Introduction

Bacterial l-asparaginases (l-asparagine amidohydrolases) are enzymes of high therapeutic value, due to their use in certain kinds of cancer therapies, mainly in acute lymphoblastic leukemia (reviewed by Ettinger et al. 1997). Screening and evaluation of the nutritional and environmental requirements of microorganisms are important steps for l-asparaginase production. Although there is extensive literature regarding the molecular structure (Aung et al. 2000; Borek and Jaskólski 2001; Kozak and Jurga 2002), catalysis (Allison et al. 1971; Roberts 1976; Kelo et al. 2002), clinical aspects (reviewed by Müller and Boos 1998) and genetic determinants involved in the regulation (Hüser et al. 1999; Ortuño-Olea and Durán-Vargas 2000; Fisher and Wray 2002) of l-asparaginase, there are few reports concerning the chemical and physiological factors required for the improvement of its production. The production of l-asparaginases has been studied in Serratia marcescens (Heinemann and Howard 1969; Khan et al. 1970), Erwinia carotovora (Maladkar et al. 1993), Escherichia coli (Barnes et al.1977; Wei and Liu 1998), Enterobacter aerogenes (Mukherjee et al. 2000), Pseudomonas aeruginosa (Abdel-Fattah and Olama 2002) and Bacillus subtilis (Fisher and Wray 2002), with various carbon and nitrogen sources, under both aerobic and fermentative conditions. The results are sometimes contradictory in terms of the effect of carbon source (mainly glucose; Khan et al. 1970; Barnes et al. 1977) and oxygen (Jennings and Beacham 1993; Mukherjee et al. 2000) on the production of this enzyme. The amount of this enzyme in E. coli is markedly increased upon a shift from aerobic to anaerobic growth, while in E. aerogenes the presence of dissolved oxygen has a significant effect in increasing l-asparaginase synthesis (Mukherjee et al. 2000). Also, glucose, a generally preferred carbon source for production of l-asparaginase, is reported to exert efficient carbon catabolic repression on the expression of this enzyme in E. aerogenes (Mukherjee et al. 2000), while in a closely related bacterium, Klebsiella aerogenes, glucose does not inhibit the production of this enzyme (Resnick and Magasanik 1976). Thus, it has been suggested that numerous mechanisms of catabolic repression and oxygen effect might exist; and two distinctly related bacteria, which exhibit comparable phenomena of carbon catabolic control, may use entirely different molecular mechanisms to achieve this control.

Vitreoscilla hemoglobin (VHb) is the first well characterized prokaryotic hemoglobin. The role of VHb in bacteria is to raise the effective dissolved oxygen tension within the cells and to scavenge and release oxygen to terminal oxidases during oxygen-limited growth conditions. The expression of the VHb gene (vgb) is regulated by oxygen in both its native host, Vitreoscilla, and in E. coli and is maximally induced under microaerophilic conditions (Dikshit and Webster 1988; Dikshit et al. 1992; Joshi and Dikshit 1994). In previous studies, we showed that bacteria engineered with vgb had 2- to 10-fold higher oxygen uptake rates than their vgb − counterparts (Geckil et al. 2001, 2003). The presence of VHb may also regulate the expression of other oxygen-regulated genes. The aim of this study was, therefore, to evaluate the effect of such a highly efficient recombinant oxygen-uptake system on l-asparaginase production in E. aerogenes under different culture conditions.

Materials and methods

Chemicals, enzymes and growth medium

Anhydrous l-asparagine, trichloroacetic acid and Nessler reagent chemicals (HgI2, KI, NaOH) were purchased from Sigma Chemicals. Luria–Bertani broth (LB; pH 7.0) medium was prepared as described by Miller (1972), except that peptone was used in place of tryptone. Glucose-supplemented LB medium was termed LBG. Where indicated, ampicillin was added to media (LBamp) at concentration of 100 µg/ml for the recombinants of E. aerogenes. Culture medium and stock solution of glucose were autoclaved separately at 120 °C for 25 min and an appropriate carbohydrate concentration was made in the final volume of cultures.

Bacterial strains

The bacterial host used throughout this study was E. aerogenes (NRRL B-427), obtained from the USDA culture collection in Peoria, Ill. Plasmid pUC8:15 (Dikshit and Webster 1988) has vgb (0.6 kb) inserted into vector pUC8 (Messing 1983) on a Vitreoscilla DNA fragment of 2.3 kb. Transformations of E. aerogenes with pUC8 or pUC8:15, respectively, were performed by the modified CaCl2 method of Cohen et al. (1972). The transformant strain bearing the plasmid pUC8:15 was designated Ea[pUC8:15], while the strain which was used as a control carried plasmid pUC8 without insert and was termed Ea[pUC8] (Geckil et al. 2003).

Cultivation in shake-flasks and anaerobic bottles

Stock cultures of E. aerogenes and recombinant strains were maintained on LB agar slants or LBamp agar slants (for recombinants). A 1/100 inoculum of overnight culture of each strain grown in LB or LBamp was transferred into varying amounts of LB (without ampicillin) in 125-ml Erlenmeyer flasks or screw-cap air-tight bottles at 37 °C. Plasmid stability was determined by transferring 30 cells from LB plates, used to titer the cell cultures at each time-point during the incubation, onto LBamp plates, followed by overnight incubation at 37 °C. High aeration/high agitation cultures were incubated in a gyratory water bath at 200 rpm in flasks containing 20 ml medium. Low aeration/low agitation cultures were grown in flasks containing 80 ml medium with 50 rpm shaking. No aeration/low agitation cultures were maintained in air-tight bottles (100 ml) containing 80 ml medium and agitated at 50 rpm. For LBG cultures, cells were grown in 20 ml LB (in 125-ml flasks) with high (1%) or low (0.1%) glucose concentrations at high agitation (200 rpm).

Samples for determining cell density (using the absorbance at 600 nm; A 600) were taken at selected intervals during incubation and were appropriately diluted (1/5 or 1/10) in the media (LB or LBG) used for growth at A 600>0.5. The pH value and ammonia concentration in the growth medium were determined in samples taken for the preparation of enzyme extracts.

Preparation of cell-free extracts for enzyme assay

Extracts for enzyme assay were prepared from cells grown to the late stationary phase. Cells harvested (10,000 rpm, 5 min) for l-asparaginase assays were washed twice with 0.05 M Tris-HCl buffer (pH 8.6); and the cell pellet was suspended in the same buffer to A 600=10 and disrupted with an ultrasonifier (Bronson sonifier 450) for 6 cycles (20 s each sonication, 40 s pause on ice). The extract was centrifuged (15,000g, 10 min, 4 °C) and the cell-free supernatant was immediately subjected to enzyme assay.

Enzyme assay

l-Asparaginase catalyzes the hydrolysis of l-asparagine to l-aspartic acid and ammonia. l-Asparaginase activity was measured by the method of Wriston (1970), utilizing Nesslerization, the most commonly used l-asparaginase assay, which is based on the determination of ammonia liberated from l-asparagine by enzyme in the Nessler reaction. The reaction mixture (2 ml) contained 1.8 ml 0.01 M l-asparagine prepared in 0.05 M Tris-HCl buffer (pH 8.6) and 0.2 ml cell-free extract. After 30 min incubation at 37 °C, the reaction was stopped by the addition of 0.1 ml 1.5 M trichloroacetic acid (TCA), the mixture was centrifuged (10,000 rpm, 5 min) and the ammonia released in the supernatant was determined colorimetrically (A 480) by adding 0.5 ml Nessler reagent to tubes containing 1 ml supernatant and 3.5 ml doubly-distilled H2O. The content in the tubes was vortexed and left to stand for 10 min at room temperature. The A 480 values were read against the blanks that received TCA before extract addition. One l-asparaginase unit of activity was defined as the amount of enzyme that liberated 1 μmol ammonia/min at 37 °C. The specific activity was expressed as units of activity per milligram of protein. The ammonia concentration produced in the reaction was determined on the basis of a standard curve previously obtained with ammonium sulfate as the standard. The limit of detection of ammonia by this method was about 10 μM. Total protein was determined colorimetrically (Lowry et al. 1951), using bovine serum albumin as the standard.

Results

Effect of VHb and aeration/agitation on l-asparaginase production

The effect of VHb on l-asparaginase production in E. aerogenes and its recombinants grown in LB under different aeration/agitation conditions is summarized in Fig. 1. The control (which contained only the vector, Ea[pUC8]) and the wild-type strain showed similar enzyme activity. In contrast, the VHb-expressing strain (Ea[pUC8:15]) had a substantially lower enzyme activity. However, with the exception of the presence of VHb in only Ea[pUC8:15], all three strains showed similar protein profiles (data not shown). In cultures grown with high aeration and agitation, the l-asparaginase activity of Ea[pUC8:15] was more than 10-fold lower than that of the non-VHb counterparts. This difference was even more pronounced (up to two orders of magnitude) in the case of low aeration/low agitation and was highest (more than two orders of magnitude) for anaerobic/low agitation cultures. The enzyme activity was down-regulated by both high aeration and anaerobic conditions. The highest enzyme activity was in cultures with low aeration and low agitation. These cultures showed up to 2-fold higher enzyme activity than high aeration/high agitation cultures and no aeration/low agitation cultures. Depending on the aeration/agitation condition, cultures also showed a distinct pattern of l-asparaginase activity during the incubation. The highest enzyme activity in high aeration/high agitation cultures was determined at 24 h of incubation and gradually decreased at further time-points. Low aeration/low agitation cultures had their highest enzyme activity at 48 h of incubation, leveling off between 72 h and 96 h of incubation. The enzyme activity in anaerobic/low agitation cultures increased up 72 h, after which it leveled off.

Fig. 1
figure 1

Effect of Vitreoscilla hemoglobin (VHb) on l-asparaginase production by Enterobacter aerogenes (triangles), Ea[pUC8] (white circles) and Ea[pUC8:15] (black circles) grown under conditions of high aeration/high agitation (a), low aeration/low agitation (b) and no aeration/low agitation (c). Each data-point is the average of three independent experiments. Error bars indicate standard deviations (σ n−1). Where not visible, they are smaller than the diameter of the data-point

Full size image

Effect of glucose on l-asparaginase production by E. aerogenes and its vgb − and vgb + recombinants

The concentration-dependent effect of glucose on l-asparaginase activity is summarized in Fig. 2. The cultures grown in LB medium supplemented with 1% glucose (high-glucose concentration) showed enzyme activity up to two orders of magnitude lower than the cultures grown with 0.1% glucose (low-glucose concentration). However, in both media and for all strains, total cell protein was not affected. In low-glucose medium, strain Ea[pUC8:15] showed up to 10-fold lower enzyme activity than the wild-type and vgb − strains, both of which showed similar enzyme activity throughout the incubation. The highest enzyme activity for cultures grown in low-glucose medium was determined to be at 48 h. Cultures grown in high glucose, however, showed a gradual decrease in enzyme activity from 24 h to the end of incubation (96 h). Furthermore, all three strains showed similar enzyme levels when grown in high-glucose medium.

Fig. 2
figure 2

l-Asparaginase production by E. aerogenes (triangles), Ea[pUC8] (white circles) and Ea[pUC8:15] (black circles) grown in LB medium supplemented with glucose at 1% (a) and 0.1% (b). Cells were cultivated in a gyratory (200 rpm) water-bath at 37 °C in 125-ml flasks containing 20 ml medium

Full size image

Recombinant strains maintained their plasmids at all time-points when grown under non-selective conditions. Both Ea[pUC8] and Ea[pUC8:15]) grown in LB or LBG with no ampicillin showed a plasmid stability of 100% at all time-points. The plasmid stability of Ea[pUC8:15] was also evident from its distinct (pink) appearance.

Cell density and ammonia release

The cell density (as A 600) and ammonia released (as x μmol ml−1 broth) in the LB culture broth are given in Fig. 3. Both parameters were determined from the same cultures used in enzyme activity studies (Fig. 1). In general, the wild-type strain had a slightly higher cell density than the recombinant strains. Cultures grown in LB under high aeration/high agitation showed similar A 600 values throughout the whole incubation period, while low aeration/low agitation cultures showed an A 600 increase up to 48 h and then leveled off. The cultures grown under anaerobic/low agitation condition had a slight increase in A 600 values throughout the incubation. The ammonia release parallels the cell density of the cultures grown at low aeration/low agitation (Fig. 3b) and at anaerobic/low agitation (Fig. 3c). Under high aeration/high agitation conditions, the wild-type strain showed lower ammonia release than the recombinant strains. The ammonia levels in culture broths from low aeration/low agitation and anaerobic/low agitation were similar for E. aerogenes and Ea[pUC8]. Ea[pUC8:15], however, showed a slightly low ammonia release under the same conditions.

Fig. 3
figure 3

Concentration of cell density (white symbols) and ammonia (black symbols) in culture broth of E. aerogenes (triangles), Ea[pUC8] (squares) and Ea[pUC8:15] (circles). Cells were grown in LB at high aeration/high agitation (a), low aeration/low agitation (b) and no aeration/low agitation (c). The data shown are from the set of experiments used in Fig. 1. Each data-point is the average of three independent experiments. A 600 Absorbance at 600 nm

Full size image

The A 600 values and ammonia levels of LBG-grown cultures (Fig. 4) were from the set of experiments used for enzyme activity studies in Fig. 2. The cultures in high-glucose (1%) medium showed a slight increase in A 600 with time, but had lower A 600 values than the cultures grown in low-glucose (0.1%) medium, which had almost steady A 600 values during the whole incubation. Similar to cultures grown in LB, the wild-type strain had a higher cell density than both recombinant strains. Glucose at high concentration promoted ammonia release. Although both low and high glucose supported a good cell density (A 600>3.0), the latter showed an ammonia release several-fold higher than the former, a result opposite to the enzyme activity determined.

Fig. 4
figure 4

Concentration of cell density (white symbols) and ammonia (black symbols) concentration in culture broth of E. aerogenes (triangles), Ea[pUC8] (squares) and Ea[pUC8:15] (circles). Cells were grown in LB medium containing glucose at 1% (a) and 0.1% (b). The data shown are from the set of experiments used in Fig. 2

Full size image

Effect of media pH on l-asparaginase activity

The production of acids was measured as the change in culture pH with time from the initial pH of 7.0. With the exception of cultures under anaerobic (no aeration)/low agitation condition, which showed an average decrease of 0.5 pH units during the first 24 h incubation and then leveled off at about pH 6.61–7.21, the pH of aerated cultures increased moderately in the first 24 h of growth and then leveled off (Table 1). The average pH of all three strains under high aeration/high agitation was pH 8.77±0.02 at 24 h incubation and then leveled off at about pH 8.81–8.98 for all three strains at all times, corresponding to an increase of 0.08–0.17 pH units. The average pH of low aeration/low agitation cultures for all three strains at 24 h was pH 7.56±0.28. The pH increase of these air- and agitation-restricted cultures during the first 24 h growth was more than 1.0 pH unit less than that of cultures grown under high aeration/high agitation for the same incubation period (24 h). At further incubation periods, however, both types of aeration/agitation condition resulted in similar pH values. In contrast to aerated cultures, the average pH of no aeration/low agitation cultures for all three strains at 24 h was pH 6.55±0.05 with an increase of 0.19–0.53 pH units at further incubation periods.

Table 1 pH values in cultures of Enterobacter aerogenes and its vgb −and vgb + recombinant strains grown in LB under different conditions of aeration and agitation. Numbers in parentheses are SDs (n=3). Values given are from shake-flask experiments shown in Figs. 1, 3. See Materials and methods for details
Full size table

In contrast to cells grown in LB or low LBG (0.1% glucose), the cultures grown in high LBG (1% glucose) showed considerable decreases in pH value, with an average pH of 4.35±0.02 from a starting culture pH of 7.0 for all three strains at all time-points (Table 2). The cultures grown in low LBG (0.1%) had an average pH value of 8.67±0.05 and, in contrast to cultures in LB, there was a slight decrease (0.19-0.55 pH units) at further incubation periods.

Table 2 pH values in cultures of E. aerogenes and its vgb −and vgb + recombinant strains grown in LB containing high (1%) and low (0.1%) levels of glucose. Numbers in parentheses are SDs (n=3). Values given are from shake-flask experiments shown in Figs. 2, 4
Full size table

Discussion

The results reported in this study establish for the first time that the expression of a highly efficient recombinant oxygen-uptake system, VHb, in E. aerogenes causes an inhibition of l-asparaginase activity by up to two orders of magnitude, compared with the wild-type strain and a control strain which carried the vector with no vgb insert. The total cell protein of all three strains, however, showed similar profiles. Of the three strains grown in LB under different aeration and agitation conditions, the VHb-expressing strain (Ea[pUC8:15]) showed substantially reduced levels (from several-fold to more than two orders of magnitude) of l-asparaginase. In general, E. aerogenes and the vgb − strain (Ea[pUC8]) showed similar enzyme activities which were reduced considerably under both full aerobic and anaerobic conditions. The highest enzyme activity was determined in cultures under low aeration and low agitation. These cultures had an about 2-fold higher enzyme activity than cultures grown under high aeration/high agitation and no aeration/low agitation conditions. In contrast to these findings (especially the substantial reduction in l-asparaginase for E. aerogenes harboring the vgb/VHb system), a similar study by Mukherjee et al. (2000) showed that dissolved oxygen level was the limiting factor for l-asparaginase production by E. aerogenes cultivated in a reactor and it was determined that the presence of dissolved oxygen significantly increased l-asparaginase synthesis. Asparaginase activity was also a function of cell age; and the maximum activity occurred in cells harvested after 24 h incubation. In general, enzyme activity peaked at 48 h, while no activity was determined in cells harvested at 10 h incubation. Similar results were also reported by Nawaz et al. (1998) in a closely related species, E. cloacae.

In addition, the effect of glucose on the production of l-asparaginase has been studied. In general, glucose was regarded as a repressor of l-asparaginase in bacteria (Heinemann and Howard 1969; Barnes et al.1977; Mukherjee et al. 2000). In contrast, glucose has been observed to enhance l-asparaginase synthesis in S. marcescens (Khan et al. 1970). In our study, both high (1%) and low (0.1%) glucose medium supported cell growth similarly, while enzyme activity was reduced substantially in cultures grown in medium with a high glucose content. The difference between the enzyme activity of cultures grown in low-glucose and high-glucose medium was even more pronounced during further incubation periods (48–72 h), suggesting that cells in low-glucose medium overcome glucose catabolic repression by utilizing almost the whole carbohydrate substrate during the first 24 h cultivation. E. aerogenes, Ea[pUC8] and Ea[pUC8:15] grown in low-glucose medium for 72 h had enzyme activities, respectively, 300-, 92- and 14-fold higher than cells grown in high-glucose medium. No inhibitory role of VHb on l-asparaginase was observed in cultures grown in high-glucose medium, as the vgb carrying strain had enzyme levels similar to non-vgb strains. This might be due to the low level of VHb expression in this medium, as it is known that glucose and oxygen are both repressors of vgb (Khosla and Bailey 1989; Joshi and Dikshit 1994). In contrast, low-glucose medium even showed a stimulatory effect on the production of this enzyme. Cultures grown in low-glucose medium had a higher enzyme activity than those grown in glucose-free medium under the same conditions. A similar catabolic repression/induction was reported by Carlsen and Nielsen (2001), although in an entirely different microorganism, the filamentous fungus Aspergillus oryzae, in which glucose acted both as a repressor (via a carbon catabolic repressor, CreA) and as an inducer for α-amylase. In this context, it has been shown that the formation of l-asparaginase in K. aerogenes is neither activated nor inhibited by glucose, but is exclusively controlled by glutamine synthetase (Resnick and Magasanic 1976). This latter observation was based on the use of an ammonia-limited medium in which the sole physiological function of l-asparaginase was suggested to be to supply cells with ammonia, aspartate and glutamate for growth. The increase in the level of glutamine synthetase that results from ammonia deprivation was suggested to be the necessary and sufficient stimulus for the formation of l-asparaginase. In our study, however, the media contained rich sources of nitrogen (peptone, yeast extract) and, thus, glutamine synthetase-dependent l-asparaginase production may not be the case. Our data regarding the inhibition of l-asparaginase in cultures grown in high glucose, however, agree with previously published results indicating that, in E. aerogenes, l-asparaginase expression is substantially reduced by glucose at certain concentrations (Mukherjee et al. 2000). l-Asparaginase production in the low-glucose medium, where the average culture pH for all strains was 8.48±0.24 during the whole incubation, increased by up to an order of magnitude, compared with cultures grown in high-glucose medium (averaging a pH value of 4.35±0.021 at all time-points of incubation). The alkalinization of the medium with low glucose concentration might be as a result of complete utilization of this carbohydrate during the initial incubation periods, followed by the release of nitrogen as NH3 through the degradation of substrates containing amino groups (such as amino acids) at further incubations, as often observed during fermentation processes. The acidity in high-glucose medium, however, might be due to the preferred use of this substrate in place of nitrogenous substrates, thus contributing to acid production. The inhibition of l-asparaginase in a high-glucose medium, however, might not be due solely to a catabolic repression-type inhibition, but also to the low pH. In this regard, it has been shown that the production of l-asparaginase in E. cloacae is down-regulated as the pH of the medium decreases (no activity was detected at acidic pH<6.0), while optimal activity is detected at neutral and alkaline pH 6.5–9.5 (Nawaz et al. 1998). It has been suggested that besides acting as a repressor when used as a sole carbon source, glucose also reduces enzyme activity by lowering the medium pH (Mukherjee et al. 2000). All three strains showed similar enzyme activity when grown in high-glucose medium while, in low-glucose medium, the enzyme activity of the vgb strain was substantially (more than 10-fold) lower than both the wild type and the vgb − strain.

Analysis of cell-free broth at intervals during cultivation showed that (with the exception of high aeration/high agitation cultures, which had maximum ammonia levels at 48 h) ammonia accumulation generally increased with incubation time. The highest ammonia levels were determined in the cultures with the lowest enzyme activity: the high aeration/high agitation cultures. Cultures grown in high-glucose medium accumulated up to 3-fold higher levels of ammonia than those in low-glucose medium. It is known that systems operating at pH values lower than 7.0 accumulate ammonia in the ionized form (NH4 +), which further contributes to the acidity of the medium. These results are in good accordance with others showing that ions have an inhibitory effect on l-asparaginase production (Tosa et al. 1971; Sun and Setlow 1991).

In conclusion, this study indicates that l-asparaginase production can be almost completely inhibited in E. aerogenes expressing VHb, which is an oxygen-buffering and highly efficient oxygen-uptake system. The aeration rate of cultures is also a determining factor for enzyme production. Cells grown at high aeration or no aeration conditions had a substantially lower enzyme activity than that of the cells at low aeration. l-Asparaginase is shown to be an enzyme of secondary metabolism as it is produced in late stationary phase cells (after 24 h incubation), but not in logarithmic cells (up to 10 h incubation). Also, glucose at high concentration (1%) in the culture medium causes almost complete inhibition of l-asparaginase activity, while low glucose (0.1%) has a stimulatory affect on the production of this enzyme. In a high-glucose medium, cells divert excess glucose mostly to the production of acidic fermentative products, such as acetate, formate and lactate, lowering the medium pH and inhibiting the formation of l-asparaginase. In contrast, at low concentrations, glucose is used both for energetic and biosynthetic purposes, causing a slight increase in medium pH that favors the production of l-asparaginase.