Introduction

In order to test our synthetic transcription factors (sTFs) and furfural converting enzymes (FCE), we had to engineer Aspergillus niger to express them. There are several methods to do this, with varying levels of complexity and time required. The easiest and quickest method is to use a transient expression platform (TEP) while a more consistent and stable solution is genomic integration. Initially, we attempted genomic integration of an expression cassette, as we wanted to characterise our proteins in more stable clones. A TEP could potentially be developed later.

Cycle 1

Design

We based our expression cassette and method of genomic integration on methods described in the book chapter: “Genomic Editing: CRISPR-Cas9” by Hoof et al. 2018. Two different types of plasmids were designed; a plasmid expressing cas9 and gRNA, as well as a donor plasmid.

Cas9- and gRNA-expressing plasmid

This plasmid has a cas9 gene and a robust gRNA expression system, see figure 1, which is compatible with multiplexing. It is structured such that gRNAs are flanked by tRNAs, enabling the release of gRNAs from the transcript by polymerase III (Nødvig et al. 2018). A hygromycin resistance gene enables selection after transformation, increasing the efficiency of the system and an AMA1 element (BBa_K3046024) is also included, enabling efficient replication in A. niger. Lastly, elements such as origin of replication and ampicillin resistance encoding genes are present for E. coli work (BBa_K3385073).

Figure 1: Cas9 plasmid, with inserted protospacer for IS-nig1 (BBa_K4129030)

Donor plasmid

The expression cassette has the promoter PgpdA (BBa_K4129024) and terminator TtrpC (BBa_K678036) and is USER compatible, thus allowing insertion of a gene of interest (GOI) in the expression cassette, see figure 2 for plasmid with inserted arz7774 (BBa_K4129007). Downstream of the expression cassette is the selection marker pyrG (BBa_K3385072), which can be used for selection in pyrG defficient A. niger during cultivation on minimal media. pyrG facilitates uridine and uracil synthesis pathways. The selection marker is flanked by double-repeat regions, facilitating the excision of the selection marker by counter-selection. Double repeat regions are unstable parts of the genome and can be spontaneously excised unless selected for. This event can also be enhanced by counter selection towards the gene flanked by the double repeats. This makes it possible to use the same selection marker for several rounds of transformations. Gene expression cassette and pyrG with double repeats are flanked by homologous regions of two kilobases each, each flanking the target-site in A. niger, allowing for precise and efficient integration into the fungal genome. The plasmid shares the same E. coli elements as the Cas9 plasmid.

Figure 2: Donor plasmid map, with inserted arz7774 (BBa_K4129018)

Build

Both of the above plasmids were developed in-house. Multiple Cas9 plasmids were assembled via USER cloning, each with a different protospacer for the gRNA. Multiple plasmids were assembled to increase the chance that at least one plasmid had an efficient gRNA. A donor plasmid for each FCE was also assembled via USER cloning.

Test

In order to save time, we did not assess the efficiency of each gRNA, but went ahead with cotransformation of Cas9 and donor plasmid. Several colonies were obtained, which were streaked clean on new plates. Tissue PCR was performed, but we could not validate the entire expression cassette was inserted, even after several attempts with changes made to the method and extra controls.

Learn

Validating the genomic insertion of the expression cassette was much harder than expected. Several changes to the system could be made to skip this step. For instance, if the expression cassette had a selection marker, this could be screened for. Alternatively we could turn to the less stable solution and develop the TEP. This requires no tissue PCR to validate the expression system, as PCR can be performed before the transformation.

Cycle 2

Design

Due to time constraints, we decided to develop the TEP as the next step, see figure 3 for plasmid map. Our TEP has several components that enable easy construction in E. coli, while also being somewhat stable in A. niger, granted it is selected for. A USER cassette flanked by the promoter PgpdA and terminator TtrpC enables easy construction using USER cloning. An AMA1 element enables A. niger to replicate the plasmid and the pyrG selection marker was also present to allow selection of correctly transformed mutants. An ori element and ampR enables stability in E. coli.

Figure 3: TEP plasmid map, with inserted Arz7774 (BBa_K4129027)

Build

The elements used to construct our TEP were amplified by PCR with USER-primers. The expression cassette was amplified from the previously described donor plasmid (figure 2) and the plasmid backbone from another in-house plasmid. Amplicons were analysed on gels and assembled to our TEP using USER cloning.

With our final TEP, we could now insert our GOIs into the expression cassette via USER cloning.

Test

The testing was two-fold as we had to test our sTFs as well as our FCE. All A. niger clones survived on minimal media meaning they were successfully transformed and had the plasmid conveying uridine and uracil synthesis pathways present.

Clones with FCE were characterised using a BioLector for cultivation and sTF clones were cultivated in 96 well plates and on agar plates. The results were of mixed success; we did not prove the FCE worked, but some sTFs did show activity. Thus the TEP worked for our initial search of usable sTFs.

Learn

While we knew a TES is not as stable as a genomic insertion, it was a simple way to express proteins. With the data we have gathered through the use of TESs, we could conclude that it was a beneficial platform to develop, as we did get some useful data out of it. However, we still needed a more robust expression platform, to further characterise our proteins.

Cycle 3

Design

While the first Cas9-based design of a genomically integratable expression cassette could not be proven to be inserted, other Cas9-based genome integration systems exist. The DIVERSIFY platform (Jarczynska et al. (2021)), developed in-house at DTU, can be used to create mutants easily validated by screening due to the presence of a marker in the expression cassette. This is useful as this circumvents the bad performance of our tissue PCR.

The DIVERSIFY system requires more time, as two transformations are needed before the desired clones have been made. This led to us to initially go with a simpler Cas9-based system, as described in the first design phase.

The DIVERSIFY system consists of two plasmids, as well as two Cas9 plasmids.

Common synthetic gene integration site (COSI) plasmid

This plasmid is used to establish the expression cassette in the genome. It has a similar structure to the donor plasmid described in the first design phase, but with additional elements, see figure 4. The same promoter and terminator encoding genes are used, but a screening gene β-glucuronidase (uidA) from E. coli is also inserted. This expression cassette is flanked by small homology regions facilitating gene expression cassette exchange. This segment is then flanked by the larger homology regions which were also described in the first design phase, facilitating genomic insertion into A. niger. Downstream of the gene expression cassette, between the small and large homology region, the selective gene pyrG is present.

Figure 4: COSI plasmid map.

This plasmid contains the expression cassette with the GOI. There are two versions of this plasmid; the one we would use for FCE includes promoter and terminator, while the second one does not. Small homology regions facilitate cassette exchange with the integrated platform.

Figure 5: GEC plasmid map with empty expression cassette.

Cas9 plasmids

The only difference between the two Cas9 plasmids needed to use the DIVERSIFY system is the gRNA used. One plasmid is used to integrate the DIVERSIFY platform into the genome, while the other is used for cassette exchange, by targeting the uidA gene.

With these plasmids, we would be able to integrate the DIVERSIFY platform and use uidA to screen for colored colonies, while later inserting our GOI with the GEC plasmid, now looking for non-colored colonies, as the uidA would be deleted (figure 6).

Figure 6: Overview of the DIVERSIFY platform. A) shows the expression cassette exchange by uidA targeting by Cas9 and GEC as the donor. Small homology regions facilitate this exchange, shown as A1 and B1. B) shows the screening process used to find clones with exchanged expression cassettes. For the integration of the DIVERSIFY platform, the screening plate would be the opposite, so control (C) would not be colored, and surrounding colonies would be. Figure 1 from Jarczynska et al. (2021).

Build

Due to time constraints, we did not get to assemble our DIVERSIFY plasmids.

References

Hoof, J.B., Nødvig, C.S., & Mortensen, U.H.. Genome Editing: CRISPR-Cas9. In: de Vries, R., Tsang, A., Grigoriev, I. (eds) Fungal Genomics. Methods in Molecular Biology, vol 1775, (2018)

Nødvig, C.S., Hoof, J.B., Kogle. M.E., Jarczynska, Z.D., Lehmbeck, J., Klitgaard, D.K., & Mortensen, U.H.. Efficient oligo nucleotide mediated CRISPR-Cas9 gene editing in Aspergilli. Fungal Genetics and Biology, 115, 78-89. (2018)

Jarczynska, Z.D., Rendsvig, J.K.H., Pagels, N., Viana, V.R., Nødvig, C.S., Kirchner, F.H., Strucko, T., Nielsen, M.L. Mortensen, U.H.. DIVERSIFY: A Fungal Multispecies Gene Expression Platform, American Chemical Society, 10, 579-588. (2021)

Introduction

The objective of this subproject was to identify, assemble and express FCE in A. niger, and thus lower the sensitivity to furfural.

Cycle 1

Design

Looking through literature, several candidate genes for conversion of furfural into other compounds were identified. Specifically, two conversion reactions were identified, in which either furfuryl alcohol or furoic acid were produced from furfural. As literature showed furoic acid might be less toxic than furfuryl alcohol, we chose two putative enzymes that could convert furfural to furoic acid (arz7774 and hmfH), and one that could convert furfural to furfuryl alcohol (fucO). The enzymes have been registered as: arz7774 (BBa_K4129007), fucO (BBa_K4129008) and hmfH (BBa_K4129009)

arz7774 is a putative aldehyde dehydrogenase from Amorphotheca resinae, another filamentous fungi and was identified by Wang. et al. (2015), fucO is an Lactaldehyde oxidoreductase from Escherichia coli identified by Koopman. et al. (2013) and hmfH is a putative oxidoreductase from Cupriavidus basilensis identified by Zheng et al. (2010)

Build

The amino acid sequence of the enzymes were identified and a DNA sequence was generated and codon optimised for A. niger by using GenScripts GenSmart Codon Optimization tool. These were ordered as gene blocks from IDT.

Two rounds of cloning were performed, as we first tried to use the genomically inserted expression cassette, as described in the section: Gene expression in Aspergillus niger. As the insertion could not be validated by tissue-PCR, we switched to the TEP. Both cloning methods involved USER cloning, thus the lab work was near identical.

With USER cloning, the plasmid with fucO could not be constructed, thus only two FECs were built with arz7774 and hmfH.

Test

Several rounds of testing were conducted using a BioLector and HPLC samples were analysed. There was no indication that our A. niger clones with FEC had a higher tolerance towards furfural compared to wildtype A. niger.

Subsequent HPLC analysis of the growth media as well as samples containing a predefined amount of furfural showed no traces of furfural, meaning that the equipment possibly had been calibrated wrongly prior to testing.

Learn

As we could not prove the FECs had a beneficial effect on A. niger, it begs the questions, why? The chosen FECs are putative and might not fold correctly in A. niger. The enzymes could also be transported to an incorrect compartment in A. niger, as the current build does not include a localization-peptide. It could also be beneficial to test the enzyme in vitro, if we can replicate the required conditions and have the necessary substrates available.

Adding a localization-peptide would be quite easy to do, using the current TEP. However, it is hard to say if that would lead to the enzymes behaving as we initially expected.

References

Koopman, F., Wierckx, N., Winde, J.h., & Ruijssenaars, H.J. Identification and characterization of the furfural and 5-(hydroxymethyl)furfural degradation pathways of Cupriavidus basilensis HMF14. Proceedings of the National Academy of Sciences. 107(11), 4919-4924 (2010)

Zheng, H., Wang, W., Yomano, L.P., Geddes, R.D.,,Shanmugam, K.T.& Ingram, L.O. Improving Escherichia coli FucO for Furfural Tolerance by Saturation Mutagenesis of Individual Amino Acid Positions. Applied and Environmental Microbiology, 79(10), 3202-3208, (2013)

Wang, X., Gao, Q. & Bao, J. Transcriptional analysis of Amorphotheca resinae ZN1 on biological degradation of furfural and 5-hydroxymethylfurfural derived from lignocellulose pretreatment. Biotechnol Biofuels 8, 136 (2015)

Introduction

The objective was to evaluate the activity of three fungal promoters (scFDH1, BBa_K4129001; anFDH, BBa_K4129002; scFDH2, BBa_K4129003) in A. niger when exposed to furfural.

Cycle 1

Design

To test the activity of the fungal promoters in A. niger, we decided to transiently express BBa_K4129004, BBa_K4129005, BBa_K4129006 in a plasmid based system. As a plasmid vector for our constructs we chose BBa_K3046021 as we had access to it in our lab.

Build

The plasmids were assembled by USER cloning and replicated in E. coli. Initial validation of the plasmids by restriction digestion showed clear discrepancies between the assembled plasmids and the maps of the designed ones. However, BBa_K3046021 had been successfully used at DTU in 2019 and Sanger sequencing showed that all essential components of our constructs (AMA1 plasmid replication region, fungal selection marker, and prokaryotic selection marker) were present. We concluded that the discrepancies between the assembled and designed plasmids were due to minor errors in the maps, and that the assembled plasmids were functionally equivalent to the designed one. Therefore we proceed to test our constructs in A. niger.

Test

Multiple attempts at transforming our plasmids into A. niger failed. After confirming that the A. niger protoplasts used were viable, we concluded that our plasmids were not functional.

Learn

After realising that the stock of BBa_K3046021 we had was not reliable, we began a new iteration of the DBTL cycle taking care of thoroughly validating each component before settling on which design to build.

Cycle 2

Design

We redesigned our plasmids using BBa_K4129006 as the backbone. BBa_K4129006 was selected as we were able to validate its sequence and successfully transform it into A. niger.

Build

The new plasmids were successfully assembled, validated, and transformed into A. niger.

Test

The A. niger strains carrying the plasmid expression systems were grown in liquid media. Subsequently, half of the samples were induced with furfural. Fluorescence and absorbance were monitored to quantify the mCherry reporter protein and biomass produced by the strains. Multiple negative and positive controls were included. Of particular interest for the learning phase of this cycle were the two positive controls which carried BBa_K3046004 either on a plasmid or as a genomic insertion as the allowed for a comparison of the two expression systems.

Learn

The results showed that the three fungal promoters tested were neither constitutive nor induced by furfural in A. niger, see the results here. As the goal of this track was to investigate the activity of these promoters during furfural exposure, the engineering cycle was not repeated. Out of the positive controls, the strain carrying a genomically integrated BBa_K3046004 showed higher average fluorescence and lower variance compared to its plasmid expression system counterpart. This indicates that when investigating promoter activity, strains with genomic insertions are preferable over transient expression systems.

Cycle 1

Design

The sTFs are a part of the synthetic expression system (SES) we developed. The design of the sTFs will thus depend on the SES. The SES are designed to be a double fungal expression system, where the sTF will be constitutively expressed from the sTF expression unit and regulate the expression of the reporter cassette. The sTF expression unit consists of the fungal promoter PgpdA (BBa_K4129024), a sTF like FunsTF05 (BBa_K4129103), and the fungal terminator TcgrA (BBa_K3669007). The other functional unit of the SES is the rapporter cassette, which consists of the synthetic promoter 6xLexO-gpdA-minimal promoter (6x-LexO-Pmin) (BBa_K4129115), mCherry (BBa_J06504) and the fungal terminator TtrpC (BBa_K678036).

Our synthetic transcription factors were designed with a modular approach. This was to allow easy engineering of the sTFs due to the fact that the sTFs are composed of three different functional domains fused together with 2 linker regions. The three domains are; 1) a DNA-binding domain, 2) a ligand sensing domain and 3) a transactivation domain. The following domains were used for the modularity: One DNA-binding domain, namely LexA, two ligand sensing domains, namely HbaR and Hmox1 and lastly two transactivation domains, namely B112 and VP16. Lastly, the linker between DNA-binding domain and ligand sensing domain can be one of two possible modules: a short version and a longer version. This gives eight possible sTFs in total. Each sTF also includes the nuclear localization sequence (NLS) SV40, which ensures transport into the cell nucleus (Garcia-Bustos et al (1991)).

It has not been experimentally proven that HbaR and Hmox1 interact with furfural. A study on Hmox1 indicated through computational analysis that Hmox1 is able to bind to furfural or vice versa (Santhakumar et. al (2021)). HbaR is a transcription factor that can sense benzoic acid and derivatives thereof (Egland et. al (2000, Castaño-Cerezo (2020))). Therefore, to increase the binding affinity of furfural to the sTF, sixteen rational mutations were designed. These additional sixteen ligand sensing domains of HbaR were named HbaR1-16. This means that, in total, 72 different sTF were designed, eight of them were labelled as original sTF and the remaining 64 as HbaR variants.

A more indept discussion about our synthetic transcription factors and promoter can be seen here.

Build

The assembly of the double expression system was carried out using USER cloning in E. coli DH5a. The assembly was standardised so that each sTF could be assembled in bulk using the same protocol. The assembly consisted of 2 rounds of assembly, with the first round of assembly being of the sTF expression units and the sTF reporter cassette. The next round was fusing the two cassettes together on our TEP.

The sTF expression units and the sTF reporter cassette were assembled independently of each other, so when a sTF expression unit was ready to be assembled into the TEP, there would be no bottlenecks.

Test

The transient expression plasmids was used to find a proof of concept. The expression cassettes were tested both on solid media plates and in liquid culture. The media used was minimal media, minimal media with 2 mM benzoic acid and minimal media with 0.6 g/L furfural.

The solid media plates with the A. niger carrying the double expression system on TEP was incubated 4 days at 30°C. The fluorescence was assessed in Gel200 Vilber Smart Imagine Fusion FX and it was shown that FunsTF02, FunsTF04, FunsTF05 and FunsTF70 did fluoresce consecutively with different efficiency. See the results here.

The liquid culture testing was performed in a CellStar 48-well cultivation plate. The optical density and fluorescence was measured in a ClarioStar plate reader. The measurements were taken after induction. Results showed that FunsTF05, FunsTF18 and FunsTF70 also displayed fluorescence in liquid.

Learn

There were some SES where the sTF showed constitutively expressed mCherry. This was a good proof of concept for our SES, but a vast majority of the tested strains did not work. The sTF needs to be further engineered to be induced in response to furfural or benzoic acid. This engineering requires a good assay and the assays performed so far could be improved, especially in liquid media.

Cycle 2

Design

The liquid assay could be improved by transferring from transient expression to genome integration, because the expression will be more stable. This system is the Gene expression in A. niger. In addition to stable expression, an improvement could be to measure the dynamics of the promoter. This could be achieved by using a BioLector, allowing further investigation of the expression over time as measurements are taken automatically at regular intervals by the machine.

In short, the SES consists of many components which can be optimised. We would focus on the sTF, because it needs to be engineered more in order to be inducible. The first step would be designing a SES without any sTF to understand whether mCherry expression from 6xLexO-Pmin was leaky or not . The next steps will be more attempts at characterization, especially to understand the ligand sensing domains. We would fuse a GFP to the sTFs to identify their cellular location. The cellular location will indicate if the sTFs, that did not work were due to the fact they were not in the nucleus or something else. It would also be interesting to remove the SV40 NLS, to see if the sTF is transported into the cell nucleus without the NLS.

References

Castaño-Cerezo S, Fournié M, Urban P, Faulon JL, Truan G. Development of a Biosensor for Detection of Benzoic Acid Derivatives in Saccharomyces cerevisiae. Front Bioeng Biotechnol. 2020 Jan 7;7:372. doi: 10.3389/fbioe.2019.00372. PMID: 31970152; PMCID: PMC6959289.

Egland PG, Harwood CS. HbaR, a 4-hydroxybenzoate sensor and FNR-CRP superfamily member, regulates anaerobic 4-hydroxybenzoate degradation by Rhodopseudomonas palustris. J Bacteriol. 2000 Jan;182(1):100-6. doi: 10.1128/JB.182.1.100-106.2000. PMID: 10613868; PMCID: PMC94245.

Erill I, Escribano M, Campoy S, Barbé J. In silico analysis reveals substantial variability in the gene contents of the gamma proteobacteria LexA-regulon. Bioinformatics. 2003 Nov 22;19(17):2225-36. doi: 10.1093/bioinformatics/btg303. PMID: 14630651.

Garcia-Bustos J, Heitman J, Hall MN. Nuclear protein localization. Biochim Biophys Acta. 1991 Mar 7;1071(1):83-101. doi: 10.1016/0304-4157(91)90013-m. PMID: 2004116.

Hirai H, Tani T, Kikyo N. Structure and functions of powerful transactivators: VP16, MyoD and FoxA. Int J Dev Biol. 2010;54(11-12):1589-96. doi: 10.1387/ijdb.103194hh. PMID: 21404180; PMCID: PMC3419751.

Ottoz DS, Rudolf F, Stelling J. Inducible, tightly regulated and growth condition-independent transcription factor in Saccharomyces cerevisiae. Nucleic Acids Res. 2014;42(17):e130. doi: 10.1093/nar/gku616. Epub 2014 Jul 17. PMID: 25034689; PMCID: PMC4176152.

Radman M. SOS repair hypothesis: phenomenology of an inducible DNA repair which is accompanied by mutagenesis. Basic Life Sci. 1975;5A:355-67. doi: 10.1007/978-1-4684-2895-7_48. PMID: 1103845.

Santhakumar P, Prathap L, Roy A, Jayaraman S, Jeevitha M. Molecular docking analysis of furfural and isoginkgetin with heme oxygenase I and PPARγ. Bioinformation. 2021 Feb 28;17(2):356-362. doi: 10.6026/97320630017356. PMID: 34234396; PMCID: PMC8225605.