Untargeted yeast bioprospecting uncovers isolates of potential industrial use
Tropical rainforests and other species-rich biomes can be found in Asia, Australia, Africa, Central and South America, Mexico, and many Pacific Islands. These biomes cover less than 20% of the Earth’s land area but may contain up to 50% of the planet’s biodiversity (Dinerstein et al., 2017; Rosa et al., 2023). Numerous studies have been carried out in Central America, South America, and Asia (Barros et al., 2023; Cadete et al., 2012; Guamán-Burneo et al., 2015), while Africa and Australasia remain underexplored sources of novel microorganisms, such as yeasts (Rosa et al., 2023). As these biomes are under threat due to deforestation (Hoang and Kanemoto, 2021) and environmental changes, it is important to catalog their microbial biodiversity before it is lost.
We used an untargeted bioprospecting approach in the sense that we isolated yeasts on agar plates with a rich nutrient medium and glucose as carbon source, allowing the growth of a wide range of yeast species, while deselecting for bacteria by including three antibiotics in the medium. Admittedly, using a single isolation growth medium, we likely missed many species that requires substantially different nutrients, pH, temperature, osmolarity, or other factors to grow and that could have been identified through e.g., a metagenomic study. However, the single isolation condition approach served an important purpose: all the isolated strains could subsequently be cultivated together on the same media plates and compared in an unbiased fashion. This makes it easy to extend the phenotypic characterization to encompass more environments of industrial interest. And while the cultivation of yeast as colonies on shared nutrient medium opens for strain-strain interactions which can influence growth estimates, it probably better resembles the natural growth mode of most yeasts.
Lactose growth capacity found in a diverse set of yeasts
Although we did not sample any lactose-rich environments, we found several yeast species from both the Ascomycota and Basidiomycota phyla that could grow on this carbon source. Lactases, the collective name of β-galactosidases and (less commonly) β-glucosidases active on lactose, are found in several glycoside hydrolase (GH) families that differ in structure, substrate specificity, catalytic mechanism, and optimal working conditions (such as pH and temperature), contributing to their functional diversity (Movahedpour et al., 2022). Most likely, these enzymes initially evolved to degrade ubiquitous plant biomass-derived oligo- and disaccharides rather than lactose, which is a relatively scare carbon source outside dairy farms. We hypothesize that upon extended exposure to lactose, yeasts that possess these enzymes with basic levels of activity towards lactose may evolve more efficient lactases over time. Supporting this hypothesis, we found strains exhibiting lactase activity but no lactose growth, and substantial variability in the lactose-growth capacity among different strains of the same species. However, obtaining a comprehensive understanding of lactose metabolism across various species and strains would necessitate genome sequencing and detailed molecular characterization of their lactose catabolic traits, which falls outside the scope of this study.
Lactose metabolism has been studied in detail in Kluyveromyces yeasts, where lactose is imported across the cell membrane via a LAC12-encoded lactose permease and subsequently hydrolyzed by a LAC4-encoded lactase (a β-galactosidase from GH family 2) (Schaffrath and Breunig, 2000). The LAC12 and LAC4 genes are co-regulated and sit next to each other in the genome, forming a metabolic gene cluster (the LAC cluster) (Varela et al., 2019). In contrast, little is known about how lactose is hydrolyzed and taken up by most other yeasts. For example, some of the best lactose-growing yeasts in our dataset, A. mycotoinivorans, P. laurentii and C. curvatum, have not yet been thoroughly characterized. Interestingly, these fast-growing basidiomycetous yeasts exhibited both intra- and extracellular lactase activities, and the GH families of these enzymes remain unknown. Future research includes characterizing these lactases, which may possess different properties in terms of optimal pH and temperature, substrate specificity, and end-product inhibition, compared to lactases from Kluyveromyces (GH2) and Aspergillus (GH1) currently dominating in industrial production of lactose-free dairies.
Surprisingly, we did not isolate any lactose-assimilating Kluyveromyces strains in our collection. Instead, the most frequently sampled species were C. intermedia and its sister species C. pseudointermedia, They belong to the Metschnikowia family of ascomycetous yeasts, and are, to the best of our knowledge, the only species in this family that display robust lactose growth (Peri et al., 2023). Our previous work on C. intermedia shows that this species possesses both the conserved LAC cluster and an additional gene cluster, the GALLAC cluster, which is essential for its lactose growth (Peri et al., 2023). The exact phylogenetic relationship between C. intermedia and C. pseudointermedia has not yet been determined, and whether C. pseudointermedia also possesses a GALLAC cluster remains to be elucidated. We also isolated multiple strains of M. carribica and the closely related yeast M. guilliermondii. In contrast to C. (pseudo)intermedia, the isolated Meyerozyma strains did not display significant lactose-consumption in whey nor robust growth in liquid MM lactose medium (although they grew on SC lactose agar plates). Genomic analysis of the type strains of both M. carribica and M. guilliermondii shows that they possess the conserved LAC cluster, confirming their genetic prerequisites for lactose metabolism (unpublished results). Follow up studies will focus on elucidating the relationship between the ability to grow on lactose, the genetic setup, and the composition of the growth medium, both for these Meyerozyma strains and for other strains displaying inconsistent lactose-growth within the collection.
Basidiomycetous yeasts – future cell factories for conversion of lactose into lipids?
Our study identified several oleaginous basidiomycetous strains capable of efficiently converting lactose into lipids. Currently, yeast lipids are mostly produced from single mono- and disaccharides, hydrolysates, and glycerol, accounting for around 70% of the feedstocks used. In contrast, whey accounts for less than one percent of the feedstocks used, although it is a sizable, available, and cost-competitive (Abeln and Chuck, 2021). This underutilization of whey can be attributed, in part, to the limited metabolic capacity of known oleaginous yeasts to metabolize lactose. In this way, our findings open new possibilities for the use of whey for lipid production in the future.
To date, over 160 yeast species have been described as having the ability to store more than 20% (w/w) of their dry weight as lipids, and thereby termed oleaginous. Literature is dominated by Cutaneotrichosporon oleaginous, Rhodotorula toruloides and Y. lipolytica, while many other of the oleaginous yeasts are still poorly characterized (Abeln and Chuck, 2021). Neither C. oleaginous nor R. toruloides were found in our screen, but we identified ten strains of Y. lipolytica. However, they did not readily consume lactose in whey, nor did they grew well in MM lactose, consistent with literature stating that Y. lipolytica is a poor lactose-grower (McKay, 1992). Instead, we found several basidiomycetous, oleaginous yeasts, including A. mycotoxinivorans (formerly Trichosporon mycotoxinivorans), P. laurentii, M. antarcticus and C. curvatum (formerly Apiotrichum curvatum), which all readily assimilated lactose in whey and produced high levels (25-40%) of lipids.
While we used MM lactose for lipid production, as this allowed us to set a defined, high carbon to nitrogen ratio to promote lipid biosynthesis and accumulation, we see no reason why cheese-whey could not serve as a substrate for these yeasts. In fact, optimizing the cultivation conditions in terms of temperature, feeding strategy, and medium composition, including fine-tuning the C/N to perfectly fit the needs of individual species and strains, holds promise to significantly enhance lipid titers and yields (Lei et al., 2024). Moreover, genetic engineering can enable the construction of lipid-overproducing strains and/or strains in which b-oxidation and lipid turnover are abolished, ensuring that produced lipids are accumulated rather than consumed (Szczepańska et al., 2022). We hope this study can help basidiomycetous yeasts gain the visibility they deserve for their superior lactose-to-lipid conversion capacities and cell factory potential.