Generation of a Novel Saccharomyces cerevisiae Strain That Exhibits Strong Maltose Utilization and Hyperosmotic Resistance Using Nonrecombinant Techniques

Abstract

A yeast strain capable of leavening both unsugared and sweet bread dough efficiently would reduce the necessity of carrying out the expensive procedure of producing multiple baker's yeast strains. But issues involving the use of genetically modified foods have rendered the use of recombinant techniques for developing yeast strains controversial. Therefore, we used strong selection and screening systems in conjunction with traditional mass mating techniques to develop a strain of Saccharomyces cerevisiae that efficiently leavens both types of dough.

Two categories of baker's yeast (Saccharomyces cerevisiae) are used in the modern baking industry. There are yeast strains optimized for use in dough containing no added sugar (unsugared dough) and yeasts that are specialized for use in sweet dough to which sugar has been added at up to 30% (wt/wt) of flour (2, 6). A yeast strain for use in a broad range of bread doughs would need to combine efficient maltose metabolism, which is relevant to fermentative activity in unsugared dough (7, 8, 15), with strong hyperosmotic adaptation, which is relevant to sweetened dough that is capable of exerting hyperosmotic stress close to the limit of growth for S. cerevisiae (2, 14). We therefore sought to generate yeast strains with these combined characteristics.

As a preliminary to strain generation, we developed selective and high-throughput screening tests for strong maltose utilization and strong hyperosmotic adaptation traits. Previous work has shown that strains with strong maltose phenotypes and efficient unsugared dough leavening also have high maltose permease and maltase protein activities when grown in medium containing galactose (8, 9). To determine whether cycling between growth on galactose and growth on maltose would enrich for growth of efficient maltose-utilizing strains, cells of strain L38 (a slow inducer of its maltose-utilizing activities) and strain NL67 (with high noninduced levels of maltose permease and α-glucosidase) were inoculated at a ratio of approximately 5:1 into yeast extract-peptone-galactose medium (8) and incubated for 3 h at 30°C. Aliquots of this culture (10 ml) were transferred to 100 ml of yeast extract-peptone-maltose medium (8) and incubated for another 3 h. This process was repeated three times, with the last incubation taking place overnight in yeast extract-peptone-maltose medium. Cells were subjected to this cycle over the course of 5 days, with aliquots of culture plated out to single colonies onto yeast extract-peptone-dextrose agar plates (8) each day. The maltose utilization phenotype of each colony was determined using an acidification power test (10) to identify the presence of either NL67 or L38 cells. The initial mixed population consisted of <20% cells with rapid maltose utilization. After the first cycle of enrichment, more than 30% of the cells exhibited NL67-like maltose utilization characteristics. The ratio of efficient maltose-utilizing cells increased over the five cycles of enrichment, so that by the fifth cycle virtually all the cells were of the NL67 type.

Osmotolerant strains produce more gas in the osmotically challenging high-sugar synthetic dough medium (HSSD) (2). Therefore, these strains would be expected to grow better in a high-sugar environment. To determine whether this was the case, we monitored the growth rates of the osmotolerant industrial strain L38 and of the strong maltose-utilizing industrial strains NL25 and NL67 after inoculation of equal numbers of cells into HSSD. After 72 h of incubation in HSSD, strain L38 showed strong growth, with an optical density of >5 compared to 0.2 for NL25. NL67 showed some growth, which correlates with its known moderate gassing ability in sweet dough compared to that of NL25 (2). These data indicate that growth in HSSD is a potentially useful protocol to select for osmotolerant strains.

The strains selected for use in the mass mating program were commercial strains used for sweet-dough (L38, L39, L52, L53, L83, L92, and L96) and unsugared-dough (NL25, NL30, and NL67) leavening. All strains were shown to efficiently produce CO₂ in their respective dough types (2). Haploid yeasts derived from these strains were mass mated essentially as outlined by Lindegren and Lindegren (11). The resultant mass mating mixture using the osmotolerant strains was subjected to high osmotic growth pressure in HSSD, where fast-growing types were enriched and selected. To further increase the genetic variability and osmotic strength of the population, the enriched population was sporulated and mass mated a second time, and the resulting hybrids were treated again to the same hyperosmotic selection. A mass mating mixture using strains NL25, NL30, and NL67 was also produced and subjected to cycling between growth in galactose and growth in maltose medium. Sporulation and enrichment were repeated, with the production of a population of hybrids enriched for strong maltose-utilizing strains resulting. Finally, sporulated haploids were produced from hybrids enriched for osmotolerance or strong maltose utilization. The pools of osmotolerant and strong-maltose-utilizing haploids were mixed and mass mated. Resultant hybrids were then enriched for strong maltose utilization. This enriched population was then enriched for osmotolerant hybrids by growth in HSSD. This protocol was repeated twice, and hybrid strains were plated onto yeast extract-peptone-dextrose agar plates for further testing.

The enriched mass-mated population was screened, using the acidification power test (10), for isolates with strong maltose-utilizing characteristics. Hybrids that acidified medium containing maltose more quickly than strain L38 were then subjected to an osmotolerance screening test, which identified hybrids that produced more gas bubbles in HSSD during 1 h at 30°C than the NL67 control. Of 2,000 isolates randomly selected from the population, 196 were determined to be regarded close to the positive controls in both osmotolerance and strong maltose utilization.

To further investigate the phenotype of the 196 isolates, their gassing abilities in HSSD and low-sugar synthetic dough were tested using a multifermentation screening system (5). This test has been shown to correlate accurately with gas production in real bread dough (2). One isolate produced as much CO₂ in low-sugar synthetic dough and HSSD as did the control strains NL67 and L38, respectively, and was designated strain NL98. The NL98 phenotype was determined to be stable over 10 generations, and its hybrid status was confirmed by PCR fingerprinting using yeast multiplex PCR primers (Bresatech, Sydney, Australia) (3).

The ability of NL98 to leaven unsugared bread dough and 18%-sugar bread dough (12, 13) was compared to that of strains used in industrial practice. This comparison showed that industrial strains could be divided into two major groups, one that had high unsugared-dough activity but low 18%-sugar dough activity and the other with the opposite activities. Strain NL98 stood out from these groups, having the highest unsugared-dough activity as well as good sweet dough activity (Fig. 1).

FIG. 1.

Comparison of NL98 gassing activities in unsugared and sweet (18%-sucrose-added) bread doughs with those of commercially used baker's yeast strains.

The nonlagging phenotype in unsugared dough has been shown to be a result of noninduced maltose permease and maltase activities (8, 15). Like the other nonlagging strains, NL98 had high noninduced levels of maltose permease and maltase activities (Table 1). Mutations in the maltose regulatory protein which lead to high noninduced maltose permease and maltase activities are numerous and are located throughout the protein (4, 9). Hence, programs that relied on mutagenesis of osmotolerant strains to produce a broad-sugar-range yeast strain would have been limited in success, due to the difficulties in producing the number of mutations needed within a defined region to give high noninduced maltose permease and maltase activities.

TABLE 1.

Maltase and maltose permease activities of baker's yeast strains grown to mid-log phase in maltose, galactose, ethanol, and glucose

Straina	Activityb
	Maltase				Maltose permease
	Mal	Gal	Eth	Glu	Mal	Gal	Eth	Glu
NL98	2,600	1,000	1,600	10	5,000	1,100	2,600	2
NL67	2,900	1,100	1,700	14	5,000	1,300	2,700	2.5
NL25	3,000	950	1,500	18	5,200	1,000	2,500	2
NL30	2,500	830	1,400	9	4,900	1,100	2,500	1.5
L38	1,900	65	130	2	2,200	30	20	0.3
L83	2,100	71	130	2.5	2,500	48	34	0.3
L96	1,800	59	120	2	2,100	25	19	0.2
L92	2,100	59	130	1.5	2,300	23	19	0.3
L53	2,000	54	90	1.5	2,300	39	26	0.4
L52	1,900	75	130	2	2,200	25	25	0.3
L39	1,800	56	100	1.5	2,200	25	20	0.3

^aNL, nonlagging phenotype; L, maltose lagging phenotype. 

^bActivities are given for strains grown in maltose (Mal), galactose (Gal), ethanol (Eth), and glucose (Glu). Maltase activity values are in nanomoles of p-nitrophenol per minute per milligram of protein. Maltose permease activity values are in picomoles of [¹⁴C]maltose taken up per minute per milligram (dry weight). All values shown are means of data derived from three experiments. Assays were carried out in triplicate with standard errors of less than 10% for maltase results and less than 15% for maltose permease results. 

In an attempt to understand the reasons for the good performance of NL98 in sweet dough, its invertase levels were measured and found to be about fourfold higher than those of other osmotolerant strains (Table 2). This result was surprising, since previously available evidence indicated that the ability of S. cerevisiae to ferment sweet dough is inversely related to the activity of invertase (6, 16). The amount of glycerol produced and retained intracellularly has also been reported to correlate with the abilities of a yeast strain to ferment sweet dough (1, 12, 14). Measurement of the glycerol parameters showed that strain NL98 synthesized and retained the polyol at higher levels in HSSD than either the osmosensitive strains or the osmotolerant strains (Table 2).

TABLE 2.

Relative enzyme activities, glycerol synthesis, and retention of yeast strains

Strainb	Dough type	Invertase activityac	Total glycerolad	Glycerol retainedad
NL98	Low or high sugar	1.3	1.8	1.3
NCYC 995	Low sugar	1.0	1.1	0.7
NL67	Low sugar	0.93	0.68	0.55
NCYC 996	High sugar	0.32	1.6	1.2
L83	High sugar	0.26	1.3	1.0

^aValues shown are means of data derived from three experiments. Assays were carried out in triplicate with standard errors of less than 15%. 

^bNCYC 995 and NL67 are known osmosensitive yeast strains, and NCYC 996 and L83 are known osmotolerant strains. 

^cValues are enzyme activities of yeast strains relative to that of the NCYC 995 control strain. 

^dValues are glycerol (in millimoles per gram [dry weight] of yeast) synthesized and retained in 90 min, following inoculation of yeast in high-sugar synthetic dough. 

Attempts to develop a true broad-sugar-range yeast strain using recombinant DNA technology have had limited success (8, 17), and these strains have not been used commercially due to the sensitive nature of the use of genetically modified organisms in the food industry. Bell et al. (2) suggested that an important factor in the lack of combined high-sugar- and low-sugar-dough phenotypes in a single yeast strain might be related to the duplication of the telomere-associated MAL and invertase (SUC) loci in yeast strains. The use of a mating strategy that provides for high genetic variability and strong selection systems appears to have circumvented this duplication and to have enabled isolation of a yeast strain that retains high invertase activity yet displays osmotolerance qualities by means of a very high capacity to produce and retain glycerol.