Arginase Activity in Human Breast Cancer Cell Lines: N^ω-Hydroxy-l-arginine Selectively Inhibits Cell Proliferation and Induces Apoptosis in MDA-MB-468 Cells¹

l-Arginine is metabolized to l-ornithine and urea by arginase, which is important in the urea cycle as well as in the biochemical pathways that are essential for cell proliferation(1, 2). Arginase has two isoforms: A I³(hepatic arginase), a cytosolic enzyme present in the liver; and A II(extrahepatic arginase), which is located within the mitochondrial matrix. A I is primarily involved in the detoxification of ammonia and urea synthesis, whereas A II is involved in biosynthetic functions,such as the synthesis of ornithine, proline, and glutamate(1). Polyamines are subsequently synthesized from ornithine, the second product of arginase reaction.

Breast tumor tissues have been reported to have 2–3-fold higher polyamine levels than surrounding normal tissue (3). The precise mechanism by which the increase in polyamines occurs in breast tumor tissue is not known. It has been suggested that estrogens modulate the growth of certain breast cancer cell lines by increasing the expression of ODC (4, 5), thereby increasing the synthesis of polyamines. However, very little is known about the role of arginase in modulating polyamine biosynthesis and the growth of breast cancer cells.

l-Arginine is also metabolized by NOS to l-citrulline and NO via an intermediate, NOHA. NOS expression has been documented in various tumor cell lines and solid tumors (6, 7, 8), including breast tumors (9),and various human breast cancer cell lines (10). However,the role of NO in tumor biology is unclear. Low concentrations of NO have been reported to stimulate tumor growth and tumor cell proliferation (1, 11) and increase the metastatic ability of the tumor (12). On the other hand, high concentrations of NO can inhibit tumor growth (13, 14) and induce apoptosis in tumor cells (15). The exact role of NO, or NOHA, in modulating the growth of HBC cells is not known. NOHA has been shown to accumulate in the cell culture medium up to 20–30% of the amounts of NO and l-citrulline generated and is a potent competitive inhibitor of arginase (16, 17, 18).

Ornithine generated during the arginase-catalyzed reaction is the precursor of polyamines and proline in the mammalian cells. Spermine generally is the most abundant polyamine in human tumors such as breast carcinomas (19) and is synthesized from ornithine via ODC,spermidine synthase, and spermine synthase, respectively(1). Spermine rarely is found in prokaryotes but is widespread in eukaryotes where its role in cell growth is well established (2). Polyamines are known to bind DNA and affect gene expression by bringing about structural changes in chromatin and thereby stimulating cell growth (20). The primary objectives of the present study were to assess the expression of arginase and NOS in selected human breast cancer cell lines and to study the possible relationship between their expression and the rate of proliferation. We also wanted to study the effect of NOHA, a potent arginase inhibitor, on cell proliferation and the possible mechanisms involved.

The following reagents were purchased from Sigma Chemical (St. Louis, MO): l-arginine HCl, l-citrulline,pepstatin A, leupeptin, sodium acetate, trahydrobiopterin, calcium chloride, phenylmethylsulfonyl fluoride, EDTA, EGTA, NADPH, flavin mononucleotide, flavin-adenine dinucleotide, DTT, and triethanolamine hydrochloride. Dowex AG 1-X8 and Dowex AG 50W-X8 were purchased from Bio-Rad (Hercules, CA). l-[2,3,4,5-³H]Arginine HCl was purchased from Amersham.

Arginase activity was calculated by measuring the formation of[¹⁴C]urea from l-[guanido-¹⁴C]arginine according to the published procedure (21). Cells(3 × 10⁶/sample) were harvested,washed with PBS, and centrifuged at 1000 × gat 4°C for 10 min, and pellets were lysed for 10 min in 500 μl of 10 mm Tris-HCl (pH 7.4) containing 0.15 mm pepstatin A, 0.2 mmleupeptin, and 0.4% (w/v) Triton X-100. Samples were centrifuged at 20,000 × g at 4°C for 10 min. Supernatants were assayed for arginase activity (21), and the results provided the maximal activity of arginase under optimal conditions.

The preparation of the antisera to A I and A II as well as the details of the immunoprecipitation procedure have been described previously(22). The two different antibody preparations were specific for corresponding arginase isoforms, and no appreciable cross-reactivity was observed in the experiments conducted.

The l-arginine-to-l-citrulline conversion assay was used to measure NOS activity in different human breast cancer cell lines. A 20% (w/v) homogenate of different cells lines was prepared in 50 mm triethanolamine-HCl (pH 7.4) containing 0.1 mm EGTA, 0.1 mm EDTA, 0.5 mm DTT, 1μ m pepstatin A, and 2 μm leupeptin at 4°C. The homogenate was centrifuged at 20,000 × g for 60 min at 4°C, and the supernatant was used to assay NOS. NOS activity was determined by measuring the formation of[³H]citrulline from[³H]arginine (23).

The RT-PCR technique was used to assess A I and A II, eNOS, iNOS, and GAPDHgene expression in different breast cancer cell lines. Tri-Reagent(Molecular Research Center, Cincinnati, OH) was used to homogenize the cells and extract the total cellular RNA. Reverse transcription was performed using 3 μg of total RNA per sample and 50 units of Moloney murine leukemia virus reverse transcriptase, and running the reaction at 42°C for 20 min. The resulting cDNA samples were amplified using the Gene Amp RNA PCR Kit, obtained from Perkin-Elmer (Norwalk,CT) in a 100-μl reaction mixture. The final reaction mixture was heat-denatured at 94°C for 5 min and then subjected to PCR,with amplification for 35 cycles at 94°C for 1 min, primer annealing at 60°C for 1 min, and extension at 72°C for 1.5 min. The amplification procedure for GAPDH consisted of 30 cycles, each undergoing denaturation at 94°C for 30 s, primer annealing at 55°C for 30 s, and extensions at 72°C for 1 min, using 1 μg of total RNA sample.

The following primers were used (Custom Primers; Life Technologies, Gaithersburg, MD): A I, sense(5′-CTCTAAGGGACAGCCTCGAGGA-3′) and antisense(5′-TGGGTTCACTTCCATGATATCTA-3′; Ref. 24); A II, sense(5′-ATGTCCCTAAGGGGCAGCCTCTCGCGT-3′) and antisense(5′-CACAGCTGTAGCCATCTGACACAGCTC-3′; Ref. 25); eNOS,sense (5′-CCAGCTAGCCAAAGTCACCAT-3′) and antisense(5′-GTCTCGGAGCCATACAGGATT-3′; Ref. 26); iNOS, sense(5′-CATGGCTTGCCCCTGGAAGTTTCT-3′) and antisense(5′-CCTCTATGGTGCCATCGGGCATC-3′; GenBank Accession No. M84373);GAPDH, sense (5′-GTGAAGGTCGGTGTCAACGGATTT-3′) and antisense(5′-CACAGTCTTCTGAGTGGCAGTGAT-3′; Ref. 27). One hundred ng of sense and antisense primers were used in each PCR in a final volume of 100 μl. The expected sizes of amplified fragments were: 794 bp (A I), 342 bp (A II), 354 bp (eNOS), 747 bp (iNOS), and 558 bp (GAPDH). Each final PCR product sample (30 μl) was loaded on a 1.5% agarose gel, electrophoresed, and visualized by ethidium bromide staining under UV light. The identities of the bands obtained after PCR were confirmed by automated DNA sequencing (data not shown).

Cells used were HS-578T, MCF-12F, MCF-7, BT-474, MDA-MB-468, ZR-75-30,and MDA-MB-231. HS-578T, MCF-7, and MDA-MB-231 were cultured in DMEM containing 10% FBS and 10 μg/ml insulin. MCF-12F was cultured in a 1:1 (v/v) mixture of DMEM and Ham’s F12 medium containing 20 ng/ml EGF, 100 ng/ml cholera toxin, 10 μg/ml insulin, and 500 ng/ml hydrocortisone in 5% horse serum. BT-474 was cultured in RPMI 1640 with 2 mm l-glutamine adjusted to contain 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mm HEPES, 1.0 mm sodium pyruvate, 10 μg/ml bovine insulin and 2 mm l-glutamine in 10% FBS. ZR-75-30 was cultured in RPMI 1640 with 1 mm sodium pyruvate in 10%FBS. MDA-MB-468 cells were cultured in DMEM containing 10 mm nonessential amino acids, 2 mml-glutamine, 10 μg/ml insulin, and 10% FBS.

Cells (1 × 10⁶) seeded in 100-mm plates were allowed to grow overnight. To the plates, NOHA at various concentrations was added with or without l-ornithine. From day 1 to day 7, cells were collected and viability was determined on a hemocytometer by the trypan blue exclusion method. Following this, the number of viable cells at each concentration and time point was determined.

Intracellular l-ornithine and spermine concentrations were determined using modified high-performance capillary electrophoresis(28). Ornithine was run using a HP 1100 series high-performance liquid chromatograph, on a 25 ×4.6 mm C₁₈ column. The mobile phase was an alkylsulfate, and the organic modifier was run at 1 ml/min. UV detection was at 191 nm. Spermine was run using a 72-cm bubble cell capillary and was quantified by UV at 191 nm. The samples were run at 30 kV with positive pressure after a 15-kV injection.

The TUNEL assay was performed using ApoAlert DNA Fragmentation Assay Kit from Clontech (Palo Alto, CA; Ref. 29). Cells(3 × 10⁶) were collected by centrifugation, washed twice with PBS, and fixed in 1%formaldehyde-PBS at 4°C for 20 min. The cells were collected, washed with PBS, resuspended in 500 μl of PBS, and stored overnight in 70%ethanol. The cells were then collected, washed, and gently resuspended in equilibration buffer. Then nucleotide mixture and terminal deoxynucleotidyl (Tdt) transferase enzyme were added, and cells were incubated at 37°C for 1 h. Cells were washed and analyzed in a Becton Dickinson Flow Cytometer.

Caspase 3 activity in MDA-MB-468 cells treated with NOHA was analyzed using the substrate Ac-DEVD-AMC (PharMingen, San Diego, CA; Ref.30). Active caspase-3 cleaves the substrate Ac-DEVD-AMC after the aspartic acid residue and before the AMC group. The released AMC becomes fluorescent, and the fluorescence is quantified in a Versofluo fluorometer with excitation at 380 nm and emission at 440 nm.

Cells (2 × 10⁶) were plated in 100-mm plates with 15 ml of medium with or without drugs. The cells were resuspended in PBS containing 1% Triton X, 0.1 mg/ml RNase A, and propidium iodide to a final concentration of 0.05 mg/ml. The cells were subjected to FACScaliber flow cytometry, and the percentage of cells in each phase of the cell cycle was obtained using Modfit software.

Cytoplasmic extracts were prepared from cells after various treatments by lysis in buffer containing 50 mm HEPES (pH 7.5), 1 mm DTT, 150 mm NaCl, 1 mm EDTA,0.1% Tween 20, 10% glycerol, 10 mm β-glycerophosphate,1 mm NaF, 0.1 mm orthovanadate, 10 μg/ml leupeptin, 10 μg/ml aprotinin, and 0.1 mmphenylmethylsulfonyl fluoride and kept at 4°C for 30 min. Lysates (30μg) were resolved electrophoretically on 10% SDS-PAGE and electrotransferred to a polyvinylidene difluoride membrane(Bio-Rad) using a tank blot procedure (Bio-Rad Mini Protean 11). The filter was incubated with 1:1000 dilution of anti-p21 purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The washed filter was incubated with a 1:1000 dilution of antimouse horseradish peroxidase-linked F(ab)₂ fragment (Amersham Corp., Piscataway, NJ) for 1 h. Immunoreactive bands were visualized with the help of the ECL detection system (Amersham). The relative intensities of the bands were quantified by densitometric analysis (Personal Densitometer SI; Molecular Dynamics).

We initially investigated the expression and biological activity of arginase and NOS in different human breast cancer cell lines. BT-474 and MDA-MB-468 were found to have high arginase activity, whereas very low activity was present in other cell lines, including ZR-75-30,MCF-12F, MCF-7, HS-578T, and MDA-MB-231 (Fig. 1). Immunoprecipitation studies with BT-474 and MDA-MB-468 cell lines using specific antibodies for A I and A II indicated that most of the arginase activity obtained in our bioassay was due to A II (data not shown).

NOS activity was detected in the ZR-75-30, MCF-12F, HS-578T, BT-474,and MCF-7 cell lines, whereas it was not detected in the MDA-MB-231 and MDA-MB-468 cell lines. The Ca²⁺/CaM-dependent eNOS and Ca²⁺/CaM-independent NOS (iNOS)activities of the various cell lines are shown in Fig. 2 . The ZR-75-30 cell line, which was found to contain relatively high eNOS activity, also expressed iNOS. iNOS expression was very low or not detectable in any other cell line. The mRNA expression of A I, A II,eNOS, and iNOS is shown in Fig. 3 . The PCR products were of the expected lengths: A I (794 bp), A II (342 bp), eNOS (354 bp), iNOS (747 bp), and GAPDH (558 bp). The identities of these PCR products were confirmed by DNA sequencing (data not shown). Direct correlations between the mRNA expression patterns of A I and A II and their enzyme activities were observed.

We wanted to assess the possible correlation between the basal proliferation rate and the expression of arginase and NOS in these cell lines. Therefore, based on our results for arginase and NOS mRNA expression and enzyme activity, we selected four cell lines for our study: (a) MDA-MB-468, because it had high arginase activity and no NOS activity; (b) BT-474, because it had high arginase activity and low NOS activity; (c) MCF-7 because it had low arginase and low NOS activity; and (d) ZR-75-30,because it had low arginase and high NOS activity. The MDA-MB-468 cell line, with high arginase and no NOS activity, had the highest proliferation rate. ZR-75-30, on the other hand, a cell line with high NOS and low arginase, had the slowest proliferation rate. The other two cell lines BT-474 and MCF-7 had intermediate proliferation rates (Fig. 4 A).

We next assessed whether NOHA specifically affected proliferation of these cell lines. In MDA-MB-468 cells, which have high arginase activity, NOHA completely inhibited cell proliferation within 3–4 days(Fig. 4,B). A NOHA concentration of 500μ m partially inhibited the proliferation of this cell line (data not shown), whereas a concentration of 1 mm produced maximal inhibition (Fig. 4,B). We therefore used this concentration of NOHA for all further studies. When 0.5 mml-ornithine was added along with NOHA, the cell proliferation rate was similar to control cells, whereas a higher concentration of l-ornithine (1 mm) significantly increased the cell proliferation rate compared with control cells (Fig. 4,B). However, putrescine (1 μm), spermidine (1μ m), or spermine (1 μm)only partially antagonized the effects of NOHA (data not shown) in this cell line. In ZR-75-30 cells, which have very low arginase and high NOS activity, NOHA did not affect the basal proliferation rate (Fig. 4,C). In MCF-7 cell lines, which had low arginase and NOS activity, NOHA produced only a very slight decrease in the rate of cell proliferation (Fig. 4 C).

The effects of NOHA on the intracellular ornithine and spermine content in MDA-MB-468 and ZR-75-30 were assessed using high-performance capillary electrophoresis. In the MDA-MB-468 cell line, the intracellular ornithine and spermine levels were significantly reduced in the presence of NOHA. The addition of l-ornithine (1 mm) in the presence of NOHA restored the intracellular ornithine and spermine levels to normal. On the other hand, NOHA had no effect on the intracellular ornithine and spermine levels in the ZR-75-30 cell line (Fig. 5).

NOHA (1 mm) caused massive DNA fragmentation as assessed by the TUNEL assay (Fig. 6). This effect was inhibited by exogenous l-ornithine (1 mm). DNA fragmentation induced by NOHA was further confirmed by measuring the caspase-3 activity. The addition of NOHA to MDA-MB-468 cells produced a 2.2-fold induction in caspase-3 activity at 48 h (Fig. 7). This effect of NOHA was not observed in the presence of l-ornithine (1 mm), indicating that induction of caspase-3 was a result of a depletion of spermine caused by NOHA.

The effect of polyamine depletion by NOHA on the cell cycle was assessed by flow cytometry using propidium iodide-stained cells. The cells were arrested in the S phase with concomitant decrease of cells in the G₀-G₁ and G₂-M phases (Fig. 8). These effects of NOHA were not observed in the presence of l-ornithine (1 mm). NOHA had no effect on the expression of cyclin D1 and cyclin E (data not shown). However, there was an increase in the expression of p21, which was detected as early as 4 h (Fig. 9) after NOHA treatment. The addition of l-ornithine attenuated the expression of p21 induced by NOHA treatment.

Our studies indicate that HBC cells can express either arginase or NOS, or in some cases, both of these enzymes. It is,however, not known whether the functional characteristics of HBC cell lines vary depending on whether they express arginase and use arginine to synthesize polyamines, or express NOS to synthesize NO and NOHA.

It appears that both arginase and NOS can use arginine to synthesize either polyamines or NO and NOHA. The K_m for l-arginine is 2–20 mm for mammalian arginase (31) and 1–20μ m for the various NOS isoenzymes(32). However, the V_maxof arginase at physiological pH is more than 1000 times that of the NOS enzymes (32). This indicates that both enzymes, arginase and NOS, can use arginine at comparable rates at low concentrations of arginine.

Our primary objective was to assess whether the constitutive presence of arginase or NOS in HBC cell lines had any relationship to their rate of proliferation. Our results indicated that cell lines that had relatively high arginase activity had significantly higher rates of proliferation compared with those that expressed NOS. This is not surprising because increased arginase activity would lead to the increased polyamine synthesis necessary for cell proliferation and because NO has cytostatic properties on numerous cell lines (33, 34), including MDA-MB-231 cells (35).

A similar relationship between the dual pathways of arginine metabolism by intratumor macrophages has been described previously(36). Specifically, i.p. tumor rejection in P815-preimmunized mice was accompanied by an upshift in intratumor macrophage arginine metabolism to the NOS pathway that yields citrulline and NO and a decrease in arginase activity, leading to a decrease in urea and ornithine. By contrast, during progressive tumor growth in naïve mice, the metabolism of arginine by the arginase pathway to yield ornithine and urea was markedly increased,and this was accompanied by a decrease in the metabolism of arginine to NO and citrulline. Our results indicate that this phenomenon may also be applicable to some breast cancer tumor cell lines as well.

The effect of NOHA in inhibiting cell proliferation was most prominent in MDA-MB-468, which had very high arginase activity and no detectable NOS activity, suggesting that this inhibition was completely due to the exogenous NOHA added. By contrast, NOHA did not have any effect on the proliferation of the ZR-75-30 cell line. This may be due to the fact that this cell line appears to be arginase independent and also has very high NOS activity. The endogenous NOHA produced in this cell line may have inhibited the low amount of arginase present in this cell line, and therefore, exogenous NOHA did not produce any further effect. On the other hand, NOHA had only a very slight effect in inhibiting the proliferation of MCF-7 cells. This cell line has very low arginase activity but also has very little NOS activity. The amount of endogenous NOHA produced by these cells, therefore, may not have been sufficient to inhibit the arginase activity completely. This may be the potential explanation as to why exogenous NOHA produced a slight decrease in the proliferation rate of this cell line. Further studies are in progress to assess whether this may be the definite mechanism.

We next focused our studies on elucidating the mechanism by which arginase activity modulates cell proliferation, and any potential interaction between the products of the NOS pathway with those of the arginase pathway in modulating proliferation of HBC cells. We selected MDA-MB-468 for these studies because this expressed high arginase activity with no NOS activity. For comparison, we used ZR-75-30, which expressed the highest amount of NOS activity and very little arginase activity. We used NOHA, an intermediate formed during the conversion of l-arginine to NO and citrulline by NOS because there is abundant literature about the mechanisms by which NO induces cytostasis in various cell lines (33, 34). Furthermore, NOHA inhibits arginase activity in many cell lines (16, 17, 18, 37, 38, 39), but there are no reports about the effects of NOHA on the proliferation of HBC cells that express either arginase or NOS. NOHA can be formed as a stable intermediate by various types of cells both in vitroand in vivo during synthesis of NO. When macrophages(17), vascular smooth muscle cells (37), and endothelial cells (38) were stimulated in vitroto induce NOS, NOHA accumulated in the culture medium. Similarly, when lipopolysaccharide was injected into rats, there was a marked elevation in the serum concentration of NOHA in addition to nitrite/nitrate(40). It is, therefore, possible that NOHA may have an important independent physiological and pathophysiological role in modulating cell proliferation different from that of NO. NOHA inhibited the proliferation and subsequently caused apoptosis of the MDA-MB-468 cells but not the ZR-75-30 cells. l-Ornithine(500 μm) antagonized the actions of NOHA on MDA-MB-468 cells, and higher concentrations of ornithine (1 mm) further increased the rate of proliferation compared with control cells. Similarly, in the presence of NOHA, the intracellular levels of ornithine and spermine were significantly reduced, whereas addition of exogenous l-ornithine was able to restore the intracellular l-ornithine and spermine levels in a concentration dependent manner.

These findings indicate that arginase activity is essential for MDA-MB-468 cells to generate polyamines and thereby proliferate and that reduction of intracellular polyamine levels can inhibit proliferation and subsequently lead to apoptosis. These cells have the capacity to bypass this enzyme and proliferate if its product ornithine is added because they have the capacity to take up ornithine from the culture medium. Other investigators (41) using this cell line have also observed a decrease in proliferation rate followed by apoptosis after inhibition of ODC and S-adenosyl methionine decarboxylase by EGF. These two enzymes are responsible for the conversion of ornithine to polyamines. The inhibitory effects of EGF could be partially reversed by putrescine or spermidine. These data,therefore, complement our own results and indicate that MDA-MB-468 has the capacity to take up ornithine and other polyamines. The lack of any effect of NOHA on cell proliferation or the intracellular polyamine content of ZR-75-30 cells may be due to the independence of this cell line from the arginase pathway for growth. The lower polyamine levels in these cells may possibly account for the slower growth rate observed.

We next correlated the inhibition of arginase activity by NOHA in MDA-MB-468 to the control of cell cycle. Our observation that the cells are arrested in the S phase of the cell cycle, following polyamine depletion, is similar to those of other investigators who observed that 2-difluoromethyl ornithine, an inhibitor of ODC, induced polyamine depletion and increased the length of the S phase within one cell cycle after seeding Chinese hamster ovary cells in the presence of the inhibitor (42). No effect on the G₁-S transition was observed until 2 days after seeding, suggesting that 2-difluoromethyl ornithine-induced lengthening of the G₁ phase occurred later than the effect on S-phase progression (42).

Our observation that inhibition of polyamine biosynthesis by NOHA did not affect the expression of cyclin D1 or cyclin E but increased the expression of p21 is similar to those of other investigators(41) who used EGF to inhibit polyamine biosynthesis in MDA-MB-468 cells. However, these investigators did not assess the cell cycle phase in which the cells were arrested following exposure to EGF. It appears that low concentrations of p21 promote both complex formation and kinase activity (43), whereas at higher concentrations p21 inhibits kinase activity (44). There are a number of cell systems where p21 induction has been associated with cell cycle progression or proliferation. Thus, the early induction of p21 seen in MDA-MB-468 cells following exposure to NOHA may have actually accelerated the transition of cells from the G₀-G₁ to the S phase. The subsequent arrest of cells in the S phase may be due to the reduction in polyamine levels. Polyamines are essential for cell viability and play important roles in cell growth and differentiation (45, 46). Polyamines interact with cellular macromolecules, such as DNA and RNA, and alter their structure and conformation(45). Spermidine and spermine are reported to stabilize chromatin, and the polyamine-depleted cells undergo changes in chromatin and DNA structure (20, 47). Furthermore,spermine protects thymocytes against apoptosis (48). It is, therefore, not surprising that after exposure of MDA-MB-468 cells to NOHA, there was initial inhibition of proliferation and arrest of cells in the S phase, followed rapidly by apoptosis.

The effect of inhibition of polyamine biosynthesis on cell cycle proteins does not appear to be consistent and may vary with the cell type under study or with the type of agent used to inhibit polyamine synthesis. Inhibition of polyamine synthesis in T-47 D human breast cancer cells by methylglyoxal bis(cyclopentylaminohydrazone)decreased gene expression of cyclin D1 (49), unlike our observations in this study and the observations of others(41), whereas similar to our study, the cells became apoptotic (49).

Our findings that NOHA inhibits cell proliferation by inhibiting arginase activity is not specific to HBC cells; similar observations have been made using the CaCo-2 human colon cancer cell line, which also expresses arginase (39). Our observation demonstrating that NOHA, an intermediate in the biosynthesis of NO,inhibits arginase in a HBC cell line and thereby reduces polyamine levels, leading to inhibition of cell proliferation and apoptosis, may have important clinical implications. Breast tumor tissues have been reported to have 2–3-fold higher polyamine levels than surrounding normal tissue (3). Macrophages, as well as myoepithelial cells at the site of breast cancer, express NOS (9) and,therefore, have the capability to generate both NOHA and NO. Thus, NOHA and NO by independent mechanisms may affect proliferation of breast cancer cells. It would be interesting to assess whether arginase-positive breast cancers in which there is infiltration of macrophages have a different prognosis compared with those breast cancers that are arginase negative.

We thank Dr. A. Leone (Oxonon) for help with the analysis of ornithine and spermine. This work is dedicated to the memory of Rajmani Devi who passed away from cancer.