Harnessing nucleotide metabolism and immunity in cancer: a tumour microenvironment perspective

For decades cancer research has focused on the seeds rather than the soil, an idea that was proposed by Stephen Paget in 1889, who hypothesised that metastasis in cancer would not occur at random but would only arise in a fertile environment [1]. The soil analogy depicts what is currently known as the tumour microenvironment (TME). Furthermore, the significant work of Hanahan and colleagues two decades ago in characterising the hallmarks of cancer is widely recognised in the world of cancer research and served as a foundation for many researchers in the field [2]. Hanahan published an updated landscape of the hallmarks of cancer in 2011 [2], in which the role of the TME in supporting immune evasion was highlighted, with tumour‐promoting inflammation and genome instability identified as ‘enabling characteristics’ [2, 3, 4] (Fig. 1).

The TME is a niche of heterogenous cell populations including tumour‐infiltrating immune cells, blood vessels, stromal cells and the ECM (Fig. 1). The TME's composition varies depending on the tumour type in which it is created, and is recruited to induce a network of molecular and physical alterations in their host tissue to encourage tumour growth [5]. In addition, the TME and its immunosuppressive elements play a crucial role not only in tumour progression, but to the response to different cancer treatments, therefore many researchers are focusing on modifying the TME to enhance the efficacy of cancer therapies [6]. This peculiar link between the host and cancer cells is primarily composed of six specialised microenvironments that collaborate to support tumour growth and represent an attractive target for anti‐tumour combination therapy. They comprise of hypoxic niches, acidic niches, innervated and mechanical niches, metabolic niches and immunological niches [7, 8].

Maintaining intracellular nucleotide pools is essential for cell growth, metabolism and proliferation. In normal dividing cells, NM is tightly regulated to maintain correct balances of pyrimidine nucleotides (cytosine, thymine and uracil) and purine nucleotides (adenine and guanine) [54]. These five nucleotides are basic units of deoxyribonucleic acid (DNA) and ribonucleic acid (RNA), which are required for cellular function and proliferation. Cancer cells often manipulate NM pathways to support enhanced proliferation rates and tumour progression, therefore understanding the intricacies of NM under health and diseased states has been paramount for understanding cancer progression, therapeutic resistance and developing effective treatment strategies [55].

Nucleotide pools are expanded and maintained in a homeostatic manner by two biosynthetic pathways [56]: (a) de novo synthesis from sugar or amino acid precursors and (b) nucleotide salvage from nucleic acid turnover or dietary consumption [57].

De novo purine synthesis was first described by Buchannan and Hartman in 1959 [58] and begins with ribose‐5‐phosphate, a pentose sugar produced from the breakdown of glucose in the pentose phosphate metabolic pathway [59]. Ribose‐5‐phosphate is converted to phosphoribosyl pyrophosphate (PRPP) and a glutamine‐derived amine group is then added to form 5‐phosphoribosylamine (5‐PRA). Subsequent ATP‐dependent reactions then convert 5‐PRA to purine precursor inosine monophosphate (IMP), involving several key catalysts such as trifunctional enzyme glycinamide ribonucleotide transformylase (GART) [60]. IMP dehydrogenase (IMPDH) and adenylosuccinate synthetase (ADSS) serve as key enzymes in the conversion of IMP to guanine and adenine monophosphate (GMP and AMP) respectively [61]. The de novo purine synthesis pathway was recently reviewed in detail [62].

De novo pyrimidines are assembled through a 6‐step pathway beginning with the l‐glutamine and l‐aspartate amino acids (Fig. 2). The first three reactions, including the rate‐limiting step, are catalysed by carbamoyl‐phosphate synthetase 2 (CAD), aspartate transcarbamylase (ATCase) and dihydroorotase (DHO) [63, 64]. The product of these reactions is dihydroorotate, which is catalysed to orotate by dihydroorotate dehydrogenase (DHODH) via the mitochondrial electron transport chain [65]. Uridine monophosphate synthetase (UMPS) then converts orotate to uridine‐5‐phosphate (UMP) in a two‐step process, which can be phosphorylated to UDP, the precursor of deoxycytidine and deoxythymidine pyrimidines [66].

Purine and pyrimidine salvage pathways utilise free nucleobases and nucleosides recycled from degraded nucleic acids or nucleotides absorbed from the bloodstream via nucleoside transporter proteins [67]. Purine salvage is catalysed by hypoxanthine‐guanine phosphoribosyltransferase (HGPRT) and adenine phosphoribosyltransferase to generate GMP and AMP respectively, whilst pyrimidine synthesis is catalysed by uridine‐cytidine kinases 1 and 2 (UCK1/2) to generate UMP and CMP. TMP may be synthesised from UMP, or from salvaged thymidine via thymidine kinases 1 and 2 in ATP‐dependent reactions [64].

Nucleotide degradation is another crucial element in the homeostasis of nucleotide pools, since relative abundance of nucleotides is key for DNA replication, cell cycle control and cell division [68]. Nucleotide pools are regulated by negative feedback loops at both the enzyme and substrate levels. Purine metabolites GMP and AMP signal negative feedback through interaction with PRPP synthetase [69], and are degraded via conversion to their relative nucleoside via nucleotidase enzymes. Pyrimidine negative feedback occurs through CAD binding, mediated by UTP signalling, with CTP synthase regulating both UTP and CTP pools [70]. Pyrimidine degradation occurs through dephosphorylation to the relative monophosphate form and is largely regulated by dihydropyrimidine dehydrogenase (DPD) [71].

Dysregulation of NM in cancer cells commonly coincides with metabolic reprogramming (Fig. 1; [32]). An increase in nucleotide availability is a key contributor to cancer initiation and progression since cancer cells rely heavily on DNA and RNA synthesis for rapid growth and proliferation. Increasing evidence suggests that there is a strong interplay between dysregulated NM, the composition of the IME and the anti‐tumour immune response [34, 56].

Purine metabolism is the most well‐studied interaction between NM and cancer immunity. Purinergic signalling, a concept coined by Geoffrey Burnstock in 1970 [72] is induced by purine analogues such as adenosine or extracellular ATP, and results in the induction or inhibition of the immune response. Adenosine is frequently elevated in solid tumours and exerts immunosuppressive activity. Adenosine receptors regulate cytokine production in macrophages, primarily mediated by the A2A receptor [73]. The adenosine receptors are G‐protein coupled and comprise of A1, A2A, A2B, and A3. Upon macrophage activation, adenosine receptors modulate cytokine production by inhibiting tumour necrosis factor alpha (TNF‐α) production, which is mediated primarily by A2A receptor [74, 75]. Furthermore, activation of adenosine receptors in neutrophils regulates ROS production and phagocytosis [74, 76]. The hypoxic tumour environment drives the expression of the transcription factor HIF1α and hence drives the expression of CD39 and CD73. CD39 catalyses the conversion of ADP and ATP into AMP, while CD73 irreversibly catalyses the conversion of AMP into adenosine in different tumour cells including stromal cells, MDSC, and Treg cells. The accumulation of adenosine in the TME is mostly the outcome of metabolising extracellular ATP to adenosine by CD39, CD73 and CD38 [73]. In injuries or tumours in which inflammation, necrosis and apoptosis take place, increased levels of ATP function as a pro‐inflammatory signal. Moreover, extracellular ATP can act on P2 receptors leading to enhanced antigen presentation and a pro‐inflammatory response. In contrast, ATP can also elicit immune suppressive activity by inducing T‐helper 2 and Treg cell differentiation [73].

The interaction between NM in the pyrimidine pathway and cancer immunity is less well characterised, however a recent study demonstrated that pyrimidine NM regulates the mtDNA‐dependent innate immunity, indicating a novel interaction pathway between NM and the immune system [77]. This interaction is driven by YME1L, a mitochondrial protease that regulates and preserves pyrimidine pools by supporting glutamine synthesis and the proteolysis of the pyrimidine transporter SLC25A33. YME1L deficiency in YME1L−/− mice was shown to drive mtDNA inflammatory response via cGAS–STING–TBK1 pathway, whereas adding exogenic pyrimidine suppresses it. In addition, inhibiting de novo pyrimidine synthesis and hence pyrimidine pool depletion in wild‐type cells induces mtDNA‐driven immune response [77]. Moreover, it has been demonstrated that urea cycle dysregulation (UCD), a metabolic hallmark of numerous cancers, increases pyrimidine production and promotes CAD activation via changes in nitrogen metabolism [78]. Increasing pyrimidine levels also drives purine to pyrimidine transversion and pyrimidine‐rich transversion mutational bias (PTMB) leading to an increase in immunogenicity [78, 79]. A study found that when compared to normal plasma, melanoma tumour interstitial fluid had higher concentrations of the mono and di‐phosphates guanosine di‐phosphate (GDP) and uridine di‐phosphate (UDP) and that this was associated with an increase in CD4⁺, CD25⁺, FOXP3‐, nuclear factor‐kappa B (NF‐_kB), and an increase in the cytokines; interferons, T‐bet, and IL17, and a decrease in IL13 expression. The authors additionally found that UDP‐treated B16 melanoma cells‐engrafted C57BL6 mice showed a high expression of anti‐tumour immune response markers; tumour‐infiltrating T‐lymphocytes (TILs) CD3⁺ CD8⁺, and CD3⁺CD4⁺, and the major histocompatibility class II high tumour‐associated macrophages (MHC II)^HI TAM, which are associated with tumour suppression in early tumour development [80].

In the complex milieu of the TME, nucleotide metabolism exerts a profound influence on various immune cells, modulating their function and shaping the cancer immunity landscape. The intersection of nucleotide metabolism with immune cell functionality is a critical aspect of cancer progression and therapeutic response.

Since cancerous cells depend on nucleotide availability for rapid and dysregulated proliferation, targeting NM represents a major vulnerability and therapeutic strategy. Inhibitors and modulators of purine and pyrimidine synthesis pathways have been in clinical use since the 1940s, and remarkably remain the standard‐of‐care treatment in many cancer types today [93, 94].

There are two main treatment modalities which target NM. The first is the use of nucleotide analogues, and the second involves inhibitors or modulators of metabolic enzymes involved in NM. Since many clinical approaches are currently based on targeting pyrimidine metabolism in standard‐of‐care treatments, and emerging clinical efforts build upon this foundation in immunotherapy combinational studies, pyrimidine‐based clinical strategies will be the focus for the remainder of this review.

Modulators of pyrimidine metabolism include pyrimidine analogues and enzymatic modulators of the pyrimidine synthesis pathway (Fig. 3). Synthetic pyrimidine analogues can mimic the activity of the naturally occurring compounds. They interfere with conventional NM by interacting with key metabolic enzymes including kinases, ribonucleotide reductase (RNR), nucleoside phosphorylases and thymidylate synthase. This leads to their assimilation or misincorporation into DNA and RNA, inhibiting canonical replication and repair of genetic material, stunting cell growth and proliferation. Enzymatic modulators of pyrimidine metabolism include direct inhibition of rate‐limiting reactions within pyrimidine biosynthesis, as well as disruptors of other critically linked pathways such as one‐carbon (1C) metabolism, encompassing both folate and methionine metabolism [95]. The two major classes of pyrimidine modulators (fluoropyrimidines and antifolates) fall into both categories, as many nucleotide analogues can have additional metabolic effects which contribute to their overall cytotoxicity [96, 97]. Therapeutic agents targeting de novo pyrimidine synthesis are summarised in Fig. 3.

Among the earliest pyrimidine analogues approved for clinical use is 5‐fluorouracil (5‐FU). 5‐FU was synthesised by Heidelberger et al. in 1957 and approved for clinical use in 1960 [98] (Table 2). Since then, 5‐FU has remained one of the most successful chemotherapeutic agents, forming the basis for standard‐of‐care therapy in multiple cancer indications including colorectal [105], breast [106], head and neck [107], gastric [108] and pancreatic [109]. Oral 5‐FU pro‐drugs were later synthesised and are now in clinical use. One such example is capecitabine, the first oral chemotherapy approved for the treatment of metastatic colorectal cancer, also now used in breast and head and neck cancers [110]. As fluorinated pyrimidines, 5‐FU and capecitabine exert anti‐cancer activity through inhibition of thymidylate synthase (TS) and exploiting DNA polymerases' inability to distinguish between genotoxic fluorouracil and normal thymine when synthesising DNA [111]. TS is the enzyme responsible for the rate‐limiting step in de novo synthesis of thymine, the inhibition of which leads to nucleotide pool imbalances, thymine depletion and a thymine‐less cell death [112]. This potent mechanism of action combined with their high tolerability among patients make TS‐targeted therapies the most widely used class of chemotherapeutics, foundational to numerous combination treatment regimens [93].

Antifolates block the synthesis of folate co‐factors needed for normal pyrimidine and purine biosynthesis. Methotrexate was among the first ever curative chemotherapeutics to be discovered, inducing complete remission in acute lymphoblastic leukaemia (ALL) in the late 1940s [113]. Currently, it is administered to ALL patients with central nervous system metastases but has also been used to treat solid tumours [114]. Pemetrexed, another antifolate that was approved relatively recently in 2004. It is used alongside carboplatin or cisplatin as a standard‐of‐care treatment for non‐small cell lung cancer (NSCLC) and mesothelioma respectively [115, 116]. Antifolates are unique to other modulators of pyrimidine metabolism as they have a dual function in targeting both pyrimidine and purine synthesis. In addition to TS, methotrexate and pemetrexed inhibit dihydrofolate reductase (DHFR), with pemetrexed also inhibiting glycinamide ribonucleotide transformylase (GART), a key mediator of purine metabolism. Although the predominant cytotoxic mechanism of pemetrexed is through TS inhibition [117], an early study demonstrated that in vitro cytotoxicity of pemetrexed could not be completely rescued solely by restoration of pyrimidine nucleotide pools, but required additional restoration of purine pools [118]. This suggests that this dual mechanism of pyrimidine and purine modulation is critical for its anti‐cancer activity. Table 2 shows a list of commonly used approved pyrimidine analogues that target different aspects of the pyrimidine metabolism pathway.

Targeted inhibition of pyrimidine metabolism can stimulate immune cell infiltration within the TME. Several studies have shown that 5‐FU can influence the TME, particularly the immunological microenvironment. Studies using colon and melanoma tumour models demonstrated the role of 5‐FU in lowering tumour burden by activating the cancer cell's intrinsic cyclic GMP‐AMP synthase and the cyclic GMP‐AMP receptor stimulator of interferon genes (cGAS‐STING) pathway, which in turn triggers anti‐tumour activity. The cGAS‐STING pathway senses the cytosolic DNA and induces IFN I expression [119]. In addition, 5‐FU induced the production of IFN I, and T‐cell infiltration, and was found that T‐cell depletion reduced the efficacy of 5‐FU [120]. Low doses of 5‐FU were also shown to cause significant MDSC depletion in the blood and the spleen of the Lewis lung carcinoma (LLC) mouse model, resulting in a reduction of tumour volume due to the increased levels of cytotoxic T‐cells [121].

Pemetrexed has also been reported to play an important role in TME immune regulation. Treatment with pemetrexed induced the expression of programmed cell death ligand 1 (PD‐L1) in NSCLC [122]. PD‐L1 is an immune checkpoint protein that regulates T‐cell activity, particularly in suppressing autoimmune responses. It binds PD‐1 on T‐cells, resulting in T‐cell inhibition and exhaustion [122]. Moreover, in a mouse model of colorectal cancer, pemetrexed activated ICD, which in turn promoted the activation of several immune associated genes [123]. A dual role for pemetrexed that involves activation of CD4⁺ infiltrating cells and APCs, as well as promoting PD‐L1 expression facilitating the activation of cytotoxic T‐cells by the action of ICB therapy [124]. The same study highlighted that TS inhibition increased PD‐L1 expression via the accumulation of ROS which activates the NF‐_kB‐PD‐L1 pathway [124]. Furthermore, it was found that sequential administration of pemetrexed and cisplatin activated the STING pathway and ICD, and enhanced immune function through elevated levels of CD8⁺ infiltrating cells, as well as PD‐L1 upregulation in comparison to combination therapy in a NSCLC mouse model [125].

The discovery of the immunomodulatory effects of pyrimidine NM inhibitors has provided strong rationale for the development of novel clinical strategies combining pyrimidine pathway inhibitors with immunotherapy. Immunotherapy was first harnessed by William Coley in the late 19th century, who treated bone sarcomas with bacterial‐based vaccines to induce an anti‐tumour immune response [126]. Their use in modern clinical efforts has expanded significantly in the 21st century due to the discovery and characterisation of immune checkpoints, which led to the development of ICB therapy [127]. The first chemo‐immunotherapy combination strategy combined methotrexate with immunisation and observed an increase in progression‐free survival compared to either single‐agent treatment [128]. Not only did this establish a precedent for chemo‐immunotherapy combination strategies for cancer treatment, but also for combining modulators of NM with immunotherapy. Table 3 contains a summary of disruptors of pyrimidine metabolism in current clinical trials with immunotherapy.

At the molecular level, most clinical investigations have been aimed at improving responses within immune hot tumours, particularly those with high microsatellite instability (MSI) status, given they respond better to immunotherapy [129, 130, 131]. The majority of patients exhibit microsatellite stable (MSS) status and for the most part do not respond to ICB therapies. More recent efforts, however, are focused on identifying MSI and MSS sub‐groups which have a greater likelihood of response to immunotherapy [132]. Combining modulators of NM with immunotherapy in this setting represents a promising clinical strategy to eliminate cancer cells whilst triggering the release of neoantigens to elicit an immune response [133]. This approach could offer significant clinical benefits for a large cohort of patients who would otherwise be ineligible for receiving immunotherapy.

The investigation of the TME continues to be a fertile ground for novel insights, particularly in the context of cancer metabolism and immune interactions. The reprogramming of cancer metabolism, including NM, is a cornerstone in supporting the aggressive growth and proliferation characteristics of cancer. This metabolic reshaping extends beyond the cancer cells themselves, exerting profound influences on the IME. The capacity of cancer cells to modulate and even undermine the host's immune defences plays a critical role in sustaining tumour growth and facilitating metastasis.

Recent studies have shed light on the intricate relationship between pyrimidine metabolism and the IME. Emerging evidence suggests that strategic modulation of pyrimidine metabolism can enhance T‐cell infiltration into the TME, bolstering the inflammatory response against the tumour. This modulation appears to counteract the tumour's immune evasion tactics, thereby reinstating the host immune system's ability to mount an effective anti‐tumour response. This insight not only provides a clearer understanding of the dynamic interactions within the TME, but also opens new vulnerabilities for therapeutic interventions.

In the field of therapeutic strategies, there is a growing trend towards leveraging the immunomodulatory effects of NM modulation. Specifically, combining pyrimidine metabolism modulators with established immunotherapies holds significant promise. This approach aims to synergise the direct anti‐tumour effects of NM modulators with the enhanced immune response elicited by immunotherapy. Such combinations could potentially improve the efficacy of standard‐of‐care treatments across various cancer types, offering a more personalised and effective approach to cancer therapy.

RDL, KAM and PMW are CV6 Therapeutics employees. MJL is a CV6 Therapeutics Consultant. HS and AE declare that there are no competing interests.

The conception and design of the work were initiated by MJL. The drafting of the manuscript was collaboratively undertaken by HS and AE, who both contributed significantly to the development and articulation of the primary content. Subsequent critical revisions, aimed at enhancing the intellectual content and ensuring accuracy, were performed by PMW, RDL, KAM and MJL. All mentioned authors (HS, AE, PMW, KAM, RDL and MJL) have also reviewed and approved the manuscript in its final form.

The authors wish to acknowledge the following funders for supporting their work, Northern Ireland Department for the Economy (NI DfE) studentship (AE); Invest NI/European Regional Development Fund (RD0914247), CV6 Therapeutics (NI) Ltd and Experimental Cancer Medicine Centre (ECMC) grant C36697/A25176.