EXPRESSION OF RECOMBINANT HUMAN ARGININOSUCCINATE SYNTHETASE IN ESCHERICHIA COLI

Escherichia coli strain BL21(DE3)/pET42a/ASS – an efficient producer of recombinant human argininosuccinate synthetase (rhASS) – was constructed, and preparations of purified rhASS were obtained using His-tag affinity chromatography. The effect of specific inhibitor, a -methyl-DL-aspartate, and nitric oxide donor, sodium nitroprusside, on the ASS specific activity was evaluated with purified rhASS protein and in mouse liver lysates. The developed expression platform is a useful tool in search for new ASS inhibitors efficient under in vitro and in vivo conditions.


INTRODUCTION
Argininosuccinate synthetase (ASS; L-citrulline, L-aspartate ligase; EC 6.3.4.5) plays an important physiological role as a rate-limiting enzyme of urea cycle and biosynthesis of arginine, a semi-essential amino acid in humans [12]. Consequently, ASS affects arginine-dependent synthesis of nitric oxide, polyamines, proline and glutamate [14]. The inherited deficiency in ASS causes citrullinemia, an autosomal recessive disorder, characterized by the elevated levels of blood citrulline and ammonia, which often results in hyperammonemic coma and early neonatal death [11]. In turn, some types of tumors, such as melanomas, renal and prostate carcinomas, mesotheliomas, some hepatocellular carcinomas exhibit downregulation of ASS expression leading to the inability to synthesize endogenous arginine, and to intrinsic dependence of such tumors on extracellular arginine supply (known as "arginine auxotrophy") [6,7]. Enzymatic arginine deprivation is a novel approach to target proliferation of tumors which lack ASS expression [8]. Several phase I/II trials with arginine-degrading pegylated bacterial arginine deiminase and human arginase I have shown clinical benefits in suppressing tumor growth and low toxicity in patients with ASS-negative tumors [5,6].
It was also observed that arginine deprivation can trigger upregulation of ASS expression in certain tumors, which, in turn, can lead to in vivo drug resis tance [6,8,12]. In such cases, application of low molecular weight ASS inhibitors may be desirable to Biol

SDS-PAGE and Western blot analysis.
For Western blot analysis proteins were separated on 10% SDS-PAGE [15], and transferred onto a polyvinylidene difluoride membrane (Millipore) [19]. The membranes were blocked in 5% non-fat dried milk in PBS containing 0.05% Tween-20 and immunoblotted with primary monoclonal antibodies to ASS (BD Transduction Laboratories) or to His-Tag (Millipore). Secondary goat horseradish peroxidase-conjugated anti-mouse antibody (Millipore) and an enhanced chemiluminescence's system ("Amersham Biosciences", USA) were used for the detection of immunoreactive protein bands.
Assay for enzymatic activity of argininosuccinate synthetase. ASS activity was assessed based on the accumulation of the product pyrophosphate as inorganic phosphate [13], following cleavage with pyrophosphatase (Sigma). ASS (1-2 μg of purified protein) was added to the reaction buffer (20 mM Tris-HCl (pH 8.0), 2 mM ATP, 2 mM citrulline, 2 mM aspartic acid, 6 mM MgCl 2 , 20 mM KCl, and 0.1 units of pyrophosphatase) to a final volume of 100 µl. The samples were incubated at 37°C in 96-well plates, and the reactions were terminated after 30 min by the addition of an equal volume of molybdate buffer (10 mM ascorbic acid, 2.5 mM ammonium molybdate, 2% sulfuric acid) and further incubation at room temperature for 30 min. In the case of mouse liver lysate, the enzymatic reaction was carried out in Eppendorf microtubes (5-10 µg of protein in a probe), and 100 µl of 10% trichloroacetic acid was added to each one, mixed and centrifuged for 10 min at 12,000 g. 100 µl of supernatants were placed in 96-well plates with further adding of molybdate buffer as described above.
Accumulation of phosphate was determined by measuring the absorbance at 630 nm on a plate reader (BioTek, Winooski, Vermont, USA), and its concentration was determined by comparison with inorganic phosphate (Pi) standards. Net ASS activity was calculated by subtracting the value at the reaction condition without substrates from that with substrates [10,13].
Inhibition of the ASS catalytic activity by a-methyl-DL-aspartate (MDLA) or sodium nitroprusside (SNP) was examined as follows. Increasing amounts of inhibitors were added to ASS fraction in reaction buffer as described in Fig. 4 legends. After incubation at 37°C for 60 min, the enzyme activity in the mixtures was estimated by the addition of substrate mixture and further incubation at 37°C for 30 min [17]; finally, the concentration of generated Pi was measured as described above. Specific activity of ASS was expressed as μmol Pi/min per mg protein.
Statistics. Each experiment was performed at least three times. Significant difference from the control was compared using Student's t-test.

RESULTS AND DISCUSSION
Expression and purification of rhASS protein. The human ASS gene comprises 16 exons which span 63 kilobases and are mapped to chromosome 9 [22]. Based on cDNA cloning, the predicted protein product consists of 412 amino acids with molecular weight of 46.4 kDa. To aid in downstream rhASS purification, we expressed rhASS in E. coli as a fusion protein with both His-and GST-N-terminal tags.
Ni-NTA agarose was used for rhASS (His) 6 -tag affinity purification as described in Materials and Methods. It was shown that 72 kDa dominant band was withdrawn from cell lysates after treatment with Ni-NTA agarose (Fig. 2, A, lines 4, 5), and its main fraction was eluted with 250 mM imidazole (Fig. 2, B). The purity of the eluted protein was approximately 90% after dialysis and concentration with centrifugal filter devices (Millipore) (Fig. 2, B, lane 3). The yield of 5-10 mg of the partially purified rhASS was obtained per liter of culture broth. Utilization of GSH-sepharose for affinity purification of rhASS via binding to GST-tag produced a lower yield (see Fig. 3

, B).
To confirm the identity of the expressed recombinant protein, Western blot analysis was carried out using two different commercial monoclonal antibodies, anti-ASS and anti-His-Tag. With anti-ASS antibody, we revealed the immunoreactive band at 72 kDa in the induced E. coli cells, transformed with pET-42a/ASS plasmid, only (Fig. 3, A, lane 4). The lysate of human hepatocellular carcinoma HepG2 cell line was loaded as a positive control (Fig. 3, A, lane 2).
With anti-His-Tag antibody, we observed no immunoreactive signal in untransformed E. coli cells, regardless of IPTG induction (Fig. 3, B, lanes 1, 2), whereas it was present in the lysates of pET-42a/ASS transformant and significantly upregulated upon IPTG treatment (Fig. 3, B, lanes 3 and 4, respectively).
We also confirmed that rhASS can be purified by affinity binding to either Ni-NTA agarose or GSH-sepharose, with higher efficacy in the former (Fig. 3, B, lanes 5, 6, respectively).
Thus, based on our Western blot analysis with anti-ASS and anti-His-Tag monoclonal antibodies, we confirmed expression of rhASS in the constructed E. coli producer. However, heterogenity of the recombinant protein product was also evident, probably due to the inefficient transcription of the unmodified heterologous gene in E. coli host  Fig. 3, B). Our attempts to overcome this problem by manipulating with media and cultivation conditions, as well as by expressing pET-42a/ASS in alternative E. coli host strains were not successful (data not shown). Inhibitory effect of MDLA and SNP on ASS activity. In order to verify, whether the expressed His (6) -ASS fusion protein is functional, we assayed its catalytic activity. After detailed analysis of the literature data, we found that for the measurement of ASS specific activity different methods were applied. In particular, for investigation of citrullinemia type I genetic variants by in vitro studies, C. Berning et al. [1] determined acti vity of wild-type and mutant ASS proteins in nickel agarose elution samples, by measuring the formation of argininosuccinate from aspartate, citrulline, and ATP using newly developed assay tandem mass spectrometry. Another accepted method to assay ASS acti vity is based on the conversion of radiolabeled aspartate to argininosuccinate [18,22]. In the work of G. Hao and co-authors [13], ASS activity was assessed based on the accumulation of its product, pyrophosphate. However, the ASS activity was represented as percentage of a control value. This method, with modifications described in Materials and Methods section, was utilized in current study. We found that purified preparations of rhASS exhibit specific activity around 30 µmol×min -1 ×mg -1 protein, whereas the ASS activity in mouse liver lysate, used as a positive control, was almost twice as high. A preheated sample (100°C, 10 min) of rhASS protein was used as a negative control and exhibited no detectable ASS activity.
An aspartate analogue, a-methyl-DL-aspartic acid (MDLA), has been previously shown to be a specific inhibitor of ASS from a variety of tissues [9,18]. For instance, to elucidate the effect of the down-regulation of ASS activity on nitric oxide (NO) production in bovine aortic endothelial cells, MDLA was used at concentration of 10 mM [18] or 8 mM [9]. Experimental procedure of Guerreiro et al [10], in which 1 mM MDLA was added into culture medium of human embryonic kidney 293 cells, or administered to spontaneous hypertensive rats at 1 mmol/kg, revealed that ASS is a functional target for a snake venom anti-hypertensive peptide. Therefore, MDLA is an ASS inhibitor in low mM range, which, however, may be too high to be used in humans as auxiliary agent of anticancer enzymotherapy (see Introduction).
Although arginine deprivation therapy, as an anticancer strategy, has been investigated for several decades [12], it is only recently was shown encouraging activity in patients with specific tumor types [21]. However, prolonged arginine starvation in human may cause vasoconstriction and thrombosis due to the deficit of arginine derivative, nitric oxide (NO), as vasodilator and disaggregant [4]. This problem can be overcome via supplementation with exogenous NO-donors in vivo, which could diminish some side effects of arginine deprivation. Sodium nitroprusside (SNP) is one of the most widely used agents for the management of hypertensive emergencies, and has been in clinical use for several decades [20]. Using the BRENDA database of the comprehensive enzyme information system (http://www.brenda-enzymes.org), nitric oxide was also found among the list of potential in vivo ASS inhibitors. Therefore, in this work we evaluated the effect of MDLA, an aspartate analogue, and SNP, an exogenous NO donor, on ASS activity in purified rhASS protein and in mouse liver lysates.
We observed that both analyzed ASS inhibitors elicited concentration-dependent inhibition of ASS (Fig. 4, A, B). The ASS enzymatic activity was completely blocked at concentration of MDLA 4 mM and higher, similarly for the purified rhASS and in mouse liver lysate (Fig. 4, A). SNP, in a range of 1-5 mM, was more effective on purified protein in comparison with mouse liver lysate (Fig. 4, B). Inhibitory concentration index (IC 50 ) toward ASS for MDLA was around 2.5 mM, for SNP -1.5 mM. The observations on SNP inhibitory effect (Fig. 4, B) are consistent with a mechanism, whereby NO donors reversibly inactivate human ASS by S-nitrosylation of its single cysteine residue [13]. It was previously shown in our laboratory that 0.1-0.2 mM SNP releases exogenous NO at the level which is close to NO physiological range and may be potentially used as an adjuvant compensatory agent upon arginine deprivation-based therapy for some types of tumors [3]. Therefore, we decided to examine the cumulative effect of SNP at 0.2 mM concentration and low dose of MDLA on ASS activity.
We observed that the combinational treatment of hrASS protein with NO donor and MDLA in low mM range revealed the opposite effect on its activity (Fig. 4, C) as compared to a separate administration of these inhibitors in a range of 1-5 mM (Fig. 4, A,  B). Apparently, 0.2 mM SNP compensates the inhibitory influence of the aspartate analogue. The mechanism of this phenomenon requires further investigation.
Taken together, the (His) 6 -and GST-tagged recombinant human ASS was expressed in the bacterial system. It was confirmed that (His) 6 -tag does not disturb the ASS enzymatic activity, but allows easy one-step purification procedure by Ni-chelating chromatography. The inhibitory effect of a-methyl-DL-aspartate, as well as NO donor, SNP, on the enzyme's activity of mouse liver lysate and purified protein was confirmed. The observed IC 50 concentration ranges for effective inhibition by MDLA and SNP were found to be in low millimolar range (Fig. 4, A, B), a relatively high level from physiological and pharmacological points of view. At lower physiological concentrations, these inhibitors do not provide a desirable ASS inhibition. Nevertheless, the developed scheme will be used in our next studies for in vitro search for more effective ASS inhibitors as components of combinational antitumor enzymotherapy.