INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Fruity aromas, due to fatty acid ethyl esters such as ethyl butyrate and ethyl hexanoate, have been detected in dairy products stored at low temperatures and spoiled by Pseudomonas fragi (3, 17). Although it may be deleterious to the food industry, P. fragi represents an obvious biotechnological interest. Indeed, this psychrotrophic microorganism hydrolyzes triglycerides of milk and butter in the 1,3-position and uses the low-molecular-weight fatty acids released to produce aroma compounds (11, 13, 21). Nevertheless, the enzymatic system involved in these reactions has been only partially identified. Endo- and exolipases have been purified (15, 16, 21) and may be responsible for ethyl ester formation. Intracellular esterases from P. fragi were also found to be involved in the biogeneration of esters (8, 19).

The aim of this work was to optimize the production of the hydrolytic and esterifying activities of P. fragi during the fermentation of whey. It has been reported that the optimal temperature for enzyme production by Pseudomonas sp. was lower than the optimal growth temperature (1, 18) and that lipase production did not increase concomitantly with the biomass. For instance, several studies with Pseudomonas fluorescens MF0 have shown that the optimal production of lipase, protease, and three periplasmic phosphatases occurred at 17.5°C, whereas its optimal growth temperature was ca. 30°C (4, 12). However, the culture temperature did not control the constitutive cell-bound esterase and cytochrome oxidase activities of P. fluorescens (12). These results indicate that the enzymatic regulation by temperature is complex and may take place at various levels. This hypothesis was challenged by Burini et al. (2), who cloned the lipase and acidic phosphatase genes in order to observe the effect of temperature. These authors demonstrated that the growth temperature regulation of those activities was posttranscriptional. Guillou et al. (5) also confirmed this conclusion and suggested that P. fluorescens MF0 lipase was inductive and sensitive to catabolic repression and that growth temperature could modulate the lipase production.

The quantity of cellular proteins from P. fragi was found to increase when the fermentation temperature was lower than the optimal growth temperature (6). About 70 proteins that represent 13.5% of the total cellular proteins were screened. These proteins were classified into five classes according to the temperature of their production. According to their class, the rates of synthesis of the proteins were found to increase or decrease with temperature. The third class contained proteins with highest expression at temperatures ranging from 15 to 20°C. According to Mérieau et al. (12), Hébraud et al. (6), and Raymond et al. (18), the ester formation from Pseudomonas sp. was highest at 12°C, but its optimum growth temperature was close to 30°C. On the basis of on these findings, we investigated the production of hydrolytic and esterifying activities from P. fragi grown at between 12 and 30°C in order to demonstrate the importance of temperature on the production of these enzymes.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Microorganism. Stock cultures of P. fragi CRDA 037 were prepared as described by Morin et al. (14).

Standardization of the inocula. A subculture was first prepared on brain heart infusion (BHI) agar and incubated at 30°C. After 24 h, the cells were suspended in 5 ml of sterile NaCl solution (0.85%) with an absorbance (A₆₀₀) adjusted to 0.2 ± 0.01 (Beckman Spectrophotometer DU-7; Beckman Instruments, Inc., Irvine, Calif.). Next, a 50-ml aliquot of BHI broth in a 250-ml Erlenmeyer flask was inoculated with the standardized bacterial suspension (1% [vol/vol]) and incubated at 30°C at 150 rpm for 24 h. Finally, an appropriate dilution of the 50-ml culture was performed in freshly prepared BHI broth in order to have an A₆₀₀ of 0.2 ± 0.01; this was then used as an inoculum for the main culture.

Fermentation medium and culture conditions. The medium used for the production of P. fragi cells was composed of rehydrated whey (6% [wt/wt]) (Les Fromageries Saputo Ltée, St.-Hyacinthe, Quebec, Canada). The pH was adjusted to 8.0 with 6 N KOH prior to autoclaving at 118°C for 20 min. The culture medium was then supplemented with 0.1% butyric acid (vol/vol) and 0.2% food-grade ethanol 95% (vol/vol), and the pH was adjusted to 6.5 prior to inoculation of the culture with 1% standardized suspension (vol/vol). Cultures of 200 and 400 ml on whey in 1-liter Erlenmeyer flasks stirred at 150 rpm were incubated at each selected temperature.

Harvesting of biocatalyst. The P. fragi resting cells used for the determination of the esterifying activity were harvested by centrifugation at 4°C at 17,700 × g for 10 min (Beckman model J2-21; Beckman Instruments, Inc., Montreal, Quebec, Canada). They were then washed three times with potassium phosphate buffer (0.05 M, pH 7.5) and centrifuged as described above. Finally, 25% (wt/vol) bacterial suspensions in potassium phosphate buffer (0.05 M, pH 7.5) were prepared. For the determination of the hydrolytic activity, cells were harvested by centrifugation at 13,490 × g for 5 min (Sigma Laboratory Centrifuge, model 113) and washed three times with sodium phosphate buffer (0.1 M, pH 7.0) before a 2.5% (wt/vol) cell suspension was prepared in the same buffer.

Esterifying activity. The esterifying activity was assayed in 250-ml Dreschel gas washing bottles outfitted with a sintered-glass gas inlet (porosity no. 3) and a gas outlet at 12°C for 6 h as described by Lamer et al. (9). The reaction mixture consisted of 90 ml of Tris-HCl buffer (0.1 M, pH 9.0), 10 ml of 25% cell suspension, 0.01 M valeric acid (Laboratoire Mat, Beauport, Quebec, Canada), and 0.02 M food-grade ethanol. The aeration of this system was set to 100 liters/min. Volatiles in the gas effluent were sampled for 2 min by connecting a stainless-steel tube (0.6-cm diameter by 17.8-cm length) filled with Tenax TA (60/80 mesh) to the reactor vent. For every assay of esterifying activity, three traps were collected and analyzed. The aroma compounds were desorbed from the Tenax traps by using a thermal AERO Trap Desorber 6000 (Tekmar, Cincinnati, Ohio) and injected into a Perkin-Elmer model 8320 gas chromatograph operating with a flame ionization detector under the conditions reported by Lamer et al. (9). The gas chromatograph analysis gave the amount of ethyl valerate in micrograms produced per 2 min. Based on an ethyl valerate calibration curve on traps (area versus ester concentration in micrograms), the esterifying activity of P. fragi was converted into micromoles of ethyl valerate formed per minute (one enzymatic unit) per liter of reaction volume with 25 g of bacterial cells per liter.

Hydrolytic activity. The hydrolytic-activity test was carried out in 96-well microplates. The hydrolytic activity of resting P. fragi cells was investigated with p-nitrophenyl valerate (Sigma Chemical Co., St. Louis, Mo.). The progress of p-nitrophenol release was measured at 405 nm (Lambda Reader EL 309 PE; Perkin-Elmer, Norwalk, Conn.). Stock solutions (48 mM) of p-nitrophenyl valerate were prepared in absolute ethanol and kept at 4°C for 1 week at most. They were diluted (1:100 [vol/vol]) just before the test in sodium phosphate buffer (0.1 M, pH 7.0). The reaction mixture consisted of 180 µl of the substrate (0.48 mM) and 20 µl of a 2.5% (wt/vol) cell suspension (22). A control reaction without the enzyme was performed in parallel to compensate for the spontaneous substrate hydrolysis. The temperature was set at 37°C and the agitation at 150 rpm. After 10 and 40 min of incubation, the A₄₀₅ was measured as follows:
where s is the A₄₀₅ of the sample after 10 and 40 min, as indicated, and b is the A₄₀₅ of the control reaction after 10 and 40 min, as indicated.

Statistical analysis. The aim of the experiment was to determine which levels of the studied factors maximized the two enzymatic activities. The studied factors were the growth temperature and the fermentation time at which samples were obtained. For the hydrolytic activity, seven equally spaced levels of temperature ranging from 12 to 30°C were tested, and seven replications of the design were completed. For the esterifying activity, five equally spaced levels of temperature ranging from 12 to 24°C were tested, and three replicates, independent of the preceding seven ones, were performed.

Analysis of variance. Before the level(s) of temperature and time favoring high activities were determined, it was necessary to determine whether or not these factors and their interaction had an effect on the responses. The SAS system procedure MIXED was used to address this problem. PROC MIXED allows the analysis of repeated measurements and takes into account all of the observations even if there are missing values, whereas PROC GLM removes observations containing missing values. The temperature-sampling-time structure of the data varied with temperature, thus MIXED was the most appropriate procedure to use (10, 20).

Effects of temperature and time on the hydrolytic and esterifying activities. For each of the two activities, the second step of the statistical analysis was to build a second-order polynomial regression model describing the effect of the two independent variables on the response variable. Such a model is useful for prediction purposes or to determine the effects of the factors on the response when an analysis-of-variance model cannot be used, as for esterifying activity. The second-order polynomial regression can be written as follows:
where is the predicted activity from the model; T and t represent the temperature and the time variables, respectively; and b_i, b_ii, and b_ij are the regression coefficients estimated by the least-squares method. The assumptions underlying the analysis are that the errors are normally and independently distributed with a mean zero value and constant variance. In order to determine which model best fit the data, the PRESS (prediction sum of squares) statistic was computed. Each observation of the data set was successively removed from the model, and the model was fitted to the data. Then, a predicted value was obtained without this observation. This allowed the computation of the PRESS statistic. A PRESS statistic can be computed for each of the tested models. It permits the quantification of the difference between the observed and predicted values and therefore it permits the comparison of the fitted models. The PRESS statistic integrates the variance and the bias of the model. The retained model is the one with the smallest PRESS statistic.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Comparison of hydrolytic and esterifying activities from P. fragi cells grown at different temperatures. Several incubation temperatures ranging from 12 to 30°C were assayed to evaluate their effect on the production of the hydrolytic and esterifying activities of P. fragi. For each fermentation, 200 or 400 ml of a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol was sampled regularly in order to measure the hydrolytic and esterifying activities. The production of both activities by P. fragi cultivated at various temperatures was measured (Fig. 1). Under such conditions, the highest amount of p-nitrophenol released (i.e., hydrolytic activity) occurred when the microorganism was cultivated between 24°C (149.2 U/liter) and 27°C (133.8 U/liter). The maximal hydrolytic activities detected at the other temperatures, i.e., 12 to 21°C, ranged from 39.6 to 53.5 U/liter. The esterifying activity was produced at a temperature lower than the hydrolytic activity. The highest esterifying activity, as measured during biotransformation of valeric acid and ethanol, was obtained when the microorganism was cultivated at 15°C. After 6 h of biotransformation, P. fragi was able to produce 0.75 µmol of ethyl valerate per min per liter of reaction volume and for 25 g (wet weight) of cells per liter. When fermentations were run at 12 and 18°C, the esterifying activities produced were 0.47 and 0.37 U/liter, respectively. The production of esterifying activity appeared to be more affected by the incubation temperature than was the hydrolytic activity. The esterifying activity produced at 24°C (0.09 U/liter) was much lower than that at 15°C (0.75 U/liter). When the cultures were grown at 27 and 30°C, very low activity and no activity, respectively, were detected.

View larger version (35K): [in this window] [in a new window]   FIG. 1.   Time course of the production of hydrolytic () and esterifying () activities from P. fragi grown on a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol. Each activity value is the mean ± the standard deviation of three (esterifying activity) or seven (hydrolytic activity) independent experiments. Both activities are expressed in enzymatic units per liter of reaction mixture and for 25 g (wet weight) of cells per liter. No esterifying activity was detected at 30°C.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Comparison of the maximal productions of hydrolytic () and esterifying () activities from P. fragi cultivated on whey-based medium at different temperatures. Each activity value is the mean ± the standard deviation of three (esterifying activity) or seven (hydrolytic activity) independent experiments. Both activities are expressed in enzymatic units per liter of reaction mixture and for 25 g (wet weight) of cells per liter. No esterifying activity was detected at 30°C.

Thermostability of hydrolytic and esterifying activities of P. fragi grown on whey at 15 and 24°C. Cultures were grown at 24°C for 36 h and at 15°C for 48 h, i.e., temperatures and incubation times used to obtain the maximal production of hydrolytic and esterifying activities, respectively.

View larger version (16K): [in this window] [in a new window]   FIG. 3.   Inactivation curves of hydrolytic ( and ) and esterifying ( and ) activities from P. fragi grown on a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol at 15°C ( and ) and 24°C ( and ). Cells were produced at 15°C for 48 h. Each activity value is the mean ± the standard deviation of two independent experiments. The activity was calculated relative to the activity measured immediately after the end of the fermentation (100%).

View larger version (17K): [in this window] [in a new window]   FIG. 4.   Inactivation curves of hydrolytic ( and ) and esterifying ( and ) activities from P. fragi grown on a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol at 15°C ( and ) and 24°C ( and ). Cells were produced at 24°C for 36 h. Each activity value is the mean ± the standard deviation of three independent experiments. The relative activity was calculated from the activity measured immediately after the end of the fermentation (100%).

Statistical models for the prediction of hydrolytic and esterifying activities. The times at which the samples were taken during the fermentation varied according to the temperature because, as the latter increased, the hydrolytic and esterifying activities reached a plateau faster.

(i) Hydrolytic activity. The analysis of variance allowed assessment of the effects of the fermentation temperature, the time, and the interaction of these two factors. According to the analysis performed on the transformed hydrolytic activities, both temperature and time have an effect on the response. Furthermore, the interaction between these two factors was also found to be significant (P < 0.0001), indicating that the effect of time on the production of hydrolytic activity depends on the temperature. In other words, these results confirm that temperature and time are interdependent with respect to hydrolytic activity.

(1)

View larger version (27K): [in this window] [in a new window]   FIG. 5.   Predicted production of hydrolytic activity as a function of temperature and time when P. fragi was grown on a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol (based on equation 1). The hydrolytic activity is expressed in enzymatic units per liter of reaction mixture and for 25 g (wet weight) of cells per liter.

(ii) Esterifying activity. Because of the imbalance in the data, PROC MIXED could not be used to estimate the effects of temperature and time and their interaction. A regression analysis on the transformed esterifying activities established that time and growth temperature were interdependent.

(2)

View larger version (23K): [in this window] [in a new window]   FIG. 6.   Predicted production of esterifying activity as a function of temperature and time when P. fragi was grown on a whey-based medium supplemented with 0.1% butyric acid and 0.2% food-grade ethanol (based on equation 2). No esterifying activity was produced at 30°C. The esterifying activity is expressed in enzymatic units per liter of reaction mixture from cells grown at 15°C (54%) and for 25 g (wet weight) of cells per liter.

Effect of enzymatic assay temperature on P. fragi hydrolytic and esterifying activities. The effect of the assay temperature on hydrolytic and esterifying activities was evaluated. Fermentations were therefore performed for 48 h at the two temperatures that allowed the maximal enzymatic productions, i.e., at 24°C for hydrolytic activity and at 15°C for esterifying activity. In order to estimate exclusively the effect of temperature on hydrolytic and esterifying activities, the cells used for the two tests were washed in the same buffer. Sodium phosphate buffer (0.1 M, pH 7.0) was selected since p-nitrophenyl valerate hydrolysis is spontaneous at alkaline pH. Enzymatic tests were performed at 37 and 12°C, which are reportedly the optimal temperatures for the determination of hydrolytic and esterifying activities, respectively (9, 22). Moreover, hydrolytic and esterifying reactions were assayed at 24 and 15°C, temperatures corresponding to the temperatures producing the highest activities for both enzymes during fermentation. The highest hydrolytic activity was obtained when the assay was performed at 37°C and was dependent on the incubation temperature for the production of P. fragi cells (Table 1). When the reaction was done at 24°C (84%) instead of at 37°C (100%), the relative hydrolytic activity was slightly decreased by using P. fragi cells grown at 24°C, whereas the ester hydrolysis was twofold lower with cells grown at 15°C (54%). As for the esterifying activity, low assay temperatures favored the highest ethyl valerate synthesis (Table 2). When the reaction occurred with cells grown at 15°C, the ester concentration detected at 15°C was not significantly higher than that detected at 12°C (P > 0.05).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The objective of the present study was to determine the effect of growth temperature on the production of hydrolytic and esterifying activities of P. fragi. Gügi et al. (4) and Mérieau et al. (12) have reported that the activities of P. fluorescens MF0 extracellular proteases and lipases were expressed to a greater extent at incubation temperature (17.5°C) that was lower than the temperature used for biomass production (30°C). The protein synthesis of P. fragi was found to be strongly influenced by the fermentation temperature, and the proteins affected were classified into five groups (6).

Our results established that the incubation temperature has different effects on the production of the esterifying and hydrolytic activities of P. fragi. The hydrolytic activity was highest when P. fragi was grown at between 24 and 27°C, whereas fermentation at 15°C enhanced fatty acid ethyl ester formation. Both of these temperatures are lower than the reported optimal growth temperature of this microorganism, i.e., 30°C. Morgan (13) had already reported the existence of two biological mechanisms involved in ethyl ester synthesis by resting cells of P. fragi. First, a lipase should catalyze the hydrolysis of triglycerides of milk into free fatty acids. Then, an esterase should use the released fatty acids with ethanol as supplied substrate to synthesize ethyl esters. Nevertheless, the optimal fermentation temperature required to produce both activities had not been identified. Our results and the statistical analysis reported here show that the highest production of esterifying activity occurred at 12 to 15°C, while 24 to 27°C would be the appropriate temperatures to produce the highest level of hydrolytic activity. To measure the hydrolytic activity, it was found to be better to perform the assays at 37°C, whereas the esterifying activity was greatly increased at low temperatures (12 to 15°C).

A potential biotechnological application of these findings is to perform a first fermentation at 24°C for 36 h to grow P. fragi cells and to hydrolyze triglycerides. Then, the temperature in the tank could be lowered to 15°C to perform the ester synthesis by supplementing the medium with food-grade ethanol and P. fragi resting cells previously grown at 15°C for 48 h.

The highest standard deviations were obtained at 24 to 27°C and at 15°C for the hydrolytic and esterifying activities, respectively. In fact, Hébraud et al. (6) classified the P. fragi proteins that were affected by the incubation temperature into five classes. Class 1 contained most of the proteins whose relative synthesis was unaffected by the growth temperature. Class 2 included proteins with the highest levels of production at low temperatures (4 or 10°C). Class 3 contained peptides that were overexpressed at intermediate temperatures (15 and/or 20°C). Class 4 contained proteins synthesized at temperatures close to the optimal growth temperature, i.e., 28 to 30°C. Class 5 included proteins with the highest production levels at the supraoptimal growth temperature (34°C). The hydrolytic activity of P. fragi CRDA 037 could belong to class 4, which encompasses proteins produced mostly at a temperature close to the optimal growth temperature. Consequently, 27°C and, above all, 24°C may represent the critical temperature areas for P. fragi to produce hydrolytic activity. The esterifying activity could be included in class 3, in which protein synthesis is increased when P. fragi is cultivated at between 15 and 20°C. Nevertheless, 15°C is a "threshold" temperature and is probably more critical for enzyme production because it is the lower temperature limit of class 3.

Our results clearly show that the hydrolytic activity is maximally produced at 24°C, while the maximal production of the esterifying activity occurs at 15°C. However, these results raise the following question. Is the difference in the maximal-production temperatures of the two activities due to different thermostabilities of the two activities? In other words, does P. fragi produce less hydrolytic activity at 15°C because this activity is less stable at 15°C, and does P. fragi produce less esterifying activity at 24°C because this activity is less stable at 24°C?

This hypothesis was tested by first growing the cells of P. fragi at 15 and 24°C, temperatures corresponding to the maximal production of esterifying and hydrolytic activities, respectively. These cells were grown at 15°C (Fig. 3) and 24°C (Fig. 4); incubated for 4, 21, and 45 h at 15°C and 24°C; and finally assayed for hydrolytic and esterifying activities at 37 and 12°C. Under all of the conditions tested, the hydrolytic activity was determined to be more stable than the esterifying activity (Fig. 3 and 4).

Interestingly, although the temperature for the maximal production of hydrolytic activity was 24°C, this activity was more stable at 15°C. Thus, P. fragi does not produce less hydrolytic activity at 15°C because this activity is less stable at 15°C.

As for the esterifying activity, it is very much affected by the incubation temperature. When the cells were grown at 15°C, the activity did not drop during the first 4 h at 15°C (Fig. 3), whereas when the cultures were grown at 24°C the activity dropped quickly after 4 h at 24°C (Fig. 4). Thus, the esterifying activity from P. fragi is not produced more at 15°C than at 24°C, but the esterifying activity is quite sensitive to an incubation temperature of more than 15°C. This might explain why less activity was detected at 24°C. Furthermore, the esterifying activity was rapidly inactivated independently of the temperature of the production (15 or 24°C) and/or of the temperature used to assess the stability (15 or 24°C) after 21 h of incubation. Although 15°C corresponded to the temperature for maximal production of the esterifying activity, rapid inactivation of the activity was also observed at 15°C. In fact, the curve of inactivation of the esterifying activity of P. fragi grown at 15°C and assayed for thermostability at 15°C parallels the curve of inactivation of the esterifying activity of P. fragi grown at 15°C and assayed for thermostability at 24°C (Fig. 3). Based on these results, the assumption that P. fragi would produce less esterifying activity at 24°C because it is less stable at 24°C might be true.

Our results regarding the temperature stability of the hydrolytic and esterifying activities of P. fragi lead to additional speculation as to whether these two activities are due to one, two, or several independent proteins. Further research is needed to determine whether hydrolytic and esterifying activities that are affected differently by the fermentation temperature are due to more than one protein, as suggested by our results on thermostability and those reported by Hébraud et al. (6).

To understand how the incubation temperature affects ester production, a synthetic medium could be developed. Hellio et al. (7) demonstrated that the proteolytic activity of P. fluorescens MF0 grown on a minimal medium was improved after the addition of several inducers (amino acids and peptides). Furthermore, the proteolytic activity was controlled by the nature of the components but also, above all, by the growth temperature. Indeed, the optimal protease activity was independent of the inducer employed and was always expressed at 17.5°C.