Although decades of research have yielded targeted therapies that are effective in eliminating or reducing many breast tumors, breast cancer remains the leading cause of morbidity and second-leading cause of death in women.(1) Recent published reports suggest that reciprocal influences exist between breast tumor cells and the tumor microenvironment and that these interactions affect the growth and energetics of the tumor. We discuss these interactions, as well as cutting-edge technologies currently used, to reveal the contributions of individual cells within a tumor to the overall disease, with the hope that new drug targets will emerge from our improved understanding of this complex “soil and seed” system.
Breast cancer remains the leading cause of morbidity and second-leading cause of death in women. Despite efforts to uncover new targeted therapies, a vast number of women die due to refractory or recurrent breast tumors. Most breast cancer studies have focused on the intrinsic characteristics of breast tumor cells, including altered growth, proliferation, and metabolism. However, emerging research suggests that the tumor microenvironment can substantially affect relapse rates and therapeutic responses. In this review, we discuss the interactions between the tumor and microenvironment in breast cancer, with regard to mutational profiles and altered metabolism that could serve as potential therapeutic targets. We also describe current technologies available to study these interactions. (Cancer Sci 2012; 103: 400–407)
Biology of the Mammary Gland in Breast Oncogenesis
The mammary gland continually remodels and develops postnatally with the cyclical influence of reproductive hormones, and it attains full epithelial-cell differentiation only with the completion of full-term pregnancy, lactation, and involution cycles.(2,3) This developmental scheme allows for the maintenance of a pool of “mammary stem cells” and bipotent progenitor cells for an extended period of time, which contributes to breast cancer being a heterogeneous disease.(4) Molecular subtyping of breast cancer, based on intrinsic gene expression profiles, lists six subtypes: luminal A (ER+, PR+, keratin 8/18); luminal B (ER+/PR+, cyclinB, keratin 8/18); ERBB2/HER2-enriched; basal-like (ER−/PR−, keratin 5/6); claudin-low (CD44+/CD24low; low in claudin 3, 4, 7, and E-cadherin; ER−; PR−; HER2−); and normal breast-like (Table 1). These subtypes substantially differ from one another in terms of incidence, survival, and response to therapy.(5)
|Gene symbol||Gene name||Gene function||Subtype marker||Expression patterns||Ref.|
|AMPK/PRKAA1||AMP-activated protein kinase||Sensor for cellular energy levels||NA||NA||63, 66|
|BRCA 1/2||Breast and ovarian cancer susceptibility protein||Tumor suppressor, DNA repair||Basal-like breast cancer (not exclusive)||Mutated in tumor||10|
|Cav-1||Caveolin 1||Scaffolding protein||Basal-like breast cancer||Down in CAF, up in basal-like tumor||43, 52, 53, 54|
|CCNB1||CyclinB||Cell cycle regulator||Luminal B breast cancer||Upregulated in tumor||5|
|CD24||Cell differentiation antigen 24||B-cell interacting protein||Claudin low breast cancer, stem cell marker||Downregulated in tumor||5|
|CD44||Cell differentiation antigen 44||Cell-surface glycoprotein||Claudin low breast cancer, stem cell marker||Upregulated in tumor||5|
|CDH1||E-cadherin||Calcium-dependent cell adhesion proteins||Claudin low breast cancer||Tumor and fibroblast cells||5|
|CLDN3/CLDN4/CLDN7||Claudins 3, 4, 7||Tight junction formation||Claudin low breast cancer||Upregulated in stroma||5|
|EGFR||Epidermal growth factor receptor||Growth factor signaling||Basal-like breast cancer||Upregulated in tumor||5, 17|
|ERα||Estrogen receptor alpha||Hormone signaling||Hormone receptor positive breast cancer||Upregulated in tumor||5, 17, 34|
|ERBB2||ERBB2/Her2||Growth factor signaling||HER2+ breast cancer||Upregulated in tumor||5, 17|
|GAPDH||Glyceraldehyde-3-phosphate dehydrogenase||Oxidoreductase activity||NA||Upregulated in tumor vessels||72|
|GLUT1/SLC2A1||Solute carrier family 2||Glucose transport||NA||Upregulated in tumor||21|
|HIF-1||Hypoxia-inducible-factor-1||Transcription factor||NA||Upregulated in hypoxic conditions||10, 32, 34, 35, 36, 43|
|KRT5A/6A||Keratin 5/6||Cytoskeleton organization||Basal-like breast cancer||NA||5, 17|
|KRT8/18||Keratin 8/18||Cytoskeleton organization||Luminal breast cancer||NA||5|
|LDHA||Lactate dehydrogenase A||Reversibly converts pyruvate to lactate||NA||Upregulated in tumor vessels||72|
|LDHB||Lactate dehydrogenase B||Reversibly converts pyruvate to lactate||NA||Upregulated in tumor vessels||72|
|MCT-1/SLC16A1||Monocarboxylate transporter 1/solute carrier family 16||Lactate transporter protein||Basal-like breast cancer||Upregulated in regions of oxygenation||24, 40, 48|
|Myc||Myelocytomatosis viral oncogene homolog||Transcription factor, oncogene||Basal-like breast cancer||Upregulated in tumor||10, 16, 26|
|PIK3CA||Phosphoinositide-3-kinase, catalytic, alpha polypeptide||Lipid phosphatase, signal transduction||More common in hormone positive tumors||Activating mutations in tumor||10, 19|
|PKM2||Pyruvate kinase splice variant 2||Energy metabolism, glycolysis||Tumors dependent on aerobic glycolysis||Upregulated in tumor vessels||28, 30, 43, 72|
|PR||Progesterone receptor||Hormone signaling||Hormone receptor positive breast cancer||Upregulated in tumor||5, 17|
|PTEN||Phosphatase and tensin homolog||Lipid phosphatase, signal transduction||More common in hormone positive tumors||Potentially mutated in stroma||12, 19, 44|
|TGFβ||Transforming growth factor, beta||Growth factor signaling||Basal-like breast cancer||Activates fibroblasts to generate CAFs||37, 41, 43, 75|
|TP53||Tumor protein 53||Tumor suppressor, transcription factor||Mutated across subtypes||Potentially mutated in stroma||58|
The architecture of the mammary gland contributes to some of this variation, with structures ranging from branched ducts and alveoli to two layers of epithelial cells encased in an outer basement membrane separating the duct from the surrounding scaffolding stroma, which is composed of fibroblasts, endothelial cells, and infiltrating leukocytes (Fig. 1).(2) The inner ductal lumen is lined by polarized, secretory epithelial cells, which are destined to hold milk during lactation and are programmed to proliferate explosively at that time.(6) This is often the site of ductal carcinoma in situ, which is typically classified into luminal A and B subtypes.(7) Although these tumors usually have a good prognosis, with low rates of metastasis, a substantial number of patients have late relapses after a long period of cancer dormancy, which poses a great therapeutic challenge.(8) The outer layer of myoepithelial cells has cytoskeletal flexibility to provide contractility to the duct, that is, to squeeze out milk from the luminal cells during lactation. These myoepithelial cells also produce the basement membrane. The basal-like intrinsic subtype harbors gene signatures similar to myoepithelial cells, suggesting that it originates from myoepithelial precursor cells with upregulated migratory and invasive machinery.(5)
Along with the various physical and hormonal interactions between the ductal cellular layers and the stroma, the paracrine effects of the stroma are also critical for maintaining the polarity, differentiation status, proliferation rate, and restricted migration potential of the luminal epithelial cells.(9) Not surprisingly, the loss of normal myoepithelium is an early sign of tumorigenesis, and once this boundary layer is compromised the tumor can become highly metastatic. The role of the loss of apical–basal cellular polarity, as well as the dramatic influence of the microenvironment on cancer cells, has become apparent from modeling breast cancer in ECM (laminin/collagen)-supported 3-D systems in vitro.
Exploiting the Link between Tumor Growth and Metabolism
Tumors arising in the breast can be considered two-compartment systems consisting of the cancer cells and the surrounding stroma. These two compartments are intricately interconnected by bridges, namely the growth and metabolic pathways of each compartment (Fig. 2). Moreover, the molecular pathways driving these two processes are not parallel but intersect frequently, making it difficult to eradicate breast cancer cells. Molecular pathways that are hyperactivated through mutations in single or multiple genes that are important for cancer cell proliferation and growth are also essential regulators of tumor metabolism.(10) Research shows that the consequences of global genetic alterations are evident in both ductal and stromal cells during tumor progression without clearly outlining the driver of this process. Oncogenic mutations such as those in p53, BRCA1 and BRCA2, RAS, PIK3CA, and, Myc alter the growth and metabolism of both the tumor and the surrounding stromal compartment.(10) There are reports of mutations in p53 and PTEN in tumor-associated fibroblasts, and several of these studies have shown a clinical association with the presence or absence of a mutation; however, these reports are controversial.(11–13)
In addition to the interplay between hormone receptors and growth factors, we are now learning how nutrients such as glucose, glutamine, serine, and even lactate can facilitate survival of cancer cells.(14,15) Evidently, different breast tumor subtypes process and use these macromolecules in different ways.(16) Thus, new parameters such as glucose versus glutamine addiction (reflective of specific mutation status) of the tumor are gaining clinical relevance, propelling new kinds of therapeutic approaches that will consider both the cancer cells and the surrounding stroma. One such approach would be to induce synthetic lethality between tumor and stromal cells by using sublethal concentrations of drugs targeting oncogenic and/or metabolic pathways in each compartment.
Hormone and mutation profiles that drive breast cancer energetics.
In breast cancer, the most traditional clinical tumor classification is based on the upregulation of key growth factor receptors (ER, PR, HER2). Other growth factor receptors, such as the epithelial growth factor receptor, androgen receptor, fibroblast growth factor receptor 2, insulin receptor, and insulin-like growth factor receptor, are also upregulated or mutationally activated in a subset of breast cancers.(17,18) Growth factors in turn regulate uptake of nutrients, affecting energetics.
The PI3K/AKT pathway is the classic example of how the mutation status of the cancer is closely linked to its metabolic profile. The PI3K/AKT pathway, the most frequently mutated pathway in breast cancer,(10,19) controls several critical aspects of cell growth and differentiation and is intricately tied to metabolism.(10,19) Furthermore, PI3K/AKT increases glucose uptake as a consequence of translocation of glucose receptors to the membrane, enhanced storage of glucose by hexokinase, and activation of phosphofructokinase activity, leading to what is often referred to as “glucose addiction.”(10,20) Moreover, expression of GLUT1, the main glucose transporter in breast cancer, is negatively correlated with ER-α expression but positively correlated with ER-β expression,(21) an observation that remains to be explained but that may suggest novel therapeutic approaches.
Cancer cells preferentially use glycolysis to generate energy, releasing lactate from the cell despite the abundance of oxygen, in a phenomenon referred to as “the Warburg effect.”(22) Breast cancer cells also use other types of nutrients to fuel their metabolic processes and ultimately generate the energy needed for survival and division. Indeed, both glucose and malate can rescue detached cells from oxidative stress(23) and lactate generated by one tumor cell can be used as a fuel for mitochondrial oxidative phosphorylation by other tumor cells as well as by stromal cells.(24) In addition, fatty acid oxidation becomes an alternative energy source for cancer cells under stress conditions.(25) Glutamine and other amino acids can enter the tricarboxylic acid cycle, providing additional alternative energy sources. The proto-oncogene Myc appears to drive the selective use of glutamine as an energy source.(26)
Thus, the mutation background of cancer cells allows them to be extremely versatile in using diverse resources for energy production. A gradual shift toward aerobic glycolysis in cancer cells is subsequently accompanied by discrete metabolic changes occurring in the mitochondria. Changes include decreased oxidation of substrates, altered expression and activity of respiratory chain subunits, increased production of ROS, mitochondrial DNA mutations, and impairment of both the respiratory chain complexes or ATP synthase organization within the inner mitochondrial membrane, eventually altering the regulation of apoptosis.(27)
Expression of PKM2 has emerged as an exclusive marker for aerobic glycolysis observed in cancer cells.(28) Association of PKM2, but not PKM1, with tumor glycolysis was validated by the observation that replacement of PKM2 with PKM1 reverses the Warburg effect and reduces tumorigenesis in vivo.(28) Indeed, earlier clinical observations found increased plasma PKM2 levels in advanced breast cancer, indicating disease activity and decreased sensitivity to chemotherapy.(29,30)
The interconnected nature of these changes emphasizes the complexity of altered metabolism in cancer cells, which may have extensive therapeutic implications.
Tumor microenvironment in growth and metabolism.
The focus of breast cancer studies on abnormalities exclusively in tumor cells has shifted as understanding of the contribution of the stroma to tumorigenesis has improved.(31) The dynamic tumor microenvironment is under physiological stresses, such as hypoxia (even in the presence of increased vasculature), the influence of ROS, inflammation, nutrient deprivation, and low pH.(32) As the tumor burden increases, the center of the tumor is frequently deprived of oxygen and nutrients, often resulting in necrosis or other forms of cell death. Additionally, acidification of the ECM, possibly resulting from increased lactic acid production and secretion by highly upregulated lactate and ATP-driven proton pumps, can alter cellular metabolism owing to the pH-dependence of many enzymes and facilitate the invasion of tumor cells.(33)
Hypoxia in the tumor microenvironment is associated with downregulation of ER, which can contribute to resistance to therapy(34) and lead to poor prognosis. Both tumor and normal cells react to hypoxic conditions through induction of HIF-1.(35) HIF-1α and many of its downstream targets are biomarkers for survival or response to therapy in breast cancer patients.(36) HIF-1α is a transcription factor that is regulated by interaction with oxygen, VHL (a tumor-suppressor gene), α-ketoglutarate (from the tricarboxylic acid cycle), and its binding partner HIF-1β.(10) In addition to playing a role in the cell cycle, proliferation, and angiogenesis, HIF-1α upregulates genes involved in metabolism, including almost all members of the glycolytic cascade, such as glucose transporters, carbonic anhydrase IX, hexokinase II, pyruvate dehydrogenase kinase 1, and LDH.(10) Upregulation of pyruvate dehydrogenase kinase 1 leads to decreased pyruvate dehydrogenase activity and, when combined with upregulation of LDH, increases the conversion of pyruvate to lactate.(20) Activation of HIF-1 in hypoxic regions of both tumor and normal cells can thus lead to the Warburg effect, with increased glycolysis and lactate release.
Fibroblasts in microenvironmental metabolic changes.
The role of fibroblasts in the tumor microenvironment in tumor initiation, progression, and particularly maintenance of the tumor phenotype is receiving increased interest. Cancer-associated fibroblasts (CAF) support tumor growth in multiple models,(37,38) whereas normal fibroblasts are either inactive or tumor inhibitory.
Other cells recruited into the microenvironment include bone marrow-derived mesenchymal stem cells, infiltrating hematopoietic cells,(39) and endothelial cells, all of which possibly interact bidirectionally with tumor cells by using the lactate produced by tumor cells as an energy source.(40)
The TGFβ pathway that activates myofibroblasts during normal wound healing activates normal fibroblasts to generate CAF.(41) Recent studies have proposed the “The Reverse Warburg Effect,” whereby CAF in the tumor environment support tumor growth and progression by altering metabolic processes in the microenvironment, including the production of nutrients such as lactate and tumor-promoting ROS.(42) Lactate produced by the fibroblasts can provide energy for endothelial cells and, potentially, tumor cells. The interaction may be bidirectional, with cancer cells inducing a feed-forward oxidative stress in fibroblasts, which can act as a metabolic and mutagenic motor to drive tumor–stroma co-evolution, DNA damage, and aneuploidy in cancer cells.(43)
Clinical Challenges in Breast Cancer
Is heterogeneity among breast cancers reflected in their metabolic profiles?
Breast cancers show intratumoral heterogeneity, which is reflected in marked differences between primary and metastatic tumors and worsened outcomes for patients with heterogeneity between primary and metastatic tumors.(44,45) Furthermore, many primary tumors appear to contain a large number of independent tumor clones.(46,47) In terms of energy metabolism, the clonal heterogeneity may be important for the cells to adapt to rapidly changing stresses in the microenvironment,(24) which in effect may reprogram tumor cells. For example, monocarboxylate transporter 1, one of the lactate transporter genes that facilitates lactate release and uptake, is upregulated in regions of oxygenation within the tumor, which could in turn contribute to regional changes in extracellular pH.(24) In breast cancer, monocarboxylate transporter 1 transporters are selectively upregulated in basal-like breast tumors, suggesting that metabolic events may vary among different tumor lineages.(48)
Indeed, a recent study elegantly reflected the metabolic heterogeneity of breast cancers. Transcriptomes from human breast cancer cell lines and primary breast tumors were assessed for changes in critical metabolic pathways associated with differential uptake of glucose analog 18F-FDG.(16) Maximum enrichment for glycolysis-related pathways was found in the high-18F-FDG-uptake group that overlapped with basal-like gene signatures, suggesting that ER+ luminal subtypes are less glycolytic than ER− basal breast cancers. The high-18F-FDG-uptake group also showed a systemic increase in Myc expression, supporting the hypothesis that Myc plays a critical role in driving the glycolysis arm of metabolism in subtypes of breast cancer.
Tumor microenvironment and response to therapeutics.
The tumor microenvironment contributes to patient outcomes. For example, the combination of epigenetic aberrations in breast tumor epithelial cells and aberrations in tumor-associated stromal cells and cancer-associated myoepithelial cells predicted a poor outcome for patients.(49) Increased carbonic anhydrase IX is associated with higher rates of relapse in patients treated with adjuvant chemotherapy.(50) Similarly, a stromal signature in breast cancers predicted resistance to neoadjuvant chemotherapy.(51) Although the mechanism by which the stroma alters chemotherapeutic responses is less known, upregulation of endothelial cell progenitors, leading to angiogenesis and tumor repopulation, has been hypothesized.(51)
In human breast cancer patients, decreased stromal cav-1 expression predicted tamoxifen resistance, early relapse, lymph node metastasis, and poor clinical outcome, independent of ER, PR, and HER2 status.(52,53) In contrast, elevated tumor cell cav-1 expression was associated with basal-like and metaplastic breast cancers and poor prognosis. This emphasizes the need to determine whether potential markers are expressed in tumor cells or in cells within the microenvironment.(54) Cav-1 is downregulated by various growth factors, which are present in increased amounts in the tumor microenvironment,(55) potentially contributing to decreased cav-1 levels in CAF. Cav-1-deficient fibroblasts have an activated phenotype with increased PKM2, HIF-1α, and TGFβ pathway signaling.(43) Cav-1 loss in CAF also leads to mitochondrial dysfunction by nitric oxide overproduction, oxidative stress, and aerobic glycolysis, resulting in release of lactate, pyruvate, and 3-hydroxy-butyrate that can provide an energy source for endothelial cells or tumor cells, contributing to both tumor growth and metastasis.(56,57) However, a recent report contradicted this, showing cav-1 enrichment in the stroma of various tumors, including breast tumors, and suggesting that increased cav-1 in CAFs could promote local invasiveness and metastasis through remodeling of the stromal ECM.(38)
Stromal mutations (e.g. p53) can sensitize tumors to doxorubicin or cisplatinum based chemotherapy, suggesting a role for stromal fibroblasts as modulators of efficacy of chemotherapy.(58)
The long-term outcomes of patients with luminal B and basal breast cancers are not substantially different owing to the propensity of hormone receptor-positive tumors to recur decades after the original tumor is removed. Indeed, the outcomes for patients who have survived 5 years is worse for patients with hormone receptor-positive tumors than for those with basal breast cancers. This delayed recurrence is thought to relate to dormancy, wherein cancer cells remain as single or small clusters of non-dividing cells for an extended period of time only to eventually reawaken as a more aggressive, metastatic phenotype.(59) An intriguing concept is that interactions between the stromal and epithelial compartments determine whether micrometastases enter into and remain in a dormant state. The key “switch” proposed in a number of studies is the acquisition of an effective tumor microvasculature and marked changes in the cellular environment.(60) However, studies of dormancy are limited by the lack of appropriate patient samples or animal models that faithfully establish the phenomenon. A key step in understanding the processes underlying entry and exit from dormancy will be characterization of the molecular and genetic events that occur in the tumor cell combined with a detailed analysis of the tumor stroma and the microenvironment under both states.(61)
Exploiting metabolism as a drug target in breast cancer: potential of treatment with metformin.
Multiple epidemiologic and clinical studies have shown that high-calorie diet- and physical inactivity-related overweight and obesity associated with the modern Western lifestyle increase diabetes and hyperinsulinemia, which in turn are linked to increased breast cancer incidence and poor outcome.(62) Treatment with metformin, a widely used antidiabetic biguanide, is associated with decreased breast cancer incidence and breast cancer-related mortality in patients with type 2 diabetes.(63,64) The contribution of metformin to this decrease, independent of insulin, which is a known tumor promoter, remains to be ascertained.
The antidiabetic activity of metformin is primarily due to its inhibition of hepatic glucose production. However, metformin can inhibit tumor cell proliferation both in vitro and in vivo in multiple cancer cell lines, including breast cancer lines,(62) potentially by altering mitochondrial function. One of the adverse effects of the use of metformin is lactic acidosis,(65) which may be caused by increased conversion of pyruvate to lactate and lactate release caused by decreased mitochondrial function.
Although many of the effects of metformin are attributed to its activation of the AMPK pathway, the central sensor for cellular energy levels, metformin does not directly affect AMPK.(66) The effects of metformin are likely secondary to changes in cellular metabolism that lead to altered ATP/AMP ratios and subsequent activation of AMPK. However, metformin can also act in an AMPK/LKB1-independent manner.(67) Other potential mechanisms behind its antitumor effects include: (i) inhibition of mTOR-dependent protein translation, cancer-related lipogenesis, local production of estrogen, and microenvironment-induced inflammation; (ii) p53-dependent cell cycle inhibition; (iii) p53-dependent induction of autophagy and apoptosis; (iv) altered control of mitosis and cytokinesis; and (v) targeting of potential cancer stem cells.(62)
Metformin, by stimulating AMPK, reportedly induced complete growth inhibition of MCF7 breast cancer cell lines.(68) Different studies have also indicated that metformin was therapeutically beneficial in ER− cell line xenografts.(69,70) A number of clinical trials are underway to investigate the effectiveness of metformin treatment and prevention of subtype-specific breast cancer.(63) However, as noted above, heterogeneity within tumors and between tumors and metastases will likely complicate the studies.(44,45) An improved understanding of the relevant targets and functions of metformin in the various breast cancer cell lineages is necessary to determine its optimal clinical utility and, in particular, to identify biomarkers that can identify patients most likely to benefit from metformin.
Current Technological Challenges and Future Directions
One of most powerful techniques currently used to study stroma–epithelium interactions is LCM, which allows selective isolation of defined cellular components from tissues, such as breast epithelial tumor cells.(71) Hill et al. combined LCM with chromatography–mass spectrometry to identify proteins that were differentially expressed in tumor vessels and normal vessels.(72) Among the proteins unregulated in tumor vessels were several involved in metabolism: GAPDH, LDHA, LDHB, and PKM2. Used together with whole gene expression profiles, LCM can be used to study gene expression patterns of specific cell types in breast cancer.(73) However, LCM comes with its own set of challenges such as the limited number of cell samples available for analysis, effects of the capture process on molecules to be studied, and the need for loading controls for the low number of cells assessed. Newer approaches, such as differential hybridization based on sequence differences between human xenografts and murine stromal cells in xenografts(74) and deep sequencing of RNA (also based on sequence differences across lineages), allow characterization of tissue-specific expression patterns without dependence on LCM.
In addition, interactions between stromal cells and tumor cells are often studied in coculture models. Culturing basal breast cancer cell lines with stromal cells increased invasive/migratory phenotype of the cancer cells specifically by increasing the expression of twist and TGFβ, interleukins, cytokines, and Stat1/Stat3.(75) In contrast, the presence of stromal fibroblasts primarily altered proliferation of luminal cell lines,(75) suggesting that this effect is cell type- or lineage-specific.
Finally, a full understanding of the tumor microenvironment requires a detailed analysis of single cells. A number of approaches can be used to analyze single cells; however, the effectiveness of those approaches remains limited. Next-generation sequencing-mediated copy number analysis is used to track cell relatedness in breast cancer.(47) The application of next-generation sequencing to RNA is beginning to produce information on either single cells or small groups of cells in terms of messenger RNA levels and fusions.(76) These tools are more powerful than earlier technologies when linked to microdissection to obtain tumor and stromal information, and they have the potential to elucidate tumor heterogeneity and tumor–stroma interactions on a single-cell level. Approaches that combine information from DNA, RNA, and protein analysis will provide clearer pictures of tumor interactions involving the microenvironment than studies of any one modality in isolation.
Analysis of energy metabolism and metabolomics in the breast tumor and microenvironment may be the key to understanding the interplay between tumor cells and their surrounding stroma. Mass spectrometry-based approaches can currently measure hundreds of metabolites and have the potential to measure thousands of metabolites in patient tumors and interstitial fluids and in model systems. Other approaches such as magnetic resonance spectroscopy and molecular imaging are beginning to provide interesting data on intact tumors in patients.
A combination of emerging genomics and metabolomics technology will greatly increase our understanding of tumor bioenergetics, in particular the interactions between the tumor and the microenvironment. This information will guide the development of biomarkers and clinical trials aimed at targeting bioenergetics and tumor–stroma interactions. These studies are in their infancy but hold great promise for improving patient outcomes.
This work was supported in part by the Kleberg Center for Molecular Markers, Susan B. Komen grant KG081099 (KSH), Komen Foundation grants KG 081694 and FAS0703849, Cancer Prevention and Research Institute of Texas, and Ovarian Cancer Research Fund (GBM). This work was further supported by a Susan G. Komen Post Doctoral Fellowship for Breast Cancer Research to SM. Funding to SC as an Odyssey Fellow was supported by the Odyssey Program and the Theodore N. Law Endowment for Scientific Achievement at the University of Texas MD Anderson Cancer Center and National Cancer Institute Breast SPORE Career Developmental Project award (CA116199). The University of Texas MD Anderson Cancer Center is supported in part by a Cancer Center Support Grant (CA016672) from the US National Institutes of Health.
The authors have no conflict of interest.
adenosine monophosphate-activated protein kinase
human epidermal growth factor receptor 2
laser capture microdissection
liver kinase B1
M2 splice variant (embryonic variant) of pyruvate kinase
reactive oxygen species
transforming growth factor