Phenotypic Profiling of Engineered Mouse Melanomas with Manipulated Histamine Production Identifies Histamine H2 Receptor and Rho-C as Histamine-Regulated Melanoma Progression Markers

Malignant melanoma is a life-threatening skin neoplasm characterized by rapid progression, poor prognosis, and continuously growing incidence in the world. Given the severity of the public health problem represented by melanoma, and the fact that an efficient therapeutic approach against advanced-stage melanomas is still lacking, during the last decades, a huge amount of research has been done on the molecular mechanisms underlying melanoma progression. This study focuses on the functional consequences of the endogenous histamine production of melanoma tumors (1).

There is a large body of experimental evidence indicating the existence of massive up-regulation of l-histidine decarboxylase (HDC) activity along with the increase of several tumors such as colon (2) and pancreatic carcinomas (3), gastric (4) and breast cancers (5), and melanoma (6, 7). HDC is the only enzyme responsible for the biosynthesis of the allergic mediator, histamine (8). Interestingly, histamine is not only produced by melanoma cells, but secreted into their environment as well (7). Furthermore, there is a broad spectrum of histamine receptors present on melanoma cells (histamine H1, H2, H3 receptors; refs. 9–11); however, the actual impact of histamine-controlled signaling loops on melanoma progression is not yet fully understood. Although histamine-mediated signals have been shown to be implicated in tumor growth (12–17), local (18–20), and systemic immunomodulation (21), a detailed unifying concept explaining the actual importance of this phenomenon is still unavailable. Finally, it should be noted that enhanced histidine catabolism in tumors is not a unique phenomenon, in as much as similar perturbations in the metabolism of several amino acids, such as massive degradation of tryptophan by the indoleamine 2,3-dioxygenase (22), or enhanced hydrolysis of arginine by arginase (23), have already been associated with the process of tumor progression.

In this study, using an in vivo approach, an attempt was made to identify the aspects of melanoma progression influenced by histamine. Mammalian expression vector systems coding for either the full-length HDC sense mRNA sequence, a mock sequence, or an antisense HDC mRNA segment, were introduced in the mouse melanoma cell line B16-F10, thus generating three types of melanoma cells with up-regulated, unmodified, and down-regulated histamine production, respectively. In order to avoid potential bias caused by random genomic insertion, all transfection procedures were done in triplicates, and nine B16-F10 variants were introduced in the experimental phase of the study. The nine variants were grafted in syngeneic C57BL/6 mice, and the progression profile of the arising tumors was compared both by traditional methods examining primary tumor growth rate or simulated metastatic lung colonization, and by techniques allowing the description of the molecular background of emerging phenotypical differences. To search the mRNA level progression profile in order to find melanoma progression markers affected by the manipulation of histamine secretion, three novel RNase protection assay (RPA) template sets were constructed measuring the expression levels of 21 well-known, or recently proposed markers coupled with different areas of tumor progression. Finally, initial conclusions drawn from mRNA level progression profiling were validated at the protein level via Western blotting.

Animals. Eight- to 12-week-old specific pathogen-free female C57BL/6 mice were purchased from the National Institutes of Oncology (Budapest, Hungary), and kept in our animal health care facility with food and water available ad libitum.

Cell culture. B16-F10 cells (American Type Culture Collection, Manassas, VA) were cultured in high-glucose DMEM, in the presence of 2% glutamine, 10% FCS, and 160 μg/mL gentamicin in a humidified 5% CO₂ atmosphere at 37°C.

Generation of B16-F10 variants with different capacities of histamine secretion. Templates for forced expression of the full-length sense murine HDC open reading frame (1,989 bp), and a partial antisense HDC segment (1,004 bp) were generated by high-fidelity PCR amplification of a plasmid construct encoding the full-length mouse HDC cDNA (kind gift from Takehiko Watanabe, Sendai, Tokyo, Japan), using appropriate linker primers (sense HDC template forward primer, GTTCGAGGTCTCACATGATGGAGCCCTGTGAATACC; reverse primer, ATATTTGGCGCGCCCTCTACACCATGGCCTCCAG; antisense HDC template forward primer, GGAGCTGGTCTCACATGTCAGCTTTTTGAAGGCACCT; reverse primer, ATTAGGCGCGCCCATCGAGTACGCTGACTCCT). Amplicons were double-digested with BsaI and AscI (New England Biolabs, Beverly, MA), and incorporated in pTriEx 1.1 Hygro expression vectors (Novagen, San Diego, CA). JM109 E. coli cells (Promega, Madison, WI) were transformed with the constructs, and PCR-screened clones were subjected to LPS-free plasmid isolation using an UltraMobius 1000 Plasmid kit (Novagen). B16-F10 mouse melanoma cells were transfected using the GeneJuice transfection reagent (Novagen) according to the manufacturer's instructions. Designated B16-F10 HDC-A1, -A2, -A3, B10-F10 HDC-M1, -M2, -M3, and B10-F10 HDC-S1, -S2, -S3: nine different novel B16-F10 variants, expressing the partial antisense HDC segment, the unmodified vector sequence (mock), or the full-length HDC sense open reading frame, respectively, were generated by 4 weeks of hygromycin B (Calbiochem, La Jolla, CA) selection (1 week at 200 μg/mL, and 3 weeks at 400 μg/mL).

Primary skin tumor model. Stably transfected B16-F10 cells (2 × 10⁵) were cultured for 1 week in the absence of hygromycin B, collected in a volume of 50 μL PBS per animal, and injected s.c. in the shaved backs of C57BL/6 mice using Hamilton LT 710 syringes (Hamilton, Bonaduz, Switzerland). At 6, 8, 11 (in one experiment, 10), 13, and 15 days after graft implantation, the longest and shortest radius (a and b, respectively) of the tumors was determined with a microcaliper, and tumor size was calculated assuming ellipsoidal tumor growth (4/3 ab² Π-formula). Three independent experiments were carried out comparing one HDC antisense-, mock-, and HDC sense-transfected B16-F10 variant in each, using animals in groups of 8 to 10. At 15 days after grafting, all mice were sacrificed, tumors and selected peripheral lymph nodes (inguinal, axillary, and brachial) were excised and frozen at −80°C for subsequent RPA studies (in two grafting experiments), or at −20°C for Western blotting (in one experiment).

Metastatic lung colonization model. C57BL/6 mice were inoculated with 2 × 10⁵ transfected B16-F10 cells per mouse, as described above, in the tail vein of the animals. At 14 days after inoculation, mice were sacrificed, necropsied, and the lungs were analyzed counting the metastatic colonies under a dissection microscope. Three independent experiments were carried out comparing one HDC antisense-, mock-, and HDC sense-transfected B16-F10 variant in each, investigating mice in groups of 8 to 10.

Construction of RNase protection assay template sets. RPA template sets were generated as described elsewhere (24). Briefly, templates for in vitro transcription of RPA-probes were generated from 1 μg of total RNA, derived from appropriate tissue samples, via RT-PCR-based extension of template-specific primers (Supplemental Materials). Amplified templates were ligated in pGEM T vectors (Promega), and the constructs were transformed into JM 109 E. coli cells (Promega). Clones harboring inserts in antisense template orientation were identified by PCR, and the chosen constructs were re-isolated with a Qiagen Plasmid Midi Kit (Qiagen, Chatsworth, CA).

RNA isolation. Total RNA isolation was carried out using the guanidinium thiocyanate-phenol-chloroform extraction method, as described (25). RNA yield was determined by spectrophotometry.

l-Histidine decarboxylase semiquantitative reverse transcription PCR. One microgram of total RNA isolated from stably transfected B16-F10 cells was reverse-transcribed using random 6-mer primers (Promega) and the Reverse Transcription system (Promega). cDNA aliquots were then amplified by Taq DNA polymerase (Promega) and mouse HDC- or glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-specific primers (HDC forward primer, CAGTACCTGAGCACTGTGCG; HDC reverse primer, CCAGCATGTCTCCTAGCAGG; GAPDH forward primer, ACCACAGTCCATGCCATCAC; GAPDH reverse primer, TCCACCACCCTGTTGCTGTA), using the following thermal program: 95°C 5 minutes, 95°C 30 seconds, 55°C 30 seconds, 72°C 1 minute-cycling, 72°C 5 minutes, for 35 and 25 PCR cycles, respectively. PCR products were visualized on agarose gels by ethidium bromide staining.

l-Histidine decarboxylase real-time PCR. cDNA templates for HDC real-time PCR were prepared and introduced in PCR reactions as described above. A mouse HDC-specific TaqMan probe set, designed to amplify mouse HDC cDNA, but not the expressed antisense HDC segments, was constructed with the help of Primer Express 2.0 software (Applied Biosystems, Foster City, CA). HDC forward primer, TCTGTGCAACGTTAGGGACTACTG; HDC reverse primer, AAAGGCCGTGCCTGCATA; 5′ 6FAM-labeled MGB TaqMan probe CCATCTGTGCCAGTGAG; amplification protocol, 95°C 10 minutes, 95°C 15 seconds, 60°C 1 minute-cycling, 40 cycles. A mouse GAPDH-specific TaqMan probe set was obtained from Applied Biosystems via the “Assays on Demand” service. Real-time primer extension was done on an AbiPrism 7000 thermal cycler (Applied Biosystems). Normalized HDC signal levels were calculated using the comparative Ct (ΔΔC_T) method, and expressed in percentages of the respective GAPDH housekeeping level.

RNase protection assay. RPAs were carried out on isolated RNA samples as described earlier (24). Briefly, RPA template sets were linearized with SalI (Promega), and used to synthesize ³³P-labeled RPA probes by the RiboQuant In vitro transcription kit (BD PharMingen, San Diego, CA) in the presence of 300 mCi [α-³³P]UTP (Institute of Isotopes, Budapest, Hungary). RPA was carried out on 10 μg of total RNA per tissue sample using RiboQuant RPA kits (BD PharMingen). Specific signals were detected by BAS-MS 2340 Imaging plates (Fuji, Nakanuma, Japan) and a FLA-3000 phosphoimager (Fuji). Signal evaluation was done by the Aida software (Raytest, Straubenhardt, Germany), mRNA expression levels were calculated as the relative percentage values of the L32 housekeeping gene expression.

Protein isolation. For protein isolation, tissue samples were lysed in a buffer containing 10 mmol/L Tris-HCl (pH 8.0), 10 mg/mL leupeptin, 0.5 mmol/L EGTA, 2% NaF, 1% Triton X-100, 25 mmol/L PMSF and 2% Na-orthovanadate. Cellular debris was removed by centrifugation, supernatants were collected, and the yield was determined by spectrophotometry.

Western blotting. For Western blotting, 10 μg heat-denatured, β-mercaptoethanol-treated protein samples were loaded on denaturing SDS-PAGE gels. After blotting to a polyvinylidene difluoride membrane (Bio-Rad, Richmond, CA), blots were probed with primary rabbit anti-rat histamine H2 receptor (1:2,000, Alpha Diagnostic International, San Antonio, TX), rabbit anti-human rho-A-B-C (1:400, Sigma-Aldrich, St. Louis, MO), goat anti-mouse matrix metalloproteinase-2 (MMP-2; 1:200, R&D Systems, Minneapolis, MN), or rat anti-yeast α-tubulin (1:4,000, Serotec, Kidlington, United Kingdom) antibodies, as stated. Then, secondary goat anti-rabbit IgG-horseradish peroxidase (1:2,500, Promega), rabbit anti-goat IgG-HRP (1:40,000, Sigma-Aldrich), or rabbit-anti rat IgGκ and λ chain-HRP (1:10,000, Sigma-Aldrich) antibodies were applied, as appropriate. Relative protein expression levels were calculated based on X-ray film densitometry data (ChemiImager 5500 software, Alpha Innotech), after normalization with the respective housekeeping α-tubulin signals, and expressed in percentages of the mean of the control group.

Histamine ELISA. Hygromycin B-selected, stably transfected B16-F10 cells (5 × 10⁵), were seeded on 25 cm² tissue culture flasks and cultured for 72 hours (∼80-90% confluency). Thereafter, 2 × 10⁶ cells, and 1 mL cell-free culture supernatant were harvested. Cells were washed in PBS, treated with 200 μL 5% trichloroacetic acid solution, centrifuged, and the debris-free cell lysates were neutralized with an equivalent volume of 0.3 mol/L NaOH solution. Finally, both cell lysates and culture supernatants were subjected to a histamine ELISA (Immunotech, Westbrook, ME) following the manufacturer's instructions.

Statistics. Statistical analyses were done using the SigmaStat software package (SPSS, Inc., Chicago, IL). One-way ANOVA, Kruskal-Wallis one-way ANOVA on ranks, two-way ANOVA, and two-way repeated measures ANOVA were used as stated, together with Tukey's all pairwise multiple comparison as post hoc test. Unless otherwise stated, all materials were purchased from Sigma-Aldrich.

Altered levels of l-histidine decarboxylase gene expression and histamine production in stably transfected B16-F10 variants. Like other melanoma cell lines, mock-transfected B16-F10 was found to express HDC and to secrete histamine per se under in vitro conditions. Sense HDC-transfected B16-F10 variants (HDC-S1, -S2, and -S3) exhibited markedly elevated levels of HDC mRNA in comparison with their mock-transfected counterparts (B16-F10 HDC-M1, -M2, and -M3) in both semiquantitative reverse transcription PCR and real-time PCR experiments (Fig. 1A and B). On the other hand, forced expression of an antisense HDC-RNA segment in the B16-F10 HDC-A1, -A2, and -A3 variants resulted in slightly diminished HDC mRNA levels (Fig. 1A and B). Results consistent with these observations were provided by histamine ELISA, showing that levels of both intracellular histamine production and histamine secretion were successfully altered in the engineered B16-F10 cells (Fig. 1C and D).

Markedly enhanced in vivo tumor growth rate and moderately elevated metastatic lung colonization potential of B16-F10 variants with higher levels of tumor histamine secretion. Primary skin tumor growth rate was analyzed by grafting the developed nine B16-F10 variants into the skin of syngeneic C57BL/6 mice. Significant and reproducible correlation was found between levels of neoplastic histamine secretion and tumor growth in three independent experiments (P < 0.001, P < 0.001, and P < 0.001; two-way repeated measures ANOVA). In comparison with mock-transfected cells, sense HDC-transfected B16-F10 variants with up-regulated histamine secretion exhibited increased tumor growth rate in all experiments (P = 0.002, P = 0.011, and P < 0.001; Tukey's test). In contrast, no significant difference was found between melanoma variants harboring antisense HDC expression vector constructs and mock vectors (Fig. 2A-C).

Moreover, neoplastic histamine secretion was found to have some limited impact on the experimental metastatic lung colonization as well. Although antisense HDC-transfected B16-F10 cells exhibited diminished metastatic colony formation in all individual experiments, no reproducible trends emerged between the mock- and sense HDC-transfected groups. Furthermore, because of the broad overlaps between all three groups, statistical analysis of the three individual experiments did not reveal consistently significant differences between them (P = 0.045, P = 0.400, and P = 0.112; Kruskal-Wallis one-way ANOVA on ranks, data not shown). However, a metaanalysis of all three independent studies identified a moderate histamine-mediated effect (P = 0.010; two-way ANOVA, Fig. 2D). According to this, melanoma grafts with antisense RNA-suppressed histamine secretion exhibit moderately diminished metastatic lung colony formation potential in comparison with mock-transfected cells (P = 0.037; Tukey's test, Fig. 2D). However, there was no significant difference between mock-transfected B16-F10 melanomas and their sense HDC-transfected counterparts.

RNA level tumor progression profiling of engineered B16-F10 melanoma grafts with different levels of histamine production. Three novel RPA template sets were constructed in order to identify melanoma progression markers related to changes in the local histamine levels, and thus possibly being affected by histamine-mediated signals. Two sets, designated ME-1 and -2, covering the progression marker profile of established primary melanoma tumors in the skin, were applied to analyze skin samples derived from tumor-grafted C57BL/6 mice. Markers associated with malignant proliferation [proliferating cell nuclear antigen (PCNA), Ki-67], angiogenesis [vascular endothelial growth factor-A (VEGF-A), basic fibroblast growth factor (bFGF), CD31, and VEGF-C], invasive matrix remodeling [MMP-2, cathepsin-B, urokinase-type plasminogen activator (uPA)], motility (rho-C), metastatic potential (MMP-2, rho-C), antigen-presenting cell (APC) maturation (B7-2), cytotoxic T cell infiltration (CD8β), and immune activation [interleukin (IL)-2 and IL-2Rα] were investigated, besides markers for histamine signaling [HDC, histamine H1 receptor (H1R), H2 receptor (H2R)] in the local tumor environment (see Supplemental Material for details). The third RPA template set, ME-3, was used in a simultaneous screen of lymph node samples isolated from the same animals in order to find evidence for a possible peripheral immunomodulation caused by melanoma histamine secretion. Markers known, or recently proposed to be coupled with mature APC immigration (B7-2), efficient antigen presentation (CD40L), clonal expansion of successfully activated T cells (IL-2 and IL-2Rα), support for Th1-type helper T cell commitment (LIGHT), support for effector T cell maturation and emigration (OX40L), APC licensing for Th-independent cytotoxic T cell activation (CD40L and 4-1BBL), and cytotoxic T cell expansion (CD8β) were investigated in the peripheral lymph nodes of the tumor-bearing animals (see Online Supplemental Material). In both approaches, respective samples, isolated from untreated healthy animals, served as controls. After cloning of the three RPA-template sets (Fig. 3A), a series of pilot RPA experiments was carried out, which showed the ability of the chosen approach to efficiently scan and semiquantitatively evaluate the expression patterns of up to 10 mRNA markers simultaneously in about 20 different samples (Fig. 3B). Subsequently, a comprehensive RPA-based progression marker-profiling was conducted on the collected skin and lymph node samples. The RPA system could detect and measure 17 out of the 21 chosen melanoma progression markers (Table 1). Judged by two-way ANOVA, the detected markers were assigned to different functional groups according to their expression patterns observed in two independent grafting experiments (Table 1; Fig. 4).

Besides HDC itself, the target of the transgenic manipulation (P < 0.001; two-way ANOVA), only two markers were reproducibly and dose-dependently affected by the transgenic modulation of the tumor HDC gene expression. Both of them showed positive correlation with melanoma histamine production, and both exhibited highly significant differences between the investigated experimental groups: histamine H2 receptor and rho-C (P = 0.006 and P < 0.001, respectively; two-way ANOVA).

Histamine H2 receptor mRNA was detected at the highest levels in sense HDC-transfected B16-F10 tumors with up-regulated histamine production, followed by moderate H2 receptor gene activity in mock-transfected melanomas, and significantly diminished H2R-signals in antisense HDC-transfected melanomas (P = 0.016; Tukey's test, Fig. 4E). Second, B16-F10 tumors with suppressed histamine production exhibited a significantly decreased rho-C gene activity as opposed to tumors with unmodified or up-regulated histamine secretion (P = 0.013 and P = 0.020, respectively; Tukey's test, Fig. 4F). Although the mean rho-C level was somewhat higher in tumors with high-level histamine production, this latter difference was not significant.

Protein level validation of histamine-affected B16-F10 melanoma progression markers. Histamine H2 receptor protein was detected at 60 kDa, and was found to be expressed in practically all investigated samples, including the control skin (Fig. 5A). In accordance with the results of RPA studies, significant differences were found between the groups investigated (P < 0.001, one-way ANOVA). There were only low H2R levels present both in healthy control skin samples and antisense HDC-transfected tumors, whereas a significant up-regulation was detected in mock-transfected tumors (P < 0.001, Tukey's test), followed by a further increase of the H2R protein amounts in the sense HDC-transfected B16-F10 melanomas (P = 0.022, Tukey's test; Fig. 5B). After all, there was a clear positive correlation between melanoma histamine production and histamine H2 receptor expression.

With respect to the rho protein levels, rho-A, -B, -C-specific antibodies identified the three closely related rho-isoforms as double bands at 27 and 24 kDa (Fig. 5A). Rho was found to remain almost undetectable in the healthy skin, but was strongly expressed in all analyzed tumor samples (P < 0.001, one-way ANOVA). In agreement with the RPA data, marked differences were observed between the different tumor groups, as both mock- and sense HDC-transfected tumors exhibited significantly higher levels of rho than antisense HDC RNA-treated B16-F10 melanomas did (P = 0.016 and P = 0.002, respectively, Tukey's test; Fig. 5B). Although in a similar trend, sense HDC-transfected tumors expressed slightly higher rho levels than their mock-transfected counterparts, this latter difference was found to be completely insignificant.

Finally, a third additional marker (MMP-2), was included in the protein level validation phase, as well. MMP-2 served as a methodologic control assessing the limits of our screening strategy, which actually extrapolated the RNA level results to the protein level. In fact, MMP-2 protein levels could not match the respective mRNA levels. Although both the inactive 74 kDa proenzyme form and the processed active 65 kDa MMP-2 could be detected by Western blotting, there was an apparent discrepancy between RNA (Fig. 4A) and protein level MMP-2 results (Fig. 5A). MMP-2 proteins were almost evenly expressed in the different samples, and no significant difference emerged between any of the investigated groups regardless of whether total, or only the newly synthesized proenzyme MMP-2 protein expression patterns, were matched against each other (P = 0.360 and P = 0.312, respectively, one-way ANOVA; (Fig. 5), detailed results not shown).

This study was inspired by the recent recognition of the complicated interplay, and the numerous regulatory circuits shared between sustained inflammatory processes, and several steps of tumor progression (26). As melanoma tumors are known to overexpress HDC, thereby releasing large amounts of histamine, an inflammatory mediator in their microenvironment, the clarification of the consequences of this phenomenon is of imperative importance. Although in previous in vitro studies, we and others have repeatedly suggested a possible involvement of neoplastic histamine secretion in melanoma tumor progression (15, 27–29), there were only sparse in vivo data available, providing rather indirect evidence for this concept (30).

Using a murine in vivo tumor model system, we were able to show that the neoplastic acquisition of autonomously regulated histamine production might be a significant benefit for progressing melanomas. We were able to show that histamine-mediated signals confer accelerated growth and moderately enhanced metastatic colonization capacity to experimental murine melanoma tumors. In order to identify the molecular events behind these complex phenotypic changes, the melanoma progression profile complex was dissected, and the individual aspects of melanoma progression, such as proliferation, invasivity, angiogenesis, etc., were replaced with molecular progression marker clusters. Two progression factors, histamine H2 receptor, and rho(-C) were found to be significantly, reproducibly, and dose-dependently affected by neoplastic histamine production both at the mRNA and the protein level.

H2R represents a histamine receptor frequently expressed by melanoma cells, which has been repeatedly suggested to support the autonomous growth of several malignancies (13–15, 17). Our data show that melanoma histamine secretion enhances the availability of H2Rs in the tumor mass, thus, theoretically further fostering its own growth-supporting effect. This in accordance with the correlation between the effects of melanoma histamine secretion exerted on tumor growth, and on H2 receptor expression. These observations give further support for the long-believed existence of histamine-controlled, H2R-mediated autocrine signaling circuits favoring the process of melanoma progression (7, 31). However, the conclusion that enhanced accessibility of H2Rs on melanoma cells can be the sole explanation for the enhanced melanoma growth, might be oversimplified and misleading. This is because there is a remarkable synergism between the suggested consequences of the up-regulation of H2Rs displayed by melanoma cells, and the effects of the up-regulation of H2Rs present on activated leukocytes, which typically infiltrate developing melanoma tumors. It is well accepted that both H2R-activation on several leukocytes (20, 32), and massive histamine releases in dermal environments (33), could cause heavy shifts in immune signaling networks. According to these data, such histamine-mediated signals are able to easily corrupt, or even abrogate an ongoing antitumoral immune response by causing either Th2-type immunomodulation, or a systemic immunosuppression. Obviously, massive H2R activation in this specific environment is able to suppress the Th1-cell mobilizing, inflammatory danger-signal effects of histamine, which might be associated rather with H1R-activation (32). Hence, histamine-mediated signals most likely support melanoma growth both via an autocrine manner, and by suppressing the immigrating leukocytes, contributing to the widely known poor immunogenicity of melanoma tumors as well.

The second histamine-affected melanoma progression marker, rho-C, represents a member of a wider family of monomeric G proteins. The rho family consists of three closely related members with exceptionally high levels of sequence conservation, and with highly similar functions (rho-A, -B, -C). Rho proteins are able to induce the remodeling of the actin cytoskeleton, which is a basic requirement of cellular motility (34), and elevated levels of rho-C activity are strongly associated with melanoma metastasis formation (35). We found a positive correlation between melanoma histamine production and rho-C mRNA or rho protein expression levels. Actually, the trends of rho(-C) expression clearly followed the trends of the metastatic colony formation potential of the same cells, measured by the experimental lung colonization assay. This is not only in complete accordance with the abovementioned literature data, but a strong proof for the reliability of the gene expression profiling results as well. Our data show that melanoma histamine secretion is possibly required for efficient lung colonization, as targeted suppression of the histamine biosynthesis in circulating melanoma cells seems to interfere with this process. However, histamine signals originating from metastatic melanoma cells in the circulation could also affect rho protein expression in another relevant context. One of the key steps of metastatic colonization is the extravasation of tumor cells from the circulation, which requires enhanced permeability of the capillary walls, and so the tumor-induced mobilization of the capillary endothelial cells. Rho proteins are deeply involved in the induction and control of endothelial motility, and interestingly, literature data suggest that histamine affects endothelial motility by activating rho-dependent signaling pathways (36).

In summary, our results suggest that histamine represents an important component of the molecular machinery inducing and directing melanoma progression, and that distinct molecular pathways can be coupled to the complex phenotypic changes induced by histamine signaling on developing melanoma tumors. However, further studies are obviously still required to narrow down the large sphere of potentially relevant targets of histamine action during the progression of malignant melanoma.

Grant support: Hungarian Medical Research Council grant ETT no. 241/2001 (to Z. Pós, A. Falus, and H. Hegyesi).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Takehiko Watanabe for his kind help and plasmid construct encoding the full-length mouse HDC cDNA; Krisztina Kovács and Anna Földes (Institute of Experimental Medicine, Budapest, Hungary) for making an FLA-3000 Phosphoimager available for this project; and finally, Zoltán Wiener (Semmelweis University, Budapest, Hungary) for his useful comments made on the design of appropriate transfection controls.