Bbt&t?d.
h g . 1994, 10, 125-128
125
Enhanced Productivity through Gratuitous Induction in Recombinant Yeast Fermentations Scott J. Napp and Nancy A. Da Silva' Department of Chemical and Biochemical Engineering, University of California at Irvine, Irvine, California 92717
The advantages of gratuitous induction for GAL-regulated cloned gene ( l a d ) product synthesis were evaluated for the yeast Saccharomyces cereuisiae. The growth, yield, and productivity of a gratuitous (gall) strain were compared with those of an otherwise isogenic, nongratuitous (GALl) strain. Batch studies clearly demonstrated the improvements possible in product synthesis when the inducer is not metabolized by the yeast cells; both 8-galactosidase specific and volumetric activities were superior for the gall strain. At equivalent metabolizable sugar concentrations, the productivity of the gratuitous strain exceeded that of the nongratuitous strain by 180%. The effects of initial inducer concentration and induction time were also examined. For the gratuitous strain, galactose:glucose ratios as low as 0.1 still gave maximum 8-galactosidase volumetric activity. A 5-fold higher ratio was necessary for full induction with the nongratuitous strain, and productivity was substantially lower relative to thegall strain. A comparison of various times for galactose addition indicated that productivity is highest when the gratuitous culture is induced for the entire batch fermentation.
Introduction In fermentations of recombinant microorganisms, inducible promoters are often employed for the regulation of cloned gene expression. Most inducible promoters rely on a natural regulatory system in the host cell and are typically activated through one of two basic techniques: a shift in culture temperature or the addition of a chemical inducer. Although both of these methods have been successful in a variety of applications, they also possess inherent drawbacks. In the case of temperature-shift induction, the microorganism is normally subjected to suboptimal conditions for growth after induction. Consequently, a reduction in growth rate and overall fermentor productivity often occurs with this technique (Caulcott et al., 1986; Da Silva and Bailey, 1989a). With chemical induction, the inducers are often natural substrates readily metabolized by the host microorganism. Therefore, high levels of cloned gene expression may be difficult to maintain as the inducer is depleted, particularly toward the final stages of a batch culture when high levels of cloned gene expression are typically desired. An alternative, general induction technique that does not suffer from these drawbacks is the use of gratuitous inducers. With this method, the cloned gene promoter is activated, but without simultaneous consumption of the inducing agent. Gratuitous induction systems can be obtained through the development of a synthetic analog of the natural inducer that cannot be metabolized or through mutation of the metabolic pathway of the inducer. With both methods, bioreactor environmental conditions can remain optimal for cell growth while the inducer concentration remains essentially constant. This combination of effects may lead to greatly improved recombinant cell productivity. In addition, gratuitous induction allows kinetic measurementsand mathematical modeling without the complications introduced by inducer depletion. Unfortunately, gratuitous induction systems have not
* Author to whom all correspondence should be addressed. Telephone: (714) 866-8288.FAX: (714)725-2541. 875&7938/94/3010-0125$04.50/0
been readily available for the yeast Saccharomyces cereuisiae. One of the most common and highly regulated promoter systems currently employed in recombinant yeast is the GAL, system based upon the galactose regulatory circuit (Johnston, 1987). When the GALl, GAL7, or GAL.10 promoter is placed upstream of the desired structural gene, the addition of galactose induces gene expression. Recent work by Hovland et al. (1989) demonstrated that gratuitous induction of the GAL. promoters can be achieved by mutation of the chromosomal GAL1 gene in S. cereuisiae. In this gall strain, galactose metabolism is deactivated early in the galactose metabolic pathway, and galactose serves as a gratuitous inducer. An additional reg1401 mutation inhibits glucose repression of the galactose regulatory circuit. The goal of this study was to evaluate the improvements in cloned gene productivity made possible through gratuitous induction in recombinant S. cereuisiae. The gall regl-501 strain developed by Hovland et al. (1989) was selected as a model system. Expression of a plasmidencoded product protein, j3-galactosidase, was under the control of the yeast GALl promoter and was induced by galactose addition. The specific objectives of this work were to compare the yields, growth rates, and j3-galactosidase specificand volumetric activities for the gratuitous strain relative to an otherwise isogenic, nongratuitous strain during batch fermentations. The effects of induction time and the galactose:glucose ratio on the final culture productivity were also studied.
Materials and Methods Yeast Strains and Plasmid. Saccharomyces cereuisiae 334 (Hovland et al., 1989), kindly provided by R. Sclafani (University of Colorado Health Sciences Center, Denver, CO), is a haploid strain with the following genotype: MATa leu2-3,112 gall ura3-52 pep4-3 prbl1122 regl-501. Important features of this strain are the reg1-501 mutation, which attenuates catabolite repression by glucose, and the gull mutation, which blocks the galactose metabolic pathway at the galactokinase reaction
0 1994 American Chemical Society and American Institute of Chemical Engineers
126
step. S. cereuisiae YM603, provided by M. Johnston (Washington University, St. Louis, MO), is a haploid yeast strain with the following genotype: MATa ade2-101 ura352 his3 lys2-801 met regl -501. Plasmid pRY121 (West et al., 1984), provided by R. Yocum (OmniGene Inc., Cambridge, MA), is a shuttle vector with a yeast 2p origin of replication and the REP3 site. The plasmid contains the Escherichia coli lac2 gene (encoding /3-galactosidase),under the control of the yeast GALl promoter, and a URA3 selection marker. Isolation of the GALl Strain. The gall mutation in S. cerevisiae 334 was isolated as a spontaneous mutation (R. Sclafani, personal communication). Therefore, a spontaneous GALl revertant of 334 was isolated to serve as the control, nongratuitous strain. Strain 334:pRY1 2 1 was spread (107-108 cells/plate) onto selective SGC medium containing galactose as the sole carbon source in order to select gal+ revertants. A colony from the SGC plates was chosen and designated R334:pRY121;this strain was employed in all subsequent experiments. The gal+ strain either reverted from gall to GALl or obtained a second site suppressor mutation allowing growth on galactose. In order to differentiate between the two possibilities, a genetic analysis was performed. R334:pRY121 was crossed with strain YM603 (GALl) and diploids selected on SD plates containing uracil. Single colonies were sporulated, and tetrad dissection onto YPD plates was performed by standard methods (Sherman and Hicks, 1991). The growth of the resulting haploid cells was then tested on SG+(galactose)plates. Strains YM603 and R334:pRY 121were used as positive controls and strain 334 as a negative control. All four haploids from each of the 22 tetrads tested grew on the galactose plates, demonstrating that strain R334 is indeed a GALl revertant. If the gal+ phenotype of R334 were due to a second site suppressor mutation, 25 5% of the haploid cells resulting from the cross should fail to grow on the galactose plates. In this genetic analysis, the only assumption made is that the possible second site suppressor mutation is not closely linked to the GALl gene. Cultivation. SD, YPD, and sporulation plates were prepared as described in Sherman et al. (1986). The SD plates were supplemented with uracil (20 mg/L). SGC plates contain galactose (20 g/L, Sigma Grade ("
lmL9*$', ,I
YJ 0 -
,
,
,
, .-.-------• I
I
I
I
0-0-0
50
/;y
0 0
10
20
30
40
50
60
70
Time (hr)
Figure 2. (a) 8-Galactosidasespecific activity and (b) 8-galactosidase volumetric activity during batch cultivation of strains 3MpRY121and R3MpRY121at an equivalent,initial total sugar concentration (-). Initial glucose and galactoseconcentrations were both 1.5g/L. The 8-galactosidasevolumetricactivity profile of gratuitoua strain 334pRY121 at an equivalent, initial metabolizable sugar concentration is shown by the dashed line (- - -). In this case, initial glucose and galactose concentrations were 3.0 and 1.5 g/L, respectively (units are AAbZolmin). studies with a similar nongratuitous reg1 -501 strain, 65 % of the galactose was consumed during continuous culture a t low dilution rate (Da Silva and Bailey, 1991). Higher inducer levels can therefore be maintained if a gratuitous strain is employed. Growth and Cell Yield. The growth profiles for the two strains were in agreement with the glucose and galactose profiles. For both strains, a short lag of 4-5 h was followed by an exponential growth phase of 6-8 h. Gratuitous strain 334:pRY121 then continued to grow slowly and reached stationary phase approximately 40 h after inoculation. This slow transition from exponential to stationary phase is commonly observed in our laboratory, particularly in semidefined and complex media. Nongratuitous strain R334pRY 121 consumed galactose and thus grew to higher cell density following glucose depletion; stationary phase was reached at 40 h for this strain as well. High yields were observed for both strains due to the semidefined nature of SDGC. Cell growth rate was not affected by the presence of the gall mutation; both strains exhibited an exponential growth rate of 0.5 h-l. The final cell yields, however, were significantly different. The gratuitous strain cannot metabolize galactose;therefore, only 50%of the total sugar present in the nutrient medium was available for growth. Consequently, the final cell yield for the gall strain was one-half that for the nongratuitous GAL1 strain (1.0 vs 2.1 g/L). Cloned Gene Product Synthesis. The performance of each yeast strain was assessed by the specific and volumetric activities of the plasmid-encoded product @-galactosidase. The &galactosidase specific activity profiles shown in Figure 2a clearly illustrate the advantages of gratuitous induction for cloned gene product synthesis. For the nongratuitous strain, two features of the 8-galactosidase specific activity profile stand out. First, the
kinetics of cloned gene expression was more rapid; the faster initial appearance of &galactosidase activity may be due to the presence of key metabolic intermediates from galactose metabolism (Johnston, 1987). Second and more important, however, is the overall nonmonotonic shape of the specific activity profile. 8-Galactosidase specific activity reached a maximum after 14 h of batch growth and then declined continuously. This decline immediately followed the onset of galactose metabolism (Figure 1)and can be attributed to inducer depletion. The @-galactosidasespecific activity profile for the gratuitous strain was substantially different. Detectable levels of 0-galactosidase activity were observed after 10 h of batch growth. @-Galactosidasespecificactivity then rose rapidly to a final value 2.7-fold higher than that for the nongratuitous strain. In this case, inducer metabolism did not occur and specific activity did not decline. The specific activity profiles for the two strains were reproducible at other glucose and galactose concentrations. Similar profiles have also been observed previously for galactose- vs temperature-induced strains (Da Silva and Bailey, 1989a). Volumetric activity is a more important parameter when comparing the overall productivities for the two strains. In both cases (Figure 2b, solid lines), activity rose monotonically from time zero and reached a final value after 2&24 h of batch cultivation. For the nongratuitous strain, this is due to the increase in biomass concentration which offsetsthe decline in @-galactosidasespecificactivity shown in Figure 2a. The gratuitous strain, however, showed significantlybetter overallproductivity. Even with a 50% deficit in relative final cell yield, the final @-galactosidasevolumetric activity was 30 % higher for the gratuitous strain. The results shown by solid lines in Figures 1and 2 were all obtained at an equivalent, total initial sugar concentration of 1.5 g/L glucose and 1.5 g/L galactose. As mentioned above, this places gratuitous strain 334:pRY 121 at a distinct growth disadvantage since only 50% of the total sugar present can be metabolized. The resulting lower cell yield unfairly biases the volumetric activity comparisons. Therefore, an experiment at equivalent, initial metabolizable sugar concentration (3.0 g/L glucose and 1.5 g/L galactose) was performed with the gratuitous strain. The initial inducer concentration was thus maintained at the same level. Under these conditions, the growth rate was still 0.5 h-l, the specific activity profile was similar to that in Figure 2a, but the cell yield increased to 1.7 g/L (more comparable to the 2.1 g/L for the nongratuitous strain). Under these conditions, the @-galactosidase volumetric activity measured for the gratuitous strain (Figure 2b, dashed line) far exceeded (by 180%) that for the nongratuitous strain. Inducer Concentration and Induction Time Effects. The effect of initial inducer concentration on the productivity of the gratuitous and nongratuitous strains was also examined. Glucose concentration was maintained at 3.0 g/L, while the galactose concentration was varied from 0.08 to 3.0 g/L. This initial galactose:glucose concentration ratio is important because the regl-501 mutation (Hovland et al., 1989) does not completely eliminate catabolite repressioneffects (Da Silvaand Bailey, 1989b). The final @-galactosidasespecific and volumetric activities are shown in Figure 3. These results confirm those in Figure 2; final specific and volumetric activities are consistently superior for the gratuitous strain. As illustrated in Figure 3, the final @-galactosidasevolumetric activities were relatively constant until critical values for the galactose:glucose ratio were reached (0.1, gratuitous; 0.5, nongratuitous). Below the critical values, significant
Biofechnoi. Prog, 1994,Vol. 10, No. 1
128 200
3s . 2
I50
.-3
;a
9
.
."
;// */yo J'
150
'I 3
100
50
0
metabolizable sugar concentration,the productivity of the gratuitous induction strain further exceeded that of the nongratuitous strain; the final ,&galactosidase volumetric activity was nearly 3-fold higher. The effectsof initial inducer concentration on the cloned gene productivities of the two strains were also examined. At any ga1actose:glucose ratio, 8-galactosidasespecificand volumetric activities were consistently superior for the gratuitous strain. Furthermore, a lower galactose concentration was needed for maximum productivity with this strain. The effect of the induction time on productivity was also investigated. Maximum @-galactosidase productivity was found when the culture was induced over the entire batch fermentation. Overall, the results from this study clearly demonstrate the improvements in cloned gene product synthesis possible through gratuitous induction during batch cultivation of recombinant S. cereuisiae. Similar improvements can be expected in continuous and fed-batch fermentations. Under these culture conditions, glucose concentrations are often extremely low, and galactose consumption can be substantial with regl strains. If this inducer metabolism is prevented, overall productivity should increase.
.-------------------------...------------.---~--------,,loo g U
-*:
-50
,
_____________o_..___-....-.-.-.-.-..~..-..-..---
IO
o&--, ~
-3
,x
21
8
2W
v
1
22
o
Gratuitous Strain Non-gratuitousStrain
o-
______--. ----
,
O
i
sb d
I
Acknowledgment The authors thank Dr. Thomas Menees (Department of Microbiology and Molecular Genetics, College of Medicine, Universityof Californiaat Irvine) for performing the tetrad dissections. Literature Cited Caulcott, C. A.; Rhodes, M. Temperature-Induced Synthesis of Recombinant Proteins. Trends Biotechnol. 1986,4,142-146. Da Silva, N.A.; Bailey, J. E. Construction and Characterization of a Temperature-Sensitive Expression System in Recombinant Yeast. Biotechnol. h o g . 1989a,5, 18-26. Da Silva, N. A.; Bailey, J. E. Effects of Inducer Concentration on GAL Regulated Cloned Gene Expression in Recombinant Saccharomyces cerevisiae. J.Biotechnol. 1989b,10,253-266. Da Silva, N. A.; Bailey, J. E. Influence of Dilution Rate and Induction of Cloned Gene Expression in Continuous Fermentations of Recombinant Yeast. Biotechnol. Bioeng. 1991,37, 309-317. Hovland, P.;Flick, J.; Johnston, M.; Sclafani, R. A. Galactose as a Gratuitous Inducer of GAL Gene Expression in Yeasts Growing on Glucose. Gene 1989,83,57-64. Johnston, M. A. Model Fungal Gene Regulatory Mechanism: the GAL Genes of Saccharomyces cerevisiae. Microbiol.Rev. 1987,51,458-476. Park, T. H.; Seo, J. H.; Lim, H. C. Optimization of Fermentation Processes Using Recombinant Escherichia coli with the Cloned Trp Operon. Biotechnol. Bioeng. 1989,34, 1167-1177. Seressiotis, A.; Bailey, J. E. Optimal Gene Expression and Amplification Strategies for Batch and Continuous Recombinant Cultures. Biotechnol. Bioeng. 1987,29,392-398. Sherman, F.; Hicks, J. Micromanipulation and Dissection of Asci. Methods Enzymol. 1991,194,21-37. Sherman, F.;Fink, G. R.; Hicks, J. Methods in Yeast Genetics; Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY, 1986. West, R. W.,Jr.; Yocum, R. R.; Ptashne, M. Saccharomyces cereuisiae G a l - G A L 1 0 Divergent Promoter Region: Location and Function of the Upstream Activating Sequence UASo. Mol. Cell. Biol. 1984,4,2467-2478. Accepted September 15, 1993." @
Abstract published in Aduance ACS Abstracts, December 15,
1993.
