FDCA-decarboxylating monooxygenase-deficient host cells for producing FDCA
20200102584 ยท 2020-04-02
Assignee
Inventors
- Johannes Adrianus Maria de Bont (Wageningen, NL)
- Harald Johan Ruijssenaars (Gorinchem, NL)
- Jan Werij (Gorinchem, NL)
Cpc classification
C12N9/0073
CHEMISTRY; METALLURGY
C12P17/04
CHEMISTRY; METALLURGY
International classification
Abstract
The invention relates to fungal cells for the production of FDCA. The fungal cell has genetic modification that reduces specific 2,5-furandicarboxylic acid (FDCA) decarboxylating monooxygenase activity in the cell, as compared to a corresponding parent cell lacking the genetic modification. The fungal cell can further be genetically modified to increase the cell's ability to oxidize furanic aldehydes to the corresponding furanic carboxylic acids. The invention also relates to a process for the production of 2,5-furan-dicarboxylic acid (FDCA) wherein the cells of the invention are used for oxidation of a furanic precursors of FDCA to FDCA.
Claims
1. A fungal cell comprising a genetic modification that reduces specific 2,5-furandicarboxylic acid (FDCA) decarboxylating monooxygenase activity in the cell, as compared to a corresponding parent cell lacking the genetic modification, wherein the genetic modification eliminates the expression of an endogenous gene encoding an FDCA decarboxylating monooxygenase by deletion of at least a part of at least one of the promoter and the coding sequence of the gene.
2. The fungal cell according to claim 1, wherein the endogenous gene in the corresponding parent cell encodes a FDCA decarboxylating monooxygenase comprising an amino acid sequence with at least 45% sequence identity to at least one of SEQ ID NO: 4.
3. The fungal cell according to claim 1, wherein at least the complete coding sequence of an endogenous gene encoding an FDCA decarboxylating monooxygenase is deleted.
4. The fungal cell according to claim 1, wherein the expression of all copies of the endogenous gene encoding an FDCA decarboxylating monooxygenase is eliminated.
5. The fungal cell according to claim 1, wherein the cell has the natural ability to oxidize HMF to FDCA.
6. The fungal cell according to claim 1, wherein the cell comprises a further genetic modification that is at least one of: a) a genetic modification that confers to the cell the ability to oxidize 5-hydroxymethyl-2-furancarboxylic acid (HMFCA) to 5-formyl-2-furoic acid (FFCA) or that increases in the cell the specific activity of an enzyme that oxidizes HMFCA to FFCA as compared to a corresponding wild type cell lacking the genetic modification; and, b) a genetic modification that confers to the cell the ability to oxidize furanic aldehydes to the corresponding furanic carboxylic acids or a genetic modification that increases in the cell the specific activity of an enzyme that oxidizes furanic aldehydes to the corresponding furanic carboxylic acids, as compared to a corresponding wild type cell lacking the genetic modification.
7. The fungal cell according to claim 6, wherein the genetic modification in a) is a modification that increases expression of a nucleotide sequence encoding a polypeptide with HMFCA dehydrogenase activity, which polypeptide comprises an amino acid sequence that has at least 45% sequence identity with the amino acid sequence of at least one of SEQ ID NO.'s: 1 and 2; and/or, wherein the genetic modification in b) is a modification that increases expression of a nucleotide sequence encoding a polypeptide having furanic aldehyde dehydrogenase activity, which aldehyde dehydrogenase has at least one of the abilities of i) oxidizing HMF to HMFCA, ii) oxidizing DFF to FFCA, and, iii) oxidizing FFCA into FDCA, which polypeptide comprises an amino acid sequence that has at least 45% sequence identity with the amino acid sequence of SEQ ID NO.: 3.
8. The fungal cell according to claim 1, wherein the cell further comprises a genetic modification selected from: a) a genetic modification that reduces or eliminates the expression of a gene encoding a short chain dehydrogenase that reduces HMF and/or FFCA to the corresponding alcohol, wherein preferably the gene is at least one of a gene encoding polypeptide comprising an amino acid sequence with at least 45% sequence identity to at least one of SEQ ID NO: 5 and 25; b) a genetic modification that increases expression of a nucleotide sequence encoding a polypeptide that transports at least one furanic compound, which polypeptide preferably comprises an amino acid sequence that has at least 45% sequence identity with the amino acid sequence of at least one of SEQ ID NO.'s: 6-10; and, c) a genetic modification that alters the expression of a gene encoding a transcriptional activator of genes involved in furan catabolism, wherein preferably the gene is a gene encoding a polypeptide comprising an amino acid sequence with at least 45% sequence identity to SEQ ID NO: 11.
9. A fungal cell having the ability to oxidize HMF to FDCA and comprising a genetic modification that increases expression of a nucleotide sequence encoding a polypeptide that transports at least one furanic compound, which polypeptide preferably comprises an amino acid sequence that has at least 45% sequence identity with the amino acid sequence of at least one of SEQ ID NO.'s: 9 and 10.
10. The fungal cell according to claim 9, wherein the cell further comprises a genetic modification selected from: a) a genetic modification that eliminates or reduces specific FDCA decarboxylating monooxygenase activity in the cell, as compared to a corresponding parent cell lacking the genetic modification; b) a genetic modification that confers to the cell the ability to oxidize HMFCA to FFCA or that increases in the cell the specific activity of an enzyme that oxidizes HMFCA to FFCA as compared to a corresponding wild type cell lacking the genetic modification; c) a genetic modification that confers to the cell the ability to oxidize furanic aldehydes to the corresponding furanic carboxylic acids or a genetic modification that increases in the cell the specific activity of an enzyme that oxidizes furanic aldehydes to the corresponding furanic carboxylic acids, as compared to a corresponding wild type cell lacking the genetic modification; d) a genetic modification that reduces or eliminates the expression of a gene encoding a short chain dehydrogenase that reduces HMF and/or FFCA to the corresponding alcohol, wherein preferably the gene is at least one of a gene encoding polypeptide comprising an amino acid sequence with at least 45% sequence identity to at least one of SEQ ID NO: 5 and 25; e) a genetic modification that increases expression of a nucleotide sequence encoding a polypeptide that transports at least one furanic compound, which polypeptide preferably comprises an amino acid sequence that has at least 45% sequence identity with the amino acid sequence of at least one of SEQ ID NO.'s: 6-8; and, f) a genetic modification that alters the expression of a gene encoding a transcriptional activator of genes involved in furan catabolism, wherein preferably the gene is a gene encoding a polypeptide comprising an amino acid sequence with at least 45% sequence identity to SEQ ID NO: 11.
10.-16. (canceled)
17. The cell according to claim 1, wherein the cell is a filamentous fungal cell selected from a genus from the group consisting of: Acremonium, Aspergillus, Aureobasidium, Cryptococcus, Fihbasidium, Fusarium, Humicola, Magnaporthe, Mucor, Mycehophthora, Neocalhmastix, Neurospora, Paecilomyces, Penicillium, Piromyces, Schizophyllum, Talaromyces, Thermoascus, Thielavia, Tolypocladium, Trichoderma, and Ustilago, preferably the cell is a filamentous fungal cell selected from a species from the group consisting of: Aspergillus niger, Aspergillus awamori, Aspergillus foetidus, Aspergillus sojae, Aspergillus fumigatus, Talaromyces emersonii, Aspergillus oryzae, Mycehophthora thermophila, Trichoderma reesei, Penicillium chrysogenum, Penicillium simplicissimum and Penicillium brasihanum; or, wherein the cell is a yeast cell selected from a genus from the group consisting of: Saccharomyces, Kluyveromyces, Candida, Pichia, Schizosaccharomyces, Hansenula, Kloeckera, Schwanniomyces, Yarrowia, Cryptococcus, Debaromyces, Saccharomycecopsis, Saccharomycodes, Wickerhamia, Debayomyces, Hanseniaspora, Ogataea, Kuraishia, Komagataella, Metschnikowia, Williopsis, Nakazawaea, Torulaspora, Bullera, Rhodotorula, and Sporobolomyces, preferably the cell is a yeast cell selected from a species from the group consisting of Kluyveromyces lactis, S. cerevisiae, Hansenula polymorpha, Yarrowia hpolytica, Candida tropicalis and Pichia pastoris.
18. A process for oxidizing HMFCA to FFCA, the process comprising the step of incubating a fungal cell according to claim 5, in the presence of HMFCA, under conditions conducive to the oxidation of HMFCA by the cell, wherein the cell expresses enzymes that have the ability to oxidize HMFCA to FFCA.
19. A process for producing FDCA, the process comprising the step of incubating a fungal cell according to claim 5, in a medium comprising one or more furanic precursors of FDCA, preferably under conditions conducive to the oxidation of furanic precursors of FDCA by the cell to FDCA, and, optionally recovery of the FDCA, wherein preferably, at least one furanic precursor of FDCA is selected from the group consisting of HMF, 2,5-dihydroxymethyl furan (DHF), HMFCA, FFCA and 2,5-diformyl furan (DFF), of which HMF is most preferred, wherein the furanic precursors of FDCA are obtained from one or more hexose sugars, preferably one or more hexose sugars obtained from lignocellulosic biomass, preferably by acid-catalyzed dehydration, and, wherein preferably the FDCA is recovered from the medium by a process comprising acid precipitation followed by cooling crystallization and/or solvent extraction.
20. The process according to claim 19, wherein the medium has a pH in the range of 2.0-3.0, wherein preferably the FDCA precipitates from the acidic medium in which it is produced and is recovered from the medium by a process comprising acid precipitation followed by cooling crystallization.
21. A process for producing a polymer from at least two FDCA monomers, the process comprising the steps of: a) preparing an FDCA monomer in a process according to claim 19; and, b) producing a polymer from the FDCA monomer obtained in a).
22. The process according to claim 21, wherein the polymer is produced by mixing the FDCA monomer and a diol monomer and bringing the mixture in a condition under which the FDCA and diol monomers polymerise.
23. Use of a fungal cell according to claim 5, for the biotransformation of one or more of furanic precursors to FDCA or a fungal cell expressing one or more bacterial enzymes with the ability to convert a furanic precursors of FDCA into FDCA, wherein preferably, at least one furanic precursor of FDCA is selected from the group consisting of HMF, DHF, HMFCA, FFCA and DFF, of which HMF is most preferred.
Description
DESCRIPTION OF THE FIGURES
[0150]
[0151]
[0152]
EXAMPLES
Introduction
[0153] In co-pending International application PCT/EP2016/072406 the inventors have described that the isolation from Dutch soil of a Penicillium brasilianum strain. This P. brasilianum, referred to P. brasilianum C1 in application PCT/EP2016/072406, was isolated by growth selection on HMF, a precursor for FDCA. Herein, the P. brasilianum C1 strain will be referred to as the wild type (WT) WT P. brasilianum strain or the WT strain. International application PCT/EP2016/072406 describes the sequencing and annotation of the genome of P. brasilianum C1, as well as the sequencing of the transcriptome of the strain grown on HMF and on citric acid. Based on the genome annotation and blasting genes against public databases as well as on differential expression RNA-sequencing results, a list has been compiled of candidate genes that are involved in encoding enzymes that are involved in the catabolism of HMF by P. brasilianum C1 via a proposed HMF to furoic acid (via FDCA) pathway, including a salicylate hydroxylase hmfK1, two alcohol dehydrogenases hmfL1 and hmfL2 and a salicylaldehyde dehydrogenase hmfN1. By heterologous expression in a Pseudomonas strain, the Examples of PCT/EP2016/072406 show that the P. brasilianum hmfK1 hydroxylase indeed acts as a decarboxylating monooxygenase on FDCA and thus is involved in the degradation of FDCA in P. brasilianum. Furthermore, by heterologous expression in yeast of the P. brasilianum hmfL1 and hmfL2 alcohol dehydrogenases and the P. brasilianum hmfN1 salicylaldehyde dehydrogenase, the Examples of PCT/EP2016/072406 show that these enzymes indeed have ability to efficiently oxidise HMF to FDCA.
[0154] In addition the P. brasilianum hmfK1, hmfL1, hmfL2 and hmfN1 genes, the Examples of PCT/EP2016/072406 described a number of further genes encoding protein or enzymes involved in involved in the catabolism of HMF in P. brasilianum, which are listed in Table 12.
TABLE-US-00026 TABLE 12 Genes in the P. brasilianum C1 genome identified as being involved in HMF catabolism Gene Function role inHMF aa aa nt name Contig (annotated) catabolism SEQ ID NO: length SEQ ID NO: hmfL1 82 alcohol HMFCA oxidation 1 351 12 dehydrogenase to FFCA Zn-binding hmfL2 153 alcohol HMFCA oxidation 2 339 13 dehydrogenase to FFCA Zn-binding hmfN1 730 salicylaldehyde HMF/FFCA oxidation 3 505 14 dehydrogenase to HMFCA/FDCA hmfK1 730 salicylate FDCA 4 427 15 hydroxylase decarboxylation FAD binding monooxygenase hmfM 273 short chain reduction of 5 245 16 dehydrogenase HMF/FFCA to the corresponding alcohol hmfT3 254 major superfamily furan transport 6 581 17 facilitator protein hmfT4 730 major superfamily furan transport 7 513 18 facilitator protein hmfT5 1 ABC transporter furan transport 8 1435 19 hmfT6 226 major superfamily furan transport 9 527 20 facilitator protein hmfT7 570 major superfamily furan transport 10 514 21 facilitator protein hmfR 730 transcriptional induction furan 11 872 22 activator catabolism genes
Example 1: Attempt to Delete hmfK1 and Identification of hmfK3
[0155] Initial attempts were made to delete the first salicylate hydroxylase gene hmfK1 using a nia1 selection marker. Although diagnostic PCR confirmed the deletion of hmfK1, RNAseq analysis revealed that hmfK1 is still expressed, indicating that the deletion was not successful. Interestingly, in the analysed presumed mutant strain yet another salicylate hydroxylase, hmfK3 (SEQ ID NO.: 23, encoded by SEQ ID NO.: 24), was upregulated that was not identified as being upregulated in the Examples of PCT/EP2016/072406. The presumed hmfK1 deletion transformants were next analysed in more detail. From this analysis we concluded that that both hmfK1 deletion transformants were not correct. Most likely the deletion construct had integrated at other site(s), possibly up/downstream of hmfK1 and/or at the nia1 gene itself (different in both mutants).
Example 2: Evaluation of RNAseq Data
[0156] We obtained RNAseq data from the WT and the presumed hmfK1 deletion transformant SH #2E4 (which turned out not to have correct hmfK1 deletion) from chemostat cultures grown on citrate and then replacing the citrate by HMF. A set of genes possibly involved in the degradation of HMF via FDCA was identified. The hmfK1 gene was the gene which showed the strongest induction in the HMF-fed culture in the WT strain. A number of HMF-induced genes are located on the same contig (no. 730) as hmfK1 is indicating the presence of a gene cluster encoding a HMF degradation pathway. As already discussed above, hmfK1 was still strongly induced in transformant SH #2E4. In addition, a second hydroxylase gene with homology to hmfK1, now termed hmfK3, was discovered to be induced in this transformant, but not in the WT strain. Since no FDCA accumulation was observed in both the WT and mutant cultures, we reanalysed the available RNAseq RPKM data for identification of potential alternative targets for improved FCDA strain design (e.g. degradation specific regulator gene).
[0157] Data re-analysis was focused on (log.sub.2) ratios of: [0158] 1. WT_HMF vs WT_Citrate [0159] 2. M_HMF vs M_Citrate [0160] 3. M_HMF vs WT_HMF [0161] 4. M_Citr vs WT_Citr
[0162] The data was first floored (+2) to be able to look at genes with ratio 0 and div0. Selections of genes were made by using arbitrary cut off values to get comparable amount (100) of induced genes (see Table 2 and the Supplementary Excel File).
[0163] In the mutant (compared to the WT) culture large sets of genes were induced and repressed similarly in both HMF and citrate conditions. This suggests the induction/repression of these genes are HMF independent and mutant and/or culture specific. An exception is the induction of the salicylate hydroxylase hmfK3.
[0164] The previously obtained results (induction of hmfK1, hmfK3, alcohol dehydrogenases, transcriptional activator, lactonase etc.) were confirmed in the HMF-treated cultures. Salicyl related genes and decarboxylases are only induced by HMF and oxidoreductases mainly regulated by HMF. HMF also induced various transporters, like g10375 (contig 570) and g5964 (contig_226), encoding Major Facilitator Superfamily proteins, which could have a role in the transport of HMF, or its derivative into and out of the cell.
[0165] The BLAST results of the hmf genes previously identified in the Examples of PCT/EP2016/072406, and possibly involved in HMF degradation via FDCA were re-evaluated. Interesting to see is that the HMF-induced gene cluster around hmfK1 (contig730) shows a clear resemblance with a fungal second metabolite cluster. The highest identity (70%) was discovered with an orthologue gene cluster in Sporothrix schenckii ATCC 58251. The hmfL1 gene (encoding an alcohol dehydrogenase lies on a different contig (82) than hmfK1 (contig 730). However, the hmfL1 orthologue in S. schenckii is in the same gene cluster. In the publicly available P. brasilianum MG11 genome (Horn et al., 2015. Genome Announc 3(5):e00724-15. doi:10.1128/genomeA.00724-15; GenBank acc. no. CDHK01000001.1) the hmf gene cluster (containing hmfK1) also includes hmfL1. Therefore we conclude contigs 82 and 730 are adjacent to each other.
[0166] The co-factor dependence of dehydrogenases in the hmf gene cluster (contigs 730 and 82) was also investigated. For an economically viable production of FDCA in fungi NAD as the co-factor is preferred, since fungi have a higher capacity to re-generate NAD than NADP.
TABLE-US-00027 Gene NCBI Annotation Co-factor hmfL1 CEJ57635 alcohol dehydrogenase NAD (contig_82) hmfN1 CEJ57637 aldehyde dehydrogenase NAD (contig_730) (NADP?)
[0167] BLAST results with hmfL1 show a preference for NAD as the co-factor. The BLAST results with hmfN1 does not discriminate for NAD and NADP dependence. Best match for hmfN1 however is with a salicylaldehyde dehydrogenase (DoxF, SaliADH, EC=1.2.1.65) involved in the upper naphthalene catabolic pathway of Pseudomonas strain C18. For DoxF NAD seems the preferred co-factor.
[0168] Based on re-evaluation results of the RNAseq data we also conclude that salicylate hydroxylases hmfK1 and secondly hmfK3 are the best initial candidates to delete in order construct a host cell that oxidises HMF to FDCA but which does not degrade the FDCA produced.
Example 3: Development of hmfK1 & hmfK3 Single and Double Mutant Strains
Gene Deletion Design and Synthesis
[0169] The WT P. brasilianum strain has been tested for phleomycin and hygromycin sensitivity to confirm the suitability of phleo or hygB antibiotic selection markers for gene disruption (see above). Gene deletion design has been performed by DDNA and is based on a split marker approach (Arentshorst et al., 2015, In: Genetic Transformation Systems in Fungi, Volume 1. van den Berg M. A., Maruthachalam K., (eds). Switzerland: Springer International Publishing, pp. 263-272) using the hygB selection marker for the hmfK1 deletion mutant and the phleo selection marker for the hmfK3 deletion mutant (
Transformation & Selection
[0170] Before the transformations were performed an FDCA degradation screen was set up for the WT P. brasilianum strain. The WT strain is able to use FDCA as a carbon source. Creating mutants that are unable to do this will not grow on media containing FDCA as the sole C-source. This assay will facilitate screening for correct hmfK1 (and hmfK3) deletion transformants (assuming deletion of these genes will impair FDCA degradation).
[0171] Corbion minimal medium was used, which contains the following per liter of demineralized water: 2.0 g of (NH.sub.4)2SO.sub.4, 0.1 g of MgCl.sub.2 6H.sub.20, 10 mg of EDTA, 2 mg of ZnSO.sub.4 0.7H.sub.20, 1 mg of CaCl.sub.2). 2H.sub.20, 5 mg of FeSO.sub.4 7H.sub.20, 0.2 mg of Na.sub.2MoO.sub.4. 2H.sub.20, 0.2 mg of CuSO.sub.4 5H.sub.20, 0.4 mg of CoCl.sub.2 6H.sub.20, and 1 mg of MnCl.sub.2 2H.sub.20. 25 mM KH.sub.2PO.sub.4 and 25 mM NaH.sub.2PO.sub.4. One or more of the following C-sources were added depending on the experiment: 0.2 g/l of yeast extract, 15 mM of FDCA, 5 mM glucose, 5 mM citric acid, 6 mM HMF. pH was set at 3.0, except when FDCA was used as sole C-source, to avoid precipitation of FDCA.
[0172] WT P. brasilianum was able to grow on Corbion minimal medium with 15 mM FDCA as sole C-source, both in liquid medium as well as on agar plates. The negative control (minimal medium without any C-source) showed a minimal amount of background growth, which must be due to the ability of this strain to grow on agar components. This is a beneficial side-affect, because in a screen with transformants this background growth is a positive control for actual transfer of spores. We decided to initially screen the transformants (see below) on agar plates with FDCA and without YE (more clear results and faster), i.e. with 15 mM FDCA as sole C-source.
[0173] Gene fragments were transformed to the WT P. brasilianum strain using a standard DDNA fungal transformation protocol (Punt and van den Hondel, 1992, Methods in Enzymol. 261:447-457). Protoplasts of the WT strain were transformed with the fragments to create hmfK1 and plated on selection plates containing HygB. In addition, protoplasts of the WT strain were transformed with fragments to create both hmfK1 and hmfK3 and plated on selection plates containing HygB alone, Phleo alone, and both HygB and Phleo. In total, 48 primary transformants were generated on the various selection plates. These primary transformants were screened based on no or reduced growth on plates containing FDCA as the sole C-source. Seventeen out of the 35 (50%) HygB+ (hmfK1) transformants showed a strong phenotype on plates (no growth on FDCA), suggesting FDCA degradation route involves an intact hmfK1 gene. Potential hmfK3 mutants showed no phenotype on plates. All transformants were additionally tested for the presence of antibiotic markers.
[0174] Next, the 17 HygB+(hmfK1) clones that showed a phenotype on plates and all hmfK1+hmfK3 transformants were purified on plates and tested in liquid medium 15 mM FDCA (YE). For all hmfK1 clones, that showed a strong phenotype on agar plates, growth was completely abolished in liquid medium with FDCA as the only C-source.
[0175] Diagnostic PCR analysis was performed with the different hmfK1 deletion transformants to discriminate between an intact and correctly deleted hmfK1 gene. Different primer combinations were tested (see Tables 13, 14 and 15) and the results showed that all hmfK1 clones showing a phenotype (no growth on FDCA) are correct (data not shown). Vice versa, hmfK1+hmfK3 transformants #31 and #33, showing no phenotype still produced WT hmfK1 PCR bands. Diagnostic PCR analysis was also performed on all hmfK1+hmfK3 transformants to test for correct hmfK3 deletion (see Tables 14 and 16). Only transformant #43 (hmfK1+hmfK3; HygB+Phleo selected) turned out to have a correct hmfK3 deletion based on PCR (data not shown). Although this transformant is HygB resistant, it does not contain a hmfK1 deletion (based on PCR results). All purified hmfK1+hmfK3 transformants were also tested in liquid cultures similarly as the hmfK1 transformants. Transformant #43 (which is a correct hmfK3 deletion) did grow on FDCA.
[0176] From the results we conclude that correct hmfK1 deletions were generated and that deletion of the hmfK1 gene results in a 100% inability to grow on FDCA as sole C-source under the conditions tested, confirming that hmfk1 is a key gene in FDCA degradation. Additionally, we can conclude that the hmfK3 deletion is correct in 1 clone (#43) and shows no phenotype, suggesting hmfk3 is not involved in FDCA degradation, at least not under the conditions tested.
TABLE-US-00028 TABLE 13 Primers for diagnostic PCR for verification of hmfK1 deletion. Primer No. Name Sequence SEQ ID NO: 180 salHupA-430 F GGTAGAAAAGGGGTTGCGAT 34 181 salHupB-1,954 F GAAGCAATCGTCGGAAGT 35 182 salHdownB-3,693 R CGCGAGGATGTATGGTATGAT 36 183 salHdownA-5,280 R GAAAGTGAGATTGTGGATGGA 37 184 salH-53 R TGTCGTTCATGGCTCCC 38 185 salH-1,583 F AAGGTGCGAACTGGAGAG 39 186 niaD-2,876 F GGAACAATCGGCAAAGAAAGTG 40 187 niaD-94 R CAACTTCTTGTGGCGTTGG 41
TABLE-US-00029 TABLE 14 Primers for diagnostic PCR for verification of hmfK1::hygB and hmfK3::phleo deletion-and split marker amplification. Primer No. Name Sequence SEQ ID NO: 212 5flanksalH1-489 F CTCGGTCGCTCTTTTGGGTA 42 213 5flanksalH1-1,499 R CAGGACATTGTTGGAGCCGA 43 214 3flanksalH1-57 F TCCGGAAGTGCTTGACATTG 44 215 3flanksalH1-1,455 R CCATCCCGAGTATTCTTTCGAG 45 216 5flanksalH2-1,350 F CCGTGCGCTATAATAACCTTCG 46 217 5flanksalH2-1,156 R GCGTCCCGGAAGTTCG 47 218 3flanksalH2-15 F GAGCGGTCGAGTTCTGGA 48 219 3flanksalH2-1,508 R GGTGAAAGAGTGATATATGAGGC 49 220 con.salH2 5flank-956 F TCTCAGGACTTGCAGATGTTG 50 221 con.salH2 3flank-2,838 R GCCAAGTCATCATCCTCGC 51 222 phleo-249 R CTTACTGCCGGTGATTCGAT 52 223 phleo-712 F CTACAGGACACACATTCATCGT 53 224 salH2-106 R TGCCAGCCCTCCTATTCC 54
TABLE-US-00030 TABLE 15 Expected sizes of PCR products of the designed primer combinations for verification of the hmfK1::hygB deletion. Primers hmfK3 (bp) wild type (bp) 212 + 215 5,504 4,580 180 + 213 3,164 183 + 214 3.142 182 + 181 2,641 1,740
TABLE-US-00031 TABLE 16 Expected sizes of PCR products of the designed primer combinations for verification of the hmfK3::phleo deletion. Primers hmfK3 (bp) wild type (bp) 216 + 219 4,900 4,566 216 + 224 1,606 220 + 222 1,752 223 + 221 2,126
Example 4: Conversion of HMF and Production of FDCA
[0177] We next set up experiments to monitor conversion of HMF and production of FDCA in time. From the results above the WT strain, confirmed hmfK1 deletions #26 and #30, and confirmed hmfK3 deletion #43 were chosen to be analysed for fermentation in shake flask cultures. Strains were grown on the Corbion minimal medium w/o yeast extract, containing 5 mM glucose or 5 mM citrate, supplemented with 6 mM HMF. Cultures were grown for 7 days and samples were collected after 1, 2, 3, 4 and 7 days. Supernatant from the samples was analysed for HMF, FDCA and intermediates by HPLC as described in the Examples of PCT/EP2016/072406.
[0178] All strains grew well on glucose containing medium. The hmfK1 deletion strains #26 and #30 did not significantly grow on citrate containing medium. Therefore, on day 4, 1% of the glucose containing culture (of #26 and #30) was used to re-inoculate the citrate containing flasks (=citr*). However, even than no (clear) further growth was observed in these cultures. The hmfK3 deletion strain #43 grew on citrate containing medium comparable as the WT strain. The pH was measured after 1 week (data not shown). Interestingly, the pH of the cultures of the hmfK1 deletion strains after 1 week incubation was lower than of the WT and hmfK3 deletion strain cultures in both media, indicating acid (FDCA) production.
[0179]
[0180] In the citrate containing cultures also no FDCA accumulation was measured in both the WT and hmK3 cultures (
[0181] From these results we conclude that hmfK1 is the key gene in FDCA degradation. Deletion of hmfK1 leads to FDCA accumulation when HMF is provided to the cultures. (Germinated) spore-based bioconversion seems possible and suggests no growth of cells is needed. However, in cultures growing in glucose containing medium FDCA accumulation is faster. Under the conditions tested, the hmfK3 gene is not involved in FDCA degradation.
Example 5: Conversion of HMF into Crystalline FDCA at Low pH
[0182] To further demonstrate the potential of the hmfK1 deletion mutant to produce FDCA under more controlled conditions, biotransformation experiments were performed in 7-L glass fermenters. Cells were cultivated at 30 C. in a mineral salts medium (per L: KCl, 1.3 g; KH.sub.2PO.sub.4, 3.8 g; MgSO.sub.4.7H.sub.20, 1.23 g; Na.sub.2EDTA.2H.sub.2O, 125 mg; ZnSO.sub.4.7H.sub.2O, 55 mg; H.sub.3BO.sub.3, 27.5 mg; MnCl.sub.2.4H.sub.20, 12.5 mg; FeSO.sub.4.7H.sub.2O, 12.5 mg; CoCl.sub.2.6H.sub.2O, 4.25 mg; CuSO.sub.4.5H.sub.2O, 4 mg; Na.sub.2MoO.sub.4.2H.sub.2O, 3.75 mg, glucose, 10 g; (NH.sub.4).sub.2SO.sub.4, 7.6 g or NaNO.sub.3, 9.8 g). Antifoam was added as needed and the dissolved oxygen tension was maintained at 20% of air saturation level. During growth, HMF was fed to a final concentration of 1 mM to induce the FDCA producing enzyme system. After the initial glucose was depleted, glucose was fed continuously until the biomass density reached 20 g/L of cell dry weight. The biotransformation phase was initiated by switching on an HMF feed at a starting rate of 2.5 mmol/L broth/h which was manually adjusted to prevent accumulation of HMFCA or HMF. Along with HMF, a feed of glucose was provided to cover the energy demands of the cells during the biotransformation. Glucose was monitored throughout, and the feed rate was manually adjusted such that no glucose accumulated. The pH was allowed to drop from the initial value (5.5) to a setpoint of 2.7 after which it was controlled by automated addition of NaOH. Alternatively the pH was allowed to drop without adjustment, in which case the value stabilized at around 2.3. During the biotransformation, HMF was transformed into FDCA which crystallized as a solid compound below pH 3.5. Transient accumulation of HMFCA was observed when HMF was fed at a too high rate; upon reducing the feed rate the accumulated HMFCA would typically be oxidized to FDCA. The biotransformations yielded FDCA up to an overall titer of 90 g/L.