CAAs represent a vital source of lipids and adipocytokines, and their essential roles in driving metabolic reprogramming and facilitating tumor progression and drug resistance are increasingly appreciated, particularly in obesity-related cancers such as prostate, breast, ovarian and colon cancers (4, 5, 159, 160). These types of cancers are prone to growth and metastasis in an adipocyte-rich microenvironment. Thus, defining the dynamic crosstalk between CAA and tumor help to address the clinical challenges confronted by obesity-related malignancies. Secreted signaling molecules from cancer cells can trigger lipolysis in CAAs and degraded lipids are smuggled to the cancer cells in a way mediated by lipoproteins, exosomes and serum albumin (105). These lipids can either be stored in LDs or fueled by β-oxidation, indicating adipocytes as an energy provider in the TME. Apart from this, CAAs can induce metabolic switch of cancer cell from glycolysis to lipid-dependent energy generation. This was demonstrated in a number of co-culture studies that the uptake and mitochondrial utilization of fatty acids in cancer were significantly promoted by CAAs (19, 160). Co-culture of ovarian cancer cells with omental adipocytes upregulated CD36 expression and accelerated FFA uptake in cancer cells (19). Elevated expression level of FABP4 is identified to be a key event in regulating adipocyte-induced lipid responses of co-cultured ovarian cancer cells, including β-oxidation, lipid peroxidation and reactive oxygen species (ROS) generation (5, 106). Both CD36 and FABP4 are regulated by PPARγ signaling and CD36 is suggested as an upstream regulator of FABP4 (106). However in breast cancers, it is the elevated expression of FABP5 rather than CD36 or FABP4 that play critical roles in acceleration of FA trafficking and nuclear activation of PPARγ, along with induction of lipolysis and ATGL-mediated utilization of accumulated lipids (4).

Though the exact mechanisms underlying CAA-cancer cell crosstalk are unclear, it is acknowledged that secretion of adipocytokines and metabolic substrates contribute to the CAA-induced modulation. Leptin, a hormone mainly derived from adipocytes, shows pleiotropic effects on regulating energy balance and cancer progression (161, 162). Leptin has been known to boost FAO and OXPHOS for energy generation (107), whereas inhibit glycolysis and shift glucose catabolism towards biosynthetic pathways to provide intermediate precursors in breast and colon cancers (107, 108). It was reported that overexpression of CPT1 and upstream activation of AMPK/PPARα signaling pathway and c-Myc/PGC-1 pathway are involved in leptin-stimulated FAO (108). Moreover, leptin-activated JAK/STAT3 pathway also takes part in activating FAO through upregulation of CPT1B in breast cancer stem cells, which is critical for self-renewal and chemoresistance (109). Under physiological conditions, enhanced FAO is usually accompanied with reduced FAS. Consistently, researcher previously proved this in HCT116 and MCF-7 cells treated with leptin for 48 h, as decreased FASN expression together with increased FAO indicated (107). However, these two pathways can occur simultaneously in cancer cells under certain conditions, such as in previously mentioned acidosis-adapted cancer cells and breast cancer cells exposed to oxidative stress (51). A recent study has suggested that leptin drives a concomitant increase in FAO and FAS in MCF-7 cells, which is ascribed to the activation of autophagy (110). Specifically, leptin-activated autophagy enhances lipolysis and the release of free FAs, which further facilitates leptin-induced FAO for energetic production. Meanwhile, the expression of SREBP-1 is upregulated by leptin via autophagy and PI3K/AKT signaling pathway, resulting in overexpressed FASN and enhanced de novo lipogenesis. Researchers showed that inhibiting or silencing SREBP-1 or FASN impeded cancer cell viability promoted by leptin. Yet, the contradictory results of FASN expression in breast cancer indicated in these two studies need further verification to see whether the effect of leptin varies in time-dependent or dose-dependent manner (107, 110). Another study also revealed that leptin elevated expression of SREBP-2 via PI3K/AKT signaling, thus upregulating acetyl-CoA acetyltransferase 2 (ACAT2) to favor the migration and invasion of breast cancer cells (111). Taken together, leptin reconstructs cancer cellular lipid metabolism by multifactorial modulating biosynthesis and utilization of lipids, which further facilitate tumor progression.

Adiponectin is another adipocyte-specific hormone, yet elicits molecular effects opposed to leptin, as it may restrict tumorigenic processes. While identification and better understanding in the effect of adiponectin is cancer-specific lipid reprogramming still need further investigation, adiponectin is putatively considered as a negative modulator of lipogenesis. Adiponectin decreased lipid uptake through suppressing CD36 and LDLr expression in cancer cells (112, 163). As both an activator of AMPK and inhibitor of PI3K/AKT pathway, adiponectin is presumed to promote FAO and restrain lipogenic biosynthesis in cancer. Yet, this hypothesis requires corresponding proofs in the context of complicated signaling network of cancer cells. In addition, adipokines like visfatin and resistin also showed the potential to stimulate lipogenesis in liver cancer cells with elevated FASN expression (113). Inflammatory cytokines released by adipocyte, such as TNF-α, IL-6, are also able to affect cancer lipid metabolism with paracrine and endocrine activity, which will be discussed later.

The role of CAFs, a major cellular component within tumor stroma, in the progression and spread of various cancers has been well-established (164). Rather than just being educated by tumor cells, CAFs have been suggested to play an active role in shaping tumor metabolism, including the lipid metabolic pathways (165). Through self-metabolic reprogramming, CAFs serve as hubs of lipids to support breast cancer cells growth with over-expression of FASN (114). FATP1 and CD36 are found as putative targets to disturb the smuggling of lipids between CAFs and breast cancer cells (114, 166). In addition to FAs, other CAF-derived metabolites like alanine are also utilized in pancreatic cancer cells as carbon sources for biosynthetic processes, including lipid synthesis (115). A recent study in PDAC cells showed that lysophosphatidylcholines (LPC) from CAFs cells supported membrane synthesis in cancer cells as well as the production of LPA by autotaxin (ATX) that was further engaged in paracrine activation of PI3K/AKT pathways (167). These CAF-derived metabolites and bioactive molecules are transported to tumor cells, at least partly, by CAF-derived exosomes (CDEs). Containing proteins, miRNAs and intact metabolites, CDEs are initially identified as a key component to support tumor growth under nutrient-deprived stress (117). CDEs could switch mitochondrial OXPHOS towards a glycolysis- and glutamine-dependent reductive carboxylation phenotype in cancer cells, thus replenishing for de novo lipogenesis (117). Interestingly, these metabolic alterations elicited by CDE-metabolite were independent of oncogenic KRAS in pancreatic cancer, suggesting CAFs’ ability to reprogram and support cancer cell metabolism independent of oncogene activation (117). The researchers later identified the glutamine anabolic pathway in CAF as a possible target to cause systemic metabolic vulnerabilities in ovarian cancers (168). Besides, miR-522 in the CDEs was found to suppress the lipid-ROS production in gastric cancer, via inhibiting the main mediator arachidonate lipoxygenase 15 (118). Ferroptosis, a form of cell death due to dysregulated membrane lipid peroxidation arising from an iron-dependent ROS accumulation, is thereby inhibited in gastric cancer, contributing to the tumor growth and decreased chemo-sensitivity to cisplatin and paclitaxel (118).

As a crucial determinant of the phenotype and function of immune cells, metabolic reprogramming in immune cells caused by TME has been fairly defined and applied in manipulation of immune responses for cancer. Lipid metabolism alterations in immune cells also play a role in coordinating immunosuppression and tumor immune escape (169). However, only few researches described the effect of immune cells and their released factors on lipid rewiring of their neighboring cancer cells.

CD8⁺ T cells take a leading position in anti-tumor response among the tumor-infiltrated immune cells and activated effector T cells in the TME mainly depend on glycolysis and FAS, while memory T cells and CD4⁺ regulatory T cells maintain their functions by enhancing FAO (169). Metabolic competition by tumor cells restricts T cells and dampen their interferon gamma (IFNγ) production and other functions, thus facilitating tumor progression. A study reported that during immunotherapy, activated CD8⁺ T cell would enhance lipid peroxidation in tumor cells and finally led to ferroptosis. Mechanistically, it was illuminated that the IFNγ released from CD8⁺ T cells restrained uptake of cystine by tumors through downregulating the expression of SLC3A2 and SLC7A11, two subunits of the glutamate-cystine antiporter system x. This further inhibits glutathione synthesis and induces a decrease in glutathione peroxidase 4 expression, which is essential in the antiperoxidant defence, consequently promoting cell lipid peroxidation. Targeting IFNγ or iron-dependent lipid peroxidation pathway in combination with checkpoint blockade therefore provides new insights for combinatorial cancer treatment (119).

Growing evidence has shown that metabolic restriction imposed by tumor cells would restrict immune cell function thus promote immunoescape and caner progression (158, 169). In gastric adenocarcinoma, FAO-dependent tissue-resident memory T cells was outcompeted by cancer cells for lipid uptake. Blockade of PD-L1 would promoting lipid uptake and resulting better survival of Trm cells, in vitro and in vivo (170). However, a recent work indicated that the suppression of immune cells is not caused by cell-extrinsic nutrient competition, evidenced by the fact that the uptake of glutamine and lipids was dominated by cancer cells instead of by immune cell in TME (171, 172). Therefore, more researches are required to clarify this issue.

In most tumors, tumor associated macrophages (TAMs) adopt an ‘alternative’ phenotype (M2-like) exerting anti-inflammatory and pro-tumoral effects. Studies have proved that enhanced FAO, elevated expression of CD36 and accumulated lipids are essential in activation and differentiation in TAMs (173). In ovarian cancers, it’s uncovered that tumor cells can scavenge cholesterol from TAMs by promoting membrane-cholesterol efflux (120). IL-4 mediated signaling such as inhibition of IFNγ-induced gene expression is promoted due to the increased cholesterol efflux in TAMs, which is associated with pro-tumoral functions (120).

The ECs within tumor also undergo phenotype alteration to provide a promoting niche for cancer cells. An in vitro experiment reported that ECs triggered coherent and non-cell line specific increase in expression of both glycerophospholipid and poly-unsaturated fatty acids (PUFA) in cancer cells, which could be associated with remodeling of cancer cellular membrane to improve cellular cross-talk or modulate signaling pathways (121). But the underlying manipulation mechanism needs further explanation.

Above mechanisms can be exploited therapeutically at the level of the stromal cells to impede tumor progression. Inhibition of lipogenesis or lipolysis in stromal cells and lipids uptake by malignant cells would be the most attractive approach to cut off lipid supply to tumor cells. However, these specific targeting strategies have not been validated in well-designed clinical trials and accompanying effects on other cells is a big challenge. Moreover, targeting CAAs alone is less likely to be effective since they are unlikely to lead to a complete reversal and ablation of tumor growth. Small-molecule inhibitors designed to target CAFs have also been developed, either by blocking the activation and trans-differentiation of stromal cells into CAFs, or silencing signaling pathways activated in CAFs and their downstream effectors (174). Blocking secretion and uptake of stroma-derived exosomes also represents a tantalizing target for clinical implementation to disrupt the level of de novo lipogenesis and oxidation. When targeting ECs, some methods are designed to normalize tumor vasculature, thus mitigating hypoxia and acidosis in TME and improving the efficacy of therapies (175). Overall, an integrated view of these cell-cell interactions needs to be further established for improved therapeutics and manageable adverse effects when simultaneously targeting multiple components of the TME.

The gastrointestinal (gut) microbiome has emerged as a pivotal microenvironmental factor for certain cancers, such as gastric cancer, colorectal cancer and HCC (179–181). The dysbiosis of gut microbiota diversity and related alterations in the level of gut microbial metabolites rewire the metabolic milieu and consequently contribute to the cancer progression. Dietary fiber is fermented by gut bacteria into short-chain fatty acids (SCFAs), including acetate, butyrate and propionate in a 3:1:1 stoichiometry (182). SCFAs are considered important substrates for energy metabolism, with propionate serving as a substrate for gluconeogenesis while acetate and butyrate serving as substrates for de novo lipogenesis.

Though acetate can also be acquired from deacetylation processes and ethanol metabolism, dietary food saccharolytic fermentation by intestinal microbiota is thought to be the main source of exogenous acetate (182). Intriguingly, acetate is specially positioned at the intersection of metabolism and genetics for its roles as both a carbon source for lipid biomass and an epigenetic regulator of post translational protein modification (129). Cancer cells exhibited activated cytosolic ACSS2-mediated acetate metabolism, which promotes tumor growth by replenishing the acetyl-CoA pool under hypoxia, acidosis and other stressed conditions (90). Consistent with this, radiolabeled acetate taken up by cancer cells has been shown to be greatly elevated than normal cells and mainly donates to the carbons in FAs under hypoxia (130). Moreover, uptake of radiolabeled acetate can reflect FASN expression levels and further the sensitivity to FASN-targeted therapy in prostate cancer cell lines as it reduces significantly with the presence of FASN inhibitors (131, 183). Therefore, [1-11C]-acetate PET based on these mechanisms is a promising non-invasive tool to diagnose cancers and even predict FASN-targeted therapy outcome and recurrence in clinical applications (184, 185). Beyond an alternative source of acetyl-CoA production, acetate is also implicated in initiating epigenetic regulation to lipogenic genes of FASN, for cancer cell survival under hypoxic stress (129). These might also provide new insights into FASN-targeted therapy.

Similarly, butyrate is mitochondrially oxidized to acetyl-CoA for lipid synthesis and represents the major energy source for normal colonocytes while cancerous colonocytes primarily undergo aerobic glycolysis (186). However, it reduces cancer cell growth while improves differentiation and this differential growth impact of butyrate on normal and neoplastic colonic cells has been known as the “butyrate paradox”. Butyrate in the glycolytic cancer cells is inefficiently metabolized and its accumulating in nucleus exerts antineoplastic effects, which are mediated, at least partially by inhibiting histone deacetylase (HDAC) activity and epigenetically regulating downstream target genes (186). The function of butyrate (either being oxidized or being a HDAC inhibitor) is related to its concentration, since low dose (<0.5 mM) meets the energy needs while high concentration (0.5~5 mM) leads to inhibitory outcome in in vitro experiments (187). With regard to its effects on cancer metabolism, though initially highly glycolytic, colon cancer cells can switch to a butyrate/glutamine-utilizing phenotype induced by butyrate, which is characterized by a lower production of lactate and raised incorporation of carbons derived from glutamine into lipids (132). Mechanistically, the expression of membrane glucose transporter 1 and cytoplasmic glucose 6 phosphate dehydrogenase decreased under the regulation of AKT signaling, leading to inhibited glucose metabolism and nucleotide synthesis and causing proliferation arrest (188). Besides, butyrate positively regulates the expression of M2-pyruvate kinase (PKM2) and PDK in colon cancer cells, significantly suppressing upstream glycolytic intermediates being diverted into biosynthetic pathway (132, 133). Thereby, butyrate contributes to acetyl-CoA pool directly or indirectly through stimulating glutamine utilization and favors a lipogenesis phenotype, as LD accumulation is observed in butyrate-treated colonic cancer cells (132). When considering butyrate as an anticancer agent, whether and how these metabolic altering contributes to the proliferation inhibitory impact of butyrate still needs further exploration, and cancer types other than colorectal cancer should also be further studied.

Given the facts that the increased risk of certain gastroenteric tumors is linked to alterations of gut microbiota species and reduced production of SCFAs, efforts have been made to elucidate the relation between dietary pattern and gut microbiota. Dietary interventions like high-fiber diets and the supplementation with polyunsaturated fatty acids, polyphenols and probiotics, which are known to regulate gut microbiota and generate SCFAs, have emerged as potential therapeutical strategy to prevent or to be used as adjuvants to conventional therapy (189). Of note, there is not a simple linear relationship between gut SCFA levels and individual dietary components or bacterial strains. Algorithms to predict individual responses to dietary and pharmaceutical interventions based on microbial metabolites is essential for any nutrition-based approaches. Alternative strategies such as increasing gut microbial production of beneficial metabolites and specific inhibitors for microbial pathways that produce harmful metabolites also has great potential in the long run.

LPA is a byproduct of the lipid biosynthesis pathway that presents at high level in several cancer patients. It functions as a growth factor stimulating oncogenesis and invasiveness through activating G protein-coupled receptors (190). It is extracellularly converted from stroma-derived LPC by a secreted lysophospholipase D, ATX. LPA has been shown to induce a glycolytic shift in cancer cells, while studies also reveal that it elicits pro-lipogenic actions in ovarian cancer by transcriptionally up-regulating SREBP-FASN and dephosphorylation of AMPK-ACC pathways, which is mediated by an LPA receptor LPA₂ (135, 136). Noteworthy, LPA was shown to activate HIF-1α via inducing a pseudohypoxic response, thus further modulating metabolism alterations in ovarian cancer mediated by HIF-1α (135). ATX/LPA axis was also found to be involved in lipid desaturation, via stimulating SCD expression, neutral lipid liver deposition and TG accumulation in HepG2 cells (137). Although it still remains to be confirmed whether ATX/LPA can have the same effect in other cancer types, the ATX/LPA axis, communicating with other lipid modified signaling, is engaged in lipid metabolism in malignant cells and is emerging as a novel cancer hallmark.

Branched chain amino acids (BCAAs; leucine, isoleucine and valine) are essential amino acids that are only available from dietary protein rather than from endogenous synthesis. The increase in plasma BCAAs has been associated with pancreatic cancer risk and cancer cells have exhibited increased BCAA uptake as well as over-expressed BCAA aminotransferase (BCAT), for proliferation (191, 192). BCAA metabolism in PDAC cells has been unveiled as critical carbon source for FA synthesis, as the knockdown of BCAT leads to dramatic reduction in levels of FAs and TGs, as well as the inhibition of tumor cell proliferation (138). Mechanistically, elevated BCAAs selectively activates mTOR signaling and triggers phosphorylation of downstream effectors, including SREBPs (139). BCAT also promotes mitochondrial biogenesis and function in a mTOR-dependent manner (140). Thence, disruption of the BCAA level or targeting pivotal enzymes indicates an exploitable therapeutic strategy for cancer therapy. However, as the most straightforward strategy, controlling dietary BCAA intake shows controversial effects in cancer therapy (193–197). An integral investigation is therefore required to find out whether this method has benefits for certain types of cancer.