NO is a potent bioactive molecule produced via enzymatic activity of NOS. Extensive activities suggest its involvement in carcinogenesis, including DNA damage, activation of oncogenes, inhibition of DNA repair enzymes and of tumor-suppressor gene functions and inhibition or promotion of apoptosis.1, 2 iNOS is synthesized by transcriptional activation under a variety of pathophysiologic conditions and produces large amounts of NO.3, 4, 5 iNOS activity is therefore thought to play important roles in carcinogenesis.6
The p53 protein acts as a gatekeeper molecule through regulation of DNA repair, the cell cycle and apoptosis to protect cells from chromosome instability; and loss of p53 function increases susceptibility to malignant transformation.7 Various biologic interactions between p53 and NO produced by iNOS have been identified. We showed previously that a high concentration of NO can inhibit p53 function by inducing conformational changes in p53.8 Li et al.9 reported that NO-induced DNA damage and apoptosis depend on p53 status. In addition, it has been reported that transcription of iNOS can be downregulated by p53,10, 11 suggesting that functional loss of p53 protein, caused by mutation and/or loss of heterozygosity of the TP53 (human p53) gene, leads to overexpression of the iNOS gene. Clinical studies have reported positive correlations between TP53 mutation and iNOS overexpression in several types of cancer,12, 13, 14, 15, 16 and administration of endotoxin to Trp53 (mouse p53)–deficient mice resulted in increased production of NO through overexpression of iNOS compared to Trp53 wild-type mice.10
Studies of tumors in Trp53-deficient mice provide valuable information relevant to human carcinogenesis since it is now generally agreed that about half of all human cancers are related to p53 dysfunction. Trp53-deficient mice spontaneously develop lymphomas of mainly thymic origin, which possess CD4+/CD8+ double-positivity and TCR-β.17, 18 Double-positive cells that have DNA damage during the process of V(D)J recombination may be eliminated by p53-dependent apoptosis, so Trp53 inactivation might facilitate the survival of cells having malignant mutations. NO is involved in the elimination process of negative selection.19 NO produced by iNOS activity in dendritic cells could apply apoptotic pressure on thymocytes having TCR-β when they make contact with MCH-I expressed on the membrane of dendritic cells.20 NO induces apoptosis in thymocytes via a p53-dependent mechanism.21 These previous findings led us to hypothesize that Trp53-deficient thymocytes are rescued from apoptotic signals by NO produced by iNOS, resulting in prolonged survival and accumulation of DNA damage in the presence of NO. Consequently, such thymocytes can develop into thymic lymphomas.
We investigated this hypothesis by creating double-mutant mice for the Trp53 and iNOS genes. iNOS homozygous disruption caused partial inhibition of the development of thymic lymphomas in Trp53–/– mice and almost complete inhibition in Trp53+/– mice. Surprisingly, lack of the iNOS gene strongly predisposed Trp53-deficient mice for development of nonthymic T-cell lymphomas, suggesting that iNOS activity may have some protective potential against nonthymic T-cell lymphomagenesis in Trp53-deficient mice.
Generation of double-mutant mice for the Trp53 and iNOS genes
Our study was approved by the IARC Animal Use and Care Committee. Both homozygous mutants, Trp53–/– mice (C57/6J-129/Sv), a line established by Donehower et al.,22 and iNOS–/– mice (C57/6J-129/Sv),23 obtained from Jackson Laboratories (Bar Harbor, ME), were maintained in a C57BL/6 background. Trp53–/– mice were bred with iNOS–/– mice to obtain Trp53+/– and iNOS+/– offspring, which were intercrossed. Mice carrying the 9 genotypes with different combinations of iNOS and Trp53 (Trp53–/–iNOS+/+, Trp53–/–iNOS+/–, Trp53–/–iNOS–/–, Trp53+/–iNOS+/+, Trp53+/–iNOS+/–, Trp53+/–iNOS–/–, Trp53+/+iNOS+/+, Trp53+/+iNOS+/– and Trp53+/+iNOS–/–) were thus created. For genotyping the iNOS locus, primers MNO20A (5′-ACAGCCTCAGAGTCCTTCATGAAGCACATGC -3′), NEO4 (5′- CAGAAGAACTCGTCAAGAAGGCGATAGAAG-3′), MNO20C (5′-GAGG AGAGA-GATCCGATTTAGAGAGTCTTGG-3′) and MNO20D (5′-TGAAGCCATGACCTTTCGCATTAGCATGG-3′) were used.24 For genotyping the Trp53 locus, primers 10480 (5′-ATGGGAGGCTGCCAGTCCTAACCC-3′), 10681 (5′-GTGTTTCATTAGTTCCCCACCTTGAC-3′), 10588 (5′-GTGGGAGGGACAAAAGTTCGAGGCC-3′) and 10930 (5′-TTTACGGAGCCCTGGCGC-TCGATGT-3′) were used.25
Mice were killed due to bad health, and autopsies were performed to determine the cause. Mice >2 years of age were also killed. Organs were cut into 2 blocks. One was fixed with 10% buffered formalin for 12 hr at 4°C for histologic analysis. The other was frozen in liquid nitrogen and stored at −80°C for analysis of protein and RNA. After tissue fixation, sections were prepared routinely and stained with HE. Histologic diagnosis was performed blind for gene status. Lymphoma was diagnosed as thymic when tumors were macroscopically observed in the thymus and the normal microscopic structure of the thymus was destroyed and as nonthymic when no tumor was observed in the thymus, the histologic structure exhibited atrophic or normal features and lymphoma cells infiltrated organs other than the thymus.
Subtypes of lymphoma were examined immunohistochemically using anti-CD3 (Dako, Carpinteria, CA; product A0452), CD45R/B220 (BD Biosciences Pharmingen, San Diego, CA; product 01122A) and F4/80 (Serotec, Oxford, UK; product mcap497) to determine T-cell, B-cell and macrophage lineages, respectively. Deparaffinized sections were washed with PBS. Antigen activation was performed by microwaves at 750 W for 20 min in 10 mM citrate buffer (pH 6.0) when anti-CD3 and anti-CD45R/B220 antibodies were used and by trypsin digestion for macrophage detection. Endogenous peroxidase activity was blocked by incubating with 0.3% H2O2, and sections were treated with 1% BSA in PBS. The antibody–antigen complex was detected using antirabbit IgG horseradish peroxidase–labeled polymer or the LSAB2 kit (Dako). Signals were developed using 3,3′-diaminobenzidine tetrahydrochloride solution (Dako). Nuclei were counterstained with hematoxylin. The same procedure without primary antibody was carried out as a negative control.
Thymus, spleen and mesenteric lymph nodes were cut into small fragments, filtered through a stainless steel screen and washed in PBS. Single-cell suspensions were stained with FITC- and PE-conjugated antibodies and analyzed by a FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA). The following antibodies were used (BD Biosciences): FITC or PE anti-CD3e (145-2C11), -CD45R/B220 (RA3-6B2), -CD11b (M1/70), -CD4 (RM4-5), -CD8a (53-6.7), -TCR-β (H57-597), -TCR-γδ (GL3), -NK-1.1 (PK136) and CD122 (IL-2/IL-15 receptor β chain) (TM-β1). The immunoreaction was performed in Dulbecco's PBS with 2% heat-inactivated FBS and 0.09% sodium azide (Pharmingen stain buffer). Nonspecific reaction in splenocytes was blocked by preincubation with antimouse IgG (Sigma, St. Louis, MO). For most FACS plots, 10,000 events were collected; dead cells were excluded by forward scatter gating. Data were analyzed with CELLQuest software (Becton Dickinson).
Total protein homogenates were prepared from frozen tissues using lysis buffer containing 10% glycerol, 1% Triton X-100, 1 mM sodium orthovanadate, 1 mM EGTA, 5 mM EDTA, 10 mM NaF, 1 mM sodium pyrophosphate, 20 mM Tris-HCl (pH 7.9), 100 μM α-glycerophosphate, 137 mM NaCl, 1 mM PMSF, 10 μg/ml aprotinin and 10 μg/ml leupeptin. Proteins (50 μg) were separated on 8% SDS polyacrylamide gels and electrotransferred to polyvinylidene difluoride membranes (ImmobilonP; Millipore, Bedford, MA). Blots were blocked overnight at 4°C with 5% nonfat dry milk and 2% BSA in TBS-T at pH 7.5 (25 mM TRIS base, 137 mM NaCl and 0.1% Tween-20), then incubated with anti-iNOS polyclonal antibody (Transduction Laboratories, Lexington, KY). After incubation with horseradish peroxidase–antirabbit IgG antibody (Bio-Rad, Hercules, CA), immunoreactive bands were visualized with enhanced chemiluminescence reagents (Amersham, Aylesbury, UK).
iNOS localization in thymus
To determine the localization of iNOS activity, NADPH-diaphorase staining was performed.26 Each thymus was immediately fixed for 4 hr with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) and embedded in OTC mounting medium (Tissue-Tek; Miles, Elkhart, IN) before storing at −80°C. Cryostat sections (7 μm) were prepared, mounted on a glass slide and air-dried. Sections were incubated in 0.01 M PBS containing 0.1% Triton X-100 for 12 hr at 4°C. After washing with 0.01 M PBS, sections were incubated in 0.01 M PBS containing 0.1% Triton X-100, 0.5 mg/ml β-NADPH (Oriental Yeast, Tokyo, Japan) and 0.1 mg/ml nitroblue tetrazolium (Sigma) for 1 hr at 37°C. The reaction was stopped in 0.01 M PBS, and sections were photographed under bright-field illumination. Immunohistochemistry was also performed to identify iNOS protein using anti-iNOS antibody (Santa Cruz Biotechnology, Santa Cruz, CA). To examine the localization of iNOS in macrophages, immunohistochemistry using anti-F4/80 antibody (Serotec) was performed using slides stained with NADPH-diaphorase. Nuclei were counterstained with methyl green.
Total RNA was isolated from frozen tissue using a Qiagen (Chatsworth, CA) kit. First-strand cDNA was synthesized with a commercial kit (Amersham) in a reaction volume of 15 μl containing 5 μg of total RNA and 0.2 μg of random hexamer primers, according to the manufacturer's instructions. To detect cytokines, first-strand cDNA was amplified by PCR with the following sets of oligonucleotide primers: IL-2 forward, 5′-AACAGCGCACCCACTTCAA-3′; IL-2 reverse, 5′-TTGAGATGATGCTTTGACA-3′; IL-4 forward, 5′-TAGTTGTCATCCTGCTCTT-3′; IL-4 reverse, 5′-CTACGAGTAATCCATTTGC-3′; IL-6 forward, 5′-GAACAACGATGATGCACTTGCAG-3′; IL-6 reverse, 5′-CCTTAGCCACTCCTTCTGTGAC-3′; IL-7 forward, 5′-GCCTGTCACATCATCTGAGTGCC-3′; IL-7 reverse, 5′-CAGGAGGCATCCAGGAACTTCTG-3′; IL-10 forward, 5′-TCAAACA-AAGGACCAGCTGGACAACATACTG-3′; IL-10 reverse, 5′-CTGTCTAGGTCCTGGAGTCCAGCAGACTCAA-3′; IL-12(35) forward, 5′-ACCTGCTGAAGACCACAGAT-3′; IL-12(35) reverse, 5′-GATTCTGAAGTGCTGCGTTG-3′; IL-12(40) forward, 5′-CTCACCTGTGACACGCCTGA-3′; IL-12(40) reverse, 5′-CAGGACACTGAATACTTCTC-3′; IL-15 forward, 5′-GTGATGTTCACCCCAGTTGC-3′; IL-15 reverse, 5′-CACATTCTTTGCATCCAGA-3′; IL-18 forward, 5′-ACTGTACAACCG-GAGTAATACGG-3′; IL-18 reverse, 5′-TCCATCTTGTTGTGTCCTGG-3′; IFN-γ forward, 5′-TGAACGCTACACACTGCATCTTGG-3′; IFN-γ reverse, 5′-CGACTCCTTTTCCGCTTCCTGAG-3′; TNF-α forward, 5′- AACTAGTGGTGCC-AGCCGAT-3′; TNF-α reverse, 5′- CTTCACAGAGCAATGACTCC-3′; GAPDH forward, 5′-TGAAGGTCGGTGTGAACGGATTTGGC-3′; GAPDH reverse, 5′-CATGTAGGCCATGAGGTCCACCAC-3′. Samples were amplified in 28 cycles for IL-18 and GAPDH and in 35 cycles for IL-4, IL-6, IL-7, IL-10, IL-12, IL-15, IFN-γ and TNF-α (denaturation at 94°C for 30 sec, annealing at 57°C for 60 sec and extension at 72°C for 120 sec). PCR products were visualized after electrophoresis through 2% agarose gels by staining with ethidium bromide.
Statistical significance of differences in tumor spectrum was determined by the χ2 test. Mean tumor latency (survival period) was calculated by the Kaplan-Meier method, and significance was determined by the log-rank test. p < 0.05 was considered significant.
iNOS expression in Trp53-deficient mice
Constitutive iNOS expression was examined in the thymus and other lymphoid organs and compared between mice with different Trp53 gene status (iNOS+/+) by Western blotting and histologic examination. Western blotting showed that the iNOS protein level was slightly higher in the thymus of Trp53–/– and Trp53+/– mice compared to Trp53+/+ mice at ages 3 and 12 weeks (Fig. 1a). iNOS protein was expressed in the spleen of Trp53–/– mice at 3 weeks of age but not at 12 weeks. In liver, iNOS expression was undetectable in Trp53+/+ and Trp53–/– mice at both ages.
We examined the localization of iNOS in the thymus. Virag et al.27 reported that NADPH-diaphorase activity in thymus is specific for iNOS and that NADPH-diaphorase staining is sensitive and useful to identify iNOS localization in the thymus of mice. In the present study, NADPH-diaphorase staining showed that positive cells were located in the corticomedullary junction of Trp53+/+ (iNOS+/+) mice and had long branching processes (Fig. 1B), indicating that they are dendritic cells.27 Numbers of NADPH-diaphorase+ cells were increased in the thymus of Trp53+/– and Trp53–/– mice (Fig. 1B). Immunohistochemistry using anti-iNOS antibody revealed a similar localization pattern of NADPH-diaphorase positivity in Trp53–/– mice (Fig. 1B). To examine the localization of iNOS activity in macrophages, double staining for NADPH-diaphorase and anti-F4/80 antibody, which is expressed in macrophages but not in dendritic cells of the mouse thymus,28, 29 was performed (Fig. 1B). In Trp53+/+ mice, NADPH-diaphorase+ cells were not positive for F4/80 antigen (Fig. 1B). However, cells with both NADPH-diaphorase activity and F4/80 antigen were observed in the thymus of Trp53–/– mice (Fig. 1B).
Distribution of lymphocytes in thymus and spleen
To investigate the effects of iNOS on differentiation of T-cell lymphocytes, subtype distributions of T-cell lymphocytes in the thymus and spleen were examined by flow cytometry since previ-ous studies suggested involvement of iNOS activity in negative selection of thymocyte differentiation.19, 20 There was no significant difference in the distribution of lymphocyte subtypes: CD4 and CD8 in thymus or spleen between iNOS–/– and iNOS+/+ mice aged 8 weeks (Fig. 2).
Effect of iNOS gene on spontaneously developing tumors
To examine the effect of the iNOS gene on spontaneous development of tumors in Trp53-deficient mice, we monitored a total of 243 mice (41 Trp53–/–iNOS–/–, 23 Trp53–/–iNOS+/–, 21 Trp53–/–iNOS+/+, 44 Trp53+/–iNOS–/–, 49 Trp53+/–iNOS+/–, 21 Trp53+/–iNOS+/+, 25 Trp53+/+iNOS–/– and 19 Trp53+/+iNOS+/–). They revealed neither deformity nor growth disadvantages. There were no significant differences in either the macro- or the microscopic appearance of the thymus, spleen and liver (data not shown). Histologic diagnoses of tumors that developed spontaneously are summarized in Table I, and the survival period (weeks) is shown in Figure 3 and Table II. Although the iNOS gene status did not affect the latencies of lymphomas in Trp53–/– or Trp53+/– mice, homozygous disruption of the iNOS gene significantly changed the tumor spectrum of lymphomas. The incidence of thymic lymphomas was highest [61.9% (13/21)] in Trp53–/–iNOS+/+ mice and decreased to 39.0% (16/41) in Trp53–/–iNOS–/– mice, though the difference did not reach statistical significance (p = 0.087). The mean latency of thymic lymphoma development was unchanged among mice with different iNOS gene status. The incidence of thymic lymphomas was significantly (p < 0.05) decreased from 19.0% (4/21) in Trp53+/–iNOS+/+ to 2.3% (1/44) in Trp53+/–iNOS–/–, and the mean latency was longer in Trp53+/–iNOS–/– than in mice with any other iNOS gene status. However, the incidence of nonthymic lymphomas showed a trend (p = 0.089), increasing from 28.6% (6/21) in Trp53–/–iNOS+/+ mice to 51.2% (21/41) in Trp53–/–iNOS–/– mice; and the mean latency was significantly (p < 0.05) shorter in Trp53–/–iNOS–/– mice. In Trp53+/– mice, the incidence of nonthymic lymphomas increased from 42.9% (9/21) in iNOS+/+ to 75.0% (33/44) in iNOS–/– (p < 0.05), though the mean latency was not significantly different among mice having different iNOS gene status.
For sarcomas, total incidence rates in Trp53–/– mice were 28.6% (6/21), 17.4% (4/23) and 26.8% (11/41) in iNOS+/+, iNOS+/– and iNOS–/–, respectively, and the corresponding rates in Trp53+/– mice were 47.6% (10/21), 40.8% (20/49) and 45.5% (20/44). Differences among mice with different iNOS gene status were not significant (p = 0.628 and 0.839 in Trp53–/– and Trp53+/– mice, respectively). Regarding subtypes of sarcoma, the respective tumor numbers of soft tissue sarcoma and osteosarcoma in Trp53+/– mice were 7 and 3 in a total of 21 iNOS+/+, 19 and 1 in a total of 49 iNOS+/– and 14 and 8 in a total of 44 iNOS–/– mice, respectively. The rate of osteosarcoma appeared to be low in iNOS+/– compared to iNOS+/+ and iNOS–/– mice.
Characteristics of lymphomas
The effects of iNOS gene status on phenotypes of lymphoma were examined by flow-cytometric analysis and histologic examination. It was previously reported that thymic lymphomas in Trp53-deficient (iNOS wild-type) mice displayed a CD4+CD8+ double-positive and TCR-β+ phenotype,17, 18 and in the present study thymic lymphomas in Trp53–/–iNOS–/– mice had the same phenotype (Fig. 4A). It has been reported that nonthymic lymphomas in Trp53-deficient mice with a C57BL/6 background are of B-cell lineage derived from spleen (marginal zone) or lymph node origin and usually do not infiltrate other organs aggressively until a late stage.17, 18 In our study, nonthymic lymphomas that developed in Trp53-deficient (iNOS+/+) mice showed similar histologic characteristics to the previous observations [Fig. 4B (lymph node and marginal zone lymphoma of spleen)] and were strongly positive for B-cell marker (Fig. 4B). In contrast, those in Trp53–/–iNOS–/– mice had a very different macroscopic and microscopic appearance; although the thymus showed an atrophic appearance or normal structure, lymphoma cells aggressively infiltrated the lung and liver (Fig. 4b). Severe hepatomegaly was always observed in Trp53–/–iNOS–/– and Trp53+/–iNOS–/– mice, even when splenomegaly was mild. Mesenteric lymph node adenopathy was less common. Lymphoma cells infiltrated into the portal triad and sinusoid space of the liver and stained more strongly with the anti-CD3 T-cell marker than with the B-cell marker in 6 of 6 mice tested immunohistochemically (Fig. 4B). Flow-cytometric analysis revealed that they were CD4+CD8–TCR-β+TCR-γδ– (Fig. 4A). Staining for NK-1.1 and IL-12R was negative (data not shown). In Trp53+/–iNOS–/– mice, half (3/6) of the nonthymic lymphomas tested stained dominantly with the T-cell marker and the others stained dominantly with the B-cell marker. Moreover, in Trp53+/+iNOS–/– mice, all (5/5) of the lymphomas tested stained dominantly with the B-cell marker rather than the T-cell marker, as observed in follicular diffuse large B-cell lymphomas, which are common in retired inbred C57BL/6 mice (Fig. 4B).30
Cytokine expression profiles
Cytokine expression profiles were examined in the spleens of mice having no lymphomas, as confirmed by histologic examination. Mice with Trp53–/–, Trp53+/– and Trp53+/+ genotypes (groups aged 16–24, 72–86 and 107–121 weeks for each genotype) were compared. Representative results from duplicate independent experiments with 2 or 3 mice are shown in Figure 5. IL-10, IL-18, IFN-γ and TNF-α were expressed in double-mutant mice for the Trp53 and iNOS genes. In addition, expression of these cytokines was associated with that of Bcl-2 mRNA.
Spontaneous lymphomas that develop in Trp53-deficient mice have been well characterized. The spectrum of spontaneous tumors depends on the background of the mice.18 A previous study showed that mice with 75% C57BL/6 and 25% 129/Sv background developed lymphomas and sarcomas (osteosarcoma+soft tissue sarcoma) (65% lymphomas vs. 30% sarcomas in Trp53–/– mice and 30% lymphomas vs. 40% sarcomas in Trp53+/– mice).18 In the present study, Trp53–/–iNOS+/+ and Trp53+/–iNOS+/+ mice developed lymphomas at higher incidence rates (almost 90% and 60%, respectively) compared to the previously published results. All of the Trp53 and iNOS mutant mice used in our study had the same background (C57BL/6/129Sv), and we maintained them with C57BL/6, leading to the high incidence of lymphomas since the C57BL/6 strain is highly prone to spontaneous lymphomas.30
Our study confirmed previous findings showing constitutive expression of iNOS in dendritic cells of the thymus in both Trp53 wild-type and -deficient mice. We also demonstrated increased iNOS expression in dendritic cells and macrophages of the thymus of Trp53-deficient mice. Thymocyte-derived IFN-γ induces iNOS expression in stromal cells,27 and the p53-dependent negative feedback loop plays an important role in the regulation of iNOS gene expression in somatic cells.10, 11 Thus, disruption of the p53-dependent negative feedback loop may increase the level of constitutive iNOS expression in the thymus.
Absence of the iNOS gene reduced the development of thymic lymphomas by 40% and 90% in Trp53–/– and Trp53+/– mice, respectively. Our results suggest that iNOS activity may enhance the development of thymic lymphomas. However, iNOS gene disruption increased the incidence of nonthymic lymphomas in Trp53+/– mice and shortened the latencies of their development in Trp53–/– mice. In addition, it has been reported that about 25% of lymphomas that developed in Trp53–/– and Trp53+/– mice were nonthymic with B-cell markers, and these did not exhibit aggressive behavior and metastasis to the liver until later stages.17, 18 However, in the present study, nonthymic lymphomas that developed in double-mutant (Trp53–/–iNOS–/–) mice aggressively infiltrated the liver. These lymphomas were frequently of T-cell lineage, suggesting that T cells, but not B cells, are susceptible to malignant transformation in mice lacking the Trp53 and iNOS genes. Furthermore, their phenotype was revealed by flow cytometry to be CD4+CD8–TCR-β+. These results suggest that the presence of the iNOS gene has a protective effect against malignant transformation of peripheral T-cell lymphocytes.
Although the precise biologic mechanisms for such a protective effect of iNOS against peripheral lymphomagenesis are unknown, the following explanations appear plausible. NO could be involved in the regulation of lymphocyte proliferation and modulation of differentiation after immune stimulation31 since iNOS suppresses lymphocyte proliferation in the spleen,32, 33 suggesting that NO might inhibit malignant transformation of peripheral lymphocytes. In addition, our results suggest that NO may mediate cytokine regulation. Deregulation of cytokine production plays a central role in lymphomagenesis,34 and p53 regulates anti- or proinflammatory cytokines in rheumatoid arthritis and cancer cell lines.35, 36 p53 functional loss can lead to autocrine IL-6 production.35, 36 Ohkusu-Tsukada et al.37 showed that accumulation of memory T cells (CD4+) was spontaneously accelerated and Th2 cytokines such as IL-4, IL-6 and IL-10 were strongly induced by antigen stimulation in Trp53-deficient mice. In our study, double-mutant (Trp53–/–iNOS–/–) mice developed nonthymic T-cell lymphomas and overexpressed Th2 cytokines, particularly IL-10, in the spleen. Previous studies demonstrated that disruption of the iNOS gene upregulated IL-10 expression.38, 39 The results obtained in the experiment using peritoneal macrophages from mice with different combinations of the Trp53 and iNOS genes showed the highest and most prolonged IL-10 mRNA expression in mice lacking both genes upon stimulation with lipopolysaccharide and IFN-γ (unpublished data). These findings suggest that iNOS activity might have an inhibitory effect on cytokine production. In addition, we found that expression of the cytokines was associated with Bcl-2 expression. Several studies have demonstrated that IL-10 induces Bcl-2 expression in various hematopoietic cells, greatly increasing their resistance to apoptotic stimuli, and overexpression of Bcl-2 has been associated with lymphomagenesis.40, 41 Cohen et al.42 showed that IL-10 rescues T cells from apoptotic cell death via upregulation of Bcl-2. Moreover, expression of IL-10 is elevated not only in patients with higher susceptibility to B-cell lymphoma as a result of HIV or hepatitis C virus infection43, 44 but also in asymptomatic HTLV-1 carriers and patients with T-cell lymphomas.45, 46, 47 Taken together, our results suggest that deregulation of IL-10 by both functional loss of p53 and lack of iNOS activity plays an important role in nonthymic lymphomagenesis.
The enhanced expression of IL-18, IFN-γ and TNF-α in Trp53–/–iNOS–/– mice that we observed suggests another mechanism for the increase of nonthymic lymphoma in Trp53–iNOS double-mutant mice. IL-18 has preventive activity against the development of murine T-cell leukemia/lymphoma EL-4 in syngeneic mice.48 IL-18 strongly induces iNOS in natural killer cells via induction of IFN-γ and TNF-α.49 However, activation of IL-18 is downregulated by iNOS activity because IL-18 is activated by IL-1β-converting enzyme, which is inhibited by S-nitrosylation by NO.50 These findings suggest that the protective effects of IL-18 against lymphomagenesis may be mediated by high levels of NO production via iNOS activity. Consequently, iNOS gene disruption might increase IL-18 expression and nonthymic lymphomagenesis in Trp53-deficient mice.
In conclusion, the role of iNOS activity in lymphomagenesis appears to differ between lymphoid organs. Our data suggest a causal association between iNOS activity and lymphomagenesis in the thymus of Trp53-deficient mice but that inhibition of iNOS activity could accelerate lymphomagenesis in nonthymic organs under conditions of p53 functional loss. Since nonthymic lymphomas of mice mimic human lymphomas histologically,51 our findings suggest that inhibition of iNOS activity by specific inhibitors might increase T-cell lymphomagenesis in humans.
We thank Ms I. Gilibert, Ms. M.-P. Cros and Ms. N. Lyandrat for technical assistance; Dr. J. Cheney for editing the manuscript; and Ms P. Collard for secretarial assistance.