Normal and cancer cell metabolism: lymphocytes and lymphoma


  • Brian J. Altman,

    1. Abramson Family Cancer Research Institute, Abramson Cancer Center, Division of Hematology-Oncology, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
    Search for more papers by this author
  • Chi V. Dang

    1. Abramson Family Cancer Research Institute, Abramson Cancer Center, Division of Hematology-Oncology, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
    Search for more papers by this author

C. V. Dang, Abramson Family Cancer Research Institute, Abramson Cancer Center, Division of Hematology-Oncology, Department of Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA
Fax: +1 215 662 3929
Tel: +1 215 662 4020


Recent studies of normal and neoplastic lymphocytes have revealed overlapping metabolic rewiring in activated T cells and Myc-transformed lymphocytes. Myc expression is attenuated in normal lymphocytes that return to the basal state, but Notch-activated or Myc-transformed lymphocytes persistently express Myc, which activates genes involved in glucose and glutamine metabolism. Although this difference could provide a therapeutic window for the treatment of cancers, the overlapping metabolic profiles suggest a potential for immunosuppression by metabolic inhibitors.


activation-induced cell death


antigen-presenting cell


B-cell lymphoma 2


breakpoint cluster region–Abelson murine leukemia


B-cell lymphoma 2-interacting mediator of cell death




farnesyltransferase inhibitor


glutaminase splice variant C


glucose transporter 1




hypoxia-inducible factor 1


isocitrate dehydrogenase




lactate dehydrogenase A


mammalian target of rapamycin


mammalian target of rapamycin complex 1


phosphatidylinositol 3-kinase


pyruvate kinase M2


pentose phosphate pathway


p53-upregulated modulator of apoptosis


tricarboxylic acid


T-cell receptor


regulatory T cell




The progression of a normal somatic cell to a fully transformed cancer cell requires the acquisition of many new traits that are not generally found in differentiated adult cells. These traits, some of which have been enumerated by Hanahan and Weinberg [1,2], include constitutive activation with respect to growth and division, even in the absence of growth factor input. Typical cancer cells also show relative insensitivity to apoptotic stresses, upregulation of nutrient transporters, and dramatic alterations in metabolism. These metabolic changes, first appreciated by Warburg [3], often include increased glucose uptake and utilization even in the presence of oxygen, and have been termed ‘aerobic glycolysis’, otherwise known as the Warburg effect. Alterations in lipid metabolism have been also documented. Santos and Schulze examine this topic in depth as part of this minireview series [4].

Recent studies have also shown that some cancer cells rely on glutamine as a primary source of energy, and have tricarboxylic acid (TCA) cycle mutations, which may reprogram metabolism and, in turn, the epigenome [5–8]. Together, these changes result in solid tumors that are profoundly different from their respective parental tissues. However, as compared with quiescent differentiated cells such as epithelial and muscle cells, activated lymphocytes exhibit many features quite similar to those of cancer cells. When activated in an immune response, T cells and B cells increase in size, begin to divide rapidly, and shift to an aerobic glycolytic phenotype. Despite their dramatic change in rate of division and metabolism, however, activated lymphocytes very rarely become cancerous. Instead, once the immune response is over and the insulting agent has been cleared or mitigated, lymphocytes will either undergo apoptosis or return to a basal transcriptional and metabolic state, and be retained as quiescent memory cells. By contrast, lymphocytic cancers such as leukemia, lymphoma and myeloma are driven by deregulation of key oncogenes and loss of tumor suppressors. In this regard, lymphocytic neoplasms are locked in a perpetual proliferative state, which may be associated with a metabolic program distinguishable from that of normal lymphocytes. In this review, we will consider the pathways that lead to lymphocyte activation (using T cells as an illustrative example) and their similarity to commonly mutated cancer pathways, and how these pathways are diminished or inactivated in lymphocytes at the end of an immune response. Understanding the complete life cycle of an activated lymphocyte may provide new insights into how to ‘deactivate’ and kill a transformed cell.

T-cell activation – pathways and metabolic changes

Mature resting (naïve) lymphocytes resemble many other somatic tissues in their energetic states and rates of division. Naïve T cells circulate in the blood or congregate in the secondary lymphoid organs, including the lymph nodes and spleen, and are reliant on the cytokine interleukin (IL)-7 for continued metabolism and survival [9,10]. These cells do not readily divide, and are characterized by a very small cell size and a highly oxidative metabolic phenotype because of complete metabolism of glucose and glutamine [11,12]. Naïve T cells also have a fairly long lifespan, being able to survive for months or years in humans [13]. This is supported in part by continued input from IL-7, which upregulates the antiapoptotic protein B-cell lymphoma 2 (Bcl-2) and thus prevents initiation of apoptosis.

When a T cell becomes activated to mount an immune response, it undergoes a dramatic shift in transcription and metabolism. Activation of T cells initially requires two distinct signals. The first is stimulation through the T-cell receptor (TCR) binding to antigen presented by the major histocompatibility complex of an antigen-presenting cell (APC), such as a dendritic cell. Dendritic cells form part of the innate immune system; they engulf and digest foreign bacteria, viruses, and other particles, and re-express a proportion of these particles on their surfaces as part of the major histocompatibility complex. In addition, T cells require a second, costimulatory signal, which is expressed on APCs and typically received on the CD28 receptor. Costimulation serves as a checkpoint to prevent generation of an immune response to self-antigens and ensure that T cells are only activated directly by APCs. Once activated, T cells require the continued presence of inflammatory cytokines such as IL-2 to remain in their activated state. Importantly, TCR-mediated, CD28-mediated and IL-2-mediated stimulation support different pathways of increased metabolism and activation, similarly to how cancer requires activation of multiple oncogenes to become fully transformed.

Activation of inflammatory T cells (effector T cells) triggers many physiological changes that are similar to hallmarks seen in cancer cells. After activation, T cells rapidly increase in size, or ‘blast’, and divide up to every 6–8 h, faster than nearly any other somatic cell type. The cells also eventually begin to express new ligand receptors, and secrete new cytokines such as IL-2 and interferon-γ for paracrine signaling to other T cells and to coordinate with other immune compartments. Most intriguingly, activated T cells undergo a dramatic shift in metabolism to become highly dependent on glucose uptake and consumption to support this highly active and proliferative phenotype [11]. Metabolic changes in T cells after activation occur extremely rapidly, as changes in calcium flux and lactate production can be observed only minutes after ligand binding [14,15]. TCR stimulation is necessary for initial glucose transporter upregulation, but maximal upregulation and localization of the transporter, and glucose metabolism, require CD28-mediated costimulation and Akt pathway activation [16,17]. IL-2 produced after activation maintains and further enhances glucose uptake [18]. Increased glucose uptake and glycolysis is accompanied by a concurrent increase in oxygen consumption, quite similar to the aerobic glycolysis phenotype observed in many cancer cells. Increased glucose uptake fluxes through the glycolytic pathway to support ATP production, as well as through the pentose phosphate pathway (PPP) to produce nucleic acid precursors, lipid precursors, and antioxidants. This glucose uptake and metabolism is wholly dependent on continued cytokine stimulation. Underscoring the importance of increased glucose uptake and catabolism, activated T cells are highly reliant on glucose for growth, survival, and production of cytokines [17]. Increased glucose uptake and metabolism in inflammatory T cells has been recently shown to be dependent on the orphan nuclear receptor estrogen-related receptor α [19]. Interestingly, the regulatory T cell (Treg) subset of activated CD4+ T cells is far less reliant on glucose than other types of activated T cell. Tregs, which serve to dampen immune responses, divide less rapidly than other T cells, and have been shown to rely on fatty acid oxidation rather than glucose for metabolism. Michalek et al. demonstrated that naïve T cells activated in the absence of glucose failed to differentiate into effector T cells, but readily differentiated into Tregs, whereas blocking fatty acid oxidation under normal activation conditions resulted in fewer Tregs but left other effector T-cell subsets unchanged [20]. Conversely, a recent study documented a role for hypoxia-inducible factor 1 (HIF-1) in favoring the development of the highly inflammatory T-helper 17 effector T cells over Tregs, suggesting that the glycolytic phenotype associated with HIF-1 may be essential for the inflammatory function of T-helper 17 cells [21]. The signaling underlying this dichotomy between the metabolism of effector T cells and Tregs may prove useful in understanding cancer metabolism as compared with somatic metabolism.

Several different signal transduction pathways are activated downstream of T-cell activation, many of which are also important in cancer. Two key pathways in T-cell activation are the Janus kinase–signal transducer and activator of transcription pathway and the phosphatidylinositol 3-kinase (PI3K)–Akt pathway. The signal transducer and activator of transcription factors are induced downstream of cytokine receptors such as the IL-2 receptor, and upregulate a number of survival and proliferative genes, including those encoding the antiapoptotic protein B-cell lymphoma extra-large, the growth and metabolism regulator c-Myc, and the cell cycle protein cyclin D1 [22,23]. Full stimulation of T cells downstream of CD28-mediated costimulation results in robust PI3K activation and phosphorylation of Akt [16,17]. Similarly, nearly all forms of cancer activate the PI3K–Akt pathway through inappropriate upstream activation [24]. Akt in activated T cells is first and foremost responsible for downstream activation of mammalian target of rapamycin complex 1 (mTORC1) to increase protein translation. mTORC1 inactivates the protein synthesis inhibitor eIF4E-binding protein to allow for cap-dependent translation, and phosphorylates and activates S6 kinase to activate the S6-ribosomal protein [25,26]. The presence of free amino acids is also critical for continued mTORC1 activity, and an extensive review of the amino acid signals upstream of mTORC1 is presented by Lamb as part of this minireview series [27]. mTORC1 activation of translation results in the upregulation of a number of transcription factors that induce expression of glucose transporter 1 (Glut1) and glycolytic genes, as well as c-Myc and HIF-1α [28]. In addition to induction of Glut1, Akt activation downstream of CD28 costimulation has been shown to be necessary for localization and maintenance of Glut1 at the plasma membrane [17,29]. Akt also regulates glucose metabolism downstream of uptake by increasing the activity and localization of hexokinases to phosphorylate and capture glucose inside the cell [30,31], activating phosphofructokinase-2 to increase glycolysis [32], and increasing the flux of glucose through the PPP to facilitate macromolecular synthesis [33]. These changes all contribute to a metabolic profile that is grossly different from that of naïve T cells and most other somatic cells, but greatly resembling that of cancer cells.

Activation of Akt and increased glucose uptake also serve to suppress apoptosis in activated T cells. The ability of Akt to localize hexokinase to the mitochondria may, in itself, prevent mitochondrial permeabilization, a key step in the activation of apoptosis [34]. Akt also indirectly controls the transcription of several proapoptotic proteins. Akt negatively regulates the stress response protein p53 to prevent apoptosis by phosphorylating and activating murine double minute 2 to ubiquitinate and promote the degradation of p53 [35]. This suppresses p53-mediated transcription of the proapoptotic p53-upregulated modulator of apoptosis (Puma) protein [36]. Interestingly, increased glucose uptake and metabolism have been shown to be important in suppressing p53 activity and preventing accumulation of Puma, which itself is also negatively regulated by Akt in a glucose-dependent manner [37,38], linking metabolism and apoptosis. Similarly, the stability of the antiapoptotic protein myeloid cell leukemia 1 is enhanced by increased glucose uptake and metabolism [39,40]. Active Akt can suppress expression of the potent proapoptotic protein Bcl-2 family protein Bcl-2-interacting mediator of cell death (Bim) by inhibitory phosphorylation of the forkhead box O family of transcription factors [41,42]. Akt also directly phosphorylates the proapoptotic Bcl-2 proteins Bcl-2 agonist of cell death and Bim, causing them to be sequestered by 14-3-3 [43–45]. Thus, Akt activation and increased glucose metabolism downstream of T-cell activation may result in a host of antiapoptotic signaling changes.

Mutation in cancer cells and similarity to lymphocytes

As discussed earlier, many signaling and morphological changes in cancer strongly resemble lymphocyte activation. Cancer cells generally divide rapidly as compared with somatic tissue. Cancer cells also exhibit strong resistance to apoptosis signals, similarly to activated lymphocytes. The p53 pathway, which responds to a variety of cell stresses and can induce apoptosis, is severely impaired in nearly all cancers [46]. Many blood cancers possess overexpressing mutations of antiapoptotic proteins such as Bcl-2 [47]. As discussed above, increased glucose uptake in activated lymphocytes and cancer cells may itself provide a strong antiapoptotic stimulus. Metabolism is altered greatly in cancer, with some strong similarities to activated lymphocytes (Fig. 1). Most cancer cells exhibit increased glucose uptake and glycolysis even in the presence of sufficient oxygen, and this is used today to visualize cancer by uptake of 2-[18F]fluoro-2-deoxy-d-glucose and positron emission tomography [48]. The PPP is an important path for glucose metabolism, and leads to the production of NADPH for reduction of reactive oxygen species and macromolecular precursors. Glucose is also processed by the hexosamine pathway for proper protein folding and sorting in the endoplasmic reticulum [49]. In fact, many cancer cells upregulate the embryonic M2 isoform of pyruvate kinase (PKM2), which mildly favors aerobic glycolysis of glucose and lactate production rather than flux through the TCA cycle [50–52]. Inappropriate expression of c-Myc in cancer or HIF-1α under hypoxia may also shunt glucose towards aerobic glycolysis and away from the TCA cycle by inactivating pyruvate dehydrogenase and upregulating lactate dehydrogenase A (LDHA) to convert more pyruvate to lactate [53–55].

Figure 1.

 Common and distinguishing metabolic features of activated lymphocytes and lymphoma. Although activated lymphocytes have increased glycolysis, glutaminolysis, oxygen consumption, and pentose phosphate pathway activity, in common with lymphoma, these processes are dependent on growth factors. On the other hand, lymphomas resulting from different oncogenic processes result in growth factor independence and constitutive ribosome biogenesis that renders lymphoma cells dependent on glucose and glutamine. Furthermore, certain lymphomas have an activated mTOR pathway and perhaps fatty acid oxidation. The pyruvate kinase splice variant PKM2 and the GLS splice variant GAC are enriched in lymphoma, and IDH mutations are present in many cancers.

Recently, it has been appreciated that cancer cells, like many lymphocytes, become dependent on extracellular glutamine as both a primary energy sourse and a carbon source [5]. This effect is, in part, dependent on inappropriate expression of the c-Myc transcription factor, and has been observed in human-derived cell lines representing glioma, prostate cancer, and Burkitt’s lymphoma [6,56]. In contrast to activated lymphocytes, cancer cells also exhibit novel mutations in the TCA cycle. For instance, some cancers, such as gliomas and leukemias, exhibit mutations in isocitrate dehydrogenase (IDH)I and IDH2. These mutant IDH enzymes are impaired in their ability to catalyze the conversion of isocitrate to α-ketoglutarate, and instead gain the ability to catalyze the conversion of α-ketoglutarate to the novel metabolite 2-hydroxyglutarate. This could inhibit oxygen-dependent dioxygenases, including prolyl hydroxylases, which would result in stabilization of HIF-1α or histone and DNA demethylases, resulting in alterations of the epigenome [8,57–59].

Cancer cells mimic the activated state of lymphocytes by expressing mutated forms of growth factor receptors or other components of the receptor signaling pathways. For instance, the Notch pathway is critical in early B-cell and T-cell development and lineage selection [60], and, not coincidentally, is mutated in > 50% of T-cell acute lymphoblastic leukemias [61]. Estrogen-related receptor α, which is critical for the altered metabolism of effector T cells [19], has been shown to be important in the growth and progression of breast cancers, particularly those negative for estrogen receptor itself [62,63]. c-Myc is critical for full activation of T cells and the observed metabolic changes, being downstream of both primary stimulation and costimulation [12]. In this regard, loss of c-Myc function through conditional knockout in T cells resulted in the diminished expression of genes involved in glucose and glutamine metabolism, resembling the changes documented in a human Burkitt’s lymphoma cell model with conditional c-Myc overexpression [6,12,64]. Chromosomal translocations and other mutations resulting in overexpression of c-Myc have been observed in many kinds of hematopoietic cancer, notably Burkitt’s lymphoma, but also including many other forms of particularly aggressive B-cell-derived cancer [65,66]. c-Myc is also expressed downstream of many oncoproteins, and has been shown to be a critical target of Notch-driven T-cell acute lymphoblastic leukemia [67–69]. The PI3K–Akt–mammalian target of rapamycin (mTOR) pathway is activated in some way in the majority of cancers [24], and can be activated by oncoproteins such as Ras, breakpoint cluster region–Abelson murine leukemia (BCR-Abl), and human epidermal growth factor receptor 2/Neu, or by loss of the phosphatase and tensin homolog tumor suppressor [70–74]. As discussed above, activation of the Akt pathway strongly upregulates glucose uptake and glycolysis and contributes to the Warburg phenotype. Constitutive activation of c-Myc, however, results in increased aerobic glycolysis and oxidative metabolism through the induction of genes involved in the metabolism of glucose and glutamine, and those involved in mitochondrial biogenesis [75]. Thus, the signaling changes common in carcinogenesis often recapitulate the signaling pathways present in activated lymphocytes.

The end of an immune response

If activated lymphocytes are similar in so many ways to cancer cells, then why is it that most immune responses do not lead to cancer? The answer lies in the fact that the immune system has devised a system for inactivating and clearing activated cells, or returning some of them to a basal memory state. Once the insulting antigen in question has been managed and the immune response is over, most effector T cells will die by apoptosis, with a few being retained as memory cells. However, the choice to become a memory cell is still not well understood, and recent studies suggest that memory T cells may divide and differentiate asymmetrically from the general effector T-cell pool during initial activation [76,77]. Interestingly, there are some striking metabolic parallels between memory T cells and Tregs. Two recent studies showed that inhibiting the mTOR pathway during initial activation, thus emphasizing oxidative over glycolytic metabolism, greatly increased the proportion of CD8+ T cells that became memory cells [78,79]. Pearce et al. also showed that treatment with metformin to activate the AMP-kinase pathway and increase fatty acid oxidation selected for increased memory T cells [79]. This suggested that restoring fatty acid oxidation and oxidative metabolism, which is normally suppressed in activated cells, resulted in a different fate from a mostly glycolytic metabolism. These data provide further evidence that the activated phenotype of both effector T cells and cancer cells may depend on increased glycolytic metabolism.

Activated T cells die by two general pathways. Effector cells that encounter continuous TCR stimulation in the absence of costimulation die by activation-induced cell death (AICD) to prevent reactivation and autoimmunity [80]. AICD generally involves death receptor signaling, which leads to apoptosis [80]. Alternatively, and more in parallel with cancer, activated T cells die from apoptosis through deprivation of growth signals such as IL-2. The proapoptotic proteins Bim and Puma have been shown to play an essential role in the apoptosis of effector T cells after an immune response [81,82]. In fact, activated T cells have several methods of negative feedback to hasten apoptosis after an immune response. Activated T cells express the cytotoxic T-lymphocyte antigen 4 receptor, which competes with CD28 to block costimulation from dendritic cells [83]. It was shown that mice expressing active Akt in lymphocytes developed autoimmunity and lymphoma over time, further suggesting that shutting down the PI3K–Akt pathway downstream of CD28 is critical at the end of an immune response [84]. In contrast, the apoptotic response is suppressed in blood cancers, rendering them insensitive to growth factor deprivation or negative feedback. Antiapoptotic proteins such as Bcl-2 are often overexpressed in follicular lymphoma and diffuse large cell lymphoma, and induction of proapoptotic proteins such as Puma is suppressed in T-cell acute lymphoblastic leukemia cells possessing activating Notch mutations [47,85]. Thus, to prevent overproliferation, which may lead to autoimmunity or cancer, effector T cells have active mechanisms in place, such as AICD and cytotoxic T-lymphocyte antigen 4 expression, and passive mechanisms, such as IL-2 deprivation, that would occur as the immune response diminishes.

How do we ‘inactivate’ cancer?

The most recent paradigm in cancer treatment, ‘targeted therapy’, has been defined as attempting to directly and specifically inhibit the activating oncoprotein, thus reducing toxicity to the patient. This would seem to parallel the death of an activated T cell when stimulating signals are no longer available. Some of these therapies have been extremely successful, such as the development of imatinib (Gleevec) to target BCR-Abl-driven blood cancers [86]. Imatinib has been especially effective in the treatment of patients with chronic myelogenous leukemia, and, in fact, many patients can manage their disease with this drug alone; however, relapse frequently occurs, and is characterized by BCR-Abl point mutations and subsequent immunity to imatinib [87]. Similarly, there are now several targeted therapies for breast cancer. Trastuzumab (Herceptin) has been very effective as a single agent and in combination with chemotherapy in the treatment of breast cancers overexpressing human epidermal growth factor receptor 2/Neu [88,89], and hormonal antagonists such as tamoxifen have been effective in the treatment of estrogen receptor-positive cancers [90]. However, ∼ 15% of breast cancer patients present with ‘triple-negative’ breast cancer that lacks these treatment targets and the progesterone receptor [91], and many patients who are treated with one of these specific therapies develop relapsed cancers demonstrating resistance [92–94]. Unfortunately, some commonly observed oncogenes present in many cancers have thus far proven very difficult to target in a clinical setting. Ras protein was targeted with inhibitors of farnesyltransferase (FTIs), which was thought essential for Ras to localize to the plasma membrane and execute its function. Despite early success with FTIs as single agents or in combination with chemotherapy to kill cancer cells in culture and in xenograft models [95,96], phase II and III clinical trials with FTIs, either alone or in combination, have shown disappointingly modest effects on human cancer [97–100]. Part of this failure stems from the fact that early FTIs were targeted against H-Ras, which is rarely mutated in human cancer [101]. K-Ras and N-Ras can be alternatively geranylgeranylated in the absence of farnesyltransferases, and although geranylgeranyltransferase inhibitors have now been developed, they show unacceptable toxicity when used in combination with FTIs [101]. Myc has long been viewed as an attractive target, as dominant-negative inhibition of Myc alone via Omomyc prevents the maintenance of lung and islet tumors in mice [102–104]. However, as Myc is not an enzyme, it has been difficult to target in a therapeutic setting [105]. Thus, although activating oncogenes are attractive targets for therapies, and some striking examples of success exist, the difficulty in targeting common oncogenes and the propensity of cancer cells to mutate have continued to frustrate clinicians.

As discussed above, altered metabolism is essential for the activated T-cell state. Cancer cells express a variety of oncogenes, and show considerable mutability and genetic elasticity in the face of targeted therapy, but the majority of cancers also have stark metabolic differences from their somatic counterparts. It may be that these metabolic differences are as inherent to cancer as they are to activated lymphocytes, and thus cancers would not be able to mutate as readily in response to metabolically targeted therapy. The idea of metabolically targeted therapy in cancer brings the field full-circle, as the first successful chemotherapy agent was the antifolate aminopterin, used by Sydney Farber in 1948 [106], and the related molecule methotrexate is still used in cancer therapy today. Glucose metabolism in cancer would appear to be an attractive target, given that cancer relies on glucose more than most somatic tissues. In fact, the nonhydrolyzable glucose analog 2-deoxyglucose (2DG) was explored as a cancer treatment as far back as 1954 [107]. Unfortunately, although 2DG showed efficacy against cancer, it also showed significant toxicity in the brain, which is highly reliant on glucose [108]. More recently, some groups have proposed using 2DG in low concentrations along with traditional chemotherapeutics [109,110]. Another possible drug that has attracted recent interest with regard to targeting aerobic glycolysis in cancer cells is dichochloroacetate (DCA), which is a known inhibitor of pyruvate dehydrogenase kinase [111,112]. Pyruvate dehydrogenase kinase phosphorylates and inhibits pyruvate dehydrogenase, thus preventing the entry of pyruvate into the TCA cycle and favoring glycolytic metabolism. A recent report by Bonnet et al. showed that DCA treatment in cancer cell lines and xenografts led to a significant increase in mitochondrial oxidation and mitochondrial NADH production, a decrease in lactate production, and eventual apoptosis, with reasonably low toxicity to the animal [112]. Other groups have experimented with combining DCA and cisplatin as combination therapy [113,114]. However, many cancer cells do not readily uptake DCA [115], and DCA is known to cause peripheral nerve damage after chronic treatment [116,117]. Nevertheless, a small clinical trial of patients with glioblastoma showed that DCA was well tolerated and caused mitochondrial depolarization and an increase in reactive oxygen species in isolated tumor samples [118]. Another emerging strategy to curb the glycolytic phenotype of cancer cells is the use of metformin to activate the AMP-kinase pathway. The AMP-kinase pathway is normally activated when ATP levels become limiting, and can increase both glucose uptake and fatty acid oxidation, as well as inhibit the mTORC1 pathway to suppress protein translation [119]. Metformin has long been used as a treatment for diabetes, and several recent retrospective studies have revealed that diabetics using metformin have a greatly reduced incidence of several kinds of cancer [120–122], pointing towards the AMP-kinase pathway as a possible metabolic target in cancer treatment. Thus, although glucose uptake and glycolysis appear to be difficult to target directly, recent advances may lead to new combination therapies.

In lieu of directly targeting the canonical glycolysis pathway, some groups have also considered targeting other mutated metabolic pathways in cancer. We and others have shown that many Myc-driven cancers are dependent on glutamine uptake and metabolism, and the glutaminase (GLS) splice variant C (GAC) is enriched in lymphoma and other cancers [6,56]. Importantly, several small molecule inhibitors of glutaminolysis have shown efficacy and specificity in cell culture models of glioma and breast cancer [123,124]. As proof-of-concept, two different inhibitors of GLS showed efficacy in vivo against a lymphoma tumor xenograft [123,125]. Another target that has generated interest is LDHA. Targeting LDHA either by genetic knockdown or with a small molecule inhibitor greatly suppressed growth in xenograft models of breast cancer, pancreatic cancer, and lymphoma [54,126]. Finally, NAD+ synthesis has become a possible target of interest, as NAD+ is critical for both glycolysis and the oxidative metabolism, as well as being a cofactor for many cellular processes [127]. A small molecule has been developed to block NAD+ synthesis, and has shown efficacy against blood and liver cancer cell lines [128–130], and has also shown promise in combination with LDHA inhibition, which itself prevents NADH from being recycled back to NAD+ [54]. As more features of cancer cell metabolism are appreciated, more opportunities to target the metabolic features of these cells will undoubtedly be uncovered.

Concluding remarks

Lymphocyte activation unleashes a host of physiological and metabolic changes that closely mirror the changes observed in cancer cells. As we continue to develop new drugs and approaches to target cancer metabolism, it is important to remember that these same treatments will probably disproportionately affect the adaptive immune system, and so proper safeguards must be ensured in patients undergoing clinical trials. Nonetheless, continued study of the transition from naïve to activated cell will help us to learn which pathways are changed in this shift, and present new possible metabolic targets in cancer.