See article by Ulmert et al., p. 320.

  • A radiolabeled antibody recognizing free PSA localizes to prostate cancer xenografts.

  • Decreased PSA production triggered by antiandrogen therapy can be tracked noninvasively.

  • Bone metastases can be distinguished from nonmalignant skeletal pathologies.

Testing for prostate-specific antigen (PSA) levels in serum is widely used as a clinical prostate cancer biomarker, albeit moderately elevated PSA-levels in blood do not reliably distinguish between normal and malignant prostate tissue and do not consistently reflect clinical outcome. Because the small percentage of tumor-associated PSA that is released into circulation is rapidly and irreversibly converted to an inactive form, Ulmert and colleagues reasoned that selective targeting of the initially produced active, “free” form of PSA using a radiolabeled monoclonal antibody (89Zr–5A10) would allow a more accurate, imaging-based clinical assessment of advanced prostate cancer. The 89Zr–5A10 radiotracer was intravenously injected and specifically taken up by the tumor tissue of mice harboring subcutaneous PSA-positive prostate cancer xenografts within 24 hours. Changes in PSA production driven either by testosterone exposure or pharmacologic androgen receptor inhibition could also be quantitatively measured by 89Zr–5A10. Furthermore, because current imaging agents cannot distinguish between metastatic bone lesions and normal skeletal repair caused by injury or degenerative disease, the authors tested whether 89Zr–5A10 localized to murine tibias that were either fractured or injected with prostate cancer cells. Positron emission tomography scans revealed that 89Zr–5A10 specifically localized to tumor-bearing hindlimbs, but did not localize to sites of bone repair like other radiotracers such as 18F-NaF or 99mTc-MDP. Together, these findings establish that a PSA-targeted radiotracer can selectively, noninvasively detect and visualize primary and metastatic prostate tumors and quantify responses to antiandrogen treatment.

See article by Ros et al., p. 328.

  • PFKFB4 is required for prostate tumor growth and cell survival in lipid-depleted conditions.

  • Silencing of PFKFB4 increases Fru-2,6-BP levels and induces ROS accumulation.

  • Exploiting a metabolic dependence on PFKFB4 may be an effective therapeutic strategy.

Cancer cells are often dependent on acquired metabolic alterations that accommodate their increased biosynthetic demands. Ros and colleagues performed a small interfering RNA-based screen of over 200 metabolic enzymes, transporters, and regulators to identify those selectively required for prostate cancer cell survival. One of the genes identified, PFKFB4, encodes an isoform of 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase (PFK2), which modulates the levels of fructose-2,6-bisphosphate (Fru-2,6-BP), a key allosteric activator of the most important regulatory enzyme of glycolysis, phosphofructokinase 1 (PFK1). Knockdown of PFKFB4 blocked prostate cancer cell growth in lipid-depleted conditions and induced regression of prostate tumor xenografts, confirming that prostate cancer cells are dependent on PFKFB4 for survival. Following PFKFB4 loss, levels of Fru-2,6-BP were significantly increased in prostate cancer cells, which indicated that metabolic intermediates may be diverted toward glycolysis and away from the pentose phosphate pathway that generates the NADPH required for lipid biosynthesis and cellular redox balance. Indeed, PFKFB4 knockdown significantly decreased cellular NADPH levels, which reduced lipid synthesis and led to an accumulation of reactive oxygen species (ROS) due to decreased antioxidant levels. Treatment with a chemical antioxidant rescued the viability of PFKFB4-deficient prostate cancer cells, further suggesting that PFKFB4 mediates ROS detoxification in cancer cells. Together, these findings suggest that prostate cancer cells are exquisitely sensitive to metabolic perturbations that affect the balance between glucose and the pentose phosphate pathway and implicate PFKFB4 as a potential therapeutic target.

See article by Martin et al., p. 344.

  • Metformin increases tumor growth and angiogenesis through VEGF-A activation.

  • Metformin disrupts ERK pathway feedback inhibition in a BRAF-mutant setting.

  • VEGF inhibitors synergize with metformin to block the growth of BRAF-mutant tumors.

Metformin, a first-line treatment for type 2 diabetes, also lowers cancer risk in diabetic patients and has demonstrated antitumor activity in preclinical testing. Metformin activates AMP-activated protein kinase (AMPK), a key regulator of cellular energy homeostasis. Suppression of mTOR signaling via AMPK-dependent TORC1 inhibition is thought to underlie the ability of metformin to suppress cancer cell growth. Given the need for improved strategies to circumvent drug resistance mechanisms in BRAF-mutant cancers, Martin and colleagues studied the effects of metformin on melanoma cells. BRAF-mutant, but not NRAS-mutant melanoma cells were resistant to metformin in vitro due to constitutive TORC1 activation mediated by RSK. However, in vivo, metformin accelerated the growth of BRAF-mutant melanoma xenografts and significantly increased the size and number of blood vessels within the tumors. The increased tumor growth and angiogenesis was concomitant with VEGF-A upregulation, which the authors deduced was ERK-dependent and caused by AMPK-mediated proteosomal degradation of DUSP6, a negative regulator of ERK. Inhibition of VEGF with a small-molecule inhibitor of VEGFR (axitinib), a neutralizing VEGF antibody (bevacizumab), or stable expression of a VEGF-A short hairpin RNA synergized with metformin treatment to reduce the growth of BRAF-mutant xenografts. These findings have clinical implications for the treatment of diabetic patients with melanoma and suggest that metformin–VEGF inhibitor combination therapy may be an effective second-line strategy for the treatment of BRAF-mutant cancers.

See article by Subbaramaiah et al., p. 356.

  • Obesity-related breast inflam-mation correlates with elevated COX-2 levels.

  • COX-2 activates a PGE2–cAMP–PKA pathway that induces aromatase expression.

  • COX-2 or aromatase inhibitors may reduce breast cancer risk in obese women with breast inflammation.

Although obesity is a known risk factor for hormone receptor-positive breast cancer in postmenopausal women, the association between excess body fat and the hormonal mechanisms that promote cancer development and progression remain unclear. Subbaramaiah and colleagues identify a link between the inflammation commonly observed in the breast adipose tissue of obese women and increased activity of aromatase, the rate-limiting enzyme for estrogen synthesis. A hallmark of breast inflammation is the presence of crown-like structures of the breast (CLS-B), which are formed by infiltrating macrophages that surround necrotic adipocytes and secrete proinflammatory molecules. The authors found that elevated levels of cyclooxygenase-2 (COX-2) and prostaglandin E2 (PGE2) in breast tissue as well as increased aromatase expression and activity were more strongly correlated with a higher CLS-B index than with a high body mass index. A strong correlation was also found between breast inflammation, levels of cyclic AMP (cAMP) and protein kinase A (PKA) activity, and PGE2, known determinants of increased transcription of the aromatase gene, CYP19. In vitro studies confirmed that saturated fatty acids, products of obesity-related lipolysis, induce a COX-2–dependent increase in PGE2 secretion by macrophages that is required for cAMP induction, PKA stimulation, CYP19 transcriptional activation, and increased aromatase activity in preadipocytes. Collectively, these findings establish a link between obesity-related inflammation and estrogen synthesis and provide a molecular rationale for the use of diet and COX-2 inhibitors or other anti-inflammatory agents as well as aromatase inhibitors to reduce the risk of hormone receptor-positive breast cancer.

See article by Birkbak et al., p. 366.

  • Cisplatin-sensitive cells harbor regions of allelic imbalance that extend to the telomere.

  • Low BRCA1 levels are associated with telomeric allelic imbalance and cisplatin sensitivity.

  • High telomeric allelic imbalance may be a readout of DNA repair deficiency.

Cancer cells with defective DNA repair mechanisms display increased sensitivity to platinum-based chemotherapeutic agents that form interstrand DNA crosslinks, stall replication forks, and induce double-strand breaks. Birkbak and colleagues analyzed single-nucleotide polymorphism array data to identify regions of allelic imbalance—an unequal contribution of maternal or paternal alleles that can occur due to defective DNA repair—and determine whether such genomic aberrations could predict cisplatin sensitivity in cancers with and without known mutations in the DNA repair genes BRCA1 and BRCA2. The number of regions of allelic imbalance extending to a telomere was specifically associated with cisplatin response in a panel of breast cancer cell lines, patients with triple-negative breast cancer treated with cisplatin in 2 separate clinical trials, and platinum chemotherapy–treated serous ovarian cancers included in The Cancer Genome Atlas. The chromosome breakpoints that marked the boundary of these regions of allelic imbalance extending to a telomere frequently occurred near regions of copy number variation, suggesting that tumors with high levels of telomeric allelic imbalance are prone to errors in DNA repair associated with replication of repetitive sequences. Consistent with this possibility, low BRCA1 expression was associated with telomeric allelic imbalance and cisplatin response in patients with sporadic triple negative breast cancer in the 2 clinical trials. These findings suggest that high levels of telomeric allelic imbalance may be an indicator of DNA repair deficiency and may identify patients likely to respond to platinum-based chemotherapy.

Note:In This Issue is written by Cancer Discovery Science Writers. Readers are encouraged to consult the original articles for full details.