Low-dose cyclophosphamide (CP) therapy induces immunogenic tumor cell death and decreases regulatory T cell (Treg) numbers in mice with transplantable tumors. Using the ret transgenic murine melanoma model that resembles human melanoma, we detected no beneficial antitumor effects with such treatment, despite a decrease in Tregs. On the contrary, low-dose CP enhanced the production of chronic inflammatory mediators in melanoma lesions associated with increased accumulation of Gr1(+)CD11b(+) myeloid-derived suppressor cells (MDSCs), which exhibit elevated suppressive activity and nitric oxide (NO) production as well as inhibition of T-cell proliferation. Moreover, the frequencies of CD8(+) T cells in the tumors and their ability to produce perforin were decreased. To study whether the observed CP-induced MDSC expansion and activation also occurs under chronic inflammatory tumor-free conditions, mice exhibiting chronic inflammation were treated with CP. Similar to tumor-bearing mice, CP-treated inflamed mice displayed elevated levels of MDSCs with enhanced production of NO, reactive oxygen species, and a suppressed in vivo natural killer (NK) cell cytotoxic activity indicating CP effects on the host immune system independent of the tumor. We suggest that melanoma therapy with low-dose CP could be efficient only when combined with the neutralization of MDSC immunosuppressive function and chronic inflammatory microenvironment.

Spontaneous melanoma models in transgenic mice are increasingly used in preclinical research as they most closely match the progression of melanoma in humans. While optical inspection only allows analysis of tumors located on the skin, the accurate measurement and growth of subcutaneous tumors have not been adequately assessed. To improve the measurement accuracy of melanoma tumors, we used a fast single-sequence MRI protocol at 9.4 Tesla for longitudinal characterization of a ret-transgenic mouse model. Repeated MRI (average acquisition time 30 min per animal) of the trunk (excluding head and distal limbs) in six siblings revealed an increase in the mean total tumor volume (TTV) from 102.0 ± 80.5 mm(3) at 35 days of age to 434.8 ± 154.9 mm(3) by 77 days. The main tumor load was located within the pelvis (>40%), followed by the proximal hind limbs and groins (>30%). The smallest detectable tumor measured 0.07 mm(3). Inter-rater reliability between a radiologist and a veterinarian analysing MRI data was 0.993 for TTV and 0.840 for number of tumors (both p < 0.001). We thus conclude that because of the high variance of TTV of same-aged mice, MRI should be used (i) to establish treatment groups matched for TTV and (ii) for longitudinal examination of the TTV in mice over the course of treatments.

Inducible NO synthase (iNOS) is a hallmark of chronic inflammation that is also overexpressed in melanoma and other cancers. Whereas iNOS is a known effector of myeloid-derived suppressor cell (MDSC)-mediated immunosuppression, its pivotal position at the interface of inflammation and cancer also makes it an attractive candidate regulator of MDSC recruitment. We hypothesized that tumor-expressed iNOS controls MDSC accumulation and acquisition of suppressive activity in melanoma. CD11b+GR1+ MDSC derived from mouse bone marrow cells cultured in the presence of MT-RET-1 mouse melanoma cells or conditioned supernatants expressed STAT3 and reactive oxygen species (ROS) and efficiently suppressed T cell proliferation. Inhibition of tumor-expressed iNOS with the small molecule inhibitor L-NIL blocked accumulation of STAT3/ROS-expressing MDSC, and abolished their suppressive function. Experiments with vascular endothelial growth factor (VEGF)-depleting Ab and recombinant VEGF identified a key role for VEGF in the iNOS-dependent induction of MDSC. These findings were further validated in mice bearing transplantable MT-RET-1 melanoma, in which L-NIL normalized elevated serum VEGF levels; downregulated activated STAT3 and ROS production in MDSC; and reversed tumor-mediated immunosuppression. These beneficial effects were not observed in iNOS knockout mice, suggesting L-NIL acts primarily on tumor- rather than host-expressed iNOS to regulate MDSC function. A significant decrease in tumor growth and a trend toward increased tumor-infiltrating CD8+ T cells were also observed in MT-RET transgenic mice bearing spontaneous tumors. These data suggest a critical role for tumor-expressed iNOS in the recruitment and induction of functional MDSC by modulation of tumor VEGF secretion and upregulation of STAT3 and ROS in MDSC.

Tumor microenvironment is characterized by chronic inflammation represented by infiltrating leukocytes and soluble mediators, which lead to a local and systemic immunosuppression associated with cancer progression. Here, we used the ret transgenic spontaneous murine melanoma model that mimics human melanoma. Skin tumors and metastatic lymph nodes showed increased levels of inflammatory factors such as IL-1β, GM-CSF, and IFN-γ, which correlated with tumor progression. Moreover, Gr1 + CD11b + myeloid-derived suppressor cells (MDSCs), known to inhibit tumor reactive T cells, were enriched in melanoma lesions and lymphatic organs during tumor progression. MDSC infiltration was associated with a strong TCR ζ-chain down-regulation in all T cells. Coculturing normal splenocytes with tumor-derived MDSC induced a decreased T-cell proliferation and ζ-chain expression, verifying the MDSC immunosuppressive function and suggesting that the tumor inflammatory microenvironment supports MDSC recruitment and immunosuppressive activity. Indeed, upon manipulation of the melanoma microenvironment with the phosphodiesterase-5 inhibitor sildenafil, we observed reduced levels of numerous inflammatory mediators (e.g., IL-1β, IL-6, VEGF, S100A9) in association with decreased MDSC amounts and immunosuppressive function, indicating an antiinflammatory effect of sildenafil. This led to a partial restoration of ζ-chain expression in T cells and to a significantly increased survival of tumor-bearing mice. CD8 T-cell depletion resulted in an abrogation of sildenafil beneficial outcome, suggesting the involvement of MDSC and CD8 T cells in the observed therapeutic effects. Our data imply that inhibition of chronic inflammation in the tumor microenvironment should be applied in conjunction with melanoma immunotherapies to increase their efficacy.

The antimalarial artesunate also exerts profound cytotoxicity toward tumor cells. Earlier investigations controversially discussed a possible immunosuppressive function of artemsinin and its derivatives. This poses the question, whether immunosuppressive activity counteracts the anticancer activity in vivo. To clarify this issue, we used a transgenic mouse spontaneous melanoma model, in which ret transgene is expressed in melanocytes under the control of metallothionein-I promoter. ret-transgenic mice were previously reported to accumulate melanoma-specific effector memory T cells and natural killer (NK) cells in the primary tumors and metastatic lymph nodes. In the present investigation, we monitored effects of artesunate on the CD4 and CD8 T cells as well as Treg and NK cells from ret-transgenic tumor-bearing mice and nontransgenic littermates in vivo. In addition, we investigated cytostatic and cytotoxic activity of artesunate on ret-tumor cells established from the mouse primary tumor. Artesunate inhibited growth of ret-tumor cells and induces their apoptosis in a concentration-dependent manner (0.1-200 micromol/l). Furthermore, we did not find considerable effects of artesunate on the immune function as measured by major cell populations of the immune system; that is, CD4 and CD8 T cells as well as Treg and NK cells both from ret-transgenic mice and nontransgenic C57BL/6 littermates treated for 2 weeks with a daily dose of 1 mg artesunate. These results indicate that the cytostatic and apoptotic effects of artesunate are not diminished by concomitant immunosuppression.

We previously reported that bone marrows of breast cancer patients contained tumor antigen-specific CD8(+) T cells with central or effector memory phenotype. Using a recently developed ret transgenic mouse melanoma model, we now show that bone marrows and tumors of transgenic mice contain high frequencies of CD8(+) T cells specific for the melanoma antigen tyrosinase-related protein 2 and showing mostly effector memory phenotype. Moreover, increased numbers of bone marrow tyrosinase-related protein-2-specific effector memory CD8(+) T cells are also detected in transgenic animals older than 20 weeks with disseminated melanoma cells in the bone marrow and lymph nodes but showing no visible skin tumors and no further melanoma progression. After a short-term coincubation with dendritic cells generated from the bone marrow and pulsed with melanoma lysates, bone marrow memory T cells from mice without macroscopic melanomas produced IFN-gamma in vitro and exerted antitumor activity in vivo after adoptive transfer into melanoma-bearing mice. Our data indicate that functionally active bone marrow-derived melanoma-specific memory T cells are detectable at the phase of microscopic tumor load, suggesting that thereby they could control disseminated melanoma cells.