Cancers co-opt cohabitants’ catabolism

Abstract

Comment on: Capparelli C, et al. Cell Cycle 2012; 11:2272-84 and Capparelli C, et al. Cell Cycle 2012; 11:2285-302.

This article refers to:

CTGF drives autophagy, glycolysis and senescence in cancer-associated fibroblasts via HIF1 activation, metabolically promoting tumor growth

Autophagy and senescence in cancer-associated fibroblasts metabolically supports tumor growth and metastasis, via glycolysis and ketone production

Autophagy is a fundamental cellular process in which surplus or modified cellular components, from individual proteins to whole organelles, are degraded in the lysosomes. Autophagy prevents the accumulation of random molecular damage in long-lived structures, particularly mitochondria, and more generally provides a means to reallocate cellular resources from one biochemical pathway to another. Consequently, it is commonly upregulated in conditions where a cell is responding to stress signals, such as starvation, oxidative stress and exercise-induced adaptation.

While autophagy is a process of damage repair, senescence might be viewed as a state of damage limitation, occurring when the level or nature of damage detected is likely to be irreparable. Triggered either by genotoxicity registered via p53 or via a secondary pathway involving p16^INK4a, the senescent state is characterized by cell cycle arrest, hypertrophy and flattening along with changes in a panel of genetic and proteomic biomarkers, especially the lysosomal protein β-galactosidase.

Although the two processes have distinct functions, recent work from Cambridge has determined that activation of autophagy is both typical of and, in some cases, sufficient to induce senescent transformation, and that inhibition of autophagy delays the acquisition of senescence.¹

Senescent cells are resistant to apoptotic signaling and are known to accumulate in many tissues during aging. Interestingly, significant numbers are frequently found in close proximity to benign tumors. This colocalization has traditionally been seen as a result of senescence programs acting successfully to terminate early-stage tumorigenesis. However, two studies^2,3 published in this issue of Cell Cycle propose a different explanation, presenting compelling evidence for a direct metabolic link between neoplastic cells and the senescent fibroblasts surrounding the tumor. Analysis of clinical data indicates that a high level of autophagy and senescence in the stroma induced in vivo, particularly through hydrogen peroxide secretion by the tumor, correlates with poor prognosis in multiple cancers; however, the mechanism underlying this connection is not well understood.

To dissect the relationship between the autophagic state and tumor growth, hTERT-immortalized fibroblast cultures were transformed with BNIP3, cathepsin B or ATG16L1. Each of the three genes employed was independently able to support a state of constitutively upregulated autophagy, as determined by a panel of biomarkers including downregulation of the membrane protein caveolin-1.

Loss of caveolin-1—which alone is sufficient to induce autophagy in some cell types,⁴ suggesting a feedforward cycle—has been shown to result in ligand-independent activation of the TGFbeta pathway, with a particularly prominent upregulation of connective tissue growth factor (CTGF). The authors demonstrate a key intracellular role for CTGF in inducing and supporting the chronic autophagic state, independent of its well-established role in enhancing extracellular matrix deposition. Specifically, CTGF overexpression is shown to promote HIF-1a signaling, resulting in increased transcription of glycolytic enzymes and components of the autophagic pathway.

The authors hypothesize that chronic autophagy, and particularly mitophagy, eventually results in a failure of mitochondrial respiration and a switch to predominantly aerobic glycolysis. As a result, the fibroblast produces a surplus of high-energy intermediates, including L-lactate and 3-hydroxy-butyrate, which escape into the microenvironment and are taken up by nearby tumor cells as a supplemental energy source. The magnitude of this nutrient transfer is shown to be significant, promoting experimental tumor growth by around 50–100%; surprisingly, metastasis rates are increased even more profoundly (up to 11-fold for ATG16L1-induced fibroblasts).

The accumulation of permanently senescent cells over time is a well-established mechanism of systemic aging, contributing to the increased incidence of a number of conditions—osteoarthritis, atherosclerosis, prostatic hyperplasia, metabolic syndrome and cancer among them.⁵ The secretion of a panel of inflammatory mediators, the “senescence-associated secretory phenotype,” or SASP, is the best-characterized mediator of the pathological effects observed, although it is not a universal feature of p16^INK4a-senescent cell populations.⁶

Although senescence is not always an irreversible phenotype, its undeniably important role in tumor suppression makes attempts to therapeutically reverse it, such as telomerase-activating drugs, particularly fraught. We have for some time advocated the more straightforward approach of selectively destroying (and, where necessary, replacing) senescent cells,⁷ and we are delighted to confirm that this approach has recently been shown to have significant rejuvenating effects in mice.⁸