Metabolic engineering and classic selection of the yeast Candida famata (Candida flareri) for construction of strains with enhanced riboflavin production

Abstract

Currently, the mutant of the flavinogenic yeast Candida famata dep8 isolated by classic mutagenesis and selection is used for industrial riboflavin production. Here we report on construction of a riboflavin overproducing strain of C. famata using a combination of random mutagenesis based on the selection of mutants resistant to different antimetabolites as well as rational approaches of metabolic engineering. The conventional mutagenesis involved consecutive selection for resistance to riboflavin structural analog 7-methyl-8-trifluoromethyl-10-(1'-d-ribityl)isoalloxazine), 8-azaguanine, 6-azauracil, 2-diazo-5-oxo-L-norleucine and guanosine as well as screening for yellow colonies at high pH. The metabolic engineering approaches involved introduction of additional copies of transcription factor SEF1 and IMH3 (coding for IMP dehydrogenase) orthologs from Debaryomyces hansenii, and the homologous genes RIB1 and RIB7, encoding GTP cyclohydrolase II and riboflavin synthetase, the first and the last enzymes of riboflavin biosynthesis pathway, respectively. Overexpression of the aforementioned genes in riboflavin overproducer AF-4 obtained by classical selection resulted in a 4.1-fold increase in riboflavin production in shake-flask experiments. D. hansenii IMH3 and modified ARO4 genes conferring resistance to mycophenolic acid and fluorophenylalanine, respectively, were successfully used as new dominant selection markers for C. famata.

Introduction

Riboflavin (vitamin B₂, RF) is a precursor of flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD), which are coenzymes of flavoproteins involved in numerous redox reactions, and more rarely in reactions with no net change in redox state (Massey, 2000, Bornemann, 2002). Most microorganisms as well as all plants and fungi synthesize RF, whereas it is not produced by animals and rare pro- and eukaryotic microorganisms. Thus, human and domestic animals must obtain RF through nutrition. Currently, vitamin B₂ is produced on a large scale by both chemical and biotechnological synthesis, with the latter gaining more significance because of the advantages of biotechnological processes such as cost effectiveness, reduction in waste and energy requirements, and the use of renewable feedstocks (Vandamme, 1992, Stahmann et al., 2000, Lim et al., 2001).

Industrial production of RF is facilitated mainly through the use of constructed strains of the bacterium Bacillus subtilis, the filamentous fungus Ashbya gossypii and the yeast Candida famata (Lim et al., 2001, Lim et al., 2003, Burgess et al., 2009). All accumulate large amounts of RF (15–20 g RF/l), however, each has its own advantages and disadvantages (Stahmann et al., 2000, Wu et al., 2007). While engineered B. subtilis is certainly the fastest-growing organism, it is subject to phagolysis and its growth-linked RF production bears the risk of selecting non-producing mutants. In A. gossypii this risk is absent because RF production occurs in stationary phase and is not growth-linked; however, growth time (the trophic phase) is lost for production (Stahmann et al., 2000). C. famata, as the unicellular yeast, is easier to distribute in a large-scale fermentor than a filamentous fungus growing in mycelial pellets. But C. famata RF overproducing industrial strain dep8 is quite unstable (Dmytruk and Sibirny, unpublished).

Biosynthesis of RF starts with guanosine triphosphate and ribulose-5-phosphate and is completed in six enzymatic steps (Fischer and Bacher, 2005). Numerous successful studies on the construction of RF overproducing strains of B. subtilis and A. gossypii using rational approaches of metabolic engineering have been described. This involved overexpression or disruption of genes encoding enzymes, which either limit important steps in the flux or catalyze undesirable reactions. For B. subtilis, these include enhancement of both gene dosages and transcription efficiency of the RF operon (Perkins et al., 1999), constitutive expression of the key gene (ribA) in RF biosynthetic pathway (Hümbelin et al., 1999), enhancing energy generation and reducing maintenance metabolism (Zamboni et al., 2003), increasing precursor supply by modulating carbon flow through the pentose phosphate pathway (Zamboni et al., 2004, Zhu et al., 2006), deregulating gapB expression by ccpN knockout based on screening B. subtilis transposon mutants (Tännler et al., 2008), and elevating the phosphoribosylpyrophosphate (PRPP) pool, thus the increasing transcript abundances of PurR-regulated genes participating in RF precursor biosynthesis (Shi et al., 2009). For A. gossypii, metabolic engineering approaches for improvement of RF production involved: increasing the flux of glycine, an early precursor of purine nucleotide synthesis, by overexpression of the glycine biosynthetic enzyme threonine aldolase (Monschau et al., 1998); reducing the flux from glycine to serine by disruption of the SHM2 gene, encoding serine hydroxymethyltransferase (Schlupen et al., 2003); conversion of glyoxylate into glycine via expression of the alanine:glyoxylate aminotransferase gene from Saccharomyces cerevisiae (Kato and Park, 2005); deregulation of genes involved in the biosynthesis of purines and glycine by C-terminal deletion in the AgBAS1 gene encoding a transcription factor (Mateos et al., 2006) and enhancing metabolic flow through the purine pathway by modification of key enzymes PRPP amidotransferase (Jiménez et al., 2005) and PRPP synthetase (Jiménez et al., 2008). Previous research on metabolic engineering of C. famata for construction of efficient yeast RF producers has not been published most likely because the molecular tools for research of C. famata have only recently been developed (Stahmann et al., 2000). Since then, methods of transformation (Voronovsky et al., 2002) and insertional mutagenesis (Dmytruk et al., 2006) have been developed and structural and regulatory genes of the RF biosynthesis pathway have been cloned from C. famata (Dmytruk et al., 2004, Dmytruk et al., 2006, Voronovsky et al., 2004). Overproducers of FMN in C. famata have been isolated earlier in our laboratory as RF overproducing strains (Sibirny et al., 2008) overexpressing the FMN1 gene coding RF kinase (Yatsyshyn et al., 2009).

In this study, we report on construction of the C. famata RF overproducing strain AF-4 using methods of classic mutagenesis and selection for resistance to several antimetabolites, guanosine and for RF overproduction at high pH (Sibirny et al., 2008). We also describe the construction of the recombinant C. famata strain overexpressing SEF1 (putative transcriptional factor) and IMH3 (IMP dehydrogenase) orthologs from D. hansenii together with native RIB1 (GTP cyclohydrolase II) and RIB7 (RF synthase) genes in RF overproducing strain isolated by conventional mutagenesis. Constructed recombinants showed significant enhancement in RF synthesis as compared to the parental RF overproducer obtained via conventional mutagenesis.