE. Bound materials were extracted in SDS sampling buffer, and these samples were applied to SDS-PAGE (4-20%) and stained with toluidine blue (0.2%). Purple-stained smears, consistent with heparin, were observed, and the density of these smears was analyzed and indicated.
Especially under 37°C induction, the purple smears became denser as IPTG concentrations increased. A representative heparin sample stained with toluidine blue is also shown.
The presence of purple staining, consistent with heparin, is highly promising data that signifies the production of highly sulfated heparosan or a heparin-like product. Either way, this data strongly suggests that we have successfully synthesized a product similar to heparin.
PAPS assay for in vitro assay (NOT YET COMPLETED)
The Assay was planned but we have not confirmed getting purified enzymes yet. Because of this, we have not done this assay yet.
What we hope to see if the constructs are working properly:
We will detect enzymatic activity by measuring PAPS consumption (PAP generation). However, since these enzymes need to work together—the substrate of each step being the product of the previous one—we cannot measure individual enzymatic activity. Instead, we must combine all enzymes from Ginkgo A1 and B1 with heparosan. This allows us to observe the activity of the entire enzymatic cascade. We incubate all purified enzymes (isolated by nickel resin) with cysH knockout cell lysate (which provides PAPS and heparosan) and measure PAP generation using a PAPS assay kit. Notably, there is likely to be endogenous free phosphate in the sample, so we need to compare readings with and without IMPAD. The difference will indicate the actual amount of PAP produced. Additionally, since background PAP may be present in the cysH knockout cell lysate, we also need to measure the PAP levels without the purified enzymes.
Expected results:
General concerns:
We learned that no single condition works universally. Even with the same host cells, we observed significant variability in expression efficacy. A great deal of optimization was required, even for bacterial transformation. Generally, electroporation is more efficient than the heat shock method. However, in our case, we were unable to achieve successful transformation of the cysH knockout Nissle strain using electroporation—only heat shock worked. This was unexpected and may be explained by PAPS overexpression affecting the bacterial phenotype.
Design Specific concerns:
Integrating a marker like GFP to easily detect the desired colonies would have saved considerable time. Furthermore, we did not add specific tags for each enzyme, so we identified them based solely on their expected molecular size. However, we also observed enzyme degradation, which could introduce uncertainty regarding the molecular size. Adding a specific tag for each enzyme would help confirm expression more precisely and evaluate whether the individual enzymes were functional.
Plasmid B1 did not have a LacI so we didn’t have as much control over expression because we were using the natural one.
Fig. 16. Diagram showing the order of genetic components included in our knockout cassette
5’ and 3’ homology arms complement the 5’ and 3’ UTR of the cysH gene locus in Nissle 1917. Primers are targeted onto the 5’ and 3’ regions of the chloramphenicol-resistant gene (CMR) (ID 42) in ID42 parts.
Name: BBa_K5254018
Part type: Device
Short description: PCR fragment Chloramphenicol resistance gene flanked with 5’ and 3’ region of cysH gene from Nissle 1917
Chloramphenicol resistance gene in ID42 parts from 2024 Distribution kit was amplified with primers directing 5’ end and 3’ end of CMR. These primers are also flanked by 40 nt from 5’ (3097442-3097403) and 3’ (3096628-3096667) UTR region of cysH gene (3096668-3097402) in Nissle 1917, which enables homologous recombination (5). PCR product should be CMR flanked with homologous arm to 5’ and 3’ UTR regions of the cysH gene (phosphoadenosine phosphosulfate reductase).
Fig. 13. CysH on GenBank
Fig. 17. CysH homolog in the Nissle 1917 genome
It is registered as phosphoadenosine phososulfate reductase, PAPS reductase, which is identical to CysH gene product. (cysH = phosphoadenosine phosphosulfate reductase)
Fig. 18. Graphical demonstration of CysH region on the Nissle 1917 genome
Fig. 19. Recombination of the knockout cassette into cysH locus by lambda Red recombinase system
The recombination works by transforming the lambda red recombinase into your host cell. The recombinase will cut up the genome and by some chance, the cmR which is designed with homology arms flanking the gene of interest(the gene you want to remove) will integrate into that position effectively knocking out the gene.
Fig. 20. PCR knockout cassette bands
PCR was performed by Super Fi II DNA polymerase. ~800 bp band were amplified. The band was cut out and PCR product was extracted by Qiagen gel extraction kit.
Fig. 21. Map of pKD46
Fig. 22. RE digestion confirmation of pKD46
Sac I is the unique site, therefore Sac I digestion gives 6.3 Kb band.
Fig. 23. CysH knockout confirmation
Nissle 1917 transformed with pKD46 was transiently induced by arabinose and prepared for electroporation. Cells were electroporated with the purified PCR product of CMR cassette. Bacteria were plated onto an LB agar plate with chloramphenicol and colonies were tested for the deletion of the cysH gene as follows.
Bacteria genome was extracted by SDS and isopropanol precipitation. 400 ul of bacterial culture was pelleted and lysed in 200 ul of 1% SDS for 5 min ambiently. 30 ul of 10 mM NaCl and 350 ul of isopropanol were added to form the precipitation. The precipitation was pelleted by centrifugation at 15000 g for 10 min. The pellets were resuspended with 200 ul water and used for the PCR reaction. Primers against 5’ UTR of CysH and mid portion of CysH were used for WT cysH gene detection. Primers against 3’UTR of CysH and mid portion of CMR were used for the detection of CMR cassette in CysH locus. Clone #7 has the band corresponding to CMR cassette integration and no band corresponding to CysH gene indicating successful Knockout of cysH gene.
Fig. 24. PAP generation is disturbed in cysH-KO/Nissle 1917
Wild-type Nissle 1917 has a higher PAP amount than CysH KO cells. The PAP amount is reflected by the difference in absorbance with and without IMPAD1. We did not have purified PAPS; therefore, cell lysates were used as the source of PAPS. We prepared a mixture of cell lysates from wild-type and CysH knockout Nissle 1917 to standardize the other background factors. The difference in absorbance in the presence of IMPAD1 increased as the wild-type Nissle cell lysate increased, while in the absence of IMPAD1, no increase was observed.
This assay does not directly measure the amount of PAPS but instead quantifies the amount of PAP generated from PAPS by CysH in E. coli. Therefore, in CysH knockout cells, we expected an accumulation of PAPS, which would correspond to a decrease in PAP levels. Notably, the assay measures the phosphate generated by IMPAD1 (added during the assay) from PAP, rather than PAP itself. Since free phosphate is present in the cell lysates from the start, we observed high background phosphate levels. To account for this, we calculated the difference in absorbance with and without IMPAD1, which reflects the phosphate newly generated by IMPAD1 and is correlated with PAP levels in the samples.
We anticipated that various factors under basal conditions could affect the measurements, so we mixed two different cell lysates (wild-type: high PAP, and CysH knockout: low PAP) in varying ratios to standardize the basal conditions for comparison. As shown in the figure, increasing the proportion of wild-type lysate led to a greater increase in phosphate after the addition of IMPAD1, while without IMPAD1, the increase in wild-type lysate did not affect the differences. These findings indicate that wild-type cell lysates contain more PAP compared to CysH knockout lysates, suggesting that CysH knockout cells have higher levels of PAPS.
Experimental design: While we successfully integrated the cmR gene into the cysH locus, we have yet to measure PAPS levels using a functional assay. We are currently using a universal sulfotransferase kit to measure PAP levels, which are produced from PAPS by sulfotransferase. Since cysH also metabolizes PAPS into PAP, lower levels of PAP in the absence of cysH were expected. However, this approach only provides indirect evidence of cysH activity, and basal PAP levels can fluctuate due to other factors. Therefore, careful experimental design, including the right controls, is required. Direct measurement of PAPS should be more reliable and provide stronger evidence of the cysH knockout, in addition to the existing genomic PCR data. It is likely that there is other method that could directly measure PAPS levels which would be easier next time.
Design: Similar to the enzyme plasmids, adding a visual marker like LacZ or GFP could facilitate easier and faster clone identification.
General Observations: An interesting observation, though not necessarily a design issue, was that knocking out cysH from the Nissle 1917 genome appeared to affect the cell's robustness, either making it stronger or weaker. After the knockout, we found that transformation via electroporation became extremely difficult, which contrasted with the ease of transforming plasmids into other strains.
After seeing the success of the knockout of cysH we wanted to venture into more uncharted territory by attempting a simultaneous knockout of cysH and knock in of our first sulfation enzyme(NDST2). If we were to successfully knock in the NDST2 gene it could also open potential into knocking in the entire pathway.
This approach will provide us the Nissle with high PAPS content and NDST2 expression. NDST2 expression will not be regulated by lac operon, instead NDST2 expression levels should be like PAPS reductase, which should be relevant to PAPS generation. Relevant expression levels of NDST2 to PAPS may be suitable for successful N sulfation. Proving a successful Knock-in could suggest
Fig. 25. Diagram showing the order of genetic components included in our knock-in cassette
Fig. 26. Plasmid design of knock-in cassette
Name: BBa_K5254003
Short description: PCR fragment containing NDST2 extracellular domain and Kanamycin resistance gene flanked with 5’ and 3’ region of cysH gene from Nissle 1917
NDST2 extracellular domain and Kanamycin resistance gene en bloc in MGC Mouse Ndst2 cDNA was amplified with primers directing 5’ end and 3’ end of NDST2 extracellular domain and Kanamycin resistance gene. These primers are also flanked by 40 nt from 5’ (3097442-3097403) and 3’ (3096628-3096667) UTR region of cysH gene (3096668-3097402) in Nissle 1917. PCR product should be NDST2+KanR flanked with homologous arm to 5’ and 3’ UTR regions of the cysH gene (phosphoadenosine phosphosulfate). Successful recombination will achieve cysH gene knockout and NDST2 integration. This cassette will replace the exact cysH gene locus, so NDST2 expression will be regulated by an intrinsic promoter.
Fig. 27. MGC mouse NDST2
Fig. 28. PCR for generation of the knock in cassette
After building and testing our knockout, we encountered several problems. Despite multiple attempts at homologous recombination, we were unable to obtain a single successful clone. This could be explained by the significant size difference between the target gene (~1kb) and the knock-in cassette (4.5kb). While many clones had kanamycin resistance (kanR) integrated, the ndst2+kanR was not in the correct genomic location (cysH locus). Increasing the homology arms of the knock-in cassette could help ensure more precise integration. Overall, the efficiency of homologous recombination was low or non-existent. Adjusting homologous arms or exploring other strategies to enhance recombination efficiency should be considered in future designs.
While our team would have liked to do many more cycles and potentially order new constructs, we were also extremely limited by time and resources. In this section, we detail potential orders we would have constructed and ordered.
Fig. 29. New knockout cassette design
We could put an FRT site in our cassette to insert genes later in vivo. We could also design the plasmid with all enzymes flanked by FRT sites. Under the expression of FLT, all enzymes will be inserted at the FRT site on the CMR knock-out cassette in vivo. This will make Nissle express all required enzymes for heparin biogenesis with access to endogenous heparosan and PAPS.
The new, improved design of plasmid B will include a lacI gene. Our original plasmid does not have lacI and relies on the natural one. Interestingly, we do not observe constitutive expression but instead see variation in IPTG induction. A GFP protein is also present to facilitate easier clone selection after transformation— something we struggled with in the original design.
We will also tag each enzyme differently, rather than using the same type of tag (His tag) for all. We will use several short sequence tags, such as the FLAG tag and Strep tag, in addition to the MBP tag. For example, 2-Ost could have the His tag, 6-Ost could use the FLAG tag, and 3-OST1 could use the Strep tag. This approach allows them to be solubilized with the MBP tag, and the distinct tags will make it easier to recognize each enzyme individually.
A GFP protein could also be present to facilitate easier clone selection after transformation— something we struggled a lot with during the build part of the cycle.
Fig. 30. New IDT order plasmid design
Main issue: The main issue was that according to IDT our order was too complex to synthesize. To fix this issue we could have made the construct using a conventional method such as cut and ligation. pMAL-c6T is a tightly regulated bacterial expression vector (lacI+) with an MBP tag. We could have generated a PCR product of NDST2 with cohesive restriction sites and ligated it into pMAL to create MBP-tagged NDST2, inducible by IPTG. This approach is straightforward, and we would likely have the plasmid by now if we had taken this route.
1. Duncan MB, Liu M, Fox C, Liu J. 2006. Characterization of the N-deacetylase domain from the heparan sulfate N-deacetylase/N-sulfotransferase 2. Biochem Biophys Res Commun 339:1232-7.
2. Badri A, Williams A, Awofiranye A, Datta P, Xia K, He W, Fraser K, Dordick JS, Linhardt RJ, Koffas MAG. 2021. Complete biosynthesis of a sulfated chondroitin in Escherichia coli. Nat Commun 12:1389.
3. Douaisi M, Paskaleva EE, Fu L, Grover N, McManaman CL, Varghese S, Brodfuehrer PR, Gibson JM, de Joode I, Xia K, Brier MI, Simmons TJ, Datta P, Zhang F, Onishi A, Hirakane M, Mori D, Linhardt RJ, Dordick JS. 2024. Synthesis of bioengineered heparin chemically and biologically similar to porcine-derived products and convertible to low MW heparin. Proc Natl Acad Sci U S A 121:e2315586121.
4. Nguyen AN, Song JA, Nguyen MT, Do BH, Kwon GG, Park SS, Yoo J, Jang J, Jin J, Osborn MJ, Jang YJ, Thi Vu TT, Oh HB, Choe H. 2017. Prokaryotic soluble expression and purification of bioactive human fibroblast growth factor 21 using maltose-binding protein. Sci Rep 7:16139.
5. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97:6640-5.
6. Praveschotinunt P, Dorval Courchesne NM, den Hartog I, Lu C, Kim JJ, Nguyen PQ, Joshi NS. 2018. Tracking of Engineered Bacteria In Vivo Using Nonstandard Amino Acid Incorporation. ACS Synth Biol 7:1640-1650.