Team:DTU-Denmark/Results

Back to Top

Overview

The goal of this project was to develop a publicly available toolbox for genetic engineering of K.phaffii due to its potential as an industrial cell factory. For this, we divided the project into three parts as the major milestones. First, we wanted to design CRISPR Cas9 tools to engineer K. phaffii. Second was to use the designed CRISPR Cas9 tools for genomic deletion and finally in genomic deletion in K. phaffii GS115 [1].

CRISPR-Cas9 tools for K. phaffii Genomic Deletion Genomic Integration

CRISPR/Cas9 tools for K. phaffii

The CRISPR-Cas9 system offers a powerful tool for genetic engineering, as it enables marker-free, scar less genome editing. We designed 16 different gRNAs for 4 integration sites and carried out USER cloning. The cloned plasmids were transformed into E. coli and colony PCR was carried out to identify transformants harbouring the correctly assembled plasmids. Based on the results of colony PCR, plasmids were purified from the E.coli transformants and sent for commercial Sanger Sequencing.

To assemble the sgRNA plasmids, 2 fragments from the pDIV153 plasmid were PCR amplified and an analytical gel was run to check the correct band sizes of these fragments. Fragment 1 was expected to have a band size of ~700 bps and Fragment 2, which are our designed gRNAs, were expected to have a band size of ~500 bps. (Fig. 1 and 2)

The PCR reactions containing our desired fragments were Dpn1 digested to remove residual methylated template DNA. The pDIV151 backbone was extracted from E. coli colonies, provided by our supervisor Louise Kastberg, following extraction these plasmids were linearized using AsiSI and Nb.BsmI. (Fig. 3)

USER cloning was carried out to assemble the pDIV151 plasmids containing fragment 1 and fragment 2. Instead of running reactions with 4 candidate gRNAs per integration site, we were advised to select 2 candidates per integration site based on the distance for the target, therefore we selected gRNA_Chr_1_Cand_1, gRNA_Chr_1_Cand_4, gRNA_Chr_2_Cand_1, gRNA_Chr_2_Cand_2, gRNA_Chr_4_Cand_1, gRNA_Chr_4_Cand_3, gRNA_Chr_3_Cand_1 and gRNA_Chr_3_Cand_2.

E. coli transformation

Transformed E. coli DHα were plated on LB-Amp plates and incubated at 37 °C overnight.

We did an initial transformation of our constructs containing gRNA_Chr_1_Cand_1 and gRNA_Chr_2_Cand_1 which resulted in 1-3 colonies. (Fig. 4) Furthermore a single colony grew on the negative control. We tested all 4 colonies using colony PCR to see if they were carrying the inserts of fragment 1 and 2.

A subsequent transformation was carried out with the remaining gRNA candidates denoted gRNA_Chr_1_Cand_4, gRNA_Chr_2_Cand_2, gRNA_Chr_4_Cand_1, gRNA_Chr_4_Cand_3, gRNA_Chr_3_Cand_1 and gRNA_Chr_3_Cand_2 on the picture above. As Fig. 5 shows, all of our constructs resulted in more than 30 colonies on each plate. There was no colony on the negative control and more than 300 colonies on the positive control. Following this, colony PCR was conducted to check if the inserts contain the correct fragments.

Colony PCR

Colony PCR was used as a screening method to find successfully assembled plasmids from the transformants prior to sequencing.

3 colonies from each plate were picked for colony PCR. However, since gRNA_Chr_1_Cand_1 had only one colony, it was subjected to colony PCR. Following this, the reaction mixtures were run on analytical gel (see Fig. 6 below). The expected band size of the correctly assembled plasmids is about 1300 bps.

Regarding gRNA_Chr_1_Cand_1 and gRNA_Chr_2_Cand_1, out of the 4 tested transformant colonies, gel check of constructs indicated that gRNA_Chr_1_Cand_1 and gRNA_Chr_2_Cand_1.2 were the successfully assembled constructs.

As it can be seen on Fig. 7 above, the following colonies also carried the correctly assembled plasmids: gRNA_Chr_2_Cand_2.2, gRNA_Chr_2_Cand_2.3, gRNA_Chr_4_Cand_3.1, gRNA_Chr_4_Cand_3.2, gRNA_Chr_4_Cand_3.3, gRNA_Chr_3_Cand_1.1, gRNA_Chr_3_Cand_1.3, gRNA_Chr_3_Cand_2.1, gRNA_Chr_3_Cand_2.2, gRNA_Chr_3_Cand_2.3.

We decided to repeat colony PCR for gRNA candidates gRNA_Chr_1_Cand_2, gRNA_Chr_1_Cand_3, gRNA_Chr_1_Cand_4 and gRNA_Chr_4_Cand_1. gRNA_Chr_1_Cand_3.2, gRNA_Chr_1_Cand_4.1, gRNA_Chr_1_Cand_4.2, gRNA_Chr_1_Cand_4.3, gRNA_Chr_4_Cand_1.1 turned out to be colonies harboring the desired constructs. (Fig. 8)

Sequencing

E. coli transformants harbouring correctly constructed plasmids based on colony PCR results were subjected to Sanger Sequencing. After successful sequencing results (see Fig. 9) cryostocks were made for all versions of the pDIV151 plasmids with gRNAs for further experimentation.

The template in the alignment above is the in silico plasmids containing the designed crRNA targeting the respective protospacer in the K. phaffii genome aligned with the sequencing result of the gRNA plasmid(visit our part collection).

In our attempt to design the toolbox for the methanotrophic K. phaffii cell factory, we were successful in identifying neutral integration sites and gRNAs which target the intergenic regions of the four chromosomes that can be utilized in the future when engineering K. phaffii.

Genomic Integration

In this step, we aimed to assemble the different repair templates, to insert into the genome of K. phaffii GS115 Δku70 strain, using Golden Gate assembly. Before we went ahead with the assembly of inserts, we checked the gRNA efficiency without providing the repair templates to the K. phaffii GS115 Δku70 strain. Once this was validated, the subgroups tried to assemble the inserts.

K. phaffii transformation without repair templates

K. phaffii transformation was an essential step for us to validate the efficiency of gRNAs since the design of the homology arms will be based on the transformation results. Since the Δku70 strain is non-homologous end-joining deficient, the transformation plates are suspected to result in none to few transformant colonies (significantly fewer colonies compared to the wild type). This was successfully validated in 6 designed gRNAs: gRNA_Chr_1_Cand_1, gRNA_Chr_2_Cand_1, gRNA_Chr_2_Cand_2, gRNA_Chr_4_Cand_3, gRNA_Chr_3_Cand_1, gRNA_Chr_3_Cand_2, thus enabling us to go ahead and design homology arms necessary to assemble the different inserts. As a positive control, wild type strain was subjected to transformation to observe the presence of colonies and the negative control was only water. (Fig. 1)

Once the gRNA efficiency was validated, the different inserts were PCR amplified before performing golden gate assembly (GGA). These included homology arms for the respective gRNAs, promoters, coding sequence and terminator.

Click on the tabs pMMO and Hemoglobin below to get more information on these two specific!

pMMO Hemoglobin

pMMO integration

For pMMO integration, homology arms, GAP1 promoter, TEF promoter, TEF terminator, PmoA, PmoB, PmoB fused to Venus and PmoC with both signal peptide and without signal peptide were amplified, as can be seen in Fig. 2,3,4,5 and 6. The PCR amplified products were purified and are ready for Golden Gate Assembly followed by E.coli transformation.

PCR amplification of homology arms for gRNA_Chr_2_Cand_1 and gRNA_Chr_2_Cand_2 was repeated to introduce BsmBI compatibility instead of BsaI, as an internal BsaI was overlooked in the ordered gBlock of all PmoB versions, which was not the same sequence as we worked with in Benchling.

Golden Gate Assembly

The aim of our Golden Gate Assembly experiments was to create the repair templates used in co-transformation of K. phaffii GS115 Δku70 with our previously constructed pDIV151_gRNA plasmids.

8 different repair templates were designed:

• Construct 1 containing PmoB_Venus
• Construct 2 containing PmoB_SP with Venus
• Construct 3 containing PmoA_SP
• Construct 4 containing PmoA
• Construct 5 containing PmoC
• Construct 6 containing PmoC_SP
• Construct 7 containing PmoB
• Construct 8 containing PmoB_SP

We attempted to integrate all the constructs with the repair templates. However, the constructs with PmoB versions failed due to unexpected BsaBI recognition site within the ordered gBlock. Results from other constructs are presented here.

E. coli transformation

DH5α competent E. coli cells were transformed and plated on YPD-NTC plates and incubated at 37°C overnight.

Looking at the construct plates (Fig. 7), we observed 7 small colonies on construct 2, 2 big colonies on construct 3 and 1 big colony on construct 6 plates, while negative control plates had no colonies and the positive control plates had many small colonies.

Based on these observations, we decided to pick the following colonies for colony PCR: 2 , 3.1, 3.2 and 6.

E. coli colony PCR

To screen for correctly assembled pGGAselect_construct 2/3/6 plasmids prior to plasmid extraction from main cultures, the following colonies were picked for colony PCR:

• pGGAselect_Construct 2
• pGGAselect_Construct 3.1
• pGGAselect_Construct 3.3
• pGGAselect_Construct 6

The expected band sizes of the constructs is the following:

Construct	Expected band size
pGGAselect_Construct 2	3014 bp
pGGAselect_Construct 3	1793 bp
pGGAselect_Construct 6	1832 bp

Fig. 8 shows the results of the colony PCR. When it comes to construct 3.2 and 6 the brightest bands can be observed around 1800 bps, which is promising, therefore we decided to send these for sequencing in the next step.

E. coli sequencing

The red lines in the sequencing results represent a couple of mismatches. This could have been due to the PCR amplification carried out prior to sequencing. It is also to be noted that, these identified mismatches have not been documented to have any effect on the expression of the gene of interest encoded to the repair template. Hence based on the sequencing results (Fig. 9 and 10) we further proceeded with amplification of repair templates before starting K. phaffii transformation.

K. phaffii transformation

Once the repair templates had been assembled in pGGAselect and success confirmed by sequencing, the repair templates were PCR amplified using primers that bind to HAU (FW) and HAD (RV)(Fig. 11), then column purified to be ready for co-transformation with a constructed site-specific Cas9-gRNA plasmid.

Previously prepared electrocompetent K. phaffii GS115 Δku70 1M sorbitol preserves with a determined competence of 6.5x10⁶ CFU/mL after electroporation were used for transformation (yeast competent cells protocol). After 3 days of incubation at 30°C colonies were obtained on plates for both constructs 3.2 and 6, respectively (Fig. 12). This result is important since there has not been a known documentation to the best of our knowledge for yeast electrocompetent cells transformation for K. phaffii making it novel. We were inspired by the S. cerevisiae electrocompetent cell transformation and successfully obtained visible results showing yeast competent cells protocol works effciently.

K. phaffii colony PCR

Following the transformation of K. phaffii, yeast colony PCR was carried out to confirm if PmoA_SP and PmoC_SP had been successfully integrated in the genome.

Based on Fig. 13, we sent PCR amplified DNA fragments from the respective integration sites in transformant colonies for sequencing, however they all turned out to be false positives.

Although we were not successful in integrating the repair templates in the K. phaffii genome, the fact that we obtained some colonies, when providing a repair template, indicates that integration is possible. Through multiple iterations it is certainly possible to obtain such an engineered strain and this will provide new opportunities in synthetic biology using K. phaffii to produce different products of interest.

Hemoglobin

For Leghemoglobin integration, homology arms for IS1C1, AOX1 promoter, AOX1 promoter 99% similarity, AOX1 promoter 89% similarity, FLD1 promoter, Leghemoglobin, Venus and TEF terminator were amplified, as seen on Fig. 14. The PCR amplified products were purified and are ready for GGA followed by E.coli transformation.

Golden Gate Assembly

The aim of our Golden Gate Assembly experiments was to create the repair templates used in co-transformation of K. phaffii GS115 Δku70 with our previously constructed pDIV151_gRNA plasmids.

4 different constructs with repair templates for Leghemoglobin were designed:

• Construct 1 containing Leghemoglobin with AOX1
• Construct 2 containing Leghemoglobin with AOX1_99
• Construct 3 containing Leghemoglobin with AOX1_89
• Construct 4 containing Leghemoglobin with FLD1

Furthermore, a fluorescent protein collection was also designed:

• Construct 5 containing codon optimized Venus with AOX1
• Construct 6 containing codon optimized Venus with AOX_99
• Construct 7 containing codon optimized Venus with AOX_89
• Construct 8 containing codon optimized Venus with FLD1
• Construct 9 containing GFP with AOX1
• Construct 10 containing mOrange with AOX1
• Construct 11 containing VENUS from kit with AOX1

E. coli transformation

DH5α competent E. coli cells were transformed and plated on LB-CAM plates and incubated at 37°C overnight.

As it can be seen on Fig. 15 and 16, we managed to obtain colonies on the following construct plates: 1,2,3,6,7,8. Therefore we proceeded to further investigate these colonies with colony PCR.

E. coli colony PCR

Colony PCR has been carried out using Q5 high-fidelity DNA polymerase. Afterwards, a gel check was performed 100V, 20 minutes on 1% agarose.

As Fig. 17 shows, we had 3 visible bands, which corresponds to construct 3, 7 and 8. The protocol was run on purified plasmids with construct 6 from 3 different colonies. From the above analysis it is clear that all three colonies possessed the desired plasmids with the inserts (Fig. 18), therefore we sent these for sequencing.

Sequencing

Sequencing results of construct 3,7 and 8 were false positives and show that the sequenced plasmids did not contain the GGA constructs. However, sequencing results from construct 6 was correctly assembled as seen in Fig. 19 and we used only this one construct for electroporation of K. phaffii.

K. phaffii transformation

With the confirmed sequencing result from construct 6, K. phaffii transformation was attempted.

The outcome of the controls were unexpected, as the positive control was not supposed to have any colonies, we suspect that we exchanged these by accident. After 3 days of incubation, we had 3 colonies on our construct 6 plates (Fig. 20), these were named 6A, 6B and 6C and were used in the next step: yeast colony PCR.

K. phaffii colony PCR

The primers used for the colony PCR target chromosome 1 and produce PCR products with the length of 719 bps, which corresponds to the band lengths for all 3 colonies, seen on the gel picture below. This means that the insertion was not successful, as that would have resulted in a band of ~3000 bps. (Fig. 21)

We aimed to integrate leghemoglobin as a proof of concept to show that the engineered K. phaffii can produce leghemoglobin which is the main flavoring and colouring ingredient in meat industry. A successful integration would be able to change the dynamics of meat industry and move the society towards circular economy. However, our attempt to do this was not fully accomplished due to the time constraints and road bumps associated with the Golden Gate Assembly of the repair template. We do believe that this approach of ours was critically evaluated at every point and with time, it is indeed possible to implement this system.

Throughout these experiments, the results we obtained have further strengthened the main goals of our project PHEAST. The design and strategy which we have currently developed can be used to dwell deeper into obtaining the methane consuming engineered K. phaffii cell-factory and bring our idea into reality.

Genomic Deletion

In this step, we aimed to understand the effect on K. phaffii upon deletion of non vital genes involved in the metabolism of methanol to formaldehyde. This experiment was designed as a proof to understand the genes which were deleted does not completely seize the growth of the cells but reduces the growth rate.

2 gRNAs each targeting the aox1, aox2, and adh2 genes were amplified and purified. (Fig. 1) Repair templates were obtained by annealing the single stranded in silico designed homology arms that target the 45 base pairs upstream and downstream of the targeted genes. Further, K. phaffii transformation was carried out and the knockouts were confirmed by colony PCR and sequencing.

PCR amplification of gRNAs for aox1, aox2, and adh2

E. coli transformation

The gRNAs were transformed into DH5α E. coli along with pDIV151 backbone using USER cloning. Colony PCR and sequencing results showed successful integration of gRNAs in the pDIV151 backbone. (Fig. 2 and 3)

Repair templates were annealed and the efficiency of gRNAs were tested. As expected, the strain did not survive without a repair template. Since repair templates were provided to the k. phaffii strain, colonies were observed on all plates. (Fig. 4)

K. phaffii transformation

K. phaffii transformation was carried out to check for knockouts of aox1, aox2, and adh2 simultaneously. Colony PCR indicated that we were successful in carrying out a single knockout in each of the genes and a double knockout of aox1 and adh2. (Fig. 5)

Although promising results from colony PCR were obtained in the form of bands expected for aox1, aox2, and adh2, sequencing results were inconclusive for aox1 and aox2. However, adh2 gene deletion was confirmed via sequencing as seen below. (Fig. 6)

Due to time constraints we did not conduct any protein assays to investigate if the K. phaffii adh2 mutant strain is incapable of producing the ADH2 protein, as expected.

Next step was to carry out the double knockout from the adh2 mutant strain and we were successful in this since aox1 gRNA candidates 1 and 2 showed successful colony PCR results. We expected to see a band at 500bps as seen in Fig. 7.

To validate the scar less deletion of aox1 the colony PCR samples were sent for Sanger Sequencing. The results we received were inconclusive, and we did not have a chance to repeat the sequencing. However, we wanted to test if the knockouts did indeed impair the growth of the K. phaffii GS115 Δku70+Δadh2+aox1 mutant strain when grown on methanol as sole carbon source.

A final Biolector experiment elucidated that the confirmed knockout of adh2 and the putative knockout of aox1 does indeed impair the growth of the K. phaffii mutant in presence of increasing levels of methanol compared to the K. phaffii GS115 Δku70 strain.

This verifies that our CRISPR-based toolbox for genomic integration/deletion performs well. Interestingly, as we can see in Fig. 8, the mutant strain shows slightly lower growth efficiency even without addition of methanol to the YP medium (look at the “MeOH: 0” graph), meaning that lack of ADH2 and possibly AOX1 expression somewhat impairs the overall performance of the strain. We came to this conclusion on the basis of the graphs which show that the growth rate for both strains strats to decline clearly at about 8% but it is very clear that the growth rate is much lower compared to the Δku70 strain.

Although we did not get to the point of understanding the difference in growth rate with 3 knockouts, we were successful in extracting and analyzing the growth rate patterns and the suspected difference as expected. This thus served as our first proof of concept that although the non vital genes were deleted, the growth rate is decline and not fully hampered and that the strain is functional.

---

References

[1]De Schutter, K., Lin, Y. C., Tiels, P., Van Hecke, A., Glinka, S., Weber-Lehmann, J., ... & Callewaert, N. (2009). Genome sequence of the recombinant protein production host Pichia pastoris. Nature biotechnology, 27(6), 561-566.

---