WO2024020252A1

WO2024020252A1 - A genetically engineered yeast producing lactic acid with improved lactic acid with improved lactic acid export

Info

Publication number: WO2024020252A1
Application number: PCT/US2023/061683
Authority: WO
Inventors: Owen RYAN; Lei Fang; Doug HERSHBERGER
Original assignee: Archer Daniels Midland Company
Priority date: 2021-07-22
Filing date: 2023-01-31
Publication date: 2024-01-25
Also published as: WO2023004336A1; WO2024020251A1

Abstract

Described herein are Schizosaccharomyces pombe yeast strains genetically engineered to produce lactic acid, wherein the best performing strains have two copies of a heterologous lactate dehydrogenase gene codon optimized for expression in S. pombe in combination with inactivation of specific alleles encoding pyruvate decarboxylase (PDC), alcohol dehydrogenase (ADH) and glycerol 3 phosphate dehydrogenase (GDP). The specific alleles that are deleted are PDC201, ADH1 ADH4, and/or GDP1. The exogenous LDH genes are expressed by the S. pombe actin promoter discovered to be a constitutive high activity promoter, which may be operably linked to any gene of interest for expression in S. pombe. Also disclosed are several variant LDH genes from a variety of species that produce high levels of lactic acid when expressed in S. pombe cells.

Description

A GENETICALLY ENGINEERED YEAST PRODUCING LACTIC ACID WITH IMPROVED LACTIC ACID WITH IMPROVED LACTIC ACID EXPORT

BACKGROUND OF THE INVENTION

The production of lactic acid by fermentation using genetically engineered yeasts has been a field of endeavor for a number of years as a sustainable method of producing a product useful in materials as diverse as foods and polymers. Candidate host yeast strains for production of lactic acid include Saccharomyces cerevisiae, Schizosaccharomyces pombe, and various species from the genera Kluyveromyces, Pichia, Candida and Hansenula. Most of the publications describing use of these yeasts for lactate production include overexpression of an exogenous lactate dehydrogenase gene from various sources in the host cell. Lactate dehydrogenase reversibly catalyzes the NAD/NADH redox conversion between pyruvate and lactic acid.

As will be evident from the detailed description of the present invention, not all lactate dehydrogenase genes work alike, even when codon optimized for expression in a selected host organism. The enzymes from various sources have different equilibrium constants, different specific activities, different rates of reaction and have different physical conformations that may make the proteins more or less difficult to express in a given strain, so there is a need in the art to discover the best lactate dehydrogenases for lactate production in a selected host strain.

RU2539092C1 describes a Schizosaccharomyces pombe yeast named VKPM Y- 4041 that is transformed with a Lactobacillus plantarum lactate dehydrogenase gene (LDH).

RU2268304C1 describes a Schizosaccharomyces pombe yeast transformed with a Rhizopus oryzae LDH gene, resulting in the strain named VKPM Y-3127.

In addition to overexpression of an exogenous lactate dehydrogenase, efforts have been made to lower or eliminate the expression of other enzymes that may siphon off the pool of pyruvate metabolites in the host cell and/or which may result in converting a portion of 3 carbon metabolites into ethanol or glycerol, which are unwanted byproducts when seeking to produce lactic acid. Enzymes of this category include pyruvate decarboxylase (PDC) that catalyzes the conversion of pyruvate to acetaldehyde, alcohol dehydrogenase (ADH) that catalyzes the reduction of acetaldehyde to ethanol, and glycerol 3 phosphate dehydrogenase (GPD) that reduces glycerone 3 phosphate to glycerol 3 phosphate which is dephosphorylated to form glycerol, thereby siphoning off three carbon metabolites that could directed toward lactate.

Most yeast, and particularly in the context of the present invention, S. pombe, contain multiple alleles for the genes mentioned above. The alleles encode different proteins with the same enzymatic activity, but the expression level and activity level of the different alleles are not the same. If all alleles for these gene were inactivated the yeast could not sustain the energy metabolism necessary for lactate production. It is therefore critical to know which alleles can be inactivated to optimize lactate production. In the summary below, alleles of various genes in S. pombe are referenced by the gene name and the open reading frame designation from an annotated curated database of the S. pombe genome called PomBase available at

PomBase identification number begin with "SP."

US9284561B2 describes a S. pombe strain transformed with a human LDH and containing a deletion of one pyruvate decarboxylase 2 gene (i.e., pdc2A , aka pdclOlA) (SPAClF8.07c) while retaining a wild type (unmodified) PDC4 (aka PDC201) (SPAC3G9.11c). This strain is therefore pdcl01APDC201+.

US10597662B2 describes a S. pombe strain containing two D-lactate dehydrogenase genes. One is from Pediococcus acidilactici and one is from either Lactobacillus bulgaricus or Lactobacillus brevis in combination with a pyruvate decarboxylase 2 gene deletion pdc2A (aka PDC101) (SPAClF8.07c) .

US9428777B2 describes a S. pombe strain containing a Lactobacillus pentosus LDH and a human lactate LDH, in combination with a deletion of a the pdc2 gene (aka pdclOlA) (SPAClF8.07c).

RU2614233C1 speculatively mentions construction of a S. pombe strain containing one or more Lactobacillus acidophilus LDH genes integrated into the chromosome and in which one or more genes of alcohol dehydrogenases (ADH) are inactivated or deleted. While the publication speculatively mentions inactivation of one or more ADH genes, the only ADH gene shown to have been deleted was the gene encoding ADH1 (SPCC13B11.01).

RU2652877C1 and RU2650669C1 similarly discloses a strain of S. pombe having an inactivated ADH1 (SPCC13B11.01) gene and expression of one more copies of a lactate dehydrogenase gens from Lactobacillus plantarum or Lactobacillus acidophilus;

The nearest prior art to the present invention is Ozaki et al, (Metabolic engineering of Schizosaccharomyces pombe via CRISPR-Cas9 genome editing for lactic acid production from glucose and cellobiose, Metabolic Engineering Communications 5 (2017) p60-67) which describes introduction of a Lactobacillus plantarum LDH gene into S. pombe and mentions a failed attempt to simultaneously engineer an ADH1 mutant, (i.e, adhlA (SPCC13B11.01)). As an alternative they overexpressed bacterial acetaldehyde dehydrogenase genes to generate acetyl-CoA. The authors further describe making a strain with deletions of two pyruvate decarboxylase genes (i.e., pdclOlA (SPAClF8.07c) a d pdc202A (SPAC13A11.06)) along with a deletion in alcohol dehydrogenase gene ADH8 i.e adh8A (SPBC1773.06c) and overexpression of the Lactobacillus plantarum LDH gene. The authors further describe deletion of what is described as a minor alcohol dehydrogenase having the systematic gene name SPBC337.il. In Pombase, SPBC337.il is annotated as a "mitochondrial membrane CH-OH group oxidoreductase, human RTN4IP1 ortholog, implicated in mitochondrial organization or tethering." Lastly the authors describe deletions of the glycerol-3-phosphate dehydrogenase gene GPD2 (SPAC23D3.04c) and another predicted glycerol-3-phosphate dehydrogenase GUT2 (SPCC1223.03c) in combination with the foregoing deletions and overexpression of the same LDH gene.

US11111482B2 describes the expression of a monocarboxylic acid/monocarboxylate transporter in a yeast strain containing a lactate dehydrogenase (LDH) gene. The resulting strain is capable of consuming lactate and producing at least 80g/L of ethanol from lactate and glucose.

US9353388B2 describes the overexpression of the ADY2 lactic acid transporter in a S. cerevisiae strain with: decreased activity of L-lactate cytochrome c oxidoreductase (CYB2) activity; decreased pyruvate decarboxylase (PDC) activity; and overexpression of a L-lactate dehydrogenase (LDH) gene.

There is a need in the art to continue to improve yeast strains, and particularly improve Schizosaccharomyces pombe strains to optimize production of lactic acid by fermentation. As will be apparent from the description that follows, it is important to select the best ADH, PDC and GPD alleles for deletion and to choose the best heterologous LDH genes for overexpression in S. pombe. SUMMARY OF THE INVENTION

The present inventors have discovered most surprisingly, that for optimal expression of lactic acid in S. pombe, it is best to inactivate each of the endogenous ADH1 (SPCC13B11.01) and ADH4 (SPAC5H10.06c) alleles, and to express two copies of a heterologous LDH having an amino acid sequence selected from the group of SEQ ID NO 9-31 provided herein. It is also best to inactivate the PDC201 (SPAC186.09) allele. Further inactivation of the GPD1 allele (SPBC215.05) reduces glycerol production and further improves lactate yields. The sharp reduction in ethanol accumulation by inactivation of PDC201 and both ADH1 and ADH4 alleles along with glycerol reduction by inactivation of the GPD1 allele results in higher lactic acid production while sustaining sufficient energy for cell growth.

The inventors have further discovered a number of variant LDH genes that produce high levels of lactic acid when operably linked to a promoter for expression of the LDH genes in S. pombe.

The inventors have further discovered that the S. pombe actin promoter is constitutive high level promoter that may be used to express any gene of interest in S. pombe.

In a one aspect, provided herein are Schizosaccharomyces pombe strains useful for the production of lactic acid by fermentation, comprising (i) an inactivated alcohol dehydrogenase 1 (ADH1) gene; (ii) an inactivated alcohol dehydrogenase 4 (ADH4) gene; and (ii) an exogenous lactate dehydrogenase (LDH) gene operably linked to a promoter to express the LDH gene in the S. pombe strain.

In particular embodiments the strains further including an inactivated glycerol 3 phosphate dehydrogenase 1 (GDP1) gene.

In particular embodiments the promoter is a S. pombe actin promoter comprising a functional portion of SEQ ID NO: 1.

In the best embodiments the strains contain two copies of the exogenous lactate dehydrogenase (LDH) gene. In various embodiments the \exogenous LDH gene encodes a LDH enzyme having an amino acid sequence selected from the group consisting of SEQ ID NOs 9-31.

In exemplary embodiments, at least one of the ADH genes is inactivated by insertion of the exogenous LDH gene at the loci of the inactivated ADH gene. In further embodiment the strains an inactivated pyruvate decarboxylase 1 gene (PDC201). In some embodiments the strains may include an inactivated glycerol 3 phosphate dehydrogenase 1 (GPD1) gene. In other embodiments both PDC201 and GPD1 are inactivated

In functional embodiments, the S. pombe strains produces at least 85 g/L of lactic acid with a yield of at least 65 % yield as determined using a DasGIP lactic fermentation protocol. In better embodiments the strains produce at least 100 g/L of lactic acid with a yield of at least 70 % as determined using a DasGIP lactic fermentation protocol.

In another aspect there is provided c nucleic acid constructs comprising a functional portion of the actinl promoter from S. pombe according to SEQ ID NO 1 operably linked to a non-native nucleic acid sequence encoding a gene of interest. In exemplary embodiments the gene of interest encodes a lactate dehydrogenase (LDH) enzyme. In exemplary embodiments the LDH gene encodes a protein sequence according to SEQ ID NOS 9-31.

In some embodiments, the strains comprise at least one copy of a nucleic acid encoding a lactate dehydrogenase gene (LDH) according to SEQ ID NOS 9-31 operably linked to a promoter that expresses said gene in the strain. In exemplary embodiments, the promoter is a functional portion of the actin 1 promoter according to SEQ ID NO. 1.

In certain embodiments, the strains contains two LDH genes each selected from the LDH genes according to SEQ ID NOS 9-31. In exemplary embodiments the promoter is a functional portion of the actin 1 promoter according to SEQ ID NO. 1.

In some embodiments the LDH gene is integrated into the genome of the strain at a locus selected from the group consisting of the locus of the alcohol dehydrogenase 1 (ADH1) and alcohol dehydrogenase 4 (ADH4) genes and the integration inactivates the ADH1 or ADH4 genes.

In conjunctive embodiments, if the ADH1 gene is inactivated by the integration the LDH gene then the ADH4 gene is also inactivated or if the ADH4 gene is inactivated by integration of the LDH gene then the ADH1 gene is also inactivated.

In any of the forgoing embodiments the strains may include an inactivated pyruvate decarboxylase 1 gene (PDC201) gene or inactivated glycerol 3 phosphate dehydrogenase 1 (GPD1) gene. In certain embodiments each of the PD201 and GPD1 genes are inactivated.

In another aspect there is provided improvements in the S. pombe strain designated Sp285 that was deposited in the United States Department of Agriculture strain depository as NRRL B-6805.

In some embodiments the improvement is directed to cellular expression of heterologous lactic acid transporters (i.e. exporters). In a strain containing a lactic acid biosynthetic pathway (i.e. Sp285) the addition of a heterologous lactic exporter results in an increased volumetric and/or specific productivity of lactic acid. An increase in export reduces the intracellular accumulation of lactic acid generated by the lactic acid production pathway. The increased export raises intracellular pH and reduces lactic acid product inhibition on the L-Lactate Dehydrogenase (LDH) enzyme by decreasing the local concentration of lactic acid in the cell. The putative S. pombe glyoxylate reductase GORI / SPACUNK4.10 gene was selected as the genomic locus for expressing the transporter library.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows lactic acid yield of several engineered S. pombe strains, each containing a unique LDH expressed in S. pombe cells. Data was ranked from highest yield (most efficient LDH) to lowest yield (least efficient LDH).

Figure 2 shows specific production resulting from different LDH variants expressed in S. pombe cells. Data was ranked from highest specific production (most efficient LDH) to lowest specific production (least efficient LDH).

Figure 3 shows the level of correlation between lactic acid yield and specific production. There is only a weak correlation between yield and specific production (R²=0.4698). This is caused by a subset of LDH variants that result in high yield but result in low specific productivity and vice versa. Though the mechanism is unknown, this indicates the underlying LDH enzyme properties between variants within this library are unique. This may be important for stacking multiple LDH variants into a single commercial strain to maximize total carbon flux from pyruvate to lactic acid. Therefore, each of these LDH variants has potential to be included in the commercial production strain and should be patent-protected. Figure 4 shows the phylogenetic relationship of 13 lactic acid transporter proteins (SEQ.ID 39-51) with the lactic acid export function and the species names from which the amino acid sequences were derived.

Figure 5 shows the volumetric productivity of lactic acid from strains containing a heterologous transporter (SEQ.ID 39-51). This shows that lactic acid is exported at a higher rate in the transporter-containing strains.

Figure 6 shows the specific productivity of lactic acid from strains containing a heterologous transporter (SEQ.ID 39-51). This shows that lactic acid is exported at a higher rate in the transporter-containing strains.

Figure 7 shows a Venn Diagram of the transporters that resulted in an increase in volumetric productivity; specific productivity; or both.

DETAILED DESCRIPTION OF THE INVENTION

The present disclosure provides examples of genetically engineered Schizosaccharomyces pombe yeast strains that ferment dextrose to produce lactic acid in an attempt to meet certain Key Performance Indicators (KPIs) believed to be desirable for large scale commercial production of lactic acid. The KPI targets for commercially suitable strains were set by the present inventor to be 140g/L of lactic acid with 87% yield at a productivity of 3.3g/L/hr, preferably with minimal pH control so that a final fermentation pH a of around 2.5 obtained. Yield is calculated as wt lactic acid/wt carbohydrate feedstock used in the fermentation.

To achieve these target KPIs, one particular pyruvate decarboxylase gene (PDC201) of the four known PDC was inactivated along with inactivation of two particular alleles of the four known S. pombe alcohol dehydrogenase (ADH1 and ADH4). In some embodiments at least one particular glycerol 3 phosphate dehydrogenase alleles (GPD2) involved in the glycerol biosynthetic pathway was also inactivated. In all embodiments at least two exogenous lactate dehydrogenase (LDH) genes having an amino acid sequence according to SEQ ID NO 9-31 were overexpressed in the strains. In preferred embodiments the overexpression was accomplished by use of the endogenous S. pombe actin 1 promoter (pActl) operably linked to the exogenous LDH genes.

As used herein "inactivated" with respect to a gene means the naturally occurring protein product of the gene is no longer expressed in the cell. Gene inactivation may be accomplished by many genetic manipulations well known to one of ordinary skill in the art including for example, by deletion of all or a portion of the coding portion of the gene, by deletion of the promoter or ribosome binding site controlling expression of the gene, by introduction of premature transcription terminator sequence into the gene, by introduction of a nonsense mutation or

5 termination codon in the coding portion of the gene, or by insertion of an exogenous genetic sequence into the endogenous gene so that the coding sequence of the gene is disrupted. In embodiments exemplified herein, the respective coding sequences were deleted by insertion of at least one of the LDH genes at the loci of the deleted gene. . In addition to inactivation of the foregoing genes, the overexpression of at

10 least two copies of a exogenous LDH genes was necessary to produce sufficient L- lactic acid to meet the KPIs.

Table I lists the alleles of the endogenous enzymes in the S. pombe strains that were untouched (+) or inactivated (-) in the present disclosure in comparison to the closest prior art, Ozaki et al, by reference to the particular systematic ORF name

15 given for the allele in in the Pombase public database of the S. genome found at https://www.pombase.org/.

Table 1

Genotypes of the Present Disclosure in comparison to Ozaki et al

20

Accordingly, the strains of the present disclosure differ from the closest prior art Ozaki et al in one or more of the following nonexclusive ways: With respect to pyruvate decarboxylase, the present strains inactivates the PDC201 gene while Ozaki et al leaves that gene intact and inactivates each the PDC101 and PDC202 genes. With respect to alcohol dehydrogenase, the present strains inactivates each of the ADH1 and ADH4 genes, while Ozaki et al leaves both these genes intact and deletes only ADH8 and the minor ADH gene. With respect to glycerol 3 phosphate dehydrogenase, the present disclosure inactivates GPD1, while Ozaki et al leaves this gene intact and deletes each of GPD2 and GUT2. With respect to two copies of the LDH gene, the present disclosure provides 23 possible heterologous LDH genes from different sources, each of which have higher activity LDH activity when expressed in S. pombe than the L. plantarum LDH described of Ozaki et al.

The LDH enzymes were expressed using a functional portion of the S. pombe actin promoter (pACTl - SEQ ID. NO 1) located upstream of the S. pombe actin open reading frame (systematic open reading frame name:. SPBC32H8.12c ) which based on transcriptomics experiments was identified by the present inventor as being a constitutive and high activity S. pombe promoter. A "functional portion" of the promoter according to SEQ ID NO: 1. means any nucleotide sequence having at least 75% sequence identity across any stretch of 20 or more nucleotides present in SEQ ID NO:24 that when inserted upstream of a target gene, is able to transcribe the target gene at least 75% as well as the whole of SEQ ID NO 24, which was used in the exemplary embodiments. Although exemplified herein with the actin promoter, any strong constitutive S. pombe promoter having similar high levels of transcriptional activity may be used in other embodiments. In exemplary embodiments, the transcriptional terminator from the ADH1 gene ((SEQ ID NO: 25) was used as the terminator for the LDH genes, but any terminator may be used in other embodiments.

Das GIP Lactic Fermentation Protocol

All strains described herein were evaluated for the production of lactic acid and other metabolites using a DasGIP fermentation protocol. This main fermentation protocol uses an Eppendorf DasGIP 1.3 liter fermenter cultured with a 10% inoculation volume containing 100 to 300 million cells per milliliter from the seed shake-flask fermentation, with media containing 10 g/L autoclave-sterilized dextrose and 15g/L autoclave-sterilized Corn Steep Liquor (CSL) obtained from Roquette (Beinheim, France). . The fermenter was cooled and temperature controlled at 33 ° C. A seed culture was inoculated into the fermenter at 10% v/v. Glucose was fed to the fermenter to maintain a concentration of 5-20 g/L throughout the run. The air flow going through the tank was controlled at 1 volume per minute (vvm) for the first 20 hours, then ramped down to 0.1 vvm and controlled the rest of the fermentation. The dissolved oxygen level of the media was controlled at 16% the first 24 hours via agitation, after which it is not controlled. The pH of the media was controlled at 3 the first 16-20 hours, after which it is not controlled and the pH dropped due to the production of lactic acid. Samples were taken throughout the fermentation to measure cell growth, cell viability, glucose and lactic acid concentration. The fermentation was stopped at 64 hours. The metabolites made by the fermentation were determined by HPLC.

Development of Strains.

S. pombe strains were genetically engineered to add one to three copies of an exogenous LDH gene and to inactivate one or more of PDC201, ADH1, ADH4 and/or GPD1 genes. For all genetic transformations, the lithium acetate method first described by Gietz, et. at. (Schiestl RH, Gietz RD. High efficiency transformation of intact yeast cells using single stranded nucleic acids as a carrier. Curr Genet. 1989 Dec;16(5-6):339-46. doi: 10.1007/BF00340712.PMID: 2692852) was performed.

First, the wild type strain had the URA4 coding sequence deleted from the genome by co-transforming a synthetic linear DNA that contains a unique molecular- barcoded DNA (BC4241) with universal primer sequences (barcode in bold): GGTTACACTGTGACAGATGCCATACGAACTGCACAGACGGTTCGCCTGACTGTTGAGCGTG ATAGACTGTGATCGACACG (SEQ ID NO 34) flanked by lkb of homology to the URA4 locus with a plasmid based CRISPR-Mad7 device using the PAM-protospacer sequence TTTGTGATATGAGCCCAAGAAGCAA (SEQ ID NO 35) within the URA4 gene. Transformants were selected on EMM + uracil + 5-FOA media.

Second, PDC201 was then deleted by replacing the PDC201 coding sequence with a synthetic S. pombe URA4 selection marker. Transformants were selected on EMM - uracil media plates. The URA4 selection marker was then replaced with a LDH expression cassette (PACTI-LCLDH-TADHI), which was genome-integrated by cotransforming linear PCR product containing the LDH expression cassette flanked by lkb of homology to the PDC201 locus with a plasmid based CRISPR-Mad7 device that targeted the PAM-protospacer sequence (SEQ ID NO 35). Transformants were recovered on EMM + uracil + 5-FOA media. Third, ADH1 was then deleted by replacing the ADH1 coding sequence with a synthetic S. pombe URA4 selection marker (SEQ ID NO 32). Transformants were selected on EMM - uracil media plates. The URA4 selection marker was then replaced with a LDH expression cassette (PACTI-LCLDH-TADHI), which was genome- integrated by co-transforming linear PCR product containing the LDH expression cassette flanked by lkb of homology to the ADH1 locus with a plasmid based CRISPR- Mad7 device that targeted the PAM-protospacer sequence (SEQ ID NO 35). Transformants were recovered on EMM + uracil + 5-FOA media.

Fourth, ADH4 was then deleted by replacing the ADH4 coding sequence with the synthetic S. pombe URA4 selection marker (SEQ ID NO 32). Transformants were selected on EMM - uracil media plates. The URA4 selection marker was then replaced with a molecular barcode (BC59) of the sequence: GGTTACACTGTGACAGATGCTGGTAGAGTTTAGCTCCTCGGACAGTCGGAGATTACATAG ATAGACTGTGATCGACACG, (SEQ ID NO 55 ) which was genome-integrated by cotransforming linear synthetic DNA with lkb of homology to the ADH4 locus with a plasmid based CRISPR-Mad7 device that targeted the PAM-protospacer sequence (TSEQ ID NO 35). Transformants were recovered on EMM + uracil + 5-FOA media. GPD1 was then deleted by replacing the GPD1 CDS with a synthetic S. pombe URA4 selection marker (SEQ I NO 32). Transformants were selected on EMM - uracil media plates. The final strain genotype is pdc201::PACTi-LcLDH-TADHi adhlA::P ACTI-LCLDH- TADHI adh4A::BC2 gpdlA::ura4 ura4A::BCl

Our initial strain used to develop the strains of the present disclosure was designated Sp38. This strain contained one copy of the LDH gene from Lactobacillus cerevisiae encoding the protein sequence according to SEQ ID NO 14 with codons optimized for expression in S. pombe under control of the actin promoter (SEQ ID NO: 1) and terminated by the ADH1 terminator (SEQ ID NO 2) which construct is designated herein as pAct:LDHt,. (reffrred to above as PACTI-LCLDH-TADHI )• The construct pActLDHt was inserted at the loci of the pyruvate decarboxylase PDC201 allele, thereby inactivating PDC201. Strain Sp38 demonstrated the following results in the Das GIP lactic production protocol:

The yield from strain Sp38 was 34% (g lactic acid/g dextrose).

Strain Sp38 was modified by inactivation of the alcohol dehydrogenase ADH1 gene (SEQ ID NO 4) by insertion of the URA4 gene (SEQ ID NO 23) as a selective marker at the loci of the ADH1 gene thereby producing a strain designated Sp45 that demonstrated the following results in the Das GIP lactic production protocol:

The significantly higher titer of lactic acid and significantly reduced titer of ethanol indicated the importance of inactivation of the ADH1 gene. The yield from strain Sp45 was 45%.

Another strain designated Spl58 was made containing the same pActLDHt construct inactivating the PDC201 gene but containing a second copy of the same LDH gene inserted at the site of the ADH1 open reading frame and driven by the endogenous ADH1 promoter, which inactivated the ADH1 gene while simultaneously adding an additional LDH gene. In addition, the ADH4 gene (SEQ ID NO 6) was inactivated in Spl58 by insertion of URA4 at the site of the ADH 4 gene. Spl58 demonstrated the following results in the Das GIP lactic production protocol:

This result demonstrated the importance of inactivating both the ADH1 and ADH4 genes to greatly reduce ethanol production, and showed that the additional copy of the LDH gene improved lactate production. The yield from strain Spl58 was 65%.

Another strain designated Spl94 was made containing the same pActLDHt construct inactivating the PDC201 gene and with a second copy of pActLDHt inserted at the site of the ADH1 inactivating that gene with the second copy of LDH driven by the actin promoter, but with no inactivation of ADH4. Instead, the glycerol 3 phosphate gene GPD1 (SEQ ID NO 8) was inactivated in Spl94 by insertion of the URA4 gene at the site of the GPD1 gene. Spl94 demonstrated the following results in the Das GIP lactic production protocol:

This result demonstrated that inactivation of GPD1 combined with ADH1 and PDC201 and expression of two copies of LDH substantially increased lactic production with a substantial reduction in glycerol production, albeit with some sacrifice in productivity due to the increase in doubling time. The yield from strain Spl94 was 81.5%.

Another strain designated Spl95 was made containing the same pActLDHt construct inactivating the PDC201 gene and also with a second copy of pActLDHt construct inserted at the site of the ADH1 inactivating that gene. In addition, inactivation of ADH4 was done by insertion of the URA4 gene at that site. Spl95 demonstrated the following results in the Das GIP lactic production protocol:

This result demonstrated that inactivation of ADH4 combined with inactivation of ADH1 and PDC201 and expression of two copies of LDH substantially increased lactic production while maintaining a substantially reduced production ethanol and dis wo without sacrificing doubling time. The yield from strain Spl95 was 73.5%.

Another strain designated Sp211 was made from parent strain Spl94. Sp211 retained all the changes made ins Spl94 but instead of inactivating GPD1 by insertion of URA4, GPD1 was inactivated by insertion of a third copy of the LDH gene under control of the actin promoter. Sp211 demonstrated the following results in the Das GIP lactic production protocol:

This result demonstrated that very high expression of LDH is highly detrimental to cell growth, greatly reducing the final titer of lactic acid even with complete loss of glycerol production and very low ethanol production. The yield from strain Sp211 was 90%.

Another strain designated Sp212 was made from parent strain Spl95 Sp212 retained all the changes made ins Spl95 but additionally contained a third copy of the LDH gene under control of the actin promoter. Sp212 demonstrated the following results in the Das GIP lactic production protocol:

The results with Sp212 confirmed again that over expression of LDH by using three copies greatly reduces growth rates and final titers of lactic acid, even with substantial reduction in ethanol accumulation. The yield from strain Sp212 was 78%.

Another strain designated Sp21was made that like Spl94 contained two copies of the LDH gene inactivating both PDC201 and ADH1 and containing the URA4 gene inactivating GPD1, bur further contained an artificial bar code marker sequence designated BC59 (SEQ ID. NO 55) inserted at the locus of the ADH4 gene, inactivating that gene as well. Sp214 demonstrated the following results in the Das GIP lactic production protocol:

These results demonstrated that inactivation of PDC201 combined with inactivation of ADH1, ADH4 and GPD1 and expression of two LDH genes resulted in a strain with relatively high titers of lactic, and very low titers of glycerol, but with an unacceptably reduced growth rate. The yield from strain Sp214 was 84%. The inventors sought to overcome the suppression in growth rate to improve seed growth and volumetric productivity in strains Sp214 by using controlled adaptive laboratory evolution (ALE) provided as a service by ALTAR (Evry, France; https://www.altar.bio/) . Sp285 is deposited as NRRL B-6805 that demonstrated the following results in the Das GIP lactic production protocol:

This demonstrated that the reduced growth phenotype could be overcome by natural selection resulting in strain with very high lactic production and very good growth rate, even though the strain makes more glycerol and ethanol than other strains containing the same inactivated PDC, ADH1, ADH1 and GPD1 genes. The yield from strain Sp285was 75%.

L-Lactate Dehydrogenase (LDH) screening and evaluation

The foregoing strain development was based on use of the LDH gene from Lactobacillus cerevisiae (SEQ ID NO 14) expressed in S. pombe cells. Both of the L. cerevisiae LDH genes were synthetic DNAs that were codon optimized for expression in S. pombe. Both heterologous LDH enzymes were expressed using a cassette that contained the constitutive and highly expressed actin promoter, encoded by the gene ACT1 (SPBC32H8.12c - SEQ ID NO 1) to drive transcription and both contained the transcriptional terminator from the endogenous ADH1 gene (SEQ ID NO 2).

The present inventors also separately screened other LDH variant genes to determine LDH variants that may be more active in S. pombe. Candidate LDH variants were selected by searching for L-Lactate Dehydrogenase in NCBI (https://www.ncbi.nlm.nih.gov/). Sequences were downloaded and aligned using Geneious software. LDH variants were selected from genera under every kingdom of life, including: 1) Animal; 2) Archaea; 3) Bacillus; 4) Fungus; 5) Lactobacillus; 6) Plant; 7) Protist; 8) Proteobacteria; and 9) Green Sulfur Bacteria. Ninety-six (96) LDH variants were selected based on sequence similarity (i.e. protein alignment tree) to capture sequence and putative functional diversity throughout the biosphere. Each of these LDH variants were codon optimized for gene expression in S. pombe cells using IDT codon optimization tool (https://www.idtdna.com/CodonOpt) and obtained as synthetic linear dsDNA from Twist Bio (https://ecommerce.twistdna.com/). Each LDH was cloned for genome integration at the PDC201 (SPAC3G9.11c) locus and expressed using the ACT1 promoter with the terminator from the ADH1 gene. The resultant genotype for each strain was pdc201A::pACTl-LDHt , meaning the PACTl-LDHt construct was inserted at the PDC201 locus to inactivate the PDC201 gene.

Each LDH-containing strain that produced lactic acid in an otherwise wild type cell was then deleted for ADH1 (SPCC13B11.01) with a synthetic URA4 selection marker and evaluated for lactic acid production in the benchtop 24 well Applikon MicroMatrixsystem The MicroMatrix is an automated fermentation instrument that contains 24 independent bioreactor wells that controlpH, dissolved oxygen, temperature, feed rate. In the MicroMatrix experiments, cells were propagated in seed fermentation conditions consisting of corn steep liquor obtained from the Archer Daniels Midland Company (5 g/L suspended solids), yeast extract (25 g/L), ammonium sulfate (5 g/L), and dextrose (30 g/L). An aliquot of each tube culture was transferred at an inoculation rate of 10% to a 250 mL baffled flask and incubated at 33° C, in a shaker ser at 250 rpm for 48 hours. For bioconversion in MicroMatrix, the cells from the seed culture are normalized to a starting OD600 of 2.5. Fermentation conditions were maintained at 33°C and 10% dissolved oxygen (DO) with Ramp feeding of 35% dextrose. Fermentation was terminated at elapsed fermentation time (EFT) of 71 hours. The final main fermentation OD600 and the final pH of the main fermentation broth was measured. After centrifugation to remove biomass the supernatant was analyzed by HPLC for lactic acid, ethanol, glycerol, pyruvic acid, acetic acid and dextrose.

Yield was calculated using the formula: (lactic acid titer)/g glucose consumed. Specific production (i.e, a measure of production per cell) was calculated using the formula: lactic acid tite r)/(optica I density at end of the experiment. Different LDH variants (SEQ 9-31) were then ranked highest to lowest based on in vivo lactic acid yield (Table 1). Table 1

Lactate dehydrogenase activity in S. pombe cells measured by yield, titer and specific production

Figure 1 is a graph that shows lactic acid yield of the above engineered S. pombe strains, each containing a unique LDH expressed in S. pombe cells. Data was ranked from highest yield (most efficient LDH) to lowest yield (least efficient LDH).

Figure 2 is a graph that shows specific production resulting from different LDH variants expressed in S. pombe cells. Data was ranked from highest specific production (most efficient LDH) to lowest specific production (least efficient LDH). Figure 3 is a graph that shows the level of correlation between lactic acid yield and specific production. There is only a weak correlation between yield and specific production (R²=0.4698). This is caused by a subset of LDH variants that result in high yield but but low specific productivity and vice versa. Though the mechanism is unknown, this indicates the underlying LDH enzyme properties between variants within this library are unique. This may be important for stacking multiple LDH variants into a single commercial strain to maximize total carbon flux from pyruvate to lactic acid. Therefore, each of these LDH variants has potential to be included in the commercial production strain for lactic acid.

The forging shows that LDHs from different sources perform differently when expressed in S. pombe in an unpredictable way. Although strain development process described herein before carried two copies of the L. cerevisiae LDH in combination with deletions of PDC201, ADH1, ADH4 and/or GPD1, the present invention can also be embodied by use of any of the LDH's according to SEQ ID NOS 9-31 in combination with the same deletions to yield similar results.

Expression of lactic acid transporters

A library of 96 heterologous proteins with the potential to transport lactic acid was determined through bioinformatics analysis. Seed sequences from 4 different classes of short chain carboxylic acid transporters were gleaned from the literature and used to identify homologues from public protein databases. JEN1 is a monocarboxylate transporter protein that has been shown to affect the yield of lactic acid in Saccharomyces cerevisiae (Turner et. al. 2019, FEMS Yeast Research 19(6) doi: 10.1093/femsyr/foz050). A key motif defining transporter activity of JEN1 has been identified as NXXXHXXQDXXXT(SEQ ID NO: 54). This motif was used along with the full sequence of JEN1 from Kluyveromyces lactis (Genbank accession ID: XP_454682.1) to perform a Pattern Hit Initiated (PHI) BLAST of the Genbank Non- Redundant (NR) Protein Database. Multiple Sequence Alignment (MSA) was performed using Clustal Omega 1.2.2 on 500 homologous sequences retrieved by the search and 63 of the proteins were chosen to maximize the diversity of the library. These first cut sequences were again aligned using MUSCLE 3.8.425 and 30 proteins chosen to maximize diversity. LacA is a proton symport lactate permease which has been shown to confer lactic acid transport capability in JEN1 deleted yeast (Skory et. al. 2010, Enzyme and Microbial Technology 46 pp. 43-50 doi: 10.1016/j.enzmictec.2009.08.014). The lacA sequence from Rhizopus delemar (Genbank accession ID: ACV53193.1) was used to perform a BLAST search of the GenBank ClusteredNR protein database. MSA was performed using Clustal Omega 1.2.2 on 100 homologous sequences retrieved by the search and 58 of the proteins were chosen to maximize the diversity of the library. These first cut sequences were again aligned using MUSCLE 3.8.425 and 30 proteins chosen to maximize diversity. ADY2 is also a monocarboxylate transporter protein that has been shown to affect the yield of lactic acid in Saccharomyces cerevisiae. The ADY2 sequence from Saccharomyces cerevisiae (Genbank accession ID: NP_009936.1) was used to perform a BLAST search of the GenBank ClusteredNR protein database. MSA was performed using MUSCLE 3.8.425 on 250 homologous sequences retrieved by the search. The sequences were separated into several major clades. Sequences from each of the clades were chosen to maximize diversity amongst the 18 sequences finally chosen.

MAE1 is an evolutionarily distinct transporter of malic acid found primarily in Schizosaccharomyces species (Grobler et. al. 1995, Yeast 15 pp. 1485-91 doi: 10.1002/yea.320111503). The MAE1 sequence from Schizosaccharomyces pombe (Genbank accession ID: NP_594777.1) was used to perform a BLAST search of the GenBank ClusteredNR protein database. MSA was performed using MUSCLE 3.8.425 on 100 homologous sequences retrieved by the search. The sequences were separated into several major clades. Sequences from each of the clades were chosen to maximize diversity amongst the 18 sequences finally chosen. A phylogenetic tree based on MSA using Clustal Omega 1.2.2 of the transporters are shown in Figure 4. The various transporters are clustered according to their basic transporter type (JEN1, lacA, ADY2 or MAE1) and their relative homology to each other is shown.

To express heterologous exporters in S. pombe cells, cloned transporter gene expression cassettes (containing the promoter, gene, terminator and genomic homology) were co-transformed with a plasmid based CRISPR-Mad7 device using the PAM-protospacer sequence TTTGGGCGGTATTGGTAAGACCATG (SEQ ID NO 36) within the GOR1/SPACUNK4.10 gene (SEQ ID NO 52, 53). Transformants were recovered on YPD + G418 (lOOmg/L) plates to select for the CRISPR-Mad7 plasmid and later inoculated in YPD media to lose the plasmid. The transporter containing gorlA mutant strains were genotype confirmed by diagnostic PCR.

To evaluate fermentation performance of the engineered strains, two promoters: 1) GHT5/SPCC1235.14 (SEQ ID 37); and 2) MAE1/SPAPB8E5.03 (SEQ ID 38) were used to express the lactic acid transporters. The heterologous transporter-containing strains were tested in 440 pl of fermentation media in deep well round bottom 96 well plates and quantified for volumetric and specific productivity of lactic acid. The fermentation media contains 15g/L Roquette CSL with 50g/L dextrose controlled to a pH 4.5. Plates were inoculated at 30 °C with 900 rpm in an INFORS HT Multitron Pro shaker. To determine the lactic acid export rate, 4 sets of 96 well plates were inoculated simultaneously but harvested at 4 different fermentation timepoints: 2 hours; 5 hours; 8 hours; and 24 hours. At each timepoint cell growth was measured by spectrophotometry (OD600) and cell-free supernatants were used to quantify lactic acid titer and residual dextrose using the Roche Cedex. The lactic acid export rate for each tested strain (volumetric productivity) was determined by fitting the lactic acid titer data (g/L/hr) against time. Strains were ranked based on the graphical slopes, which represent the volumetric productivity. Overall, 9 strains showed improved LA export rate (3.5%-28% increase) (Figure 5).

For specific productivity, LA titer was divided by OD and plotted against time (g/L/h/OD) (Figure 6). Strains were ranked based on specific productivity. In comparison to the control Sp285, 10 strains showed improved specific productivity (l.lx to 3.4x). In total, 13 strains had improved lactic acid export over the control strain with 3 strains having improved volumetric productivity, 4 strains having improved specific productivity and 6 strains having both improved volumetric and specific productivity (Figure 7). The transporter class distribution of the 13 transporters selected from these screens were: lacA (2); JEN1 (1); ADY2 (5); and MAE1 (5).

Claims

CLAIMS What is claimed is:

1. A S Schizosaccharomyces pombe strain useful for the production of lactic acid by fermentation, comprising (i) an inactivated alcohol dehydrogenase 1 (ADH1) gene; (ii) an inactivated alcohol dehydrogenase 4 (ADH4) gene; (ii) an exogenous lactate dehydrogenase (LDH) gene operably linked to a promoter to express the LDH gene in the S. pombe strain; and (iv) a carboxylic acid transporter gene capable of transporting lactic acid from the S. pombe strain operably linked to promoter to express the carboxylic acid transporter gene in the S. pombe strain.

2. The S. pombe strain of claim 1, further including an inactivated glycerol 3 phosphate dehydrogenase 1 (GDP1) gene.

3. The S. pombe strain of claim 1 wherein the promoter is a S. pombe actin promoter comprising a functional portion of SEQ ID NO: 1.

4. The S. pombe strain of claim 1 comprising at least two copies of the exogenous lactate dehydrogenase (LDH) gene.

5 The S. pombe strain of claim 1, wherein the exogenous LDH gene encodes a LDH enzyme having an amino acid sequence selected from the group consisting of SEQ ID NOS 9-31.

6 The S. pombe strain of claim 1, wherein at least one ADH genes is inactivated by insertion of the exogenous LDH gene at the loci of the inactivated ADH gene.

7. The S. pombe strain of claim 1, further including an inactivated pyruvate decarboxylase 1 gene (PDC201).

8. The S. pombe strain of claim 2, further including an inactivated pyruvate decarboxylase 1 gene (PDC201).

9. The S. pombe strain of claim 1, wherein said strain produces at least 85g/L of lactic acid with a yield of at least 65 % yield as determined using a DasGIP lactic fermentation protocol.

10. The S. pombe strain of claim 1, wherein said strain produces at least lOOg/L of lactic acid with a yield of at least 70 % yield as determined using a DasGIP lactic fermentation protocol.

11. A nucleic acid construct comprising a functional portion of the actinl promoter from S. pombe according to SEQ ID NO 1 operably linked to a non-native nucleic acid sequence encoding a gene of interest.

12. The nucleic acid construct according to claim 11 wherein the gene of interest encodes a lactate dehydrogenase (LDH) enzyme.

13. The nucleic acid construct according to claim 11 wherein the LDH gene encodes a protein sequence according to SEQ ID NOS 9-31.

14. A S. pombe strain comprising at least one copy of a nucleic acid encoding a lactate dehydrogenase gene (LDH) according to SEQ ID NOS 9-31 operably linked to a promoter that expresses said gene in the strain.

15. The S. pombe strain according to claim 14 wherein the promoter is a functional portion of the actin 1 promoter according to SEQ ID NO. 1.

16. The S. pombe strain according to claim 14 wherein the strain contains two LDH genes each selected from the LDH genes according to SEQ ID NOS 9-31.

17. The S. pombe strain according to claim 16 wherein the promoter is a functional portion of the actin 1 promoter according to SEQ ID NO. 1.

18. The S. pombe strain according to claim 14 wherein the LDH gene is integrated into the genome of the strain at a locus selected from the group consisting of the locus of the alcohol dehydrogenase 1 (ADH1) and alcohol dehydrogenase 4 (ADH4) genes and the integration inactivates the ADH1 or ADH4 genes.

19. The S. pombe strain according to claim 18 wherein if the ADH1 gene is inactivated by the integration the LDH gene then the ADH4 gene is also inactivated or if the ADH4 gene is inactivated by integration of the LDH gene then the ADH1 gene is also inactivated.

20. The S. pombe strain according to claim 19 further including an inactivated pyruvate decarboxylase 1 gene (PDC201) gene or inactivated glycerol 3 phosphate dehydrogenase 1 (GPD1) gene.

21. The S. pombe strain according to claim 21 where each of the PD201 and GPD1 genes are inactivated.

22. The S. pombe strain designated Sp285 on deposit as NRRL B-6805.

23. The S. pombe strain of claim 1, wherein the carbocyclic acid transporter gene is derived from at least one protein sequence according SEQ ID NOS 39-51.

24. The S. pombe strain of claim 1 wherein the promoter operably linked to the carbocyclic acid transporter is comprised of a functional portion of SEQ ID NO: 37 that ordinarily drives expression of a S. pombe glucose hexose transporter gene is S. pombe.

25. The S. pombe strain of claim 23 wherein the promoter operably linked to the carbocyclic acid transporter is comprised of a functional portion of SEQ ID NO: 38 that ordinarily drives expression of a S. pombe malic acid transporter gene S. pombe.