De novo synthesis of arginine and ornithine from citrulline in human colon carcinoma cells: metabolic fate of l -ornithine

Biochimica et Biophysica Acta 1425 (1998) 93^102

De novo synthesis of arginine and ornithine from citrulline in human colon carcinoma cells: metabolic fate of L-ornithine Mohamed Selamnia, Ve¨ronique Robert, Camille Mayeur, Serge Delpal, Franc,ois Blachier * Laboratoire de Nutrition et Se¨curite¨ Alimentaire, Institut National de la Recherche Agronomique, F-78352 Jouy-en-Josas Cedex, France Received 30 March 1998; revised 13 May 1998; accepted 19 May 1998

Abstract In human colon carcinoma cells (HT-29 cells), L-arginine is the common precursor of L-ornithine which generates polyamines strictly necessary for cellular growth, and nitric oxide which has a strong antiproliferative activity. We show here that proliferative HT-29 cells possess the capacity for de novo synthesis of L-arginine from L-citrulline, but not from L-ornithine. L-Ornithine is apparently not an L-arginine precursor due to the absence of any detectable ornithine carbamoyltransferase activity. In contrast, the newly synthesized L-arginine was competent for urea and thus L-ornithine production in a context of a high putrescine production in the ornithine decarboxylase pathway and a low degradation of this polyamine in the diamine oxidase pathway. However, cells grown in an arginine-free culture medium containing added L-citrulline were unable to reach confluency. Furthermore, the low amount of nitric oxide produced from L-arginine by these cells was apparently not involved in the control of cell growth since inhibition of nitric oxide synthase activity was without effect. On the other hand, the capacity of more differentiated and less proliferative HT-29 cells for de novo L-arginine synthesis from L-citrulline was increased. It is concluded that L-citrulline is a precursor of L-arginine and L-ornithine in proliferative HT-29 cells and that the metabolic fate of L-ornithine in these cells is mainly devoted to polyamine synthesis. The similarity between differentiated HT-29 cells and the enterocytes of newborn animals in terms of L-arginine metabolism is finally discussed. ß 1998 Elsevier Science B.V. All rights reserved. Keywords: Carcinoma cell; Colon; Metabolism; Proliferation; Citrulline; Ornithine; Arginine

1. Introduction Cells originating from human colon carcinoma cells have provided a useful tool for the study of the proliferation and di¡erentiation of intestinal cells [1]. Among them, HT-29 Glc3=‡ cells can spontaneously di¡erentiate after con£uency in enterocyte-like cells equipped with a typical brush-border membrane, and expressing enzymes associated with this struc-

* Corresponding author. Fax: +33 (1) 3465-2311.

ture, such as dipeptidyl peptidase IV [2,3]. In these cells, polyamines have been shown to be strictly necessary for growth. Indeed, inhibition of polyamine biosynthesis in HT-29 cells can lead to an arrest of cell proliferation [4]. Polyamine synthesis involves the conversion of L-ornithine into putrescine and CO2 by ornithine decarboxylase (ODC). This short-lived protein involved in cell growth and transformation is considered as a proto-oncogene [5]. Putrescine can lead to spermidine synthesis in HT-29 cells through the action of spermidine synthase which uses decarboxylated S-adenosylmethionine as a cosubstrate [6].

0304-4165 / 98 / $ ^ see front matter ß 1998 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 1 6 5 ( 9 8 ) 0 0 0 5 6 - 7

BBAGEN 24651 4-9-98

94

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

Although the mechanism of polyamine action is not fully understood, putrescine has been shown to stimulate DNA synthesis in intestinal cells [7] and increase the transcription of growth-associated genes in human colon carcinoma cells [8]. Furthermore, in HT29 cells, ODC appears to be closely related in terms of catalytic activity to the proliferative-di¡erentiated cellular state with an inverse relationship between ODC activity and the apparition of the di¡erentiated phenotype [9]. Therefore, the determination of the di¡erent sources of L-ornithine for polyamine synthesis appears to be an important point in clarifying how cell growth is regulated by amino acids metabolism. In our culture conditions, L-arginine is a precursor of L-ornithine in HT-29 cells. This amino acid is present at a relatively high concentration, i.e. 400 WM. Indeed, HT-29 cells possess an arginase activity which enables L-arginine to be converted into L-ornithine and urea [9]. L-Ornithine, although present at a low concentration in the culture medium, i.e. 6 WM, does, however, represent a qualitatively major contributor to putrescine synthesis since extracellular L-ornithine is apparently more e¤ciently utilized in the ODC pathway than L-ornithine derived from L-arginine [9] and complete inhibition of the £ux of L-arginine through arginase in HT-29 cells induces a relatively modest e¡ect on cellular growth [10]. In contrast, L-glutamine is a poor precursor of L-ornithine in HT-29 cells, and does not lead to putrescine or spermidine synthesis [9]. In addition to being a precursor for polyamine synthesis, L-arginine is also a substrate for nitric oxide synthase (NOS) which catalyzes the conversion of this amino acid into NO and L-citrulline. Low NOS activity is detected in both proliferative and di¡erentiated HT-29 cells [9]. When the NO-donor sodium nitroprusside was added at micromolar concentration to the culture medium, it was shown to severely decrease HT-29 cell growth [11]. Thus L-arginine is the precursor of metabolites having opposite e¡ects on the cellular state. Hence, our study was undertaken to resolve the following questions: (1) is L-citrulline present in the culture medium a precursor of L-arginine in HT-29 cells? (2) is newly synthesized Larginine used for the production of urea and L-ornithine? (3) is this metabolism di¡erent when HT-29

cells move from a proliferative state to a more di¡erentiated phenotype? and (4) is arginine-derived NO exerting a control on HT-29 cell growth? 2. Materials and methods 2.1. Materials 14 L-[ureido- C]Citrulline; 3

[1,4-14 C]putrescine; D,L-K[3,4- H]di£uoromethylornithine and L-[guanido14 C]arginine were purchased from New England Nuclear. L-[U-14 C]Ornithine and L-[U-14 C]arginine were obtained from Amersham. D,L-K-Di£uoromethylornithine (DFMO) was a gift from the Marion Merrell Dow Research Institute. Gly-Pro p-nitroanilide, p-nitroanilide, N-g-nitro-L-arginine (L-NNA), carbamoylphosphate and aminoguanidine were purchased from Sigma. 2.2. Cell culture The human adenocarcinoma cell line HT-29 was established in permanent culture in 1975 [12]. HT-29 Glc3=‡ cells used in this study were selected by Zweibaum et al. [13] from parental cells by growing them in a glucose-free medium for 36 passages, then leaving them to grow at 37³C under a 10% CO2 atmosphere in a Dulbecco's modi¢ed Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum containing 25 mM D-glucose, 100 U/ml penicillin, 100 Wg/ml streptomycin and 100 Wg/ml fungizone. HT-29 Glc 3=‡ were used between passages 47 and 85 (one passage every 7 days) and were seeded at a density of 0.08U106 cells/ml on day 0. The culture medium was changed every day. DMEM and L-arginine-depleted DMEM were purchased from Gibco. L-Ornithine (chloride form) and L-citrulline were obtained from Sigma. Cell isolation was performed with phosphate bu¡er saline containing 1 g/l EDTA and 0.25 g/l trypsin. Cells were counted on a hematocymeter. Isolated cell viability was estimated by a Trypan blue exclusion test. Protein content of HT-29 cells was determined by the method of Lowry et al. [14]. Nitrite cell content was measured using Griess reagent [15].

BBAGEN 24651 4-9-98

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

2.3. Electron microscopy Day 7 or day 20 cells were ¢xed in 2% glutaraldehyde in 0.1 M cacodylate bu¡er, pH 7.4 for 30 min, post-¢xed in 1% OsO4 in the same bu¡er, dehydrated in increasing concentrations ethanol solutions and embedded in Epon 812 [16]. Semi-thin sections were stained with Azur II^Methylene blue [17] and observed with a polyvar microscope. Thin sections were contrasted with uranyl and lead citrate and observed with an EM-10 electron microscope (Zeiss). 2.4. Cell incubation in the presence of radioactive precursors Isolated cells were resuspended in a Krebs^Henseleit bicarbonate-bu¡ered medium (pH 7.4) saturated with a mixture of O2 /CO2 (19:1, v/v), containing 10 mM HEPES, 1.3 mM CaCl2; 2 mM MgCl2 , enriched with 10 mg/ml bovine serum albumin (incubation medium). The cells were incubated at 37³C in 120 Wl in the presence of radioactive substrates. L-Arginine was separated from radioactive L-citrulline or L-ornithine using a partisphere C-18 column (Whatman); the bu¡ers and gradient were described previously [18]. The same system was used for the separation of urea from L-[guanido-14 C]arginine. For the separation of urea from L-[ureido-14 C]citrulline and that of 14 L-citrulline from L-[U- C]arginine, a Kromasil C-18 column (AIT) was used after o-phthaldialdehyde derivatization for 2 min as described [19]. 2.5. ODC protein content This method is based on irreversible ¢xation of a substrate analog on ODC, namely K-di£uoromethylornithine (DFMO), and was adapted from previous studies [20,21]. Cell homogenates were incubated for 20 min at 37³C in a ¢nal volume of 140 Wl of 100 mM, pH 7.2 Tris-HCl bu¡er containing 0.1 mM EDTA, 2.5 mM dithiothreitol, 40 WM pyridoxal phosphate and in the presence or the absence of unlabeled 10 mM DFMO in order to deduce non speci¢c binding. The latter represented 51 þ 9% (n = 3) on day 7 and 86 þ 3% (n = 3) on day 20 of total binding. Then 20 Wl of [3,4-3 H]DFMO (¢nal concentration 0.35 WM) was added and incubated at 37³C for 5 h. The incubation medium was then transferred

95

into collodion bags (Sartorius, Goettingen, Germany) and dialyzed against a 100 mM, pH 7.2 Tris-HCl bu¡er. After 24 h dialysis at 4³C, the contents were counted for radioactivity. ODC protein content was calculated assuming an equal stoichiometry of DFMO to enzyme and a molecular weight of 53 000 for ODC [20]. 2.6. Uptake studies The uptake of radioactive L-citrulline by 0.2^ 0.4U106 cells was carried out by cell centrifugation through an oil layer as described [9]. Brie£y, cells were incubated in a 120 Wl incubation medium containing 2 mM D-glucose and L-[ureido-14 C]citrulline and then centrifuged (12 000Ug, 3 min) through silicon oil (TAI Lubricants, Hockessin, USA). In all experiments, the calculated net uptake was the di¡erence between the uptake performed at 37³C and that at 4³C. It was corrected for the apparent volume of distribution of 30 WM L-[1-14 C]glucose performed at both temperatures [10]. The uptake at 4³C represented 8 þ 1% (n = 6) of that measured at 37³C. 2.7. Enzyme assays Enzyme assays were performed at 37³C on sonicated HT-29 cells. Dipeptidyl peptidase IV activity was assayed using the method of Nagatsu [22] on scraped cells. Gly-Pro-p-Nitroanilide 1.5 mM was used as the substrate. Ornithine carbamoyl-transferase (OCT) activity was determined in a Tris-HCl bu¡er (50 mM, pH 8.3) containing 1 mM 14 L-[U- C]ornithine and 0.5 mM carbamoylphosphate. Radioactive L-citrulline was separated by HPLC. Diamine oxidase (DAO) activity was measured using the Okuyama and Kobayashi method [23]. [1,4-14 C]putrescine (0.1 mM) was used as substrate for 1 h incubation at 37³C. Diamine oxidase activity was calculated as the aminoguanidine (0.1 mM) sensitive conversion of radioactive putrescine into [v1-14 C]pyrroline [24]. 2.8. Data analysis The production of radioactive metabolites was calculated by reference to the speci¢c activity of the precursors in the incubation medium. Results were

BBAGEN 24651 4-9-98

96

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

HT-29 cells isolated at di¡erent passages (n). The statistical signi¢cance of di¡erences between mean values were assessed by Student's t-test. 3. Results 3.1. Metabolism of L-citrulline in proliferative HT-29 cells

Fig. 1. Evolution of dipeptidyl peptidase IV activity in HT-29 cells recovered at di¡erent times after seeding. HT-29 cells were seeded at a concentration of 0.40U106 cells per 25 cm2 on day 0 and maintained for up to 23 days in culture. The culture medium was changed every day and cells were scraped from £asks, homogenized and tested for dipeptidylpeptidase activity and protein content as described in Section 2. Mean values ( þ S.E.) are shown and represent four experiments.

expressed as the mean ( þ S.E.) together with the number of individual experiments performed with

In these experiments, cells were isolated 7 days after seeding (day 7) which corresponded to undi¡erentiated cells shown by the activity of the brush-border membrane marker dipeptidylpeptidase IV used as a biochemical marker of cell di¡erentiation (Fig. 1). At this stage, morphological examination using an electronic microscope revealed that most cells did not show the typical brush-border membrane characteristics of enterocyte-like cells (Fig. 2). The HT-29 cells which were isolated on day 7 were in the exponential growth phase before con£uency which was reached on day 9 (Fig. 3). From the metabolic study, we found that cells were equipped to ensure the sequential conversion of L-citrulline into L-arginine. In-

Fig. 2. Morphological aspect of HT-29 cells recovered on day 7. The cells recovered in the exponential growth phase were observed using a polyvar microscope (A) and an electronic microscope EM-10 (B). For magni¢cation scale bar.

BBAGEN 24651 4-9-98

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

Fig. 3. E¡ect of the NO-synthase inhibitor N-g-nitro-L-arginine on HT-29 cell growth. The cells were seeded at a concentration of 0.40U106 cells per 25 cm2 at day 0 and culture medium containing or not 1 mM N-g-nitro-L-arginine was added at day 1 and changed every day thereafter. The cells were isolated and counted after EDTA-trypsin treatment as indicated in Section 2. Mean values ( þ S.E.) are shown and represent three to six experiments.

deed, a measurable amount of radioactive L-arginine was produced from 1 mM L-[ureido-14 C]citrulline (Table 1). This represented a net production of L-arginine since a large part of the neosynthetized labeled L-arginine was converted into radioactive urea in the arginase pathway (Table 1). It can be noted that L-ornithine derived from L-[ureido-14 C]citrulline is not radioactive since the 14 carbon is lost in urea after arginase catalysis. Since urea is a metabolic end product, we approximated that HT-29 cells were able to synthesize L-arginine from L-citrulline at 602 pmol/106 cells/90 min. The net uptake of L-citrulline did not limit L-arginine synthesis since total L-arginine de novo synthesis represented 5% of the precursor net uptake (Table 1). D-Glucose, when tested at a concentration of 5 mM, was found to more than double the amount of L-citrulline converted into L-arginine. In fact, in the presence of the

97

hexose, the L-arginine net production represented 234 þ 41% (n = 4, P 6 0.05) of the control value determined in the absence of the hexose. This D-glucose e¡ect was not related to L-aspartate synthesis used in the conversion of L-citrulline into argininosuccinate catalyzed by argininosuccinate synthase, since in the presence of 1 mM L-aspartate, the conversion of L-citrulline into L-arginine represented 84 þ 6% (n = 4) of the control value. The stimulatory e¡ect of D-glucose on L-arginine synthesis coincided with the inhibitory e¡ect of the hexose on the L-arginine £ux through the arginase pathway in HT-29 cells. Indeed, the basal production of radioactive urea from 1 mM L-[guanido-14 C]arginine was inhibited by 28 þ 7% (n = 5, P 6 0.025) with 5 mM D-glucose. A higher D-glucose concentration i.e. 25 mM exerted the same inhibitory e¡ect (data not shown). We were unable to detect any L-arginine production from 1 mM L-[U-14 C]ornithine (less than 70 pmol/106 cells/90 min, Table 1) or L-citrulline production from 1 mM L-[U-14 C]arginine (less than 80 pmol/ 106 cells/90 min). This corresponded to the fact that ornithine carbamoyltransferase (OCT) activity was not detectable in HT-29 cell homogenates. Using isolated rat enterocytes as a positive control for OTC activity, we calculated that ornithine carbamoyltransferase would represent less than 2% of that found in isolated rat enterocytes (i.e. less than 280 pmol/mg protein/15 min) [25]. Diamine oxidase (DAO), the enzyme responsible for the degradation of putrescine was found to have a very weak activity in proliferative HT-29 cells with an average value of 6.75 þ 0.86 pmol/mg protein/60 min (n = 7). On day 7, a radioactive DFMO-binding test was used on HT-29 cells homogenates (Table 2). This allows us to determine the amount of ODC protein in the proliferative cells (i.e. 2.3 þ 0.7 ng ODC protein per mg protein).

Table 1 L-Citrulline metabolism in HT-29 cells isolated in a proliferative state (day 7) pmol/106 cells/90 min 1 mM L-[ureido-14 C]citrulline net uptake Production of L-arginine from 1 mM L-[ureido-14 C]citrulline Production of urea from 1 mM L-[ureido-14 C]citrulline Production of L-arginine from 1 mM L-[U-14 C]ornithine Production of L-citrulline from 1 mM L-[U-14 C]arginine

12 073 þ 1967 (n = 5) 159 þ 29 (n = 9) 443 þ 134 (n = 3) Not detectable Not detectable

HT-29 cells isolated after 7 days of culture were incubated at 37³C in the presence of radioactive citrulline, ornithine or arginine. Mean values ( þ S.E.) are shown together with the number of experiments (n).

BBAGEN 24651 4-9-98

98

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

day 7, the cell viability in L-arginine-containing medium (i.e. 86 þ 3%, n = 4) was identical in the L-arginine-depleted L-citrulline-containing medium (i.e. 84 þ 5%, n = 4) and L-ornithine-containing medium (i.e. 85 þ 7%, n = 4). 3.3. E¡ect of an inhibitor of NOS activity on HT-29 cell growth

Fig. 4. E¡ect of L-citrulline and L-ornithine on HT-29 cells grown in an L-arginine-depleted culture medium. HT-29 cells were seeded at a concentration of 0.40U106 cells per 25 cm2 at day 0 in an L-arginine (0.4 mM) containing standard culture medium, or in an L-arginine-depleted culture medium, or in an L-arginine-depleted medium containing 0.4 mM L-citrulline or 0.4 mM L-ornithine. Mean values ( þ S.E.) are shown and represent four experiments. Note the break in the vertical axis.

3.2. E¡ect of L-citrulline and L-ornithine on HT-29 cell growth in an L-arginine-depleted medium As indicated in Fig. 4, no cell growth was observed in the L-arginine-depleted medium when compared with control growth in the standard culture medium, i.e. in the presence of 0.4 mM L-arginine. The addition of 0.4 mM L-citrulline to the L-arginine-depleted medium provoked only poor cellular growth. L-Ornithine was without e¡ect (Fig. 4) and the addition of both L-citrulline and L-ornithine exerted the same e¡ect on cell growth as L-citrulline alone (data not shown). The lack of L-arginine in the culture medium with L-citrulline, L-ornithine or both, had no e¡ect on the viability of HT-29 cells recovered in the £asks. Indeed, as shown by a Trypan blue exclusion test on

The addition of 1 mM N-g-nitro-L-arginine to inhibit endogenous NO production was without e¡ect on HT-29 cell growth up to 20 days after seeding (Fig. 3). Due to the high endogenous cellular nitrite content measured on day 7, i.e. 668 þ 195 pmol/106 cells (n = 5), it was not feasible to measure increased nitrite synthesis in the cells or in the culture medium in response to 1 mM L-citrulline added in the culture medium at day 6 for 24 h. 3.4. Metabolism of L-citrulline in di¡erentiated HT-29 cells Cells isolated at late con£uency (day 20) were characterized by an increase in dipeptidyl peptidase IV activity (Fig. 1). An increased number of HT-29 cells showing typical brush-border membranes was observed together with the formation of spherical structures (Fig. 5). On day 20, the HT-29 cells were in a linear growth phase (Fig. 3). The production of 14 L-arginine from 1 mM L-[ureido- C]citrulline had more than doubled on day 20 compared with day 7 representing 377 þ 72 pmol/106 cells/90 min (n = 8, P 6 0.025 vs. day 7). This increase was not related to increased net uptake of L-citrulline, which amounted to 4403 þ 866 pmol/106 cells/90 min (n = 3) on day 20. In contrast, as in day 7 cells, no production of radioactive L-arginine from 1 mM 14 L-[U- C]ornithine was recorded (n = 3). The newly

Table 2 Amount of ornithine decarboxylase protein in proliferative (day 7) or di¡erentiated (day 20) HT-29 cells 3

Amount of bound [3,4- H]DFMO (fmol/mg protein)

Day 7

Day 20

42.5 þ 10.6 (n = 3)

8.9 þ 3.0 (n = 3)a

Cellular homogenates were incubated in the presence or absence of 10 mM DFMO for 20 min at 37³C. Then, 0.35 WM [3,43 H]DFMO was added to the incubation medium for 5 h at 37³C. Finally, cell homogenates were dialyzed for 24 h at 4³C against a Tris-HCl (100 mM, pH 7.2) bu¡er and radioactivity was counted by liquid scintillation. The amount of non-speci¢c binding was corrected from the total DFMO-binding. a P 6 0.05 vs. day 7.

BBAGEN 24651 4-9-98

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

99

Fig. 5. Morphological aspect of HT-29 cells recovered at day 20. The cells recovered at the late con£uency were observed using a polyvar microscope (A) and an electronic microscope EM-10 (B). For magni¢cation see scale bar.

synthesized L-arginine derived from L-citrulline was competent for L-ornithine and urea production since the latter represented 648 þ 269 pmol/106 cells/90 min (n = 3), when measured from 1 mM L-[ureido-14 C]citrulline. However, the amount of bound radioactive DFMO was severely decreased in HT-29 cells homogenates recovered on day 20 when compared with those on day 7 (Table 2). Using this data, we calculate an amount of ODC protein in di¡erentiated cells equal to 0.5 þ 0.2 ng ODC protein per mg protein. The diamine oxidase activity on day 20 (i.e. 7.18 þ 0.94 pmol/mg protein/60 min, n = 7) was as low as that on day 7. Finally, 5 mM D-glucose increased basal L-arginine production from 1 mM L-[ureido-14 C]citrulline on day 20 HT-29 cells (132 þ 7% of basal value, n = 5, P 6 0.01) coinciding with an inhibition of the £ux of 1 mM L-[guanido-14 C]arginine in the arginase path-

way (31 þ 8% inhibition, n = 4, P 6 0.05 vs. basal value). 4. Discussion Our data clearly demonstrate that proliferative HT-29 cells possess the metabolic capacity for L-citrulline net uptake and conversion into L-arginine. Using the L-citrulline concentration in fetal bovine serum in our culture medium [26], we calculated that this amino acid was present at a ¢nal concentration of 9 WM. This newly synthesized L-arginine was competent for urea and thus ornithine synthesis indicating that this amino acid is a substrate for arginase. Since L-ornithine is the precursor of polyamines, L-citrulline is therefore a potential precursor of polyamines. Polyamines are considered as growth

BBAGEN 24651 4-9-98

100

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

factors involved in cell proliferation and di¡erentiation, and since ODC is very highly expressed in proliferative HT-29 cells [9], we consider that it is important to list all the precursors of L-ornithine in these cells. When used at a 1 mM concentration, L-citrulline produced an amount of urea equal to 443 pmol/106 cells/90 min incubation. Since the end-product urea and L-ornithine are coproducts of L-arginine hydrolysis catalyzed by arginase, the latter value represents also L-ornithine production. This production is not vastly di¡erent to that of L-ornithine generated from 1 mM L-arginine [9], indicating that L-citrulline can be e¤ciently used for the production of L-ornithine. When compared with the maximal capacity for putrescine synthesis in day 7 cells, i.e. 443 pmol/106 cells/90 min [9], the amount of L-ornithine generated from 1 mM L-citrulline represents the maximal capacity of proliferative HT-29 cells to synthetize putrescine. Thus, although L-citrulline is present at a relatively low concentration in the culture medium, it probably represents a signi¢cant contributor to L-ornithine synthesis. In these proliferative cells, the metabolism of L-ornithine is apparently, through high ODC activity, devoted to putrescine and spermidine synthesis. Indeed, in HT-29 cells unlike what is found in normal intestinal absorptive cells [25], we demonstrate that L-ornithine is not converted into Lcitrulline due to the absence of a detectable amount of OCT activity. Furthermore, from previous work [9,10], we found that the amount of L-[U-14 C]ornithine converted to putrescine was similar to that of L-[1-14 C]ornithine which was decarboxylated by HT-29 cells indicating that little L-ornithine was used for succinyl coenzyme A production and that little putrescine was degraded by diamine oxidase. The latter point coincides with the very low amount of DAO activity recorded in HT-29 cells homogenates. Once again, it is in sharp contrast with the high DAO activity recorded in normal intestinal absorptive cells [24]. Furthermore, we have also previously shown [9] that the putrescine content was very low in these cells and that proliferative cells exported high amounts of putrescine. This shows that in HT-29 cells the release plays a major role in the regulation of the diamine cell content in the presence of low DAO activity. The L-arginine newly synthesized from L-citrulline

is a potential precursor of protein synthesis. The experiments dealing with the culture medium lacking Larginine, but containing L-citrulline, indicate that de novo synthetized L-arginine probably does not meet with L-arginine requirements for protein synthesis. Indeed, the amount of L-arginine produced from L-citrulline in our present work is low when compared with proliferative HT-29 cells protein synthesis capacity [10] and would thus explain the limited effect of L-citrulline on cell growth. Our data show an inhibitory e¡ect of D-glucose on the L-arginine £ux through arginase; this amino acid would then become more available for protein synthesis. When L-ornithine was added to the L-arginine-depleted medium, it had no e¡ect on cell growth when compared with cells cultured in the L-arginine free medium. This corresponds to the fact that, due to the absence of detectable amount of OCT activity, L-ornithine cannot be converted to L-arginine. Finally, as recently reported [27,28], L-arginine could have been utilized in the NO-synthase pathway. Since proliferative HT-29 cells possess detectable NO-synthase activity [9] and since low concentrations of NO are able to severely limit HT-29 cell proliferation [11], we tested the e¡ect of an inhibitor of NO-synthase activity, namely N-g-L-nitro-L-arginine (L-NNA) which is transported in mammalian cells by system L [29] and which e¤ciently inhibits HT-29 cell NO-synthase activity when used at 0.1 mM concentration [9]. L-NNA (1.0 mM) failed to exert any detectable e¡ect on cell growth indicating that there was insu¤cient NO generated from L-arginine to exert any control over cell growth. Therefore, the de novo synthetized L-arginine appears to be primarily related to L-ornithine production. The part of the study related to less proliferative and more di¡erentiated enterocyte-like HT-29 cells indicated that the biosynthesis of L-arginine from L-citrulline is still operative with production of urea and L-ornithine from L-arginine. However, due to the spectacular decrease in ODC activity [9] and ODC protein content (this study), L-ornithine would be used to a much lesser extent in polyamine synthesis. A number of analogies in terms of L-arginine metabolism were found between HT-29 cells and newborn enterocytes isolated from animal models. First, as in HT-29 cells, newborn pig enterocytes are able to convert L-citrulline into L-arginine, but nearly lose

BBAGEN 24651 4-9-98

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

101

apparently not produced in su¤cient amounts to regulate HT-29 cell growth. Acknowledgements The authors wish to thank Mrs. M.-C. Kopka for secretarial help and Mrs. K. Archbold-Zenon from the Translation Department at INRA for revising the manuscript. Fig. 6. Schematic view of the de novo synthesis of L-arginine from L-citrulline and channelling of L-arginine in the polyamine, nitric oxide or protein synthesis pathways in proliferative day 7 HT-29 cells. 999 s , major metabolic pathways; - - - - s , minor metabolic pathways. ARG, L-arginine; AS, argininosuccinate; ORN, L-ornithine; PUTR, putrescine; CITR, L-citrulline ; SPMD, spermidine ; v1-P, v1 pyrroline; ASP, L-aspartate; FUM, fumarate; GLN, L-glutamine; GLU, L-glutamate.

this capacity after weaning [30]. Secondly, in HT-29 cells and newborn pig enterocytes, D-glucose decreases the £ux of L-arginine through arginase, which is not the case in weaned pig enterocytes [30]. Third, NO-synthase activity was equally low in HT-29 cells and newborn pig enterocytes. This activity increases during the suckling period and after weaning [31]. Finally, ornithine decarboxylase activity is relatively high in enterocytes isolated at birth and falls to undetectable levels in the suckling period and after weaning [32,33]. Although we made comparisons between absorptive intestinal cells from di¡erent origins, and with this reservation in mind, the metabolic analogies listed above reinforce the view that di¡erentiated HT-29 cells should be regarded as fetal enterocyte-like cells as previously proposed [34^36]. In conclusion (Fig. 6), the data from the present study demonstrate that the polyamine precursor L-ornithine in HT-29 cells can (in addition to extracellular source and intracellular synthesis from L-arginine) be provided by the stepwise conversion of L-citrulline. However, L-arginine derived from L-citrulline is not produced, even in the presence of Dglucose which limits the £ux of L-arginine in the arginase pathway, in su¤cient quantities to ful¢l arginine needs for protein synthesis in an arginine-free culture medium. Finally, endogenously formed NO is

References [1] A. Zweibaum, M. Laburthe, E. Grasset, D. Louvard, in: M. Field and R.A. Frizzell (Eds.), Handbook of Physiology, The Gastrointestinal System, Vol. IV, American Physiological Society, Washington, DC, 1991, pp. 223^255. [2] T. Lesu¥eur, A. Kornowski, C. Augeron, E. Dussaulx, A. Barbat, C. Laboisse, A. Zweibaum, Int. J. Cancer 49 (1991) 731^737. [3] D. Darmoul, M. Lacasa, L. Baricault, D. Marguet, C. Sapin, P. Trotot, A. Barbat, G. Trugnan, J. Biol. Chem. 267 (1992) 4824^4833. [4] L. Gamet, Y. Cazenave, V. Trocheris, C. Denis-Pouxviel, J.C. Murat, Int. J. Cancer 47 (1991) 633^638. [5] M. Auvinen, A. Paasinen, C.A. Leif, E. Ho«lta«, Nature 360 (1992) 355^358. [6] A.E. Pegg, Cancer Res. 48 (1998) 759^774. [7] D.D. Ginty, D.L. Osborne, E.R. Seidel, Am. J. Physiol. 257 (1989) G145^G150. [8] P. Celano, S.B. Baylin, R.A. Casero, J. Biol. Chem. 264 (1989) 8922^8927. [9] F. Blachier, M. Selamnia, V. Robert, H. M'Rabet-Touil, P.H. Due¨e, Biochim. Biophys. Acta 1268 (1995) 255^262. [10] M. Selamnia, V. Robert, C. Mayeur, P.H. Due¨e, F. Blachier, Biochim. Biophys. Acta 1379 (1997) 151^160. [11] F. Blachier, V. Robert, M. Selamnia, C. Mayeur, P.H. Due¨e, FEBS Lett. 396 (1996) 315^318. [12] J. Fogh, G. Trempe, in: J. Fogh (Ed.), New Human Tumor Cell Lines, Plenum Press, New York, 1975, pp. 115^141. [13] A. Zweibaum, M. Pinto, G. Chevalier, E. Dussaulx, N. Triadoux, B. Lacroix, K. Ha¡en, J.L. Brun, M. Rousset, J. Cell. Physiol. 122 (1985) 21^29. [14] O.H. Lowry, N.J. Rosenbrough, A.L. Farr, R.J. Randal, J. Biol. Chem. 193 (1951) 265^275. [15] L.C. Green, D.A. Wagner, J. Glogowski, P.L. Skipper, J.S. Wishnok, S.R. Tannenbaum, Anal. Biochem. 126 (1982) 131^138. [16] J. Luft, J. Biophys. Biochem. Cytol. 9 (1961) 409^414. [17] C.R. Leeson, T.S. Leeson, Can. J. Zool. 48 (1970) 189^190. [18] N. Seiler, B. Kno«dgen, J. Chromatogr. 221 (1980) 227^235. [19] J. Morineau, M. Azoulay, F. Frappier, J. Chromatogr. 467 (1989) 209^216.

BBAGEN 24651 4-9-98

102

M. Selamnia et al. / Biochimica et Biophysica Acta 1425 (1998) 93^102

[20] B.G. Erwin, J.E. Seely, A.E. Pegg, Biochemistry 22 (1983) 3027^3032. [21] D.D. Ginty, M. Marlowe, P.H. Pekala, E.R. Seidel, Am. J. Physiol. 258 (1990) G454^G460. [22] T. Nagatsu, M. Hino, H. Fuyamada, T. Hayakawa, S. Sakakibara, Y. Nakagawa, T. Takemoto, Anal. Biochem. 74 (1976) 466^476. [23] T. Okuyama, Y. Kobayashi, Arch. Biochem. Biophys. 95 (1961) 242^250. [24] G. Guihot, V. Colomb, A. Jobert-Giraud, M.T. Morel, O. Corriol, P.H. Due¨e, C. Ricour, F. Blachier, J. Parenter. Enter. Nutr. 21 (1997) 259^265. [25] G. Guihot, F. Blachier, V. Colomb, M.T. Morel, P. Raynal, O. Corriol, C. Ricour, P.H. Due¨e, J. Parenter. Enter. Nutr. 21 (1997) 316^323. [26] P. Healy, J. Denis, R. Rawlinson, L. Andersson, Res. Vet. Sci. 55 (1993) 271^274. [27] A.K. Nussler, T.R. Billiar, Z.Z. Liu, S.M. Morris, J. Biol. Chem. 269 (1994) 1257^1261. ë nggard, J.R. [28] M. Hecker, W.C. Sessa, J.J. Harris, E.E. A Vane, Proc. Natl. Acad. Sci. USA 87 (1990) 8612^8616.

[29] K.K. McDonald, R. Rouhani, M.E. Handlogten, E.R. Block, O.W. Gri¤th, R.D. Allison, M.S. Kilberg, Biochim. Biophys. Acta 1324 (1997) 133^141. [30] F. Blachier, H. M'Rabet-Touil, L. Posho, B. Darcy-Vrillon, P.H. Due¨e, Eur. J. Biochem. 216 (1993) 109^117. [31] H. M'Rabet-Touil, F. Blachier, M.T. Morel, B. Darcy-Vrillon, P.H. Due¨e, FEBS Lett. 331 (1993) 243^247. [32] F. Blachier, H. M'Rabet-Touil, L. Posho, M.T. Morel, F. Bernard, B. Darcy-Vrillon, P.H. Due¨e, Biochim. Biophys. Acta 1175 (1992) 21^26. [33] H. M'Rabet-Touil, F. Blachier, N. Hellio, V. Robert, C. Cherbuy, B. Darcy-Vrillon, P.H. Due¨e, Mol. Cell. Biochem. 146 (1995) 49^54. [34] M. Rousset, Biochimie 68 (1986) 1035^1040. [35] M. Hekmati, S. Polak-Charcon, Y. Ben-Shaul, Cell Di¡er. Dev. 31 (1990) 207^218. [36] A. Zweibaum, N. Triadou, M. Kedinger, C. Augeron, S. Robine-Le¨on, M. Pinto, M. Rousset, K. Ha¡en, Int. J. Cancer 32 (1983) 407^412.

BBAGEN 24651 4-9-98