GLYCOENGINEERING OF THERMOTHELOMYCES HETEROTHALUCA

FIELD OF THE INVENTION

The present invention relates to genetically-modified Thermothelomyces heterothallica (formerly Myceliophthora thermophila) in which protein glycosylation pathways have been engineered to produce proteins with N-glycans similar to those of human, companion animal and other mammalian proteins.

BACKGROUND OF THE INVENTION

Most therapeutic proteins require glycosylation to ensure proper folding, function, and activity. Glycosylation of therapeutic proteins is also particularly important for their immunogenicity. Therefore, such proteins cannot be produced in standard prokaryotic expression systems, which lack the necessary glycosylation machinery. Since glycosylation and other post-translational modifications are essential for therapeutic glycoproteins, most of them are currently produced in mammalian cells. However, fermentation processes based on mammalian cell culture (e.g., CHO, murine, or human cells) are typically very slow, require expensive nutrients and cofactors (e.g., bovine fetal serum or specific growth factors), often yield low product titers, and also are susceptible to infections which may contaminate the resulting protein product. Thus, there is a growing shift to serum-free expression systems. In particular, yeasts and fungi are being developed as alternative protein expression systems.

As eukaryotic organisms, yeast and fungi are able to perform post-translational modifications, including N- and O-glycosylation, but protein glycosylation in yeast and fungi is quite different from that in mammalian cells. To overcome these problems, the possibility of reengineering the N-glycosylation pathway has been explored, especially in the species most frequently used for the production of heterologous proteins (e.g., 5. cerevisiae, Pichia pastoris, Yarrowia lipolytica, Hansenula polymorpha, and Aspergillus and Trichoderma species). However, protein yields still need improvements, and particularly there is a need to improve the glycosylation pattern, such that a high percentage of the produced proteins carry the desired glycoforms, namely, glycoforms of mammalian (particularly human and companion animals) proteins.

Parsaie Nasab et al., 2013, Appl Environ Microbiol., 79(3): 997-1007 describe a synthetic N-glycosylation pathway to produce recombinant proteins carrying human N- glycans in Saccharomyces cerevisiae. A Δalg3 Δalg11 double mutant strain was used, which was further genetically modified to express an artificial flippase, a protozoan oligosaccharyltransferase and Golgi-targeted human N-acetylglucosaminyltransferases I and II. The results confirmed the presence of the complex human N-glycan structure GlcNAc ₂ Man ₃ GlcNAc ₂ on a secreted monoclonal antibody recombinantly expressed in the mutant strain. However, due to the interference of Golgi apparatus-localized mannosyltransferases, heterogeneity of N-linked glycans was observed.

The work by Parsaie Nasab et al. is also described in US 2011/0207214, disclosing cells modified to express lipid-linked oligosaccharide (LLO) flippase activity that is capable of flipping LLO comprising 1 mannose residue, 2 mannose residues and 3 mannose residues, from the cytosolic side to the lumenal side of an intracellular organelle, and further reviewed along with other related studies in De Wachter et al., 2018, Engineering of Yeast Glycoprotein Expression. In: Advances in Biochemical Engineering/Biotechnology. Springer, Berlin, Heidelberg.

US 7,029,872, US 7,326,681, US 7,629,163, US 7,981,660 disclose cell lines having genetically modified glycosylation pathways that allow them to carry out a sequence of enzymatic reactions, which mimic the processing of glycoproteins in humans. Eukaryotes such as unicellular and multicellular fungi, which ordinarily produce high-mannose- containing N-glycans, are modified to produce N-glycans such as Man ₅ GlcNAc ₂ or other structures along human glycosylation pathways.

US 7,449,308, US 7,935,513 disclose eukaryotic host cells having modified oligosaccharides which may be modified further by heterologous expression of a set of glycosyltransferases, sugar transporters and mannosidases to become host-strains for the production of mammalian, e.g., human therapeutic glycoproteins. N-glycans made in the engineered host cells have a Man ₅ GlcNAc ₂ core structure which may then be modified further by heterologous expression of one or more enzymes, e.g., glycosyltransferases, sugar transporters and mannosidases, to yield human-like glycoproteins.

US 7,795,002 discloses eukaryotic host cells such as yeast and filamentous fungi producing human-like glycoproteins characterized as having a terminal β-galactose residue and essentially lacking fucose and sialic acid residues. Further disclosed is a method for catalyzing the transfer of a galactose residue from UDP-galactose onto an acceptor substrate in a recombinant eukaryotic host cell, which can be used as a therapeutic glycoprotein. US 8,986,949 discloses genetically engineered strains of non-mammalian eukaryotes expressing calalytically active endomannosidase genes to enhance the processing of the N-linked glycan structures with the overall goal of obtaining a more human-like glycan pattern. In addition, cloning and expression of a novel human and mouse endomannosidase are disclosed.

US 9,359,628 discloses genetically engineered strains of Pichia capable of producing proteins with smaller glycans. In particular, the genetically engineered strains are capable of expressing either or both of an α-1,2-mannosidase and glucosidase II. The genetically engineered strains can be further modified such that the OCH1 gene is disrupted. Methods of producing glycoproteins with smaller glycans using such genetically engineered stains of Pichia are also provided.

US 9,695,454 discloses compositions including filamentous fungal cells, such as Trichoderma fungal cells, having reduced protease activity and expressing fucosylation pathway. Further described are methods for producing a glycoprotein having fucosylated N- glycan, using genetically modified filamentous fungal cells, for example, Trichoderma fungal cells, as the expression system.

Thernwthelomyces heterothallica (Th . heterothallica) strain C1 (recently renamed from Myceliophthora thermophila, which was renamed from Chrysosporium lucknowense) is a thermo-tolerant ascomycetous filamentous fungus producing high levels of cellulases, which made it attractive for production of these and other enzymes on a commercial scale.

For example, US Patents 8,268,585 and 8,871,493 disclose a transformation system in the field of filamentous fungal hosts for expressing and secreting heterologous proteins or polypeptides. Also disclosed is a process for producing large amounts of polypeptide or protein in an economical manner. The system comprises a transformed or transfected fungal strain of the genus Chrysosporium, more particularly of Chrysosporium lucknowense and mutants or derivatives thereof. Also disclosed are transformants containing Chrysosporium coding sequences, as well expression-regulating sequences of Chrysosporium genes.

Wild type C1 was deposited in accordance with the Budapest Treaty with the number VKM F-3500 D, deposit date August 29, 1996. High Cellulase (HC) and Low Cellulase (LC) strains have also been deposited, as described, for example, in US Patent 8,268,585. For example: strain UV13-6, deposit no. VKM F-3632 D, strain NG7C-19, deposit no. VKM F- 3633 D, strain UV18-25, deposit no. VKM F-3631 D. Additional improved C1 strains that have been deposited include (i) HC strain UV18-100f (ΔalplΔpyr5) - deposit no. CBS141147; (ii) HC strain UV18-100f (ΔalplΔpep4Δalp2Δpyr5Δprtl) deposit no. CBS141143; (iii) LC strain W1L#100I (ΔchilΔalpΔAalp2Δpyr5) - deposit no. CBS141153; and (iv) LC strain W1L#100I (ΔchilΔalplΔpyr5) - deposit no. CBS141149.

US 9,175,296 discloses a fungal host strain of Chrysosporium lucknowense. Also disclosed is a method for homologous and/or heterologous production of a pure protein with a purity of higher than 75%, a method for production of artificial protein mixes and a method for simplified screening of strains functionally expressing a desired enzyme. US 9,175,296 further discloses an isolated promoter sequence suitable for the transcriptional control of gene expression in Chrysosporium lucknowense and a method for isolating a fungal host strain of Chrysosporium lucknowense wherein the protease secretion is less than 20% of the protease secretion of Chrysosporium lucknowense strain UV 18-25.

EP 2505651 discloses an isolated fungus that has been mutated or selected to have low protease activity, wherein the fungus has less than 50% of the protease activity as compared to a non-mutated fungus. The fungus is of the genus Chrysosporium, preferably it is a strain of Chrysosporium lucknowense.

There is a need for an expression system for producing recombinant human, companion animal and other mammalian proteins that is able to produce high yields of glycoproteins with N-glycans of mammalian proteins, particularly human and companion animal proteins, such that the proteins are suitable for therapeutic use in humans, companion animals and other mammals.

SUMMARY OF THE INVENTION

The present invention provides Thermothelomyces heterothallica (formerly Myceliophthora thermophila) genetically modified to produce glycoproteins with N-glycans of mammalian proteins, particularly N-glycans of human, companion animal and other mammalian proteins. The genetic modification of the Th. heterothallica of the present invention comprises deletion or disruption of the alg3 gene, deletion or disruption of the alg11 gene and over-expression of an endogenous flippase or expression of a heterologous flippase. The genetic modification of the Th. heterothallica may also further comprise expression of heterologous GlcNAc transferase 1 (GNT1) and GlcNAc transferase 2 (GNT2). In some embodiments, the genetic modification further comprises expression of the STT3 subunit of a heterologous oligosaccharyltransferase. In additional embodiments, the genetic modification further comprises expression of a heterologous galactosyltransferase.

The present invention is based in part on the finding that Th. helerolhallica genetically-modified as disclosed herein produces glycoproteins in which the desired mammalian/human N-glycans constitute over 90% of the N-glycans found on the glycoproteins, and in some cases even over 95% of the N-glycans. In addition, when further modified to express a heterologous mammalian glycoprotein (e.g., an antibody), the Th. heterothallica genetically-modified as disclosed herein produces high levels of the heterologous glycoprotein, with the desired mammalian/human N-glycans constituting over 85% of its N-glycans. This is in contrast to hitherto described expression systems, which produce large variation in the obtained N-glycans. Remarkably, no major negative effects on cell viability have been observed with any of the modifications done.

It is noted that Th. heterothallica , unlike most fungi and yeast, does not have hypermannosylated N-glycans, which may contain up to 50 mannose residues, but rather has “oligo mannose” N-glycans, containing between 3-9 mannose residues, and hybrid type N-glycans, containing mannose and HexNAc residues, whose structure is not fully characterized. Since the structure, as well as the synthesis pathway, of the hybrid N-glycans is not fully characterized, it was unclear that such glycans can be eliminated using the genetic modifications described herein. Surprisingly, the genetic modifications according to the present invention resulted in essential elimination of these structures, with over 90% of the N-glycoforms being the desired mammalian/human N-glycans.

Advantageously, the Th. heterothallica cells of the present invention produce high yields of proteins. The protein levels obtained using the Th. heterothallica cells of the present invention are much higher than those obtained using mammalian cells, such as CHO cells, or yeasts.

The present invention therefore provides an efficient system for producing glycoproteins with desired N-glycans, suitable for therapeutic use in humans.

According to one aspect, the present invention provides a Thermothelomyces heterothallica genetically modified to produce glycoproteins with mammalian N-glycans, wherein the genetic modification comprises:

(i) deletion or disruption of the alg3 gene such that the Th. heterothallica fails to produce a functional alpha- 1,3- mannosyltransferase; (ii) deletion or disruption of the alg11 gene such that the Th. heterothallica fails to produce a functional alpha- 1,2-mannosyltransferase; and

(iii) over-expression of an endogenous flippase or expression of a heterologous flippase.

In some embodiments, the genetic modification further comprises expression of heterologous GlcNAc transferase 1 (GNT1) and GlcNAc transferase 2 (GNT2).

In some embodiments, heterologous GNT1 and GNT2 according to the present invention are animal-derived.

In some embodiments, an animal-derived GNT1 according to the present invention comprises a heterologous Golgi localization signal.

In some embodiments, an animal-derived GNT1 according to the present invention is human GNT1. In some embodiments, an animal-derived GNT1 according to the present invention is human GNT1 comprising a heterologous Golgi localization signal.

In some embodiments, an animal-derived GNT1 according to the present invention is bovine GNT1. In some embodiments, an animal-derived GNT1 according to the present invention is bovine GNT1 comprising a heterologous Golgi localization signal.

In some embodiments, a heterologous Golgi localization signal according to the present invention is a yeast Golgi localization signal. In some embodiments, the yeast Golgi localization signal is from the yeast protein KRE2. In some particular embodiments, a human GNT1 fused to a Golgi localization signal from the yeast protein KRE2 comprises the amino acid sequence set forth in SEQ ID NO: 7. In additional particular embodiments, a human GNT1 fused to a Golgi localization signal from the yeast protein KRE2 is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 8.

In some embodiments, a heterologous Golgi localization signal according to the present invention is a Th. heterothallica Golgi localization signal. In some embodiments, the Th. heterothallica Golgi localization signal is from the Th. heterothallica protein KRE2a.

In some particular embodiments, a human GNT1 fused to a Golgi localization signal from the Th. heterothallica protein KRE2a comprises the amino acid sequence set forth in SEQ ID NO: 47. In additional particular embodiments, a human GNT1 fused to a Golgi localization signal from the Th. heterothallica protein KRE2a is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 46.

In some particular embodiments, a bovine GNT1 comprising a Golgi localization signal from the Th. heterothallica protein KRE2a comprises the amino acid sequence set forth in SEQ ID NO: 98. In additional particular embodiments, a bovine GNT1 comprising a Golgi localization signal from the Th. heterothallica protein KRE2a is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 97.

In some embodiments, an animal-derived GNT2 according to the present invention is human GNT2. In some particular embodiments, the human GNT2 comprises the amino acid sequence set forth in SEQ ID NO: 9. In additional particular embodiments, the human GNT2 is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 10.

In some embodiments, an animal-derived GNT2 according to the present invention is rat GNT2. In some particular embodiments, the rat GNT2 comprises the amino acid sequence set forth in SEQ ID NO: 11. In additional particular embodiments, the rat GNT2 is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 12.

In some embodiments, the animal-derived GNT1 is human GNT1 and the animal- derived GNT2 is human GNT2. In some embodiments, the human GNT1 comprises a yeast Golgi-localization signal. In some embodiments, the yeast Golgi localization signal is from the yeast protein KRE2. In other embodiments, the human GNT1 comprises a Th. heterothallica Golgi localization signal. In some embodiments, the Th. heterothallica Golgi localization signal is from the C1 protein KRE2a.

In some embodiments, the animal-derived GNT1 is human GNT1 and the animal- derived GNT2 is rat GNT2. In some embodiments, the human GNT1 comprises a yeast Golgi-localization signal. In some embodiments, the yeast Golgi localization signal is from the yeast protein KRE. In other embodiments, the human GNT1 comprises a Th. heterothallica Golgi localization signal. In some embodiments, the Th. heterothallica Golgi localization signal is from the Th. heterothallica protein KRE2a.

In some embodiments, the animal-derived GNT1 is bovine GNT1 and the animal- derived GNT2 is rat GNT2. In some embodiments, the bovine GNT1 comprises a Th. heterothallica Golgi localization signal. In some embodiments, the Th. heterothallica Golgi localization signal is from the Th. heterothallica protein KRE2a.

In some embodiments, the Th. heterothallica according to the present invention is genetically modified to express a heterologous flippase, wherein the heterologous flippase is the yeast FLC2p flippase. In some particular embodiments, the yeast FLC2p flippase comprises the amino acid sequence set forth in SEQ ID NO: 3. In additional particular embodiments, the yeast FLC2p flippase is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 4.

In some embodiments, the Th. heterothallica according to the present invention is genetically modified to overexpress the endogenous Th. heterothallica RFT1 flippase. In some particular embodiments, the over-expressed endogenous Th. heterothallica RFT1 flippase comprises the amino acid sequence set forth in SEQ ID NO: 5. In additional particular embodiments, the over-expressed endogenous Th. heterothallica RFT1 flippase is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 6.

In some embodiments, the genetic modification according to the present invention further comprises expression of the STT3 subunit of a heterologous oligosaccharyltransferase (heterologous STT3). In some embodiments, a heterologous STT3 according to the present invention is Leishmania major STT3. In some particular embodiments, the Leishmania major STT3 comprises the amino acid sequence set forth in SEQ ID NO: 79. In additional particular embodiments, the Leishmania major STT3 is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 78.

In some embodiments, the genetic modification according to the present invention further comprises expression of a heterologous galactosyltransferase. In some embodiments, the heterologous galactosyltransferase is an animal-derived galactosyltransferase.

In some embodiments, the animal-derived galactosyltransferase is a Xenopus tropicalis galactosyltransferase. In some embodiments, an animal-derived galactosyltransferase according to the present invention is a Xenopus tropicalis galactosyltransferase comprising a heterologous Golgi localization signal, for example, comprising the S. cerevisiae KRE2 Golgi-localization signal. In some particular embodiments, a Xenopus tropicalis galactosyltransferase comprising the S. cerevisiae KRE2 Golgi-localization signal comprises the amino acid sequence set forth in SEQ ID NO: 13. In additional particular embodiments, a Xenopus tropicalis galactosyltransferase comprising the S. cerevisiae KRE2 Golgi-localization signal is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 14.

In some embodiments, the animal-derived galactosyltransferase is a human galactosyltransferase. In some embodiments, an animal-derived galactosyltransferase according to the present invention is a human galactosyltransferase comprising a heterologous Golgi localization signal, for example, comprising the S. cerevisiae KRE2 Golgi-localization signal. In some particular embodiments, a human galactosyltransferase comprising the S. cerevisiae KRE2 Golgi-localization signal comprises the amino acid sequence set forth in SEQ ID NO: 102. In additional particular embodiments, a human galactosyltransferase comprising the S. cerevisiae KRE2 Golgi-localization signal is encoded by an exogenous polynucleotide introduced into the Th. heterothallica which comprises the sequence set forth in SEQ ID NO: 101.

In some embodiments, the Th. heterothallica is Th. heterothallica C1.

In some embodiments, the C1 is a strain modified to delete one or more genes encoding an endogenous protease.

In some embodiments, the C1 is a strain modified to delete a gene encoding an endogenous chitinase.

In some embodiments, the C1 is a strain selected from the group consisting of: wild type C1 deposit no. VKM F-3500 D, UV13-6 deposit no. VKM F-3632 D, NG7C-19 deposit no. VKM F-3633 D, UV18-25, deposit no. VKM F-3631 D, W1L#100I (prt- Δalpl Δchil Δalp2Δpyr5) deposit no. CBS141153, UV18-100f (prt -Δalpl, Δpyr5) deposit no. CBS 141147, W1L#100I (prt-Δalpl Δchil ΔpyrS) deposit no. CBS 141149, and UV18- 100f (prt -ΔalplApep4Δalp2AprtlΔpyr5) deposit no. CBS 141143. Each possibility represents a separate embodiment of the present invention.

In some embodiments, the Th. heterothallica is further genetically modified to express a heterologous mammalian glycoprotein. In some embodiments, the heterologous mammalian glycoprotein is an antibody or an antigen-binding fragment thereof.

According to another aspect, the present invention provides a method for generating a Th. heterothallica that produces glycoproteins with mammalian N-glycans, comprising: (a) deleting or disrupting the alg3 gene of the Th. heterothallica such that the Th. heterothallica fails to produce a functional alpha- 1,3- mannosyltransferase;

(b) deleting or disrupting the alg11 gene of the Th. heterothallica such that the Th. heterothallica fails to produce a functional alpha- 1 ,2-mannosyltransferase; and

(c) introducing into the Th. heterothallica: an exogenous polynucleotide encoding an endogenous flippase to induce over-expression of said endogenous flippase in the Th. heterothallica', or an exogenous polynucleotide encoding a heterologous flippase to induce expression of said heterologous flippase in the Th. heterothallica.

According to a further aspect, the present invention provides a method for producing a glycoprotein with mammalian N-glycans, the method comprising:

(A) providing a Th. heterothallica genetically modified according to the present invention;

(B) culturing the Th. heterothallica under conditions suitable for expressing the glycoprotein; and

(C) recovering the glycoprotein.

In some embodiments, the glycoprotein is a heterologous mammalian glycoprotein recombinantly expressed in the Th. heterothallica. In some particular embodiments, the glycoprotein is a human protein recombinantly expressed in the Th. heterothallica. In other embodiments, the glycoprotein is a protein of a companion animal recombinantly expressed in the Th. heterothallica. In some embodiments, the heterologous mammalian glycoprotein is an antibody or an antigen-binding fragment thereof.

According to a further aspect, the present invention provides a recombinant glycoprotein produced by the Th. heterothallica genetically modified according to the present invention, wherein the glycoprotein comprises GlcNAc ₂ Man ₃ GlcNAc ₂ (G0) glycans.

According to a further aspect, the present invention provides a recombinant glycoprotein produced by the Th. heterothallica genetically modified according to the present invention, wherein the glycoprotein comprises Gal ₁ GlcNAc ₂ Man ₃ GlcNAc ₂ (Gl) glycans, Gal ₂ GlcNAc ₂ Man ₃ GlcNAc ₂ (G2) glycans or a combination thereof.

In some embodiments, the recombinant glycoprotein produced by the Th. heterothallica genetically modified according to the present invention is a pharmaceutical grade glycoprotein. These and further aspects and features of the present invention will become apparent from the detailed description, examples and claims which follow.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1A. Lipid-linked oligosaccharide biosynthesis pathway and transfer of the oligosaccharide to a nascent polypeptide at the membrane of the endoplasmic reticulum (ER) in eukaryotic cells. The proteins deleted in C1 according to the present invention, ALG3 and ALG11, are marked with a square.

Figure IB. Illustration of yeast N-glycan structure and Th. heterothallica N-glycan structure.

Figure 2. Important target N-glycan structures of the present invention.

Figure 3. Glycan patterns on native proteins produced in a non-glycomodified C1 strain (A) and in the alg3 deletion strain M2533 (B). Hybrid glycans are designated Man ₂ - ₈ HexNAc ₃ in (A).

Figure 4. Abundance of different glycan forms on proteins secreted from the alg3+alg11 deletion/GNT1+2 expression strains.

Figure 5. Fluorograms of N-glycans of strains M2546 = human GNT1 with yeast kre2 localization signal + human GNT2 (A); M2658 = human GNT1 with yeast kre2 localization signal + rat GNT2 (B); and M2660 =human GNT1 with C1 Kre2a localization signal + rat GNT2 (C).

Figure 6. Abundance of different glycan forms in the strains expressing the yeast Flc2p or the C1 RFT1 flippase. The sum means the total fluorescence found in all the peaks.

Figure 7. Glycan analysis of G1/G2 glycan-producing C1 strains. (A) Glycan analysis results of total secreted proteins from the M3697 strain expressing X. tropicalis galactosyl transferase (M3584-557-2 Xt-GalT); (B) Glycan analysis results of total secreted proteins from the M3699 strain expressing X. tropicalis galactosyl transferase (M3586-557- 5 Xt-GalT); (C) Glycan analysis results from total secreted proteins produced by M3692 strain expressing human galactosyl transferase (M3563-559-4 hu-GalT); (D) Glycan analysis results from total secreted protein produced by M3693 strain expressing human galactosyl transferase (M3563-559-8 hu-GalT).

Figure 8. Released N-glycans of Protein A affinity purified Nivolumab from M3291 (C1 strain with alg3 deletion). Figure 9. Released N-glycans of Protein A affinity purified Nivolumab from M3544 (C1 strain with alg3 and alg11 deletions and expression of GNT1, GNT2, RFT1 and STT3). Figure 10. Glycan analysis from protein A column purified Nivolumab expressed in M3563.

Figure 11. Released N-glycans of Protein A affinity purified Nivolumab from M4112 cultivated in shake flasks (C1 strain with alg3 and alg11 deletions and expression of bovine GNT1, rat GNT2, C1 RFT1 and Lm STT3).

DETAILED DESCRIPTION OF THE INVENTION

The present invention is directed to genetic modification of the fungus Thermothelomyces heterothallica, particularly the strain C1, to produce glycoproteins with N-glycans of mammalian proteins, particularly N-glycans of human, companion animal and other mammalian proteins.

The glycoproteins produced by Th. heterothallica genetically-modified as described herein are suitable for therapeutic use in humans, companion animals such as dogs, cats and horses, and other mammals.

Protein glycosylation, namely, the covalent attachment of oligosaccharides to side chains of newly synthesized polypeptide chains in cells, is an ordered process in eukaryotic cells involving a series of enzymes that sequentially add and remove saccharide moieties. N-glycosylation is the process in which an oligosaccharide is attached to the side chain of an asparagine residue, particularly an asparagine which occurs in the sequence Asn-Xaa- Ser/Thr, where Xaa represents any amino acid except Pro.

N-glycosylation initiates in the endoplasmic reticulum (ER), where the oligosaccharide Glc ₃ Man ₉ GlcNAc ₂ is assembled on a lipid carrier, dolichol-pyrophosphate, and subsequently transferred to selected asparagine residues of polypeptides that have entered the lumen of the ER. Figure 1A illustrates the biosynthesis pathway of the lipid- linked oligosaccharide and the transfer of the oligosaccharide to a nascent polypeptide at the membrane of the ER in eukaryotic cells. The biosynthesis of the lipid-linked oligosaccharide requires the activity of several specific glycosyltransferases (ALG1, ALG2 ALG3 and so forth, see Figure 1A). It begins at the cytoplasmic side of the ER membrane and terminates in the lumen where oligosaccharyltransferase (OST) selects N-X-S/T sequons of a nascent polypeptide and generates the N- glycosidic linkage between the side chain amide of asparagine and the oligosaccharide. The flipping of the lipid-linked oligosaccharide from outside the ER to the inside is carried out by a flippase located at the ER membrane. Following transfer to the nascent polypeptide, the oligosaccharide is typically trimmed by glucosidases and mannosidases and the nascent glycoprotein is then transferred to the Golgi apparatus for further processing.

The synthesis of the dolichol pyrophosphate-bound oligosaccharide is essentially conserved in all known eukaryotes. However, further processing of the oligosaccharide as the glycoprotein moves along the secretory pathway varies greatly between lower eukaryotes such as fungi or yeasts and higher eukaryotes such as animals and plants. Thus, the final composition of a sugar side chain is different between various organisms, and depends upon the host.

In microorganisms such as yeasts, typically additional mannose and/or mannosylphosphate sugars are added, resulting in “high-mannose" type N-glycans which may contain up to 30-50 mannose residues (see Figure 1B).

In animal cells, including human, companion animal and other mammalian cells, the nascent glycoprotein is transferred to the Golgi apparatus where mannose residues are removed by Golgi-specific 1,2-mannosidases. Processing continues as the protein proceeds through the Golgi by a number of modifying enzymes including N-acetylglucosamine transferases (GnT I, GnT II, GnT IIΙ, GnT IV GnT V GnT VI), mannosidase II and fucosyltransferases that add and remove specific sugar residues. Finally, the N-glycans are acted on by galactosyl tranferases (GalT) and sialyltransferases (ST) and the finished glycoprotein is released from the Golgi apparatus. The N-glycans of animal glycoproteins have bi-, tri-, or tetra-antennary structures, and may typically include galactose, fucose and N-acetylglucosamine. Commonly the terminal residues of the N-glycans consist of sialic acid.

Th. heterothallica, unlike most fungi and yeast, does not have hypermannosylated N-glycans, but rather has “oligo mannose” glycans - Mans to Man _8-9 - and hybrid type glycans containing both Man and HexNAc residues (Man ₃ HexNac-Man ₈ HexNac) (see Figure IB). The exact structure of these hybrid glycans is not completely known. The hybrid glycans have the typical mannose residues but in addition an unknown HexNAc attached via a yet uncharacterized bond.

Since the structure, as well as the synthesis pathway, of the hybrid glycans is not fully characterized, it was unclear that such glycans can be eliminated using the genetic modifications described herein. Surprisingly, the genetic modification according to the present invention resulted in essential elimination of these structures, with over 90% of the N-glycoforms being the desired mammalian/human glycans.

The present invention is directed to genetic modification of the N-glycosylation pathway in Th. heterothallica such that it produces high percentage of glycoproteins with mammalian N-glycans, particularly human N-glycans, such as GlcNAc ₂ Man ₃ GlcNAc ₂ (“G0”), GlcNAc ₂ Man ₃ GlcNAc ₂ (Fuc) (“FG0”), Gal _1-2 GlcNAc ₂ Man ₃ GlcNAc ₂ ( “G1”/“G2”) and Gal _1-2 GlcNAc ₂ Man ₃ GlcNAc ₂ (Fuc) (‘TGI”/ “FG2”).

In particular, in some embodiments, the genetic modification of the N-glycosylation pathway in Th. heterothallica comprises the following:

1. Deletion of the C1 alg3 gene (encoding alpha-1,3- mannosyltransferase);

2. Deletion of the C1 alg11 gene (encoding alpha- 1,2-mannosyltransferase);

3. Expression of a heterologous flippase or over-expression of the endogenous flippase;

4. Expression of a heterologous GlcNAc transferase 1 (GNT1); and

5. Expression of a heterologous GlcNAc transferase 2 (GNT2).

The deletion of alg3 and alg11 terminates the synthesis of the N-glycan precursor at

Man ₃ GlcNAc ₂ .This glycan serves as the substrate for GNT1 and GNT2 that are introduced to the Th. heterothallica. Additional genetic modifications may include introduction of additional enzymes from the human, companion animal and other mammalian glycosylation pathways, such as galactosyltransferase and/or fucosyltransferase.

The above-described heterologous enzymes are expressed with targeting peptides, such that the expressed enzymes are targeted to specific cell compartments.

As used herein, when an enzyme is mentioned, it encompasses enzymatically-active fragments thereof and enzymatically-active variants thereof.

The present invention is particularly directed to engineering of the N-glycosylation pathway of Th. heterothallica. It is noted that O-glycans may be present or removed or altered by further genetic modifications of the Th. heterothallica.

It is to be understood that the genetic modifications according to the present invention are such that the genetically-modified Th. heterothallica is able to grow at sufficient rates suitable for its intended use.

As used herein "C1" or "Thermothelomyces heterothallica C1" or “Th. heterothallica C1”, all refer to Thermothelomyces heterothallica strain C1. Description of the genus Thermothelomyces and its species can be found, for example, in Marin-Felix Y (2015. Mycologica 107(3): 619-632) and van den Brink J et al. (2012, Fungal Diversity 52(1): 197-207).

It is noted that the above authors (Marin-Felix et al., 2015) proposed splitting of the genus Myceliophthora based on differences in optimal growth temperature, morphology of the conidiospore, and details of the sexual reproduction cycle. According to the proposed criteria C1 clearly belongs to the newly established genus Thermothelomyces, which contain former thermotolerant Myceliophthora species rather than to the genus Myceliophthora, which remains to include the non-thermotolerant species. As C1 can form ascospores with some other Thermothelomyces (formerly Myceliophthora ) strains with opposite mating type, C1 is best classified as Th. heterothallica strain C1, rather than Th. thermophila C1.

It must also be appreciated that the fungal taxonomy was also in constant move in the past, so the current names listed above may be preceded by a variety of older names beyond Myceliophthora thermophila (van Oorschot, 1977. Persoonia 9(3):403), which are now considered synonyms. For example, Thermothelomyces heterothallica (Marin-Felix et al., 2015. Mycologica, 3:619-63), is synonymized with Corynascus heterotchallica (von Arx et al., 1983), Thielavia heterothallica (von Klopotek, 1976. Archives of Microbiology 107(2), 223-224), Chrysosporium lucknowense and thermophile (von Klopotek, 1974. Archives of Microbiology 98(1), 365-369) as well as Sporotrichium thermophile (Alpinis 1963. Nova Hedwigia 5:74).

It is further to be explicitly understood that the present invention encompasses any strain containing a ribosomal DNA (rDNA) sequence that shows 99% homology or more to SEQ ID NO: 48, and all those strains are considered to be conspecific with Thermothelomyces heterothallica.

SEQ ID NO: 48 is 99.98% identical with the rDNA sequence found on chromosome 7 of Th. heterothallica/thermophila (listed as Myceliophtora thermophilica) ATCC 42464 rDNA sequence (ncbi.nlm .nih.gov/nucleotide/CP003008.1 ) .

Th. heterothallica strain C1 (as Chrysosporium lucknowense strain C1) was deposited in accordance with the Budapest Treaty with the number VKM F-3500 D, deposit date August 29, 1996.

The above terms also encompass genetically modified sub-strains derived from the wild type strain, which have been mutated, using random or directed approaches, for example, using UV mutagenesis, or by deleting one or more endogenous genes. For example, the C1 strain may refer to a wild type strain modified to delete one or more endogenous genes encoding an endogenous protease and/or one or more genes encoding an endogenous chitinase. For example, C1 strains (sub-strains) which are encompassed by the present invention include UV18-25, deposit No. VKM F-3631 D; strain NG7C-19, deposit No. VKM F-3633 D; and strain UV13-6, deposit No. VKM F-3632 D. Further C1 strain that may be used according to the teachings of the present invention include HC strain UV18- lOOf deposit No. CBS141147; HC strain UV18-100f deposit No. CBS141143; LC strain W1L#100I deposit No. CBS 141153; and LC strain W1L#100I deposit No. CBS 141149.

Th. heterothallica fungi in general and strain C1 in particular show higher biomass production compared to yeast strains when grown in suitable conditions. Th. heterothallica fungi can grow in large volumes of 3 dimensions (3D) liquid cultures as well as on solid medium. Several strains developed by the Applicant of the present invention are less sensitive to feedback repression by glucose and other fermentable sugars present in the fungal growth medium as carbon source compared to conventional yeast and other fungi, and can tolerate high feeding rate of the carbon source leading to high yields. Furthermore, some of these strains provide significantly reduced medium viscosity when grown in commercial fermenters compared to the high viscosity obtained with non-glucose repressed wild type Th. heterothallica fungi or with other filamentous fungi known to be used for proteins production. The low viscosity may be attributed to the morphological change of the strain from having long and highly interlaced hyphae in the parental strain(s) to short and less interlaced hyphae in the developed strain(s). Low medium viscosity is highly advantageous in large scale industrial production in fermenters. For example, the Th. heterothallica C1 strain UV18-25, deposit No. VKM F-3631 D, which shows reduced sensitivity to glucose repression, has been grown industrially to produce recombinant enzymes at volumes of more than 100,000 liters.

In some embodiments, the C1 strain of the present invention is a strain modified to delete a plurality (i.e., at least two) genes encoding endogenous proteases. In some embodiments, the C1 strain is a strain modified to delete at least four genes encoding endogenous proteases. In additional embodiments, the C1 strain is a strain modified to delete at least five genes encoding endogenous proteases. In some particular embodiments, the C1 strain is a strain modified to delete at least six genes encoding endogenous proteases. In additional particular embodiments, the C1 strain is a strain modified to delete at least eight genes encoding endogenous proteases. It is to be explicitly understood that the teachings of the present invention encompass mutants, derivatives, progeny, clones and analogous of the Th. heterothallica C1 strains, as long as these derivatives, progeny, clones and analogous, when genetically modified according to the teachings of the present invention, are capable of growing and producing a protein with N-glycans as described herein.

It is to be explicitly understood that the term “derivative” with reference to fungal line encompasses any fungal parent line with modifications positively affecting product yield, efficiency, or efficacy, or affecting any trait improving the fungal derivative as a tool to produce heterologous proteins with N-glycans of mammalian proteins, particularly of human, companion animals and other mammalian proteins, as described herein. As used herein, the term “progeny" refers to an unmodified descendant from the parent fungal line, such as cell from cell.

As used herein, “glycan” refers to an oligosaccharide chain that can be linked to a carrier such as an amino acid, peptide, polypeptide, lipid or a reducing end conjugate. The present invention particularly relates to N-linked glycans (“N-glycan”) conjugated to a polypeptide N-glycosylation site such as -Asn-Xxx-Ser/Thr- by N-linkage to side-chain amide nitrogen of asparagine residue (Asn), where Xxx is any amino acid residue except Pro. The present invention may further relate to glycans as part of dolichol-phospho- oligosaccharide (Dol-P-P-OS) precursor lipid structures, which are precursors of N-linked glycans in the endoplasmic reticulum of eukaryotic cells. The precursor oligosaccharides are bound by their reducing end to two phosphate residues on the dolichol lipid.

The monosaccharides typically constituting N-glycans found in mammalian glycoproteins, include, without limitation, N-acetylglucosamine (abbreviated “GlcNAc”), mannose (abbreviated “Man”), glucose (abbreviated “Glc”), galactose (abbreviated “Gal”), sialic acid (abbreviated “Neu5Ac”) and fucose (abbreviated “Fuc”).

N-glycans share a common pentasaccharide referred as the “core” structure Man ₃ GlcNAc ₂ (abbreviated “Mans”, see Figure 2). Important target glycan structures of the present invention include N-glycans which have one GlcNAc residue on the terminal 1,3 mannose arm of the core structure and one GlcNAc residue on the terminal 1,6 mannose arm of the core structure. Such N-glycans include: GlcNAc ₂ Man ₃ GlcNAc ₂ (termed “G0” glycoform), Gal _1-2 GlcNAc ₂ Man ₃ GlcNAc ₂ (termed “G1” or “G2” glycoform according to the number of galactose residues), and their core fucosylated glycoforms: GlcNAc ₂ Man ₃ GlcNAc ₂ (Fuc) (“G0F’ or “FG0”) and Gal _1-2 GlcNAc ₂ Man ₃ GlcNAc ₂ (Fuc) (“G1F’ and “G2F’, or “FG1” and “FG2”) (Figure 2).

The term “ alg3 gene” refers to the gene encoding alpha- 1,3- mannosyltransferase. The term “alpha- 1,3- mannosyltransferase” refers to dolichy 1-P-Man : Man ₅ GlcN Ac ₂ -PP- dolichol alpha- 1 ,3-mannosyltransferase (EC 2.4.1.258) , which is an ER-resident enzyme that catalyzes the reaction: dolichyl beta-D-mannosyl phosphate + D-Man-alpha-( 1 ->2)-D-Man-alpha-( 1 ->2)- D-Man-alpha-( 1 ->3)-[D-Man-alpha-( 1 ->6)]-D-Man-beta-( 1->4)-D-GlcNAc-beta-( 1 ->4)- D-GlcNAc-diphosphodolichol

D-Man-alpha-(1->2)-D-Man-alpha-( 1 ->2)-D-Man-alpha-( 1->3)-[D-Man-alpha-( 1 - >3)-D-Man-alpha-( 1 ->6)] -D-Man-beta-( 1 ->4)-D-GlcNAc-beta-( 1 ->4)-D-GlcN Ac- diphosphodolichol + dolichyl phosphate

In some particular embodiments, “ alg3 gene” is the gene encoding alpha- 1,3- mannosyltransferase of C1 (ortholog of JGI M. thermophila genome (mycocosm.jgi.doe.gov) accession no. 2310419).

The Th. heterothallica of the present invention is genetically modified by deletion or disruption of the alg3 gene such that the Th. heterothallica fails to produce a functional alpha- 1,3- mannosyltransferase. The Th. heterothallica of the present invention does not display a detectable alpha- 1,3- mannosyltransferase activity.

The term “ alg11 gene” refers to the gene encoding alpha- 1 ,2-mannosyltransferase. The term “alpha- 1 ,2-mannosyltransferase” refers to GDP-Man:Man ₃ GlcNAc ₂ -PP-dolichol alpha- 1 ,2-mannosyltransferase (EC 2.4.1.131), which is an ER-resident enzyme catalyzes the reaction:

2 GDP-alpha-D-mannose + D-Man-alpha-( 1 ->3 )-[D-Man-alpha-( 1 ->6)] -D-Man- beta-(1->4)-D-GlcNAc-beta-(1->4)-D-GlcNAc-diphosphodol ichol

2 GDP + D-Man-alpha-( 1->2)-D-Man-alpha-( 1 ->2)-D-Man-alpha-( 1 ->3)-[D-Man- alpha-( 1 ->6)]-D-Man-beta-( 1 ->4)-D-GlcNAc-beta-( 1->4)-D-GlcNAc-diphosphodolichol In some particular embodiments, “ alg11 gene” is the gene encoding alpha- 1,2- mannosyltransferase of C1 (ortholog of JGI M. thermophila genome (mycocosm.jgi.doe.gov) accession no. 2312819).

The Th. heterothallica of the present invention is genetically modified by deletion or disruption of the alg11 gene such that the Th. heterothallica fails to produce a functional alpha- 1,2- mannosyltransferase. The Th. heterothallica of the present invention does not display a detectable alpha- 1,2- mannosyltransferase activity.

The term “flippase” (EC 7.6.2.1) refers to an enzyme that transfers the lipid-linked glycan precursor during its synthesis in the ER from the cytosolic side to the luminal side of the ER.

The term “GlcNAc transferase 1”, abbreviated “ GNT1” (also “GnTI”), refers to alpha- 1,3-mannosyl-glycoprotein 2-beta-N-acetylglucosaminyltransferase (EC 2.4.1.101), which is a Golgi-resident enzyme that transfers a GlcNAc residue from UDP-GlcNAc to the acceptor substrate Man ₅ GlcNAc ₂ , to produce GlcNAcMan ₅ GlcNAc ₂ . In the present invention the synthesis of the N-glucan precursor terminates at Man ₃ GlcNAc ₂ in view of the deletion of alg3 and alg11, therefore the glycan Man ₃ GlcNAc ₂ serves as the substrate for GNT1, to produce GlcNAcMan ₃ GlcNAc ₂ .

The term “GlcNAc transferase 2”, abbreviated “GNT2” (also “GnTII”), refers to alpha- 1,6-mannosy 1-glycoprotein 2-beta-N-acelylglucosaminyltransferase (EC 2.4.1.143), which is a Golgi-resident enzyme that transfers a GlcNAc residue from UDP-GlcNAc to the free terminal mannose residue in GlcNAcMan ₃ GlcNAc ₂ , to produce GlcNAc ₂ Man ₃ GlcNAc ₂ .

The terms "STT3 subunit of oligosaccharyltransferase", "STT3 protein" or simply "STT3" are used herein interchangeably to refer to dolichyl-diphosphooligosaccharide- protein glycosyltransferase subunit (EC:2.4.99.18). It is the catalytic subunit of the oligosaccharyltransferase (OST) complex that catalyzes the initial transfer of a defined glycan (Glc ₃ Man ₉ GlcNAc ₂ in eukaryotes) from the lipid carrier dolichol-pyrophosphate to an asparagine residue within an Asn-X-Ser/Thr consensus motif in nascent polypeptide chains, the first step in protein N-glycosylation. STT3 protein catalyzes the reaction:

Dolichyl diphosphooligosaccharide + L-asparaginyl-[protein] Dolichyl diphosphate + H ⁺ + N ⁴ -(oligosaccharide-(1→4)-N-acetyl-β-D- glucosaminyl-(1→4)-N-acetyl-β-D-glucosaminyl)-L-asparagin y-[prote-in]

The term “galactosyltransferase” (EC 2.4.1.38) refers to a Golgi-resident enzyme that transfers β-linked galactosyl residues to terminal N-acetylglucosamine.

The term “heterologous”, when referring to a gene, enzyme, protein or peptide sequence such as a subcellular localization signal, is used herein to describe a gene, enzyme, protein or peptide sequence that is not naturally found or expressed in C1. When referring to a subcellular localization signal, the term also describes a subcellular localization signal that is different from the one naturally found in the respective protein.

The term “endogenous”, when referring to a gene, enzyme, protein or peptide sequence such as a subcellular localization signal, refers to a gene, enzyme, protein or peptide sequence that is naturally present in C1.

The term “exogenous”, when referring to a polynucleotide, is used herein to describe a synthetic polynucleotide that is exogenously introduced into the C1 via transformation. The exogenous polynucleotide may be introduced into the C1 in a stable or transient manner, so as to produce a ribonucleic acid (RNA) molecule and subsequently a polypeptide molecule.

Expression constructs and vectors

The terms “expression construct”, “DNA construct” or “expression cassette” are used herein interchangeably and refer to an artificially assembled or isolated nucleic acid molecule which includes a nucleic acid sequence encoding a protein of interest and which is assembled such that the protein of interest is expressed in a target host cell. An expression construct typically comprises appropriate regulatory sequences operably linked to the nucleic acid sequence encoding the protein of interest. An expression construct may further include a nucleic acid sequence encoding a selection marker.

The terms “nucleic acid sequence”, “nucleotide sequence” and “polynucleotide” are used herein to refer to polymers of deoxyribonucleotides (DNA), ribonucleotides (RNA), and modified forms thereof in the form of a separate fragment or as a component of a larger construct. A nucleic acid sequence may be a coding sequence, i.e., a sequence that encodes for an end product in the cell, such as a protein. A nucleic acid sequence may also be a regulatory sequence, such as, for example, a promoter. The terms “peptide”, “polypeptide” and “protein” are used herein to refer to a polymer of amino acid residues. The term “peptide” typically indicates an amino acid sequence consisting of 2 to SO amino acids, while “protein" indicates an amino acid sequence consisting of more than 50 amino acid residues.

A sequence (such as a nucleic acid sequence and an amino acid sequence) that is “homologous” to a reference sequence refers herein to percent identity between the sequences, where the percent identity is at least 75%, preferably at least 80%, at least 85%, at least 90%, at least 95%, at least 98% or at least 99%. Each possibility represents a separate embodiment of the present invention. Homologs of the sequences described herein are encompassed within the present invention. Protein homologs are encompassed as long as they maintain the activity of the original protein. Homologous nucleic acid sequences include variations related to codon usage and degeneration of the genetic code. Sequence identity may be determined using nucleotide/amino acid sequence comparison algorithms, as known in the art.

Nucleic acid sequences encoding the polypeptides of the present invention may be optimized for expression. Examples of such sequence modifications include, but are not limited to, an altered G/C content to more closely approach that typically found in Th. heterothallica, and the removal of codons atypically found in this fungus, commonly referred to as codon optimization.

The phrase “codon optimization” refers to the selection of appropriate DNA nucleotides for use within a structural gene or fragment thereof that approaches codon usage within the organism of interest, and/or to a process of modifying a nucleic acid sequence for enhanced expression in the host cell of interest by replacing at least one codon (e.g., about or more than about 1, 2, 3, 4, 5, 10, 15, 20, 25, 50, or more codons) of the native sequence with codons that are more frequently or most frequently used in the genes of that host cell while maintaining the native amino acid sequence. Various species exhibit particular bias for certain codons of a particular amino acid. Codon bias (differences in codon usage between organisms) often correlates with the efficiency of translation of messenger RNA (mRNA), which is in turn believed to be dependent on, among other things, the properties of the codons being translated and the availability of particular transfer RNA (tRNA) molecules. The predominance of selected tRNAs in a cell is generally a reflection of the codons used most frequently in protein synthesis. Accordingly, genes can be tailored for optimal gene expression in a given organism based on codon optimization. Therefore, an optimized gene or nucleic acid sequence refers to a gene in which the nucleotide sequence of a native or naturally occurring gene has been modified in order to utilize statistically- preferred or statistically-favored codons within the organism. The present invention explicitly encompasses polynucleotides encoding the enzyme of interest as disclosed herein which are codon optimized for expression in Th. heterothallica.

The term “regulatory sequences” refer to DNA sequences which control the expression (transcription) of coding sequences, such as promoters and terminators.

The term “promoter” is directed to a regulatory DNA sequence which controls or directs the transcription of another DNA sequence in vivo or in vitro. Usually, the promoter is located in the 5’ region (that is, precedes, located upstream) of the transcribed sequence. Promoters may be derived in their entirety from a native source, or be composed of different elements derived from different promoters found in nature, or even comprise synthetic nucleotide segments. Promoters can be constitutive (i.e. promoter activation is not regulated by an inducing agent and hence rate of transcription is constant), or inducible (i.e., promoter activation is regulated by an inducing agent). In most cases the exact boundaries of regulatory sequences have not been completely defined, and in some cases cannot be completely defined, and thus DNA sequences of some variation may have identical promoter activity.

The term “terminator” is directed to another regulatory DNA sequence which regulates transcription termination. A terminator sequence is operably linked to the 3' terminus of the nucleic acid sequence to be transcribed.

The terms “Th. heterothallica promoter” and “Th. heterothallica terminator” indicate promoter and terminator sequences suitable for use in Th. heterothallica, i.e., capable of directing gene expression in Th. heterothallica. In some particular embodiments, C1 promoters and C1 terminators are used, which indicate promoter and terminator sequences capable of directing gene expression in C1.

According to some embodiments, the Th. heterothallica promoter/terminator is derived from an endogenous gene of Th. heterothallica. According to other embodiments the Th. heterothallica promoter/terminator is derived from a gene exogenous to Th. heterothallica.

Suitable constitutive promoters and terminators include, for example, those of C1 glycolytic genes such as phosphoglycerale kinase gene (PGK) (Uniprot: G2QLD8, NCBI Reference Sequence: XM_003665967), glyceraldehyde 3-phosphate dehydrogenase (GPD) (Uniprot: G2QPQ8, NCBI Reference Sequence: XM_003666768), phosphofructokinase (PFK) (Uniprot: G2Q605, NCBI Reference Sequence: XM_003659879); or the β- glucosidase 1 gene bgl1 (Accession number: XM_003662656); or triose phosphate isomerase (TPI) (Uniprot: G2QBR0, NCBI Reference Sequence: XM_003663200); or actin (ACT) (Uniprot: G2Q7Q5, NCBI Reference Sequence: XM_003662111); or the C1 cbhl promoter (GenBank AX284115) or C1 chil promoter (GenBank HI550986). Additional promoters that can be used are Aspergillus nidulans gpdA promoter; and synthetic promoters described in Rantasalo et al. (2018 NAR 46(18):el 11). Synthetic promoters that can be used with the present invention are further described in WO 2017/144777. As exemplary terminators, the terminator of the C1 chitinase lgene chil (GenBank HI550986), cellobiohydrolase 1 cbhl (GenBank AX284115) can be used, or the yeast adhl terminator.

Exemplary promoter and terminator sequences, and promoter/terminator pairs, are provided in the Examples section that follows. In some embodiments, promoter sequences for use with the present invention are selected from the group consisting of: promoter-8, bgl8 promoter, promoter-9, promoter-3, promoter- 1 and TEF1A promoter, as exemplified hereinbelow. Each possibility represents a separate embodiment of the present invention.

The term "operably linked" means that a selected nucleic acid sequence is in proximity with a regulatory element (promoter or terminator) to allow the regulatory element to regulate expression of the selected nucleic acid sequence.

The terms "localization signal”, "localization sequence", “subcellular targeting peptide/signal/sequence” and the like are used herein interchangeably and refer to a short peptide sequence (usually 5-30 amino acids long) included within a protein sequence (typically present at one terminus of the protein such as the N-terminus) that directs the protein to a particular subcellular localization within the cell. For example, a Golgi localization signal targets the protein to the Golgi apparatus. A “heterologous localization signal”, for example, a “heterologous Golgi localization signal”, indicates a localization signal that is not the one naturally found in the protein. In some embodiments, “heterologous” refers to a localization signal from another organism.

In some embodiments, localization signals of proteins expressed in Th. heterothallica according to the present invention are derived from endogenous genes of Th. heterothallica. For example, in some embodiments, a Golgi localization signal from the C1 protein KRE2a (ortholog of JGIM. thermophila genome (mycocosm.jgi.doe.gov) accession no. 2300989) is used. The amino acid sequence of the Golgi localization signal of the C1 protein KRE2a is set forth in SEQ ID NO: 1.

In other embodiments, localization signals of proteins expressed in Th. heterothallica according to the present invention are derived from genes exogenous to Th. heterothallica (heterologous localization signals). For example, in some embodiments, animal-derived enzymes expressed in Th. heterothallica according to the present invention are expressed with their own naturally-occurring Golgi localization signals. As another example, in some embodiments, a Golgi localization signal from yeast proteins, such as the S. cerevisiae protein KRE2 (Genbank accession no. CAA44516) is used. The amino acid sequence of the Golgi localization signal of the S. cerevisiae protein KRE2 is set forth in SEQ ID NO: 2.

As used herein, the term "in frame", when referring to one or more nucleic acid sequences, indicates that these sequences are linked such that their correct reading frame is preserved.

Expression constructs according to some embodiments of the present invention comprise a Th. heterothallica promoter sequence and a Th. heterothallica terminator sequence operably linked to a nucleic acid sequence encoding an enzyme, such as a flippase, GNT1 or GNT2. In some particular embodiments, expression constructs of the present invention comprise a C1 promoter sequence and a C1 terminator sequences operably linked to a nucleic acid sequence encoding an enzyme, such as a flippase, GNT1 or GNT2.

A particular expression construct may be assembled by a variety of different methods, including conventional molecular biology methods such as polymerase chain reaction (PCR), restriction endonuclease digestion, in vitro and in vivo assembly methods, as well as gene synthesis methods, or a combination thereof. Exemplary expression constructs and methods for their construction are provided in the Examples section below.

Deletion of alg3 and alp 11

Gene deletion techniques enable the partial or complete removal of a gene, thereby eliminating its expression. In such methods, deletion of the gene may be accomplished by homologous recombination using a plasmid that has been constructed to contiguously contain the 5' and 3' regions flanking the gene.

Gene deletion may also be performed by inserting into the gene a disruptive nucleic acid construct, also termed herein a deletion construct. A disruptive construct may be simply a selectable marker gene accompanied by 5' and 3' regions homologous to the gene. The selectable marker enables identification of transformants containing the disrupted gene. Alternatively or additionally, the disruptive nucleic acid construct may comprise one or more polynucleotides encoding heterologous proteins to be expressed in the host cell.

Exemplary deletion constructs for alg3 and algll and procedures for carrying out the deletion are provided in the Examples section below. As described herein, the alg3 gene is deleted using a disruptive construct comprising a selectable marker, as shown in Example 1 below. As further described herein, the alg11 gene is deleted using a disruptive construct comprising polynucleotide sequences encoding heterologous GNT1 and GNT2, to express the heterologous GNT1 and GNT2 from the alg11 locus (shown in Example 2 below).

The deletion(s) may be confirmed using PCR with appropriate primers flanking the disruptive construct(s).

Flippases

In some embodiments, the Th. heterothallica of the present invention is genetically modified to express a heterologous flippase. In some particular embodiments, the heterologous flippase is a yeast flippase. In additional particular embodiments, the yeast flippase is the S. cerevisiae mutant flippase FLC2p, which is a C-terminally truncated version of the S. cerevisiae ER-localized flippase FLC2. The amino acid sequence of FLC2p is set forth in SEQ ID NO: 3. FLC2p may be expressed in Th. heterothallica according to the present invention by the introduction of an exogenous polynucleotide encoding FLC2p, comprising the nucleic acid sequence encoding FLC2p operably linked to regulatory sequences operable in Th. heterothallica. An exemplary nucleic acid sequence encoding FLC2p for use according to the present invention is set forth in SEQ ID NO: 4.

In additional embodiments, the Th. heterothallica of the present invention is genetically modified to over-express the endogenous Th. heterothallica RFT1 flippase. In some particular embodiments, Th. heterothallica C1 is genetically modified to over-express the endogenous C1 RFT1 flippase. The amino acid sequence of C1 RFT1 is set forth in SEQ ID NO: 5. Over-expression of RFT1 in Th. heterothallica according to the present invention may be performed by the introduction of an exogenous polynucleotide encoding RFT1, comprising the nucleic acid sequence encoding RFT1 operably linked to regulatory sequences operable in Th. heterothallica. The genomic sequence encoding C1 RFT1 is set forth in SEQ ID NO: 6.

In some exemplary embodiments, the Th. heterothallica is genetically modified to delete or disrupt alg3 and alg11 and to express the yeast FLC2p flippase, and further genetically modified to express an animal-derived GNT1 comprising a heterologous Golgi localization signal and an animal-derived GNT2, for example to express human GNT1 comprising a Golgi localization signal from the yeast protein KRE2 and human GNT2.

In some exemplary embodiments, the Th. Heterothallica is genetically modified to delete or disrupt alg3 and alg11 and to over-express the endogenous Th. heterothallica RFT1 flippase, and further genetically modified to express an animal-derived GNT1 comprising a heterologous Golgi localization signal and an animal-derived GNT2, for example, to express human GNT1 comprising a Golgi localization signal from the yeast protein KRE2 and human GNT2.

In additional exemplary embodiments, the Th. heterothallica is genetically modified by: deletion or disruption of alg3 and alg11 ; over-expression of the endogenous Th. heterothallica RFT1 flippase; expression of human GNT1 comprising Th. heterothallica KRE2a Golgi-localization signal and rat GNT2; and expression of Leishmania major STT3. In some embodiments, such a Th. heterothallica is further genetically modified by expression of Xenopus tropicalis galactosyltransferase or human galactosyltransferase.

In additional exemplary embodiments, the Th. heterothallica is genetically modified by: deletion or disruption of alg3 and alg11 ; over-expression of the endogenous Th. heterothallica RFT1 flippase; expression of bovine GNT1 comprising Th. heterothallica KRE2a Golgi-localization signal and rat GNT2; and expression of Leishmania major STT3.

GNT1 and GNT2

In some embodiments, the Th. heterothallica of the present invention is genetically modified to express heterologous GNT1 and GNT2. In some embodiments, the heterologous GNT1 and GNT2 are animal-derived. As used herein, “animal-derived” encompasses mammalian origin including for example companion animals such as dogs and cats and additional mammals such as horses. As exemplified herein below, animal-derived includes for example a rat origin. The term “animal-derived” further encompasses human-derived, as further exemplified hereinbelow. The heterologous GNT1 and GNT2 may be expressed in Th. heterothallica according to the present invention by the introduction of one or more exogenous polynucleotide encoding the GNT1 and GNT2, comprising nucleic acid sequences encoding the GNT1 and GNT2 operably linked to regulatory sequences operable in Th. heterothallica. In some embodiments, the nucleic acid sequences encoding the GNT1 and GNT2 are included in a single polynucleotide that is introduced into the Th. heterothallica. In other embodiments, the nucleic acid sequences encoding the GNT1 and GNT2 are included in two different polynucleotides that are introduced into the Th. heterothallica.

As described herein and shown in Example 2 below, the nucleic acid sequences encoding the heterologous GNT1 and GNT2 are included in a single polynucleotide that is designed for insertion into the alg11 locus, to obtain deletion of alg11 and at the same time expression of the heterologous GNT1 and GNT2 from the alg11 locus.

In some embodiments, the GNT1 is expressed in the Th. heterothallica with its own naturally-occurring Golgi-localization signal. In other embodiments, the GNT1 is expressed in the Th. heterothallica with a heterologous Golgi-localization signal.

In some embodiments, the heterologous Golgi-localization signal is a yeast Golgi- localization signal. In some particular embodiments, the heterologous Golgi-localization signal is from the yeast protein KRE2 alpha- 1,2-mannosyltransferasc. In some exemplary embodiments, the heterologous Golgi-localization signal is from the KRE2 of S. cerevisiae. The amino acid sequence of the Golgi-localization signal from the KRE2 of S. cerevisiae is set forth in SEQ ID NO: 2.

In other embodiments, the heterologous Golgi-localization signal is from a filamentous fungus. In some embodiments, the heterologous Golgi-localization signal is from Th. heterothallica. In some particular embodiments, the heterologous Golgi- localization signal is from the C1 homolog of the yeast protein KRE2. The amino acid sequence of the Golgi-localization signal from the KRE2 of C1 is set forth in SEQ ID NO: 1.

In some embodiments, the GNT 1 is human GNT 1. In some embodiments, the human GNT1 that is introduced into the Th. heterothallica comprises a heterologous Golgi- localization signal. In some embodiments, the GNT1 is human GNT1 comprising a yeast Golgi-localization signal. In some particular embodiments, the GNT1 is human GNT1 comprising the Golgi-localization signal from the protein KRE2 of S. cerevisiae. The amino acid sequence of human GNT1 comprising the Golgi-localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 7. An exemplary nucleic acid sequence for use according to the present invention encoding human GNT1 comprising the Golgi- localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 8.

In some embodiments, the GNT2 is human GNT2. GNT2 is typically expressed with its own naturally-occurring Golgi localization signal. The amino acid sequence of human GNT2 is set forth in SEQ ID NO: 9. An exemplary nucleic acid sequence of a polynucleotide for use according to the present invention encoding human GNT2 is set forth in SEQ ID NO: 10.

In other embodiments, the GNT2 is rat GNT2. GNT2 is typically expressed with its own naturally-occurring Golgi localization signal. The amino acid sequence of rat GNT2 is set forth in SEQ ID NO: 11. An exemplary nucleic acid sequence of a polynucleotide for use according to the present invention encoding rat GNT2 is set forth in SEQ ID NO: 12. Exemplary combinations of GNT1 and GNT2 according to the present invention include:

- human GNT1 with yeast KRE2 Golgi-localization signal and human GNT2;

- human GNT1 with yeast KRE2 Golgi-localization signal and rat GNT2;

- human GNT1 with C1 KRE2a Golgi-localization signal and human GNT2;

- human GNT1 with C1 KRE2a Golgi-localization signal and rat GNT2;

- bovine GNT1 with C1 KRE2a Golgi-localization signal and rat GNT2.

Each combination represents a separate embodiment of the present invention.

Galactosvltransferase

In some embodiments, the Th. heterothallicaof the present invention is genetically modified to express a heterologous galactosyltransferase. In some embodiments, the heterologous galactosyltransferase is animal-derived.

A galactosyltransferase may be expressed in Th. heterothallica according to the present invention by the introduction of an exogenous polynucleotide encoding the galactosyltransferase, comprising the nucleic acid sequence encoding the galactosyltransferase operably linked to regulatory sequences operable in Th. heterothallica.

In some embodiments, the galactosyltransferase is expressed in the Th. heterothallica with its own naturally-occurring Golgi-localization signal. In other embodiments, the galactosyltransferase is expressed in the Th. heterothallica with a heterologous Golgi-localization signal, as described above. In some embodiments, the galactosyltransferase is from Xenopus tropicalis (XtGalTl). In some embodiments, the galactosyltransferase from Xenopus tropicalis that is introduced into the Th. heterothallica comprises a heterologous Golgi-localization signal. In some particular embodiments, the galactosyltransferase from Xenopus tropicalis comprises the Golgi-localization signal from the protein KRE2 of S. cerevisiae. The amino acid sequence of Xenopus tropicalis galactosyltransferase with the Golgi-localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 13. An exemplary nucleic acid sequence of a polynucleotide for use according to the present invention encoding Xenopus tropicalis galactosyltransferase with the Golgi-localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 14.

In some embodiments, the galactosyltransferase is a human galactosyltransferase (huGalTl). In some embodiments, the human galactosyltransferase that is introduced into the Th. heterothallica comprises a heterologous Golgi-localization signal. In some particular embodiments, the human galactosyltransferase comprises the S. cerevisiae KRE2 Golgi- localization signal. The amino acid sequence of human galactosyltransferase with the Golgi- localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 102. An exemplary nucleic acid sequence of a polynucleotide for use according to the present invention encoding human galactosyltransferase with the Golgi-localization signal from the protein KRE2 of S. cerevisiae is set forth in SEQ ID NO: 101.

STT3 oligosaccharvltransferase

In some embodiments, the Th. heterothallica of the present invention is genetically modified to express a heterologous STT3 subunit of oligosaccharyltransferase. In some particular embodiments, the heterologous STT3 is-Leishmania major STT3-The amino acid sequence of Leishmania major STT3 is set forth in SEQ ID NO: 79. Leishmania major STT3 may be expressed in Th. heterothallica according to the present invention by the introduction of an exogenous polynucleotide encoding Leishmania major STT3, comprising the nucleic acid sequence encoding Leishmania major STT3 operably linked to regulatory sequences operable in Th. heterothallica. An exemplary nucleic acid sequence encoding Leishmania major STT3 for use according to the present invention is set forth in SEQ ID NO: 78. Genetically-engineered Th. heterothallica

Th. heterothallica cells genetically engineered to produce glycoproteins with N- glycans of mammalian proteins (particularly human and companion animal proteins) according to the present invention are generated by modifying, such as deleting, two endogenous genes of Th. heterothallica, alg3 and alg11, such that the genes fail to produce functional proteins, and expressing exogenous polynucleotides encoding various enzymes.

It is to be understood that the genetic modification of Th. heterothallica as disclosed herein does not necessarily requires that each and every cell of the genetically-modified Th. heterothallica be modified, as long as the desired outcome disclosed herein of production of glycoproteins with N- glycans of mammalian proteins (particularly human and companion animal proteins) is obtained.

In some embodiments, the Th. heterothallica is further genetically modified to express a heterologous mammalian glycoprotein. In some embodiments, the heterologous mammalian glycoprotein is an antibody or an antigen-binding fragment thereof.

In some exemplary embodiments, the Th. heterothallica is genetically modified to express Nivolumab or an antigen-binding fragment thereof. In additional exemplary embodiments, the Th. heterothallica is genetically modified to express Nivolumab light chain with Th. heterothallica CBH1 signal sequence and Nivolumab heavy chain with Th. heterothallica CBH1 signal sequence. An exemplary sequence encoding Nivolumab LC with Th. heterothallica CBH1 signal sequence is set forth as SEQ ID NO: 103. An exemplary sequence encoding Nivolumab HC with Th. heterothallica CBH1 signal sequence is set forth as SEQ ID NO: 105. In some exemplary embodiments, the Th. heterothallica expressing Nivolumab comprises the expression constructs set forth as SEQ ID NO: 66 and SEQ ID: 67.

In some embodiments, the present invention provides a Th. heterothallica cell genetically modified as disclosed herein.

The deletion of the endogenous genes is described above and is also demonstrated in the Examples section below.

The expression of an exogenous polynucleotide is carried out by introducing into Th. heterothallica cells, particularly into the nucleus of Th. heterothallica cells, an expression construct comprising a nucleic acid encoding a protein to be expressed in C1. In particular, the genetic modification according to the present invention means incorporation of the expression construct to the host genome. Introduction of an expression construct into Th. heterothallica cells, i.e., transformation of Th. heterothallica, can be performed by methods for transforming filamentous fungi, for example, using the protoplast transformation method described in the Examples section below.

To facilitate easy selection of transformed cells, a selection marker may be transformed into the Th. heterothallica cells. A "selection marker" indicates a polynucleotide encoding a gene product conferring a specific type of phenotype that is not present in non-transformed cells, such as an antibiotic resistance (resistance markers), ability to utilize a certain resource (utilization/auxotrophic markers) or expression of a reporter protein that can be detected, e.g. by spectral measurements. Auxotrophic markers are typically preferred as a means of selection in the food or pharmaceutical industry. The selection marker can be on a separate polynucleotide co-transformed with the expression construct, or on the same polynucleotide of the expression construct. Following transformation, positive transformants are selected by culturing the C1 cells on e.g., selective media according to the chosen selection marker. In some cases, a split marker system is used, where the selection marker is split into two plasmids and a functional selection marker is formed only when the two plasmids are co-transformed and joined together via homologous recombination.

When the synthetic expression system is used, an expression cassette coding for a suitable synthetic transcription factor (sTF) is introduced into the host cell.

The transformed DNA may integrate into Th. heterothallica chromosomes through homologous recombination or non-homologous end joining. To facilitate targeted integration into a specific locus in the genome, sequences corresponding to the target locus are incorporated into the same polynucleotide with the expression construct.

Selected clones are then grown and examined for the production of protein with the desired N-glycoforms. The genetically-modified Th. heterothallica is cultured under suitable conditions, for example, as exemplified in the Examples section that follows. According to certain embodiments, the fungus is grown at a temperature in the range of from about 20°C to about 45°C and at a medium pH of from about 4.0 to about 8.0. Particular media types may be selected according to regulatory requirements of the end product. The produced glycoproteins may be isolated and analyzed. Expression of GNT1, GNT2 and optionally additional enzymes such as a galactosyltransferase in the Th. heterothallica may be determined by structural analysis of N-glycans produced by the C1, for example as described in the Examples section below.

A Th. heterothallica genetically modified according to the present invention produces G0 (Man ₃ GlcNAc ₂ ) as a final N-glycan structure or an intermediate N-glycan structure.

In some embodiments, a Th. heterothallica genetically modified by deletion or disruption of alg3, deletion or disruption of alg11, expression of a heterologous flippase or over-expression of an endogenous flippase, expression of heterologous GNT1 and GNT2 and optionally expression of a heterologous STT3 oligosaccharyltransferase produces secreted glycoproteins wherein G0 constitutes at least 80% of the N-glycans on the secreted glycoproteins, preferably at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94% or even at least 95% of the N-glycans on the secreted glycoproteins. Each value represents a separate embodiment of the present invention.

In some embodiments, a Th. heterothallica which is further genetically modified to express a heterologous galactosyltransferase produces secreted glycoproteins wherein G1 and G2 (total of both G1 and G2) constitute at least 75% of the N-glycans on the secreted glycoproteins, preferably at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94% or even at least 95% of the N-glycans on the secreted glycoproteins. Each value represents a separate embodiment of the present invention.

In some embodiments, the genetic modification of the Th. heterothallica does not include expression of a heterologous oligosaccharyltransferase (OST).

The following examples are presented in order to more fully illustrate certain embodiments of the invention. They should in no way, however, be construed as limiting the broad scope of the invention. One skilled in the art can readily devise many variations and modifications of the principles disclosed herein without departing from the scope of the invention.

EXAMPLES

Example 1 - Deletion of C1 alg3 gene

The first step of glycoengineering in C1 was the deletion of the gene encoding the mannosyltransferase ALG3 in order to terminate the glycan precursor synthesis at the level of DolP-GlcNAc ₂ Man ₅ , and therefore to eliminate larger fungal type glycans from the C1 glycan pattern.

The deletion was carried out by transforming a deletion construct into C1. The sequence of the deletion construct is set forth in SEQ ID NO: 15. The deletion construct includes an alg35’ flanking sequence (positions 5 -1,006 of SEQ ID NO: 15), a direct repeat sequence (positions 1,016 - 1,367 of SEQ ID NO: 15), marker genes (positions 1,375 - 8,164 of SEQ ID NO: 15), and an alg3 3’ flanking sequence (positions 8,173 - 9,172 of SEQ ID NO: 15).

The different elements of alg3 were amplified from C1 genomic DNA and cloned with the marker genes into a backbone vector (pRS426) by Gibson cloning with NEBbuilder

TM HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions. The deletion construct was excised from the plasmid and transformed into the C1 strain DNL115 having 6 deletions of protease genes according to the following protocol: 50 ml of complete medium (medium components: 70mM NaNO ₃ , 7mM KC1, 11mM KH ₂ PO ₄ , 1% Glucose, 2mM MgSO ₄ , 0.1% Cas amino acids, Trace element solution (lOOOx: 174mM EDTA, 76mM ZnSO ₄ .7H ₂ O, 178mM H3B0 ₃ , 25mM MnSO ₄ .H ₂ O, 18mM FeSO ₄ .7H ₂ O, 7.1mM CoCL ₂ .6H ₂ O, 6.4mM CuSO ₄ .5H ₂ O, 6.2mM Na ₂ Mo0 ₄ .2H ₂ O), 2.5g/l yeast extract, Penicillin/Streptomycin (lOOOx: 20g/l Pencillium, 50g/l Streptomycin), pH 6.5) was inoculated with 2 ml frozen mycelium stock in a 250 ml flat-bottom Erlenmeyer. The culture was incubated at 35°C and 200 RPM for 18 hours. Mycelium was harvested by centrifuging in 50 ml tubes at 4000 rpm at room temperature for 10 min after which the mycelium was washed with 0.6M NaC1 - 0.27M CaC1 ₂ solution and re-centrifuged. The washed mycelium was weighted and 2-3 grams of mycelium used per each transformation. Lysing mix was prepared by dissolving 20mg lysing enzymes from Trichoderma harzianum and 15mg Driselase of Basidiomycetes sp. per gram wet mycelium in final suspension. Lysing enzymes were dissolved in 5 ml of 0.6M NaC1 - 0.27M CaC1 ₂ solution per gram mycelium and mixed using magnetic stirrer at +30°C for 20-30 minutes. The solution was sterile filtrated (0.45 pm filter) before adding it on the washed mycelium. Mycelium was carefully suspended in the lysing enzyme solution and incubated at +30°C and 70 RPM, for around two hours until enough protoplasts had formed. Protoplasting was followed under microscope 1-3 times during the incubation period. The protoplasts were then filtered by using Schott glass sinter covered with sterile Miracloth, to remove mycelial fragments. Sinter was moisturized with 0.6 M NaC1 - 0.27 M CaC1 ₂ solution before starting the filtration of protoplasts. Protoplasts were gently stirred in the filter with a cotton stick and washed with a small volume of room temperature STC (1.2 M sorbitol - 50mM CaCI ₂ - 35mM NaC1 - 10mM Tris, pH 7.5) solution. A minimum of 25 ml of cold STC was added to the tube to reach a total volume max 50 ml. The solution was mixed gently by inverting the tube. The protoplasts were spun down at 1500g for 10 min, at +4°C, re-suspended gently with a pipette into 50 ml cold STC and centrifuged again. Possible parallel tubes were combined to one before adding STC. The pellet was re-suspended gently in ~500 μl cold STC.

Protoplasts were divided into 200 μl aliquots containing 10 ⁶ -10 ⁷ protoplasts. 5 - 10μg of transforming DNA was added in maximum 50 μl volume onto 200 μl of protoplasts. 4μl 250mM ATA (Aurintricarbonic acid, Sigma, A36883) was added onto protoplasts and mixed gently with a pipette tip/shaking. A 60 % PEG4000 solution was added in three batches: 500 μl PEG400 after which solution was gently mixed by tapping/shaking, then another 500 μl PEG4000 followed by gently mixing and finally 1000 μl PEG4000 followed by gently mixing. The protoplasts were incubated at room temperature for 20 minutes. After the incubation, STC (at room temperature) was added on top of protoplasts to reach total volume of ~11 ml. Protoplasts were carefully mixed and then centrifuged at 1500 g (~2500 rpm) for 10 min at room temperature. Supernatant was removed and the pellet resuspended in ~500 μl STC and divided over 3-4 selective transformation plates. The plates were incubated at +35°C until colonies emerged (~l-2 weeks).

The transformant colonies growing on the selection medium plates were streaked out again on the selective medium. Identification of correct transformants was carried out by PCR. Mycelium from the transformant streaks was dissolved in 20 mM NaOH and incubated at 100°C to lyse the cells. 1-2μ1 of this solution was used as template for PCR with Phire Plant PCR kit ™ (Thermo Fisher). The oligonucleotide primers used in this PCR are shown in Table 1.

The integration of the deletion construct into the alg3 locus was shown by two PCR reactions. Integration at the 5’ end of the gene was shown by a reaction with the primers set forth as SEQ ID NO: 23 and SEQ ID NO: 22. A 1445 bp fragment was amplified and this showed successful integration to alg3 locus. Integration at the 3’ end of alg3 was shown with the primers set forth as SEQ ID NO: 25 and SEQ ID NO: 24. A 1134 bp fragment was amplified and this showed successful integration to alg3 locus. Table 1. Oligonucleotide primers for screening of alg3 deletion

Transformants positive for both of these PCR reactions were further analyzed by quantitative PCR with the primers set forth as SEQ ID NO: 26 and SEQ ID NO: 27 and with the primers set forth as SEQ ID NO: 28 and SEQ ID NO: 29 to demonstrate that the alg3 gene had been completely deleted from them. The transformant C1 strain, positive for integration of the construct into alg3 locus and negative in the qPCR test for presence of alg3 gene, was stored at -80°C and given the strain number M2533.

The alg3 deletion strain M2533 and its parental strain DNL115 were cultivated in 250 ml flasks in 50 ml of a medium with 35 mM (ΝΗ ₄ )2SO ₄ , 7 mM NaC1, 55 mM KH ₂ PO ₄ ,

0.1 % casamino acids, 10 mM uracil, 2 mM MgSO ₄ , 0.5 % glucose, 4 μg/l biotin, lx trace element solution, 20 g/1 penicillin, 50 g/1 streptomycin, pH 6.5 at 35°C and 200 RPM for 4 days. Mycelia were removed by centrifugation at 4000 RPM for 15 minutes. The total protein concentration in the supernatants was measured and equal amounts of total protein was used for N-glycan analysis for the different strains. Glycan analysis from total protein present in the supernatants was done with the GlycoWorks™ RapiFluor-MS™ N-Glycan Kit (Waters) according to manufacturer’s protocols.

The analysis showed that high mannose fungal type glycans and hybrid glycans were eliminated from the glycan pattern of the alg3 deletion strain as compared with the parental strain (Figure 3A+B). GlcNAc ₂ Man ₃ glycan (“M3” in the figures), a precursor for making mammalian/human glycans, became much more abundant in the alg3 deletion strain. Example 2 - Deletion of alg11 and expression of animal-derived

GlcNAc transferases 1 and 2 from the alg11 locus

The next C1 glycoengineering step was to delete the alg11 mannosyltransferase in the alg3 deletion strain and simultaneously express GNT1 and GNT2 GlcNAc transferases in the fungus. The alg11 deletion terminates the glycan precursor synthesis at the level of DolP-GlcNAc ₂ Man ₃ and thus eliminates glycan forms larger than GlcNAc ₂ Man ₃ . Expression of GNT1 and GNT2 results in the addition of GlcNAc residues on both branches of the GlcNAc ₂ Man ₃ glycan and thus in production of the G0 glycan.

The DNA constructs designed to delete alg11 and simultaneously express GNT1 and GNT2 were constructed in two parts into two separate plasmids. The first plasmid contained the alg11 5' flanking region fragment for integration, an expression cassette for GNT1 where human GNT1 gene fused to a Golgi localization signal (from C1 or from yeast) is between bgl8 promoter and bgl8 terminator, and the first 2/3 of the amdS marker gene. The second plasmid contained the last 2/3 of the amdS marker, an expression cassette for the GNT2 gene where either human or rat GNT2 is between bgl8 promoter and chil terminator, and the alg11 3' flanking region fragment for integration. The amdS marker fragments in these two plasmids overlap with each other, and this region is planned to undergo homologous recombination in C1 between the plasmids at the same time as the 5 ^* and 3’ flanking region fragments recombine with genomic DNA on both sides of the alg11 gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the alg11 5' flanking region fragment for integration and an expression cassette for GNT1 is set forth in SEQ ID NO: 16. The 5 ^* flank sequence corresponds to positions 1-988 of SEQ ID NO: 16. A nucleic acid sequence encoding human GNT1 fused to S. cerevisiae KRE2 Golgi-localization signal corresponds to positions 2,389- 3,912 of SEQ ID NO: 16, where positions 2,389-2,688 encode the KRE2 localization signal and positions 2,689- 3,912 encode the human GNT1. The nucleic acid sequence encoding human GNT1 fused to S. cerevisiae KRE2 Golgi-localization signal is also set forth as SEQ ID NO: 8. This sequence was obtained by codon-optimizing the human GNT1 gene for C1, and synthesis by Genscript. The synthetized sequence contained 40 bp flanks for bgl8 promoter and terminator. In the synthesized sequence the region of amino acids 1-100 is replaced by die S. cerevisiae KRE2 Golgi localization signal (amino acids 1-100). The first 2/3 of the amdS marker gene corresponds to positions 5,045- 6,462 of SEQ

ID NO: 16.

The fragments described above and the backbone vector pRS426 were assembled together by Gibson assembly to get the 5’ arm vector.

A construct containing the last 2/3 of the amdS marker, an expression cassette for the GNT2 gene where human GNT2 is between bgl8 promoter and chil terminator, and the alg11 3' flanking region fragment for integration is set forth in SEQ ID NO: 17. The 2/3 of the amdS marker gene corresponds to positions 1-1,722 of SEQ ID NO: 17. The sequence encoding human GNT2 corresponds to positions 2,885- 4,228 of SEQ ID NO: 17. This sequence was obtained by codon-optimizing the human GNT2 gene for C1, and synthesis by Genscript. It includes 40 bp flanks for bgl8 promoter and terminator.

The 3’ flank sequence corresponds to positions 5,629- 6,592 of SEQ ID NO: 17. The fragments described above and the backbone vector pRS426 were assembled together by Gibson assembly to get the 3 ^* arm vector.

Variations of the GNT1 and GNT2 expression cassettes were made by using a different localization signal for GNT1 and using the rat GNT2 instead of human GNT2.

The alternative localization signal for GNT1 was from a C1 homologue of the yeast KRE2 gene. The alternative localization signal was designated C1 KRE2a. The fragment harboring the alternative localization signal was amplified by PCR from C1 genomic DNA. The construct containing human GNT1 with the Golgi localization signal of C1 KRE2a is set forth in SEQ ID NO: 18.

The 5 ^* flank sequence corresponds to positions 1-988 of SEQ ID NO: 18. The nucleic acid sequence encoding human GNT1 fused to C1 KRE2a Golgi-localization signal corresponds to positions 2,389-3,877 of SEQ ID NO: 18, where positions 2,389- 2,653 encode the KRE2a localization signal and positions 2,654- 3,877 encode the human GNT1. The 2/3 of the amdS marker gene corresponds to positions 5,010- 6,427 of SEQ ID NO: 18.

To get a 3 ^* arm vector variant for expression of the rat GNT2, the rat GNT2 gene was codon-optimised for C1 and synthesized by Genscript. The synthetized fragment contained 40 bp flanks for bgl8 promoter and terminator. This fragment was assembled together with the 3’ alg11 flanking fragment, bgl8 promoter fragment, chil terminator fragment and amdS marker gene fragment into the backbone vector pRS426, to get the 3’ arm vector. The sequence encoding rat GNT2 is set forth in SEQ ID NO: 19. To obtain a G0 glycoforai-producing strain, the constructed plasmids were transformed into the alg3 deletion strain M2533. In each transformation a pair of one 5’ arm vector and one 3’ arm vector was used. The pairs were:

- human GNT1 with yeast kre2 localization signal + human GNT2;

-human GNT1 with yeast kre2 localization signal + rat GNT2;

- human GNT1 with C1kre2a localization signal + human GNT2; and

- human GNT1 with C1 kre2a localization signal + rat GNT2.

The C1 transformation and transformant selection were done as described in

Example 1. The transformants were screened by PCR to find clones where alg11 gene had been replaced by the construct. The primers used for the screening are shown in Table 2. The primers set forth as SEQ ID NO: 34 and as SEQ ID NO: 35, each in combination with the primer set forth as SEQ ID NO: 38, were used to show that correct integration had taken place at the 5’ end. The primers set forth as SEQ ID NO: 36 and as SEQ ID NO: 37, each in combination with the primer set forth as SEQ ID NO: 38 were used for showing correct integration at the 3’ end. Combinations of primer pairs SEQ ID NO: 30 + SEQ ID NO: 31 and SEQ ID NO: 32 + SEQ ID NO: 33 were used to show complete deletion of the alg11 gene as described in Example 1 for alg3 gene.

Table 2. Oligonucleotide primers used for showing correct integration and loss of ale 11 gene Transformants showing the correct integration of the construct and loss of the alg11 gene were stored at -80°C and were named as follows:

- M2546, human GNT1 with yeast kre2 localization signal + human GNT2;

- M2658, human GNT1 with yeast kre2 localization signal + rat GNT2;

- M2660, human GNT1 with C1 Kre2a localization signal + rat GNT2. Transformants expressing human GNT1 with C1 Kre2a localization signal + human

GNT2 were less effective than those expressing rat GNT2, therefore only the latter were kept and stored for further assay (M2660).

The constructed C1 strains were grown in 24- well plates in the liquid medium used for the alg3 deletion strain in Example 1. The cultures were made at 35°C, 800 RPM for 4 days. Mycelia were removed as described in Example 1 and supernatant samples were subjected to glycan analysis as described in Example 1. The results, summarized in Figures 4-5, show that very high G0 glycan levels were reached, above 90% and even above 93%.

Example 3 - Expression of flippase genes in a G0 glycan-producing C1 strain

The flipping of glycan precursors from outside the ER to the inside may not work properly in alg11 deletion strains. This is because the normal substrate for the flippase is DolP-GlcNAc ₂ Man ₅ structure, and in alg11 deletion strains this structure is not formed and the substrate for the flippase is DolP-GlaNAc ₂ Man ₃ . In order to improve the flipping step of glycan precursor synthesis, two flippases were over-expressed in the G0 glycan- producing C1 strain M2546 described in Example 2. The flippases were the native C1 flippase RFT1 and the yeast mutant flippase FLC2p.

The RFT1 and FLC2p flippases were expressed in C1 from an expression vector containing the phosphinitricin resistance (bar) marker, the C1 promoter-8 and yeast adhl terminator and the 5 ^* and 3’ flanking region fragments for integration into the srp7 protease locus.

The C1 flippase RFT1 including overlaps to the promoter-8 and adhl terminator was amplified from C1 genomic DNA. The protein-coding region of the 5. cerevisiae FLC2 gene flanked by overlaps to promoter-8 and adhl terminator was codon-optimized and synthesized by Genscript. The inserts were cloned into the expression vector by Gibson assembly using the NEBuilder® kit (New England Biolabs) according to manufacturer’s protocols. The resulting plasmids were analyzed by digestions and sequencing and correct clones were collected and stored. The sequence encoding yeast FLC2p is set forth in SEQ ID NO: 20. The promoter- 8 sequence corresponds to positions 1-1,020 of SEQ ID NO: 20. The sequence encoding FLC2p corresponds to positions 1,021- 2,379 of SEQ ID NO: 20 (this sequence is also set forth as SEQ ID NO: 4). The adhl terminator sequence corresponds to positions 2,380- 2,566 of SEQ ID NO: 20.

The sequence encoding C1 RFT1 flippase is set forth in SEQ ID NO: 21. The promoter-8 sequence corresponds to positions 1-1,020 of SEQ ID NO: 21. The sequence encoding RFT1 corresponds to positions 1,021- 2,937 of SEQ ID NO: 21 (this sequence is also set forth as SEQ ID NO: 6). The adhl terminator sequence corresponds to positions 2,938- 3,124 of SEQ ID NO: 21.

The flippase expression plasmids were digested to release the insert from the vector and transformed into the C1 strain M2546 described above, which is a G0-producing strain with algS and alg11 deletions, and expressing human GNT1 with yeast KRE2 localization signal and human GNT2. Transformation was done as described in Example 1 and transformants were selected on plates with 280 mg/1 glyphosate. Transformants were streaked on new selective plates and analyzed by PCR for correct integration into the srp7 locus and total loss of the srp7 gene. 5 ^* integration was verified with the primers set forth as SEQ ID NO: 39 and SEQ ID NO: 40, 3 ^* integration with the primers set forth as SEQ ID NO: 41 and SEQ ID NO: 42 and the loss of srp7 with the primers set forth as SEQ ID NO: 43 and SEQ ID NO: 44 as described in Example 1.

Table 3. Oligonucleotide primers used for showing correct integration

The C1 transformants showing the correct results in the PCR screenings were stored at -80°C and named M2670 (yeast FLC2p flippase) and M2671 (C1 RFT1 flippase). They were grown in 24-well plate cultures for 4 days as described in Example 2. Mycelia were removed as described in Example 1 and supernatant samples were subjected to N-glycan analysis from the total secreted protein as described in Example 1. The results, summarized in Figure 6, show that the flippase-expressing strains produce a very high percentage of G0 glycans, above 94% (almost 95% in the strain expressing the yeast Flc2p). Furthermore, the total glycosylation level (sum of area of all the glycan peaks) of the proteins has increased from the parental strain M2546 (Figure 2). It is about 4-fold higher in the strain expressing yeast Flc2p and about 9-fold higher in the strain expressing C1 RFT1 as compared with the parental strain M2546.

Example 4 - Construction of a G1/G2 clycan-producing C1 strain

This example used G0-producing Nivolumab expressing strains described in Example 6-7 for generating a strain that produces G1 and G2 glycans (G0 + one or two terminal galactoses). The strains were genetically modified to express a heterologous galactosyllransferase, either XtGalT 1 or huGalT 1.

The part of the gene encoding amino acids 55-362 of the galactosyltransferase of Xenopus tropicalis (XtGalT 1) (SEQ ID NO: 45) was codon-optimised and synthesized by Genscript. The generated sequence was cloned with a Golgi localization signal of the 5. cerevisiae protein KRE2 into an appropriate plasmid for transformation into C1. The amino acid sequence of KRE2- XtGalT 1 is set forth in SEQ ID NO: 13. The localization signal corresponds to positions 1-100 of SEQ ID NO: 13.

The part of the gene encoding amino acids 78-398 of the human galactosyltransferase (huGalTl) was codon-optimised and synthesized by Genscript. The generated sequence was cloned with a Golgi localization signal of the S. cerevisiae protein KRE2 into an appropriate plasmid for transformation into C1.

A construct containing the srp75' flanking region fragment for integration and an expression cassette for KRE2-XtGalTl is set forth in SEQ ID NO: 49 (pMYT0557). The 5’ flanking sequence corresponds to positions 9-962 of SEQ ID: 49. A nucleic acid sequence encoding Xenopus tropicalis GalTl fused to S. cerevisiae KRE2 Golgi-localization signal corresponds to positions 1,990- 3,216 of SEQ II) NO: 49, where positions 1,990-2,289 encode the KRE2 localization signal and positions 2,290- 3,216 encode the XtGalTl. The sequence encoding XtGalTl fused to KRE2 is also set forth as SEQ ID NO: 14. This sequence was obtained by codon optimizing the Xenopus tropicalis GalTl gene for C1, and synthesis by Genscript. The synthetized sequence contained 40 bp flanks for S. cerevisiae KRE2 and terminator 8.

The bar marker cassette corresponds to positions 3,721- 5,919 of SEQ ID NO: 49. Finally, the srp73' flanking region fragment for integration corresponds to positions 5,928- 6,903 of SEQ ID NO: 49.

A construct containing the srp75' flanking region fragment for integration and an expression cassette for KRE2-huGalTl is set forth in SEQ ID NO: 50 (pMYT0559). The 5 ^* flank sequence corresponds to positions 9-962 of SEQ ID50. A nucleic acid sequence encoding human GalTl fused to S. cerevisiae KRE2 Golgi-localization signal corresponds to positions 1,990- 3,255 of SEQ ID NO: 50, where positions 1,990-2,289 encode the KRE2 localization signal and positions 2,290- 3,255 encode the huGalTl. The sequence encoding huGalTl fused to KRE2 is also set forth as SEQ ID NO: 101. The amino acid sequence of huGalTl fused to KRE2 is set forth as SEQ ID NO: 102. This sequence was obtained by codon optimizing the human GalTl gene for C1, and synthesis by Genscript. The synthetized sequence contained 40 bp flanks for S. cerevisiae KRE2 and terminator 8.

The bar marker cassette corresponds to positions 3,760- 5,958 of SEQ ID NO: 50. Finally, the srp73' flanking region fragment for integration corresponds to positions 5,967- 6,942 of SEQ ID NO: 50.

The fragments described above and a backbone vector (pRS426) were assembled together by Gibson assembly to get the XtGalTl or huGalTl expression vectors. The expression constructs were used to transform the G0 strains described above and transformants were plated on Acetamid transformation plates (volume 25ml) supplemented with 70μ1 of [lOOmg/ml] Glyphosate. The C1 transformants were tested for integration of the transformed DNA into srp7 locus by PCR using oligos described in Table 4. Primers SEQ ID NO: 43 (oMYT0697) and SEQ ID NO: 51 (oMYT0698) were used to show the loss of the srp7 PRF. Primers SEQ ID NO: 39 (oMYT1264) and SEQ ID NO: 52 (oMYT1231) were used to show correct integration at the 5 ^* end. Primers SEQ ID NO: 42 (oMYT0696) and SEQ ID NO: 53 (T798) were used to show correct integration at the 3’ end. The C1 transformants showing deletion of srp7 and correct integration of the GalT expression construct were stored at -80 °C and named M3697 and M3699 (X. tropicalis GalT), and M3692 and M3693 (human GalT). The strains were cultured in 24-well plates for four days as described in Example 1. Mycelia were removed and supernatant samples were subjected to N-glycan analysis from the total secreted protein, as described in Example 1. Table 4. Oligonucleotide primers used for showing correct integration and loss of srv7 gene

The results of the glycan analysis are shown in Figures 7A-D. High levels of G2 glycans were observed in the glycan patterns of all the strains. In addition, 12-16% of G1 glycans were detected.

Example 5 - Generating an alg3 deletion strain expressing Nivolumab antibody

The first step of glycoengineering in C1 was the deletion of the gene encoding the mannosyltransferase ALG3 in order to terminate the glycan precursor synthesis at the level of DolP-GlcNAc ₂ Man ₅ , and therefore to eliminate larger fungal type glycans from the C1 glycan pattern. In order to construct a strain producing G0 glycans and expressing Nivolumab, the alg3 gene was disrupted from an 8-fold protease deletion strain with the pyr4 selection marker. Pyr4 marker can be removed from C1 by counter-selection, and therefore it allows multiple successive transformations with the same marker.

The DNA constructs designed to delete alg3 were constructed in two parts into two separate plasmids. The first plasmid contained the alg3 5' flanking region fragment for integration and the first 2/3 of the pyr4 marker cassette. The second plasmid contained the last 2/3 of the pyr4 marker cassette, a direct repeal sequence targeted to the end of alg3 5’ flanking region and the alg3 3' flanking region fragment for integration. The pyr4 marker fragments in these two plasmids overlap with each other, and this region is planned to undergo homologous recombination in C1 between the plasmids at the same time as the 5’ and 3’ flanking region fragments recombine with genomic DNA on both sides of the algS gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the alg35' flanking region fragment for integration and the first 2/3 of the pyr4 marker gene is set forth in SEQ ID NO: 54 (Al/pMYT0649). The 5’ flanking sequence corresponds to positions 9 - 1010 of SEQ ID NO: 54. The first 2/3 of the pyr4 marker gene corresponds to positions 1019 - 2798 of SEQ ID NO: 54.

A construct containing the last 2/3 of the pyr4 marker, a direct repeat for pyr4 marker removal and the alg33' flanking region fragment for integration is set forth in SEQ ID NO: 55 (A2/pMYT0648). The 2/3 of the pyr4 marker gene corresponds to positions 9 - 1265 of SEQ ID NO: A2/55. The alg3 direct repeat sequence corresponds to positions 1274 - 1773 of SEQ ID NO: 55. The alg33 ^* flank sequence corresponds to positions 1782 - 2781 of SEQ ID NO: 55.

The different elements of alg3 were amplified from C1 genomic DNA and cloned with the marker gene fragments into a backbone vector (pRS426) by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 5’ and 3’ deletion vectors.

The deletion constructs were excised from the plasmids and transformed together into a C1 strain having 8 deletions of protease genes and a deletion of the pyr4 gene, according to the protocol described in the Example 1.

The transformant colonies growing on the selection medium plates were streaked out again on the selective medium. Identification of correct transformants was carried out by PCR. Mycelium from the transformant streaks was dissolved in 20 mM NaOH and incubated at 100°C to lyse the cells. 1-2μ1 of this solution was used as template for PCR with Phire Plant PCR kit ™ (Thermo Fisher). The oligonucleotide primers used in the PCR and qPCR reactions are shown in Table 5.

The integration of the deletion construct into the alg3 locus was shown by two PCR reactions. Integration at the 5’ end of the gene was shown by a reaction with the primers set forth as SEQ ID NO: 23 (A-pri-1) and SEQ ID NO: 56 (A-pri-2). A 1159 bp fragment was amplified and this showed successful integration to alg3 locus. Integration at the 3’ end of alg3 was shown with the primers set forth as SEQ ID NO: 57 (A-pri-3) and SEQ ID NO: 24 (A-pri-4). A 1786 bp fragment was amplified and this showed successful integration to alg3 locus. Purity of the strain was confirmed by standard PCR reaction with the primers set forth as SEQ ID NO: 26 (A-pri-5) and SEQ ID NO: 29 (A-pri-6). A 756 bp fragment was not amplified from the deletion strain whereas the product was seen from the parental strain indicating the deletion of alg3 gene. The transformant C1 strain, positive for integration of the construct into alg3 locus and negative in the qPCR test for presence of alg3 gene, was stored at -80°C and given the strain number M2897.

Table 5. Oligonucleotide primers for screening of als3 deletion

The removal of pyr4 selection marker using the deletion cassettes described above is based on two features: a) a functional pyr4 gene converts 5-Fluoroorotic acid into 5- Fluorouracil, a toxic metabolite, thus clones which have lost a functional pyr4 gene are able to grow in the presence of 5-FOA; and b) under 5-FOA selection pressure the direct repeat sequence in the deletion construct enables the clones to remove the pyr4 selection marker based on homologous recombination event between the alg3 5 ^* flanking region and the direct repeat. Successful recombination loops out the complete pyr4 marker enabling the correct clones to grow in the presence of 5-FOA.

The pyr4 marker removal from M2897 was carried out according to the following protocol: a small portion of fresh mycelium from a plate was suspended into 0.9% NaC1, 0.025% Tween20 solution. Dilutions of the suspension were prepared. Varying amounts of mycelial suspension were spread onto 5-Fluoroorotic acid (5-FOA) containing plates (medium components of 5-FOA plates: 7mM KC1, llmM KH ₂ PO ₄ , 0.1% Glucose, 10 mM Uracil, 10 mM Uridine, 2mM MgSO ₄ , 10 mM Proline, Trace element solution (lOOOx: 174mM EDTA, 76mM ZnSO ₄ .7H ₂ O, 178mM H3B0 ₃ , 25mM MnSO ₄ .H ₂ O, 18mM FeSO ₄ .7H ₂ O, 7.1mM CoCL ₂ .6H ₂ O, 6.4mM CuSO ₄ .5H ₂ O, 6.2mM Na ₂ Mo0 ₄ .2H ₂ O), 4 mM 5-Fluoroorolic acid, 20g/l agar granulated, pH 6.0). Plates were incubated at +35°C until colonies emerged.

The colonies growing on the 5-FOA medium plates were streaked out again on the same selective medium. Identification of clones with pyr4 marker removed was carried out by PCR. Mycelium from the transformant streaks was dissolved in 20 mM NaOH and incubated at 100°C to lyse the cells. 1-2μ1 of this solution was used as template for PCR with Phire Plant PCR kit ™ (Thermo Fisher). The oligonucleotide primers used in the PCR reactions are shown in Table 6.

The removal of the pyr4 selection cassette from the alg3 locus was shown by two PCR reactions. Loss of pyr4 marker gene was shown by a reaction with the primers set forth as SEQ ID NO: 58 (A-pri-7) and SEQ ID NO: 59 (A-pri-8). A 816 bp fragment was not amplified from the loopout strain whereas the product was seen from the parental strain indicating the removal of pyr4 cassette. Loss of pyr4 marker gene was further shown by a reaction with the primers set forth as SEQ ID NO: 60 (A-pri-9) and SEQ ID NO: 61 (A-pri- 10). A 817 bp fragment was amplified from the loopout strain whereas a much larger product (3462 bp) was seen from the parental strain indicating the removal of pyr4 cassette.

Table 6. Oligonucleotide primers for screening of removal of pyr4 marker from alg3 deletion strain C1ones with positive pattern in PCR reactions were further analyzed by quantitative PCR with the primers set forth as SEQ ID NO: 62 (A-pri-11) and SEQ ID NO: 63 (A-pri- 12) and with the primers set forth as SEQ ID NO: 64 (A-pri-13) and SEQ ID NO: 65 (A- pri-14) (Table 6) to demonstrate that the pyr4 gene had been completely removed from them. To verify the clones had also completely lost the alg3 gene, the clones were additionally analyzed by quantitative PCR with the primers set forth as SEQ ID 26: (A-pri- 15) and SEQ ID NO: 27 (A-pri-16). The loopout C1 strain, positive for removal of the pyr4 marker cassette from alg3 locus and negative in the qPCR test for presence of alg3 or pyr4 gene, was stored at -80°C and given the strain number M2972.

The strain expressing human monoclonal antibody Nivolumab was created by transforming two expression cassettes to the C1 alg3 deletion stain M2972.

The DNA constructs designed to express Nivolumab were constructed in two parts into two separate plasmids. The first plasmid contained the cbhl 5' flanking region fragment for integration, bgl8 promoter, Nivolumab light chain LC with C1 CBH1 signal sequence, bgl8 terminator for LC and the first 85% of the nial-HygR double marker cassette. The second plasmid contained the last 25% of the nial-HygR double marker cassette, a direct repeat sequence targeted to the end of bgl8 terminator, chil terminator in reverse orientation, C1 codon optimised Nivolumab heavy chain (HC) with C1 CBH1 signal sequence in reverse orientation, bgl8 promoter for expression of Nivolumab HC in reverse orientation and the cbhl 3' flanking region fragment for integration. The HygR marker fragments in these two plasmids overlap with each other, and this region undergoes homologous recombination in C1 between the plasmids at the same time as the 5 ^* and 3 ^* flanking region fragments recombine with genomic DNA on both sides of the cbhl gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the cbhl 5 ' flanking region fragment for integration, the bgl8 promoter, Nivolumab LC, bgl8 terminator and the first 85% of the nial-HygR double marker cassette is set forth in SEQ ID NO: 66 (Bl/pMYT0546). The 5’ flanking sequence corresponds to positions 9 - 1,965 of SEQ ID NO: 66. The bgl8 promoter corresponds to positions 1,974 - 3,365 of SEQ ID NO: 66. C1 CBH1 signal sequence corresponds to positions 3,366 - 3,416 of SEQ ID NO: 66. Nivolumab LC corresponds to positions 3,417 - 4,061 of SEQ ID NO: 66. The bgl8 terminator corresponds to positions 4,062 - 4,528 of SEQ ID NO: 66. The 85% of the nial-HygR double marker gene corresponds to positions 4,545 - 10,368 of SEQ ID NO: 66. The sequence encoding Nivolumab LC fused to C1 CBH1 signal sequence is also set forth as SEQ ID NO: 103. The amino acid sequence of Nivolumab LC fused to C1 CBH1 signal sequence is set forth as SEQ ID NO: 104.

A construct containing the last 25 % of the nial -HygR double marker, a direct repeal for nial-HygR double marker removal, chil terminator, Nivolumab HC, bgl8 promoter and the cbhl 3' flanking region fragment is set forth in SEQ ID NO: 67 (B2/pMYT0547). The 25% of the nial-HygR double marker gene corresponds to positions 9 - 1,732 of SEQ ID NO: 67. The bgl8 terminator direct repeat sequence corresponds to positions 1,749 - 2,073 of SEQ ID NO: 67. The chil terminator sequence (in reverse) corresponds to positions 2,733 - 2,090 of SEQ ID NO: 67. The Nivolumab HC sequence corresponds to positions 4,066 - 2,744 of SEQ ID NO: 67. C1 CBH1 signal sequence (in reverse) corresponds to positions

4,117 - 4,067 of SEQ ID NO: 67. The bgl8 promoter sequence (in reverse) corresponds to positions 5,509 - 4,118 of SEQ ID NO: 67. The cbhl 3 ^* flank sequence corresponds to positions 5,518 - 6,274 of SEQ ID NO: 67. The sequence encoding Nivolumab HC fused to C1 CBH1 signal sequence is also set forth as SEQ ID NO: 105. The amino acid sequence of Nivolumab HC fused to C1 CBH1 signal sequence is set forth as SEQ ID NO: 106.

The Nivolumab LC and HC fragments, both containing C1 CBH1 secretion signal sequence in the N-terminus, were codon optimised for C1 and ordered from GenScript. The different elements of cbhl, bgl8 and chil were amplified from C1 genomic DNA and cloned with the marker gene fragments into a backbone vector (pRS426) by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions, to give the 5’ and 3’ expression vector backbones. The synthetic DNA fragments for Nivolumab LC and HC with C1 CBH1 signal sequence were excised from their original plasmids and cloned to the forementioned backbone vectors by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 5’ and 3’ expression constructs SEQ ID NO: 66 and SEQ ID: 67. The expression constructs were excised from the plasmids and transformed together into the C1 strain M2972 described above according to the protocol described in Example

1.

The transformant colonies growing on the selection medium plates were streaked out onto a rich selective medium. Identification of correct transformants was carried out by PCR. Mycelium from the transformant streaks was dissolved in 20 mM NaOH and incubated at 100°C to lyse the cells. 1-2μ1 of this solution was used as template for PCR with Phire Plant PCR kit ^TM (Thermo Fisher).

The integration of the expression construct into the cbhl locus was shown by PCR reactions with the primers shown in Table 7. Integration at the 5 ^* end of the gene was shown by a reaction with the primers set forth as SEQ ID NO: 68 (B-pri-1) and SEQ ID NO: 69 (B-pri-2). A 3592 bp fragment was amplified and this showed successful integration tocbhl locus. Integration at the 3 ^* end ofcbhl was shown with the primers set forth as SEQ ID NO: 70 (B-pri-3) and SEQ ID NO: 71 (B-pri-4). A 2384 bp fragment was amplified and this showed successful integration tocbhl locus. Purity of the strain was confirmed by standard PCR reaction with the primers set forth as SEQ ID NO: 72 (B-pri-5) and SEQ ID NO: 73 (B-pri-6). A 498 bp fragment was not amplified from the deletion strain whereas the product was seen from the parental strain indicating the deletion ofcbhl gene.

Table 7. Oligonucleotide primers used for showing correct integration of Nivolumab LC+HC and loss ofcbhl gene The strain M3291 was grown in 24- well plates in the liquid medium used for the alg3 deletion strain in Example 1. The culture was made at 35°C, 800 RPM for 4 days. Mycelia were removed as described in Example 1 and part of the supernatant used in Protein A affinity purification of Nivolumab using Akta Start system and Mab Select Sure 1 ml columns according to manufacturer's instructions. Peak fraction was subjected to glycan analysis as described in Example 1.

The results showed that high mannose fungal type glycans and hybrid glycans were not delected in the glycan pattern of the target Mab in the alg3 deletion strain (Figure 8), similar to the N-glycans of total secreted proteins in Example 1. GlcNAc ₂ Man ₃ glycan (“M3” in the figures), a precursor for making mammalian/human glycans such as G0 glycans, is also abundant on the Mab from alg3 deletion strain.

Example 6 - Generation of a strain producing Nivolumab with G0 glycans bv expressing glvcan modifying enzvmes from the ale 11 locus

The next C1 gly coengineering step was to delete the alg11 mannosyltransferase in the alg3 deletion strain and simultaneously express GNT1 and GNT2 GlcNAc transferases, RFT1 flippase and STT3 oligosaccharyltransferase from the alg11 locus. The alg11 deletion terminates the glycan precursor synthesis at the level of DolP-GlcNAc ₂ Man ₃ and thus eliminates glycan forms larger than GlcNAc ₂ Man ₃ . Expression of C1 flippase RFT1 improves the transfer (flipping) of glycan precursor DolP-GlcNAc ₂ Man ₃ from the cytosolic side of the ER membrane to the luminal side. Expression of Leishmania major STT3 oligosaccharyltransferase improves the transfer of glycan precursor DolP-GlcNAc ₂ Man ₃ from the DolP-lipid to the target protein. Expression of GNT1 and GNT2 results in the addition of GlcNAc residues on both branches of the GlcNAc ₂ Man ₃ glycan and thus in production of the G0 glycan.

The DNA constructs designed to delete alg11 and simultaneously express GNT1, GNT2, RFT1 and STT3 were constructed in two parts into two separate plasmids. The first plasmid contained the alg11 5' flanking region fragment for integration, an expression cassette for GNT1 where human GNT1 gene fused to a Golgi-localization signal from C1 is between bgl8 promoter and bgl8 terminator, an expression cassette for RFT1 where C1 RFT1 flippase is between promoters and terminators, and the first 2/3 of the amdS marker gene. The second plasmid contained the last 2/3 of the amdS marker, a direct repeat to amdS 5’ part, an expression cassette for the STT3 oligosaccharyltransferase gene in reverse orientation where Leishmania major STT3 is between promoter9 and terminator 1, an expression cassette for GNT2 in reverse orientation where rat GNT2 is between promoters and bgl8 terminator, and the alg11 3' flanking region fragment for integration. The amdS marker fragments in these two plasmids overlap with each other, and this region undergoes homologous recombination in C1 between the plasmids at the same time as the 5 ^* and 3’ flanking region fragments recombine with genomic DNA on both sides of the alg11 gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the alg11 5' flanking region fragment for integration and an expression cassette for GNT1 and C1 RFT1 is set forth in SEQ ID NO: 74 (C1/pMYT0689). The 5’ flank sequence corresponds to positions 9 - 996 of SEQ ID NO: 74. The bgl8 promoter sequence corresponds to positions 1,005 - 2,396 of SEQ ID NO: 74. The sequence encoding truncated human GNT1 fused to C1 KRE2a Golgi-localization signal corresponds to positions 2,397 - 3,885 of SEQ ID NO: 74, where positions 2,397 - 2,661 encode the C1 KRE2a Golgi-localization signal and positions 2,662 - 3,885 encode the N-terminally truncated human GNT1. The bgl8 terminator sequence corresponds to positions 3,886 - 4,352 of SEQ ID NO: 74. The promoters sequence corresponds to positions 4,353 - 5,372 of SEQ ID NO: 74. The sequence encoding C1 RFT1 flippase corresponds to positions 5,373 - 7,289 of SEQ ID NO: 74. The terminators sequence corresponds to positions 7,298 - 7,793 of SEQ ID NO: 74. The 2/3 of the amdS marker gene sequence corresponds to positions 7,802 - 9,868 of SEQ ID NO: 74.

The nucleic acid sequence encoding human GNT1 fused to C1 KRE2a Golgi- localization signal used in this example is also set forth as SEQ ID NO: 75 (C3/C1 KRE2a- huGNT1 nt). This sequence was obtained by codon-optimizing the human GNT1 gene for C1, and was synthesized by Genscript. In the fusion construct the region of amino acids 1- 38 of native human GNT1 is replaced by the C1 KRE2a Golgi-localization signal (amino acids 1-70 in the construct). The full amino acid sequence of C1 KRE2a - GNT1 used in this example is set forth as SEQ ID NO: 76 (C4/C1 KRE2a-huGNT1 aa).

The different elements of the expression vector and C1 KRE2a secretion signal were amplified from C1 genomic DNA, human GNT1 with N-terminal truncation was amplified from a plasmid containing the C1 codon-optimised GNT1 gene and cloned with the marker gene fragment into a backbone vector (pRS426) by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 5’ expression vector SEQ ID NO: 74.

A construct containing the last 2/3 of the amdS marker, a direct repeat to amdS 5’ end, the expression cassettes of STT3 and GNT2 and the algl 1 3’ flanking region fragment is set forth in SEQ ID NO: 77 (C2/pMYT0825). The 2/3 of the amdS marker gene corresponds to positions 9 - 1,722 of SEQ ID NO: 77. The direct repeat to amdS 5’ end corresponds to positions 1,739 - 2,238 of SEQ ID NO: 77. The terminatorl sequence (in reverse) corresponds to positions 2,826 - 2,247 of SEQ ID NO: 77. The Leishmania major STT3 gene sequence (in reverse) corresponds to positions 5,408 - 2,835 of SEQ ID NO: 77. The promoted sequence (in reverse) corresponds to positions 6,417 - 5,409 of SEQ ID NO: 77. The bgl8 terminator sequence (in reverse) corresponds to positions 6,884 - 6,418 of SEQ ID NO: 77. The rat GNT2 gene sequence (in reverse) corresponds to positions 8,213 - 6,885 of SEQ ID NO: 77. The promoters sequence (in reverse) corresponds to positions 9,219 - 8,214 of SEQ ID NO: 77. The 3’ flanking region sequence corresponds to positions 9,228 - 10,191 of SEQ ID NO: 77.

The nucleic acid sequence encoding Leishmania major STT3 gene is also set forth as SEQ ID NO: 78 (C5/Lm STT3 nt). This sequence was obtained by codon-optimizing the L. major STT3 gene for C1, and was synthesized by Genscript. The full amino acid sequence of L major STT3 is set forth as SEQ ID NO: 79 (C6/Lm STT3 aa).

The nucleic acid sequence encoding rat GNT2 is also set forth as SEQ ID NO: 12 (C7/rat GNT2 nt). This sequence was obtained by codon-optimizing the rat GNT2 gene for C1, and was synthesized by Genscript. The full amino acid sequence of rat GNT2 is set forth as SEQ ID NO: 11 (C8/rat GNT2 aa).

The different elements of the expression vector were amplified from C1 genomic DNA, L. major STT3 was amplified from a plasmid containing the C1 codon-optimised STT3 gene, rat GNT2 was amplified from a plasmid containing the C1 codon-optimised GNT2 gene and cloned with the marker gene fragment into a backbone vector (pRS426) by Gibson cloning with NEBbuilder ^TM HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 3’ expression vector SEQ ID NO: 77.

The expression constructs were excised from the plasmids and transformed together into the Nivolumab producing alg3 deletion strain M3291 (Example 5). The C1 transformation and transformant selection were done as described in Example 1. The transformants were screened by PCR to find clones where algl 1 gene had been replaced by the construct pair. The primers used for the screening are shown in Table 8. Integration at the 5’ end of the alg11 locus was shown by a reaction with the primers set forth as SEQ ID NO: 34 (C-pri-1) and SEQ ID NO: 38 (C-pri-2). A 1176 bp fragment was amplified and this showed successful integration to alg11 locus. Integration at the 3 ^* end of alg11 was shown with the primers set forth as SEQ ID NO: 80 (C-pri-3) and SEQ ID NO: 36 (C-pri-4). A 1718 bp fragment was amplified and this showed successful integration to alg11 locus. Purity of the strain was confirmed by standard PCR reaction with the primers set forth as SEQ ID NO: 32 (C-pri-5) and SEQ ID NO: 81 (C-pri-6) or with SEQ ID NO: 30 (C-pri-7) and SEQ ID NO: 33 (C-pri-8). In correct transformants the PCR-fragment was not amplified from the deletion strain whereas the product was seen from the parental strain showed the deletion of alg11 gene. Additional purification step was required for a few clones with correct integration signals to ensure final clone purity. The clone purification was carried out according to the protocol described in Example 5. Purified clones were analyzed for the correct integration signals and for the loss of alg11 signal by the PCR reactions described above.

Purified clones with positive patter in PCR reactions were further analyzed by quantitative PCR with the primers set forth as SEQ ID NO: 82 (C-pri-9) and SEQ ID NO: 83 (C-pri-10) or with the primers set forth as SEQ ID NO: 84 (C-pri-11) and SEQ ID NO: 85 (C-pri-12) (Table 8) to demonstrate that the alg11 gene had been completely deleted from them. Two transformant C 1 strains, positive for integration of the constructs into algl 1 locus and negative in the qPCR test for presence of alg11 gene, were stored at -80°C and given the strain numbers M3544 and M3545.

Table 8. Oligonucleotide primers used for showing correct integration and loss of alg11 gene

The constructed C1 strains were grown in 24- well plates in the liquid medium used for the alg3 deletion strain in Example 1. The cultures were made at 35°C, 800 RPM for 4 days. Mycelia were removed as described in Example 1 and part of the supernatant from both cultivations were used in Protein A affinity purification of Nivolumab using Akta Start

(GE Healthcare) HPLC system and HiTrap MabSelect SuRe™ (GE Healthcare) 1 ml columns according to manufacturer’s instructions. Peak fraction from both samples was subjected to glycan analysis as described in Example 1. The results, summarized in Figure 9, show that very high G0 glycan levels on target Mab were reached by using the algll targeted expression vectors, above 87.6 % of G0.

Example 7 - Construction of Nivolumab/G0 strains bv inserting glvcoenzvme genes to alp6 locus and subsequent deletion of alg11

After the alg3 deletion, the next C1 glycoengineering step was to integrate GNT1 and GNT2, and flippase RFT1 and oligosaccharyltransferase STT3 in the alp6 locus. Then the Nivolumab antibody under control of bgl8 promoter is integrated into the cbhl locus. Finally, the alg11 gene was deleted and this terminates the glycan precursor synthesis at the level of DolP-GlcNAc ₂ Man ₃ and thus eliminates glycan precursors larger than GlcNAc ₂ Man ₃ . Expression of GNT1 and GNT2 results in the addition of GlcNAc residues on both branches of the GlcNAc ₂ Man ₃ glycan and thus in production of the G0 glycan.

The DNA constructs designed to integrate in alp6 and simultaneously express GNT1 and GNT2, as well as RFT and STT3, were constructed in two parts into two separate plasmids. The first plasmid contained the alp65' flanking region fragment for integration, an expression cassette for GNT1 where human GNT1 gene fused to C1 kre2 Golgi localization signal is between bgl8 promoter and bgl8 terminator, and the C1 RFT1 between promoter and terminator of the ubiquitin-like protein gene (JGI M. thermophila genome database ID 2315548), and the first 2/3 of the amdS marker gene. The second plasmid contained the last 2/3 of the amdS marker, reversed expression cassettes for STT3 from Leishmania major between the promoter and terminator of a C1 hypothetical protein (JGI M. thermophila genome database ID 2315630), the GNT2 gene from rat between bgl8 promoter and bgl8 terminator, and finally the alp6 3' flanking region fragment for integration. The amdS marker fragments in these two plasmids overlap with each other, and this region undergoes homologous recombination in C1 between the plasmids at the same time as the 5’ and 3’ flanking region fragments recombine with genomic DNA on both sides of the alp6 gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the alp65' flanking region fragment, the expression cassettes for GNT1 and C1 RFT1 and the first 2/3 of the amdS marker is set forth in SEQ ID NO: 86 (pMYT0879). The 5 ^* flank sequence corresponds to positions 9-1157 of SEQ ID NO: 86. A nucleic acid sequence encoding human GNT1 fused to C1- KRE2 Golgi-localization signal corresponds to positions 2,558- 4,046 of SEQ ID NO: 86, where positions 2,558-2,822 encode the KRE2 localization signal and positions 2,823- 4,046 encode the human GNT1. The nucleic acid sequence encoding human GNT1 fused to C1- KRE2 (JGI M. thermophila genome database ID 2300989) Golgi-localization signal was obtained by codon-optimizing the human GNT1 gene for C1, and synthesis by Genscript. The synthetized sequence contained 40 bp flanks for bgl8 promoter and terminator. In the synthesized sequence the region of amino acids 1-100 is replaced by the C1- KRE2 Golgi localization signal (JGI M. thermophila genome database ID 2300989) (amino acids 1-70).

The C1 -RFT gene (JGI: 2307799) corresponds to positions 5,534-7,450 of SEQ ID NO: 86. The first 2/3 of the amdS marker gene corresponds to positions 5,045- 6,462 of SEQ ID NO: 86.

The fragments described above and the backbone vector pRS426 were assembled together by Gibson assembly to get the 5’ arm vector.

A construct containing the last 2/3 of the amdS marker, a reversed expression cassette for the STT3 from Leishmania major, a reversed expression cassette for the rat GNT2 gene between bgl8 promoter and bgl8 terminator, and the alp63' flanking region fragment for integration is set forth in SEQ ID NO: 87 (pMYT0880). The 2/3 of the amdS marker gene corresponds to positions 9-1,554 of SEQ ID NO: 87. The sequence encoding STT3 from Leishmania major, obtained by codon-optimizing the STT3 gene for C1, and synthesis by Genscript, corresponds to positions 5,408- 2,835 of SEQ ID NO: 87. The sequence encoding rat GNT2 corresponds to positions 9,885- 8,213 of SEQ ID NO: 87. This sequence was obtained by codon-optimizing the human GNT2 gene for C1, and synthesis by Genscript. It includes 40 bp flanks for bgl8 promoter and terminator. The alp63 ^* flank sequence corresponds to positions 9,614- 10,669 of SEQ ID NO: 87. The fragments described above and the backbone vector pRS426 were assembled together by Gibson assembly to get the 3’ arm vector.

The second transformation step introduced Nivolumab antibody expression into the strain by constructing two plasmids containing heavy and light chain of Nivolumab under control of bgl8 promoter designed to integrate into the cbhl locus. A construct containing the cbhl 5' flanking region fragment, the expression cassettes for Nivolumab light chain (LC), the nial marker and the first 2/3 of the hygromycin resistance marker is set forth in

SEQ ID NO: 66 (pMYT0546). The 5 ^* flank sequence corresponds to positions 9-1965 of SEQ ID NO: 66. A nucleic acid sequence encoding expression construct for Nivolumab LC between bgl8 promoter and bgl8 terminator corresponds to positions 3,366- 4,062 of SEQ ID NO: 66, where positions 3,366-3,416 encode the cbhl signal sequence and positions 3,417- 4,062 encode the Nivolumab LC.

A construct containing the last 2/3 of the hygromycin marker, a reversed expression cassette for the Nivolumab HC between bgl8 promoter and cbhl terminator, and the cbhl 3' flanking region fragment for integration is set forth in SEQ ID NO: 67 (pMYT0547). The 2/3 of the hygromycin marker gene corresponds to positions 9-1,740 of SEQ ID NO: 67. The sequence encoding Nivolumab heavy chain (HC) corresponds to positions 4,117- 2,744 of SEQ ID NO: 67, where positions 4,117-4,067 encode the cbhl signal sequence and positions 4,068- 2,744 encode the Nivolumab HC.

Finally, a alg11 deletion construct was made, comprising of alg115 'flanking region together with the first 2/3 of the pyr4 marker gene and the second 2/3 of the pyr4 marker gene together with alg11 3 'flanking region. The alg11 5’ flank sequence corresponds to positions 9- 996 of SEQ ID NO: 88 (pMYT0853). The first 2/3 of the pyr4 marker corresponds to the positions 1,012- 2,791of SEQ ID NO: 88. The second 2/3 of the pyr4 marker corresponds to the positions of 17- 1,273 SEQ ID NO: 89 (pMYT0854) .The alg11 3’ flank sequence corresponds to positions 1,290- 2,760 of SEQ ID NO: 89.

The fragments described above and the backbone vector pRS426 were assembled together by Gibson assembly to get the 3’ arm vector.

To obtain a G0 glycoform-Nivolumab producing strain, the constructed plasmids were transformed consecutively into the alg3 deletion strain M2972 (Example 1). In each transformation a pair of one 5 ^* arm vector and one 3 ^* arm vector was used. The pairs were: -First round: alp6 5 ’flanking region-human GNT1 cassette-C1 RFT1 cassette-2/3 amdS + 2/3 amdS-LmSTT3 cassette-rat GNT2 cassette-alp63’ flanking fragment; -Second round: cbhl 5’ flanking region-Nivolumab light chain cassette-nia 1-2/3 HygR + 2/3 HygR-heavy chain cassette- cbh1 3’ flanking region;

-Third round: alg11 -5 'flanking region-2/3 pyr4 + 2/3 pyr4-algl 1 -3 'flank.

The C1 transformation and transformant selection were done as described in Example 1. The transformants were screened by PCR to find clones where alp6 (first round), cbhl (second round) or alg11 (third round) gene had been replaced by the construct. The primers used for the screening are shown in Table 9. For the first round: The primers SEQ ID NO: 90 (oMYT1990) and SEQ ID NO: 38 (oMYT0087) were used to show that correct integration had taken place at the 5’ end. The primers SEQ ID NO: 38 (oMYT0087) and SEQ ID NO: 91 (oMYT1994) were used for showing correct integration at the 3’ end. Primers SEQ ID NO: 92 (oMYT1992) and SEQ ID NO: 93 (1993) were used to show complete deletion of the alp6 gene.

For the second round: The primers SEQ ID NO: 68 (oMYT0127) and SEQ ID NO: 38 (oMYT0087) were used to show that correct integration had taken place at the 5’ end. The primers SEQ ID NO: 38 (oMYT0087) and SEQ ID NO: 71 (oMYT0129) were used for showing correct integration at the 3’ end. Primers SEQ ID NO: 72 (oMYT0157) and SEQ ID NO: 94 (oMYT0389) were used to show complete deletion of the cbhl gene.

For the third round: The primers SEQ ID NO: 34 (oMYT0934) and SEQ ID NO: 56 (0MYTOIO6) were used to show that correct integration had taken place at the 5’ end. The primers SEQ ID NO: 57 (oMYT0027) and SEQ ID NO: 36 (oMYT0936) were used for showing correct integration at the 3’ end. Primers SEQ ID NO: 30 (0MYTI866) and SEQ ID NO: 95 (oMYT1871) were used to show complete deletion of the alg11 gene. Table 9. Oligonucleotide primers used for showing correct integration and loss of alp6. cbhl and alg11 gene

Transformants of the first round showing the correct integration of the construct and loss of the alp6 gene were stored at -80°C and were named M3297 and M3298. M3297 was used for second round of transformation:

The transformant of the second round showing the correct integration of the construct and loss of the cbh1 gene was stored at -80°C and named M3529.

Transformants of the third round showing the correct integration of the construct and loss of the alg11 gene were stored at -80°C and were named M3562 and M3563. The constructed final C1 strains M3563 was grown in shake flasks in the liquid medium described in Example 1. The cultures were grown at 35°C, 200 RPM for 4 days. Mycelia were removed as described in Example 1 and supernatant samples were subjected to purification of Nivolumab with the HiTrap MabSelectSuRe™ protein A affinity chromatography columns (GE Healthcare) in an AKTA Start automated HPLC system (GE Healthcare) according to manufacturer’s protocols. The N-glycans on the purified Nivolumab were analysed as described in Example 1. The results, summarized in Figure 10. Very high G0 glycan levels were reached, up to 94.1% as shown for M3563.

Example 8 - Generation of Nivohunab/G0 strains modified with bovine GNT1

The DNA constructs designed to delete alg11 and simultaneously express GNT1, GNT2, RFT1 and STT3 were constructed in two parts into two separate plasmids, essentially similar to the constructs described in Example 6. The first plasmid contained the alg11 5' flanking region fragment for integration, an expression cassette for GNT1 where bovine ( Bos taurus ) GNT1 gene fused to a Golgi-localization signal from C1 is between bgl8 promoter and bgl8 terminator, an expression cassette for RFT1 where C1 RFT1 flippase is between promoters and terminators, and the first 2/3 of the amdS marker gene. The second plasmid contained the last 2/3 of the amdS marker, a direct repeat to amdS 5’ part, an expression cassette for the STT3 oligosaccharyltransferase gene in reverse orientation where Leishmania major STT3 is between promoterl and terminator 1, an expression cassette for GNT2 in reverse orientation where rat GNT2 is between TEF1A promoter and terminators, and the alg11 3' flanking region fragment for integration. The amdS marker fragments in these two plasmids overlap with each other, and this region is planned to undergo homologous recombination in C1 between the plasmids at the same time as the 5’ and 3’ flanking region fragments recombine with genomic DNA on both sides of the alg11 gene. The recombination between the selection marker fragments makes the marker gene functional and enables the transformants to grow under selection.

A construct containing the alg11 5' flanking region fragment for integration and an expression cassette for bovine GNT1 and C1 RFT1 is set forth in SEQ ID NO: 96 (pMYT1090). The 5’ flank sequence corresponds to positions 9 - 996 of SEQ ID NO: 96.

The bgl8 promoter sequence corresponds to positions 1,005 - 2,396 of SEQ ID NO: 96. The sequence encoding truncated bovine GNT1 fused to C1 KRE2a Golgi-localization signal corresponds to positions 2,397 - 3,891 of SEQ ID NO: 96, where positions 2,397 - 2,661 encode the C1 KRE2a Golgi-localization signal and positions 2,662 - 3,891 encode the N- terminally truncated bovine GNT1. The bgl8 terminator sequence corresponds to positions 3,892 - 4,358 of SEQ ID NO: 96. The promoters sequence corresponds to positions 4,364 - 5,383 of SEQ ID NO: 96. The sequence encoding C1 RFT1 flippase corresponds to positions 5,384 - 7,300 of SEQ ID NO: 96. The terminators sequence corresponds to positions 7,309

- 7,804 of SEQ ID NO: 96. The 2/3 of the amdS marker gene sequence corresponds to positions 7,813 - 9,879 of SEQ ID NO: 96.

The nucleic acid sequence encoding bovine GNT1 fused to C1 KRE2a Golgi- localization signal is also set forth as SEQ ID NO: 97 (C1 KRE2a-boGNT1 nt). This sequence was obtained by codon-optimizing the bovine GNT1 gene for C1, and synthetised by GenScript. In the fusion construct the region of amino acids 1-38 of native bovine GNT1 is replaced by the C1 KRE2a Golgi-localization signal (amino acids 1-70 in the construct). The full amino acid sequence of C1 KRE2a - GNT1 is set forth as SEQ ID NO: 98 (C1 KRE2a-boGNT1 aa).

This expression vector was modified from the SEQ ID NO: 74 (C1/pMYT0689) by first creating a plasmid without any GNT1 and thereafter by inserting the C1 KRE2a - boGNT1 into this intermediate vector. The C1 KRE2a secretion signal was amplified from a previous plasmid and bovine GNT1 with N-terminal truncation was released with a restriction enzyme from a plasmid containing the C1 codon-optimised bovine GNT1 gene. These fragments were cloned into the intermediate vector originating from SEQ ID NO: 74 by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 5’ expression vector SEQ ID NO: 96.

A construct containing the last 2/3 of the amdS marker, a direct repeat to amdS 5’ end, an expression cassette for STT3 gene in reverse orientation where Leishmania major STT3 is between promoterl and tcrminatorl, an expression cassette for the GNT2 gene in reverse orientation where rat GNT2 is between TEF1A promoter and terminators, and the alg11 3' flanking region fragment for integration is set forth in SEQ ID NO: 99 (pMYT1085). The 2/3 of the amdS marker gene corresponds to positions 17 - 1,738 of SEQ ID NO: 99. The direct repeat to amdS 5’ end corresponds to positions 1,747 - 2,246 of SEQ ID NO: 99. The terminatorl sequence (in reverse) corresponds to positions 2,834 - 2,255 of SEQ ID NO: 99. The Leishmania major STT3 gene sequence (in reverse) corresponds to positions 5,416 - 2,843 of SEQ ID NO: 99. The promoterl sequence (in reverse) corresponds to positions 6,425 - 5,417 of SEQ ID NO: 99. The terminators sequence (in reverse) corresponds to positions 6,934 - 6,431 of SEQ ID NO: 99. The rat GNT2 gene sequence (in reverse) corresponds to positions 8,263 - 6,935 of SEQ ID NO: 99. The TEF1A promoter sequence (in reverse) corresponds to positions 9,320 - 8,264 of SEQ ID NO: 99. The 3’ flank sequence corresponds to positions 9,327 - 10,290 of SEQ ID NO: 99.

The nucleic acid sequence encoding Leishmania major STT3 gene is also set forth as SEQ ID NO: 78. This sequence was obtained by codon-optimizing the L major STT3 gene for C1, and synthetised by Genscript. The full amino acid sequence of L· major STT3 is set forth as SEQ ID NO: 79.

The nucleic acid sequence encoding rat GNT2 is also set forth as SEQ ID NO: 12. This sequence was obtained by codon-optimizing the rat GNT2 gene for C 1 , and synthetised by Genscript. The full amino acid sequence of rat GNT2 is set forth as SEQ ID NO: 11.

The different elements of the modified expression vector were either amplified by PCR or released from previous plasmids with restriction enzymes and cloned together by Gibson cloning with NEBbuilder ™ HiFi DNA Assembly (New England Biolabs) according to manufacturer’s instructions to give the 3’ expression vector SEQ ID NO: 99.

The expression constructs were excised from the plasmids and transformed together into the Nivolumab producing C1 strain M3291 (Example 5) (having also 8 deletions of protease genes, deletion of the pyr4 gene and a deletion of the alg3 gene with the pyr4 selection marker removed). The C1 transformation and transformant selection were done as described in Example 1. The transformants were screened by PCR to find clones where alg11 gene had been replaced by the construct pair. The primers used for the screening are shown in Table 10. Integration at the 5’ end of the algll locus was shown by a reaction with the primers set forth as SEQ ID NO: 34 (C-pri-1) and SEQ ID NO: 38 (C-pri-2). A 1176 bp fragment was amplified and this showed successful integration to alg11 locus. Integration at the 3’ end of alg11 was shown with the primers set forth as SEQ ID NO: 36 (C-pri-4) and SEQ ID NO: 100 (C-pri-13). A 1225 bp fragment was amplified and this showed successful integration to alg11 locus. Purity of the strain was confirmed by standard PCR reaction with the primers set forth as SEQ ID NO: 32 (C-pri-5) and SEQ ID NO: 81 (C-pri-6) or with SEQ ID NO: 30 (C-pri-7) and SEQ ID NO: 33 (C-pri-8). In correct transformants the PCR-fragment was not amplified from the deletion strain whereas the product was seen from the parental strain indicating the deletion of alg11 gene. Additional purification step was required for a few clones with correct integration signals to ensure final clone purity. The clone purification was carried out according to the protocol described in Example 5. Purified clones were analyzed for the correct integration signals and for the loss of alg11 signal by PCR reactions described above.

Purified clones with positive pattern in PCR reactions were further analyzed by quantitative PCR with the primers set forth as SEQ ID NO: 82 (C-pri-9) and SEQ ID NO: 83 (C-pri-10) or with the primers set forth as SEQ ID NO: 84 (C-pri-11) and SEQ ID NO:

100 (C-pri-12) (Table 10) to demonstrate that the alg11 gene had been completely deleted from them. Two transformant C1 strains, positive for integration of the constructs into algll locus and negative in the qPCR test for presence of alg11 gene, were stored at -80°C and given the strain numbers M4112 and M4113.

Table 10. Oligonucleotide primers used for showing correct integration and loss of alg11 gene

The constructed C1 strains were grown in 250 ml shake flasks in 50 ml of a liquid medium as described in Example 1. The cultures were carried out at 35°C, -200 RPM for 4 days. Mycelia were removed by centrifuging and the supernatant from both cultivations were used in Protein A affinity purification of Nivolumab using Akta Start system and Mab Select Sure 1 ml columns according to manufacturer’s instructions. Peak fraction from both samples was subjected to glycan analysis as described in Example 1. The results, shown in Figure 11, show that very high G0 glycan levels on target Mab were reached by using the alg11 targeted expression vectors, above 98% of G0.

The foregoing description of the specific embodiments will so fully reveal the general nature of the invention that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without undue experimentation and without departing from the generic concept, and therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. The means, materials, and steps for carrying out various disclosed chemical structures and functions may take a variety of alterative forms without departing from the invention.