Identifying the stromal cell type that contributes to tumor aggressiveness associated with carbonic anhydrase IX.

Otto Warburg was the first to postulate a role for cell metabolism in carcinogenesis. Hanahan and Weinberg recently updated their seminal review to include metabolic reprogramming as a hallmark of cancer. While normal cells predominantly depend on mitochondrial oxidative phosphorylation for their energy needs, cancer cells favor aerobic glycolysis, also known as the Warburg effect. This unique metabolic shift provides a survival advantage to the cancer cells in the developing tumor microenvironment and, paradoxically, provides oncologists with potential therapeutic targets. Indeed, metabolic changes have been described as the “Achilles’ heel” of cancer. One such metabolic change is the acidification of the tumor microenvironment by carbonic anhydrases (CAs), especially CAIX. CAIX expression is regulated by the pro-survival transcription factor hypoxia-inducible factor-1α (HIF-1α). CAIX is overexpressed in many tumor types and has been linked to poor prognosis, purportedly due to its involvement in the breakdown of extracellular matrix, protease, and growth factor activation and augmentation of metastatic potential. Previous research has focused predominantly on the metabolic and molecular features of tumor cells, but there is an increasing awareness that stromal cells recruited to the tumor microenvironment are important contributors to the development, progression, and aggressiveness of tumors.

In the June 1, 2013 issue of Cell Cycle, Chiarugi and colleagues demonstrated the role of CAIX-expressing cancer-associated fibroblasts (CAFs) in regulating the epithelial–mesenchymal transition (EMT) of prostate cancer cells. They report that normal human prostate fibroblasts (HPFs) do not express CAIX; however, exposing HPF to conditioned media (CM) from prostate cancer (PCa) cells activates HPF cells to CAFs. CAIX expression was also induced in prostate cancer (PCa) cells treated with CM from CAFs, highlighting the cross-talk between the tumor and its microenvironment (Fig. 1). Interestingly, CAIX was expressed at similar levels in CAFs and serum-starved PCa cells, but PCa cells treated with CM from CAFs expressed higher CAIX levels than CAFs themselves. However, CAIX activity was higher in CAFs compared with PCa cells treated with CAF CM. CAIX expression in both PCa cells and CAFs was HIF-1α-dependent despite these experiments being conducted under normoxic conditions; this observation further supports that the activation of HIF1α signaling was mediated by redox-based stabilization of HIF1α. CAIX inhibition decreased extracellular acidification thereby demonstrating that CAIX is necessary and sufficient for such acidification.

Figure 1. Tumor microenvironmental cross-talk mediates epithelial–mesenchymal transition. (1) Malignant transformation induces the secretion of growth factors; these growth factors activate resident and recruited fibroblasts to CAFs and increase CAIX expression in the tumor microenvironment. (2) Increased expression and activity of CAIX causes extracellular acidification. (3) Extracellular acidification induces MMP secretion. (4) MMP secretion degrades the extracellular matrix and drives epithelial–mesenchymal transition leading to (5) invasion and metastasis.

The role of matrix metalloproteinases (MMPs) in aggressive/metastatic disease and their response to low pH are well documented. Consequently, the authors investigated the link between CAIX and MMP expression. CAIX-induced acidosis increased the expression of MMPs in CAFs, and inhibition of CAIX decreased the secretion of MMP-2 and MMP-9. Inhibition of MMPs reduced the invasiveness of PCa cells. Addition of recombinant MMPs to CAIX inhibited CM rescues ability of PCA cells to undergo EMT. In immune-compromised mice, inhibition of CAIX in CAFs reduced the ability of PCa cells to form viable tumors and effectively metastasise to the lung.

The cellular and mechanistic insights provided by this article are exciting and timely, but it is important that these insights be applied in patient samples to understand the clinical significance of the findings. We have previously reported, in 2 independent head and neck cancer cohorts, that stromal CAIX levels are more strongly associated with poor survival than tumor CAIX., High-stromal CAIX was also associated with increased nodal metastasis. However, we did not identify the specific contributing stromal cell-types. In the future, co-staining tissue micro-arrays with α-smooth muscle actin (a specific marker for CAFs) would potentially improve the definition of the stromal contribution to CAIX expression and association with prognosis. Chiarugi and colleagues report CAFs as the main protagonists in the CAIX-induced tumor aggressiveness, but the role of other cell types in the tumor microenvironment should be investigated. Furthermore, the direct effect of CAIX inhibition in PCa cells needs to be determined.

CAIX is an attractive therapeutic target, because its expression is relatively tumor specific, several low-toxicity pharmaceuticals are available, and novel analogs of existing inhibitors are currently being tested. Given the disappointing results of MMP inhibition trials, targeted reduction of MMP2 and MMP9 by inhibition of CAIX may provide an alternative strategy. Also, the effects of CAIX inhibition on other MMPs and the potential for regulation by other compensatory mechanisms should be addressed.