Background
Prostate cancer is a major cause of cancer-related mortality and morbidity in males [
1]. Although the majority of prostate cancers present as low malignant, indolent tumors, there is also an aggressive subset. For example in Germany, about 60,000 new cases of prostate cancer are diagnosed every year, and still about 11,000 patients die from their disease [
2]. The common pre-operative parameters including Gleason grade, tumor extent in biopsies, and preoperative prostate specific antigen (PSA) levels are statistically powerful prognosticators, however insufficient to allow for optimal treatment decisions in individual patients. Accordingly, there is a considerable need for improved diagnostic tools to early distinguish these patients requiring aggressive therapy with all its associated side effects from the majority of patients who will not. It is hoped, that advances in basic prostate cancer research will eventually lead to novel prognostic biomarkers and better therapeutic options.
The growing interest in mitochondrial function and dysfunction reflects the potential role of mitochondria for cancer development [
3]. Loss of proliferation control in cancer cells eventually results in cellular bulks that extend beyond the capacity of their vasculature, resulting in oxygen and nutrient deprivation. Accordingly, tumor growth is accompanied by cellular adaptations to overcome these limitations. Mitochondria are key organelles for energy production including glucose metabolism and oxidative phosphorylation with a critical role in cell survival and apoptosis. Amount and activity of mitochondria may hence play a critical role in tumor initiation and progression [
4], and it is not surprising that mutations of mitochondrial genes or alterations of the mitochondrial content have been suggested to play an important role in various cancer types [
5‐
8]. As a consequence, an increasing number of anti-cancer drugs is under development [
9‐
11] targeting mitochondria and associated structures. Some studies have even suggested that intracellular accumulation of mitochondria (the so-called mitochondrion-rich phenotype) might represent an important adaptive mechanism in rectal and breast cancer [
12,
13].
Mutations of mitochondrial DNA were also identified in prostate cancer [
14‐
22] and deregulated mitochondrial metabolism has been suggested to play a relevant role in prostate carcinogenesis [
23‐
25]. Based on these reports, we hypothesized that also the quantity of mitochondria present in prostate cancer cells might be of clinical relevance, and that the cellular mitochondria content might vary between prostate cancer subgroups harboring different key molecular alterations that might influence cell metabolism.
The antibody MTC02 (mouse monoclonal to mitochondria) recognizes a 60 kDa non glycosylated protein component of mitochondria found in human cells, and has been used to determine the mitochondrial content of tumor cells in a variety of previous studies. For example, earlier studies used MTC02 to determine the molecular genetic alterations [
13] and the frequency, morphology and clinical features of mitochondrion-rich breast cancers [
26]. A tissue microarray (TMA) containing 11,152 prostate cancer specimens with clinical follow-up information and an attached molecular database was analyzed in order to evaluate the clinical significance of mitochondria content, and to search for possible associations with molecularly defined cancer subgroups. Our study demonstrates that “mitochondrion-rich phenotype” is strongly associated with molecular cancer features and strongly linked to poor prognosis in ERG negative prostate cancers.
Discussion
The results of our study show, that the mitochondria content is tightly linked to various pathological, molecular features of prostate cancer. This data highlight the prominent importance of mitochondrial function for prostate cancer development and progression.
Immunohistochemical detection of a 60 KDa non-glycosylated protein component of mitochondria was utilized in this project to quantitate mitochondria in cancer cells on TMAs. The TMA approach is optimal for the identification of subtle staining differences of proteins that are abundantly present in cancer, such as mitochondrial components, because TMAs enable maximal experimental standardization. In this study, more than 10,000 prostate cancer specimens were analyzed in one day in one experiment using one set of reagents at identical concentrations, temperatures and exposure times. Moreover, all TMA sections were cut on one day immediately before staining in order to avoid unequal decay of a tissues reactivity to antibody binding [
37]. Finally, one pathologist interpreted all immunostainings at one day to enable maximal standardization of staining interpretation. In earlier studies, our TMA enabled us to validate several biomarkers with importance for prostate cancer, such as p53 expression [
38],
PTEN inactivation [
32], CRISP3 overexpression [
39] or deletions at 6q15 [
31] and 5q21 [
30].
The data derived from this approach demonstrate a marked increase of mitochondria content from normal prostate epithelial cells to cancer cells. A further increase was observed with increasing tumor grade and stage, suggesting that higher numbers of mitochondria are necessary or supportive for cancer development and progression. This is also supported by the observation that the mitochondrial content was linked to increased cell proliferation. Our findings are consistent with recent studies suggesting a prominent role of mitochondria content in cancer. For example, Ambrosini-Apaltro et al. [
12] detected oncocytic, mitochondrion-rich modifications in adenocarcinoma cells after radiochemotherapy and Ragazzi et al. [
26] described a link between mitochondrion-rich and undifferentiated breast cancers. Despite the early belief that cancer metabolism is primitive and inefficient, it has now become evident that cancer cells actively reprogram their metabolism activity [
40]. Adaptation of cellular metabolism towards macromolecular synthesis is critical to supplying sufficient amounts of nucleotides, proteins, and lipids for cell growth and proliferation, which are fundamental to cell growth and proliferation [
40]. Accordingly, previous studies described interactions between the mitochondrial metabolism and the activity of growth signaling pathways involving key human oncogenes such as Myc, Ras, Akt and phosphoinositide 3-kinase (Pi3K) [
41‐
43]. Activated PI3K/Akt leads to enhanced glucose uptake and glycolysis [
44,
45] by induction of glucose transporters, mitochondrial enzymes involved in the glycolytic metabolism and glucose carbon flux into biosynthetic pathways [
46‐
49]. Downstream of PI3K/Akt, the well-characterized cell growth regulator mTORC1 also has many effects intertwined with mitochondrial metabolism [
50‐
52]. Taken together, these findings demonstrate that the reprogramming of mitochondrial metabolism is a central aspect of PI3K/Akt associated oncogenic activity.
The large number of tumors analyzed in this study enabled us to separately analyze cancer subgroups defined by molecular features, the most common of which is the
TMPRSS2:ERG gene fusion. Gene fusions involving the androgen-regulated gene
TMPRSS2 and
ERG, a member of the
ETS family of transcription factors, occur in about 50% of prostate cancers, especially in young patients, and result in strong ERG protein overexpression [
53‐
55]. Our data demonstrate that high mitochondrial content is significantly linked to fusion type prostate cancer. Finding this association by two independent approaches for
ERG fusion detection (IHC/FISH) largely excludes a false positive association due to inefficient immunostaining for both MTC02 and ERG in a subset of damaged non-reactive tissues. This finding strongly argues for generally increased energy demands of ERG positive as compared to ERG negative cells. ERG expression causes massive deregulation of the global expression patterns in prostate cells. Several studies analyzing the transcriptomes of
ERG positive and
ERG negative tumors revealed that multiple energy-consumptive signaling pathways are activated as a result of ERG expression, including ER-, TGF-ß, WNT, PI3K/Akt and Myc signaling [
56‐
60] all of which involve multiple ATPases and ATP-dependent kinases. Particularly PI3K/Akt and Myc signaling also directly activates glycolysis [
43] and induces transcription of numerous glycolytic enzymes [
4] in cancer cells.
That mitochondrial content has a different role and function in ERG positive and ERG negative cancers is further supported by our ERG-stratified analysis of disease outcome. Mitochondrial content had a prognostic role in ERG negative but not in ERG positive cancers. This striking difference may be caused by the substantial increase of cellular mitochondrial content by ERG rearrangements, which by themselves do not have any prognostic impact on prostate cancers. The magnitude of ERG-induced molecular and cellular changes, at least most of which are unrelated to cancer progression, may lead to an increased mitochondrial content in “fusion-type” prostate cancers, that masks demand for higher mitochondria content caused by specific molecular “progression events” requiring more mitochondrial function. The strong prognostic impact of mitochondria content in ERG negative prostate cancers fits well with models suggesting, that in a surrounding with low mitochondria content, “progression events” requiring more mitochondrial function would rather lead to a detectable increase of the mitochondria count, than in an environment with high mitochondria content.
Deletions of PTEN, 5q21 and 6q15 represent such “progression events” in prostate cancer as all of them are strongly linked to tumor growths and adverse clinical features. It seems likely that a shortage of nutrients and oxygen typically occurring during tumor expansion will eventually trigger additional adaptation steps, and increase of the mitochondrial content might be one of these. That such an increase of the mitochondrial content was not observed for 3p13 deletions may be due to the low number of analyzed ERG negative tumors for this deletion. Alternatively, it might be due to the small number of genes affected by these small 3p13 deletions, none of which may lead to additional “energy demand” in case of inactivation. A role of PTEN inactivation as a “progression event” associated with higher requirements for mitochondrial function is further supported by the observation that high mitochondrial content loses its prognostic relevance in PTEN deleted ERG negative cancers.
It is of note that the relationship between all analyzed deletions (
PTEN, 3p13, 5q21, 6q15) and the mitochondria content tended to invert within ERG-positive cancers. The causes for this observation cannot be deducted from our data. It might be speculated, that non-vital ERG induced mitochondria production is restrained under a different cellular environment driving towards tumor progression including more rapid tumor cell growth. More specifically, it may be possible that specific molecular events caused by chromosomal deletions interfere with ERG induced general upregulation of number of mitochondria. It has indeed been shown, that
PTEN inactivation can directly trigger both glycolysis [
61] and mitochondrial respiratory capacity [
62] through AKT/mTOR signaling activation.
The marked prognostic relevance of mitochondrial abundance found in the subset of ERG negative cancers may suggest “mitochondria content” as a biomarker with potential clinical utility. This notion is further supported by the fact that the prognostic impact of mitochondria content was found on a TMA containing just one 0.6 mm cancer sample per patient. This approach of analyzing molecular features closely models the molecular analyses of core needle biopsies where comparable amounts of tissues are evaluated. Various models of multivariate analyses applied in this project indeed suggested an independent predictive role of mitochondria content for prognosis if only parameters were utilized that are available before radical prostatectomy. These data must be interpreted with caution, however, because the MTC02 immunostaining was done on tissue from radical prostatectomies and not on the core needle biopsies that were used to determine the preoperative Gleason grading. It is obvious, that potential prognostic biomarkers should be evaluated on preoperative needle biopsies but from a practical point of view such analyses are hardly feasible. This is because needle biopsies are usually done at many different facilities and not accessible for studies. Moreover, if such precious core needle biopsies were available, they would be exhausted after only few studies. Independent of this, it might be rewarding to further consider mitochondria content as a potential feature in multiparametric prognostic prostate cancer tests.
In summary, the results of our study highlight a different role of mitochondrial content in ERG fusion-positive and -negative cancers and identify “mitochondrial abundance” as a potential prognostic feature in ERG-negative cancers. Strong associations between chromosomal deletions and the cellular mitochondrial content further highlight the important role of mitochondria content as an adaptation process during cancer progression.
Competing interests
The authors declare that they have no competing interest.
Authors’ contribution
GS, TS, JI, SS, RS and SM designed research; KG, KJ, MCT, CK, WW, AQ and MA performed the experiments and analyzed the data. GS, RS, KG, HH and MG wrote the manuscript. All authors read and approved the manuscript.