The Aged Microenvironment of Melanoma Feeds Escape from Targeted Therapy

Nongenetic mechanisms play important roles in tumor progression and drug resistance. Two sources of such variation are via alterations in cellular metabolism and the tumor microenvironment (TME). In the article by Alicea and colleagues in this issue of Cancer Discovery, the authors demonstrate that the aged TME can be a source of resistance to BRAF/MEK inhibition in melanoma, and make the surprising link to enhanced lipid uptake as the causative mechanism (1). This work adds to a growing body of literature showing the importance of lipid uptake in melanoma and other cancers, and highlights exciting new avenues for clinical therapies aimed at such uptake.

It has become apparent that cancer cells are not in isolation and have extensive communication with neighboring cells in the TME through signaling and nutrient exchange. Weeraratna and colleagues previously demonstrated that aged dermal fibroblasts have altered secretion of the WNT antagonist sFRP2 from aged fibroblasts (2) as well as decreases of extracellular matrix components such as HAPLN1 (3), which leads to increased melanoma metastasis and decreased T-cell motility. Because fibroblasts are a key component of the TME it is vital that we increase understanding of how age alters the communication between melanoma cells and dermal fibroblasts in the aged TME.

In the current study by Alicea and colleagues, the authors show that age-related metabolic changes in the fibroblasts are a driving factor in supporting the communication between fibroblasts and melanoma cells. They illustrate that aged fibroblasts have increased expression of FASN (Fig. 1; ref. 1), the initiating enzyme in the de novo lipogenesis pathway. This coincides with increased secretion of several lipid species into the microenvironment, increasing the local availability of lipids. Several studies have demonstrated the importance of nutrient exchange in supporting cancer cells metabolic demands (4).

This work expands our current knowledge on the contribution of dermal fibroblasts to nutrient exchange in the aged setting, where they can provide several lipids species including neutral lipids, polyunsaturated phosphatidylglycerols, and ceramides, which are vital for maintaining tumor growth. A question not fully answered in the current study is the mechanism by which aging increases FASN and de novo lipogenesis in the dermal fibroblasts. What about aging is driving this metabolic switch? Furthermore, what other aspects of cellular metabolism are altered in the fibroblast (i.e., glycolysis, nucleotide, and amino acid metabolism?) and how does this affect the metabolic exchange between these two cell types? Fully delineating these mechanistic insights will be an interesting area for further exploration to help understand how age-related alterations in normal microenvironmental cells support cancer progression.

Most cells rely on extrinsic uptake of fatty acids and other lipids, although this can be circumvented by increases inde novo fatty acid synthesis. There are numerous mechanisms by which cells can take up these fatty acids, including the scavenger protein CD36, the FABP fatty acid binding proteins, and the FATP fatty acid transporters. More than a decade ago, it was observed that ovarian cancers frequently metastasize to the adipocyte-rich omental tissue, and that these adipocytes could act as fatty acid donors to the tumor cell, mainly linked to FABP4 expressed on the tumor cells (5). More recently, CD36 has been highlighted as a more general mechanism for tumor cells to take up lipids in metastasis (6), and a number of anti-CD36 strategies are in development. Finally, recent work in melanoma has demonstrated that melanoma cells in contact with subcutaneous adipocytes upregulate FATP1, which promotes their invasiveness via fatty acid uptake (7).

So where does the work by Alicea and colleagues fit into this? Given the increased lipids in the fibroblasts and subsequent secretion into the TME, they hypothesized that, much like the effect with adipocytes, the melanoma cells could take up these lipids to promote invasion. They demonstrated that this occurs using both lipidomics and fatty acid tracing. They then go on to show that in the context of the aged TME, the melanoma cells specifically upregulated FATP2, a protein similar to FATP1 that can both transport lipids and act as a synthetase (Fig. 1). Using genetic knockdown of FATP2, along with inhibition using lipofermata (which blocks both FATP1 and FATP2), they showed that this inhibited lipid accumulation in the aged TME both in vitro and in vivo using engineered YUMM mouse melanoma cells (1).

These observations add to the intriguing, but still not fully resolved, role of FATP proteins in cancer. They were originally isolated as putative fatty acid transporters in 3T3-L1 cells, and subsequent work has shown there are six family members which may have overlapping but distinct functions in terms of fatty acid transport versus lipid modification roles. In addition to the role for FATP1 in melanoma, and now FATP2 in aged melanoma, only a few other studies have mechanistically linked other family members such as FATP2, FATP4, and FATP6 to cancer progression. It is also highly likely that expression of the FATP proteins on TME cells plays a role as well. For example, an interesting recent study showed that high levels of FATP2 are found on myeloid-derived suppressor cells (MDSC), which can suppress their function, and that blocking FATP2 restores antitumor immunity (8). The study by Alicea and colleagues extends this observation about FATP2 and shows that when they treated with lipofermata alone [in the absence of BRAF/MEK inhibitor (BRAF/MEKi)], the effects were not necessarily on the tumor cell, but instead most likely on the TME.

This leaves open the very important but as yet unanswered question of what ultimately controls FATP protein expression in cancer, and in which cells is that expression most essential? It is known that insulin acts as a negative regulator of mouse FATP expression, which raises the possibility that secreted hormones in the aged TME or systemic milieu could be key regulators. It is also likely that whatever regulates the FATPs in cancer may also affect other fatty acid transport mechanisms like CD36 or FABPs, although the study by Alicea and colleagues did not find changes in at least CD36. The identification of such regulatory factors is imperative, because if the FATPs are dynamically regulated and then subject to feedback alterations in their expression or that of other transporters, then targeting them with small molecules such as lipofermata may have only short-term effects that ultimately are hampered by resistance. Because it is clear that FATPs also play roles in nontumor cells in the TME, these regulatory mechanisms are highly likely to be cell type–specific, and unraveling this will be key to understanding in which patients FATP blockade is likely to be useful.

Several groups have demonstrated that altered lipid metabolism can enable cancer cells to escape therapy, resulting in resistance. These studies show that alterations primarily involve expression of FASN in cancer cells, enabling them to be less reliant on lipid uptake. Interestingly here, Alicea and colleagues provide evidence that increased expression of FASN in aged fibroblasts allows for lipid uptake via FATP2 on the melanoma cell, which in turn mediates resistance to BRAF and MEK inhibition (1). Spheroids treated with aged conditioned media are less sensitive to BRAF and MEK inhibition, and this is reversed with lipofermata or FATP2 knockdown. Melanoma cells cultured in aged conditioned media have increased expression of CPT1 and higher levels of oxygen consumption, and are more sensitive to etomoxir, suggesting increased beta oxidation in the aged environment. When treated with BRAF and MEK inhibitors there is an increase in beta oxidation, which is blocked by FATP2 inhibition. Interestingly in a small cohort (eight samples), increased FATP2 expression was correlated with poorer survival, although there was no analysis of age in this cohort because of the small sample size. FATP2 staining increases in tumors that have become resistant to treatment, and their genetic data support this as causal. Finally, addition of ceramides to the conditioned media from young fibroblasts promotes resistance to BRAF/MEKis, suggesting that ceramides are key contributors to resistance similar to what has been shown previously. How the ceramides provide a survival advantage in the setting of drug resistance is not yet clear. The authors suggest that it might be through increased conversion of ceramides to acylceramides, a phenomenon known to allow cancer cells to escape therapy (9). They provide evidence that there is a small increase in acylceramides in the aged condition with a corresponding increase in the required enzyme DGAT2. These data are suggestive and open up new investigations to better understand the precise mechanisms by which these “donated” fatty acids provide an advantage in the resistance setting.

The collective work from the Weeraratna laboratory is beginning to paint a picture in which older patients do poorly in melanoma not simply because of the characteristics of the tumor, but equally so because of their poor response to therapy. The majority of therapies in melanoma either target the MAP kinase pathway (with combined BRAF/MEKis) or activate immune checkpoints (with anti–CTLA4/PD-1/PD-L1). Although the role of age in response to immune checkpoints is still not entirely clear, increasing evidence has shown that increased age is associated with decreased efficacy with BRAF/MEKis (10). Because resistance to such targeted therapies can occur through a multitude of nongenetic mechanisms (most of which reactive the MAP kinase pathway), finding new orthogonal approaches to combat age-related drug resistance is of the utmost clinical need. Prior work has suggested that melanoma cells can escape BRAF inhibition by upregulation of oxidative phosphorylation, which in part depends on inflammatory lipids. One of the most striking results in this study by Alicea and colleagues is the extent to which combined BRAF/MEK inhibition plus lipofermata suppresses tumor growth in aged mice. By implanting YUMM1.7 cells into young or old mice and then treating with BRAF/MEKi, the aged mice relapsed very quickly, but the addition of lipofermata obliterated tumor growth, with no evidence of relapse even after 60 days (1). However, it is not straightforward that lipofermata's ability to block FATP2 alone is the sole mechanism involved here, because lipofermata alone in aged mice can suppress tumor growth, whereas short hairpin RNA to FATP2 alone does not, suggesting that the drug may target other cells (like MDSCs) in the TME or that the drug has other targets (like FATP1) that could be relevant.

So, will blocking FATPs be useful in melanoma, or cancer more generally? Will it be specifically most useful in older patients? These are major unanswered questions in the field. One complication of this problem is the dearth of highly specific FATP small-molecule inhibitors. Although lipofermata and a similar compound called grassofermata are clearly useful tool compounds and seemingly well tolerated, they are by no means specific to FATP2. Perhaps antibody-based approaches, rather than small molecules, could be helpful here. We also do not know much about their potential long-term toxicity, especially in older patients who may have other comorbidities. And even if we were to find highly specific inhibitors (small molecule or antibody based), because of the possibility of cross-regulation and compensatory increases in other lipid transporters, it may be very challenging to find a therapeutic window without the rapid development of resistance. These are all significant issues, but the study by Alicea and colleagues will help to drive others to investigate these questions in the near future.

Overall, this study clearly supports the notion that metabolic changes in the aged TME promote drug resistance in melanoma. Increased expression of FASN in the fibroblasts allows them to be secreted into the TME and subsequently taken up by the FATP2 transporter on the melanoma cell. Although cancer incidence increases with age, most preclinical studies in mice involve younger animals. Alicea and colleagues' work addresses age as a driving factor in tumor progression and provides a novel combinatorial therapeutic approach to combat resistance in the aging population.

R.M. White reports personal fees from N-of-One/Qiagen outside the submitted work, has a provisional patent filed on use of FATP inhibitors in cancer pending to MSKCC, and is a scientific advisory board member for Consano, a nonprofit crowdfunding agency, but receives no royalties for this. No potential conflicts of interest were disclosed by the other author.