SEARCH

SEARCH BY CITATION

Keywords:

  • pSer2448-mTOR;
  • TMPRSS2:ERG;
  • PTEN;
  • prostate cancer;
  • tissue microarray

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Prevalence and clinical significance of mammalian target of rapamycin (mTOR) phosphorylation at the serine 2448 is disputed in prostate cancer. A tissue microarray containing 3,261 prostate cancers and 49 normal prostate samples with clinical follow-up data was analyzed for pSer2448-mTOR expression by immunohistochemistry. Moderate to strong pSer2448-mTOR staining was found in all (n = 49) normal prostate tissues, but was lost in 24% or weak in 29% cancers. Moderate and strong staining was found in 36 and 11% of tumors. Loss of pSer2448-mTOR staining was significantly linked to advanced stage (p = 0.0027), high-grade (p = 0.0045), nodal positive cancers (p = 0.0483), early tumor recurrence (p < 0.0001, independently from stage and grade, p = 0.0016), lack of Ets-related gene (ERG) fusion (p < 0.0001), reduced androgen receptor expression (p < 0.0001 each) and increased cell proliferation (p = 0.0092) in all cancers and in the subset of ERG-fusion-positive cancers. Loss of pSer2448-mTOR expression was linked to tumor metastasis (p = 0.0275) in ERG-fusion-positive cancers only. Molecular subset analysis using pre-existing phosphatase and tensin homolog (PTEN) deletion data revealed that loss of pSer2448-mTOR expression is of prognostic relevance and defines a subpopulation of PTEN-deleted and ERG-fusion-positive cancers with a particular poor outcome. The results of our study strongly suggest that loss of pSer2448-mTOR expression is a marker for activated AKT/mTOR signaling. Tumors with concomitant PTEN deletion and activated mTOR signaling indicated by loss of pSer2448-mTOR expression characterize a small (4%) but clinically significant subset of prostate cancers that might optimally benefit from anti-mTOR therapies.

Prostate cancer is the most frequent cancer type in males. About 500,000 patients are diagnosed with this disease every year, and almost 30,000 eventually die from their cancer alone in the United States.1 Although autopsy studies suggest that 50% of males will develop prostate cancer during their lifetime, it is estimated that only a small fraction suffers from potential life-threatening disease that requires radical treatment.2 Accordingly, an important goal of prostate cancer research is to establish molecular markers, enabling distinction between the indolent and the aggressive forms of the disease, and to develop efficient therapies for advanced and metastatic cancers. Prostate cancer is frequently characterized by genomic rearrangements and deletions. The most common rearrangement fuses the androgen-responsive TMPRSS2 serine protease to the ETS family transcription factor Ets-related gene (ERG).3 The TMPRSS2:ERG fusion is the prototype of a number of similar rearrangements resulting in oncogenic activation of ETS genes (reviewed in Ref.4), which have led to a molecular subclassification into fusion-negative and fusion-positive prostate cancers.

Deletion of the phosphatase and tensin homolog (PTEN) gene occurs in 20–70% of prostate cancers, and has been linked to rapid tumor progression and early recurrence.5–7 As PTEN acts as a gatekeeper of the AKT pathway by limiting phosphatidylinositol 3-kinase (PI3K) signaling activity, impaired PTEN function is an important cause of AKT-dependent activation of cell growth and proliferation.8 The mammalian target of rapamycin (mTOR) serine–threonine kinase is the downstream target of AKT signaling. Upon activation, mTOR initiates protein synthesis and cell growth through activation of S6 kinase and eukaryotic initiation factor 4E-binding protein 1.9 Inhibitors of mTOR, for example rapamycin, that have been widely used to suppress transplant rejection and are discussed as potent anticancer drugs,9 for example in renal cell carcinoma,10 neuroendocrine tumors,11 malignant lymphomas,12 breast13 and colorectal cancer.14

Little is known about the prevalence and clinical relevance of pSer2448-mTOR expression, or about its association with ERG fusion, PTEN deletion and mTOR activation in prostate cancer. Phosphorylation of mTOR at serine 2448 has been described as reasonable marker for activated mTOR signaling in some studies.15–18 To comprehensively search for associations between clinical and molecular features of prostate cancer and pSer2448-mTOR expression, we made use of a large prostate cancer tissue microarray (TMA) with attached molecular data on ERG fusion and PTEN deletion.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Patients

A set of prostate cancer TMAs was used in our study containing one tissue core each from 3,261 consecutive radical prostatectomy specimens from patients undergoing surgery at the Department of Urology, and the Martini Clinic, Prostate Cancer Center, University Medical Center Hamburg-Eppendorf as described in detail elsewhere.19 All surgeries were performed according to our institutional standard.20 The composition of the TMA is summarized in Supporting Information Table 1. The database attached to this TMA contains results on ERG expression in 2,874 tumors,21 ERG break apart fluorescence in situ hybridization (FISH) analysis in 1,440 tumors (extended from 21), Ki67 labeling index (LI) in 2,629,22 androgen receptor (AR) expression in 2,665 tumors21 and PTEN deletion or inactivating rearrangement in 1,406 cancers.23 In addition, 49 tissue spots containing normal prostate glands were analyzed. Lack of tumor cells in these samples had been confirmed by immunostaining with the basal cell marker β34E22 (Supporting Information Fig. 1).

Immunohistochemistry

Freshly cut TMA sections were analyzed in 1 day and in one experiment. A phospho-specific mTOR antibody (ser2448, rabbit monoclonal, Cell Signaling Technology, Beverly, MA, 49f9; 1/50; cat#2976) was used for the detection of phosphorylated pSer2448-mTOR. Slides were deparaffinized and exposed to heat-induced antigen retrieval for 5 min in an autoclave at 121°C at pH 2. Bound primary antibody was visualized using the DAKO EnVision Kit (Dako). Staining was typically cytoplasmic and often also membranous (Fig. 1f). Cytoplasmic and/or membranous pSer2448-mTOR expression was scored according to the following system: The staining intensity (0, 1+, 2+ and 3+) and the fraction of positive tumor cells was recorded for each tissue spot. A final score was built from these two parameters as follows. Negative: staining intensity of 0; weak: intensity of 1+ in ≤70% of tumor cells or staining intensity of 2+ in ≤30% of tumor cells; moderate: intensity of 1+ in >70% of tumor cells, or staining intensity of 2+ in >30% but ≥70% of tumor cells or staining intensity of 3+ ≤30% of tumor cells; strong: intensity of 2+ in >70% of tumor cells or staining intensity of 3+ in >30% of tumor cells.

thumbnail image

Figure 1. Representative images of pSer2448-mTOR immunostaining in prostate cancer and prostate normal tissue. (a) Negative, (b) weak, (c) moderate, (d) strong staining in cancers, (e) strong staining in normal prostate epithelium and (f) high magnification of prostate glands showing cytoplasmic and membranous (arrowheads) localization of pSer2448-mTOR.

Download figure to PowerPoint

Statistics

For statistical analysis, the JMP 9.0 software (SAS Institute, NC, USA) was used. Contingency tables were calculated to study the association between pSer2448-mTOR expression and clinicopathological variable, and the Chi-square (Likelihood) test was used to find significant relationships. Kaplan–Meier curves were generated for prostate-specific antigen (PSA) recurrence free survival. The log-Rank test was applied to test the significance of differences between stratified survival functions. Cox proportional hazards regression analysis was performed to test the statistical independence and significance between pathological, molecular and clinical variables.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Technical aspects

PSer2448-mTOR analysis was successful in 2,874/3,261 arrayed patient cancers (88.1%). Immunohistochemistry (IHC) was not informative in 387 (11.9%) tumors because the lack of tumor cells in the tissue spots or missing tissue spots on the TMA section.

PSer2448-mTOR expression in prostate cancer

PSer2448-mTOR immunostaining was moderate to strong in normal prostate epithelium but often reduced in cancers (Fig. 1). pSer2448-mTOR immunostaining was found in 2,178 out of the 2,874 analyzable tumors (76%). pSer2448-mTOR immunostaining was weak in 29%, moderate in 36% and strong in 11% of tumors. The remaining 696 (24%) tumors were pSer2448-mTOR negative. Representative images are shown in Figure 1. Loss of pSer2448-mTOR staining was significantly linked to tumors with advanced stage (p = 0.0027), high Gleason grade (p = 0.0045) and the presence of lymph node metastases (p = 0.0483). PSer2448-mTOR staining was unrelated to the status of the surgical margin (p = 0.7762). Separate analysis of these parameters in subsets of 1,469 ERG-fusion-positive and 1,405 ERG-fusion-negative cancers revealed that these associations held true in ERG-fusion-positive cancers, but were nonsignificant in ERG-fusion-negative tumors. All results are summarized in Table 1.

Table 1. Associations between pSer2448-mTOR immunostaining results and prostate cancer phenotype1
inline image

Associations with other molecular markers

Tumors were grouped into pSer2448-mTOR negative and positive (including weak, moderate and strong staining) to study the associations between loss of pSer2448-mTOR expression and other molecular markers of prostate cancer including ERG fusion, PTEN alterations, Ki67LI and AR expression. Loss of pSer2448-mTOR staining was more frequent in ERG-fusion negative (412/1,405; 29.3%) as compared to ERG-fusion positive (285/1,469; 19.4%, p < 0.0001) if ERG was detected by IHC analysis. Comparable findings were made if genomic ERG rearrangement was detected by FISH using an ERG break apart probe.21 Loss of pSer2448-mTOR staining was more frequent in tumors without ERG rearrangement (132/504; 26.2%) as compared to tumors with ERG rearrangement (78/476; 16.4%, p = 0.0016, Fig. 2a). pSer2448-mTOR expression was unrelated to the presence of PTEN inactivation (p = 0.6422; data not shown). Comparison between expression of AR21 and pSer2448-mTOR revealed a strong positive association (p < 0.0001, Fig. 2b). Loss of pSer2448-mTOR expression was also linked to increased tumor cell proliferation as measured by Ki67LI22 in all cancers (p = 0.0092) as well as in ERG-fusion-positive tumors, (p = 0.0002), but not in ERG-fusion-negative cancers (p = 0.6982, Supporting Information Fig. 2).

thumbnail image

Figure 2. Associations between loss of pSer2448-mTOR immunostaining and (a) ERG fusion by IHC and FISH analysis and (b) AR expression.

Download figure to PowerPoint

Prognostic relevance of pSer2448-mTOR expression

Follow-up data were available from 2,613 cancers with pSer2448-mTOR results. Loss of pSer2448-mTOR staining was significantly linked to early biochemical (PSA) recurrence in all cancers (p < 0.0001, Fig. 3a). This association became even more evident in ERG-fusion-positive cancers (p < 0.0001, Fig. 3b), but was lost in the subset of ERG-fusion-negative tumors (p = 0.2219, Fig. 3c). This association was independent from established prognostic risk factors including pT stage, Gleason grade and nodal stage (p = 0.0001, Table 2) if all cancers were jointly analyzed. To search for possible cooperative effects of mTOR phosphorylation and PTEN, which is a strong known prognosticator and upstream regulator of mTOR activation, we directly compared the prognostic effects of pSer2448-mTORexpression and PTEN deletion alone and in combination. For this analysis, we grouped all cancers according to their pSer2448-mTOR status (negative vs. any positivity) and the PTEN status (deleted vs. normal copy numbers). Loss of pSer2448 expression identified a subset of PTEN-deleted cancers with a particular poor outcome. This was true in all cancers (PTEN-deleted and pSer2448-mTOR-negative vs. PTEN-deleted and pSer2448-mTOR-positive p = 0.0248; Fig. 3d) as well as in ERG-fusion-positive cancers (p = 0.0077; Fig. 3e) but not in ERG-fusion-negative cancers (Fig. 3f).

thumbnail image

Figure 3. Association between of pSer2448-mTOR immunostaining and biochemical (PSA) recurrence in (a) all cancers (n = 2,613), (b) ERG-fusion-positive cancers (n = 1,207), and (c) ERG-fusion-negative cancers (n = 1,148). Combined prognostic impact of pSer2448-mTOR immunostaining (grouped into negative and positive including weak, moderate and strong staining) and PTEN deletion in (d) all cancers z(n = 1,192), (e) ERG-fusion-positive cancers (n = 645) and (f) ERG-fusion-negative cancers (n = 542).

Download figure to PowerPoint

Table 2. COX regression multivariate analysis for the predictive factor biochemical recurrence
inline image

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The results of our study demonstrate that loss of pSer2448-mTOR staining is strongly linked to adverse features of the disease in ERG-fusion-positive prostate cancers. In particular, loss of pSer2448-mTOR was linked to advanced tumor stage, high Gleason grade and the presence of lymph node metastases, increased cell proliferation as well as early tumor recurrence independently from other established markers.

The comparison between normal and cancerous tissues revealed that pSer2448-mTOR expression was often reduced in cancer tissues, and that reduced pSer2448-mTOR levels were linked to adverse tumor features in our study. Three previous studies have used the same antibody as in our study to analyze pSer2448-mTOR expression in normal, benign and cancerous prostate tissues.18, 24, 25 In line with our findings, Evren et al. reported decrease in pSer2448-mTOR expression from high-grade PIN through Gleason 7 to high-grade tumors in 179 prostate cancers.24 Kremer et al. found that pSer2448-mTOR staining was significantly downregulated in cancer as compared to normal prostate tissues,25 corroborating our data. However, conflicting results were published by Dai et al.,18 who reported a positive association between pSer2448-mTOR staining and Gleason grade, as well as higher pSer2448-mTOR staining levels in 182 prostate cancers as compared to 20 cases of benign prostate hyperplasia.

As PTEN is an upstream regulator of mTOR activity,26 we studied the relationship between pSer2448-mTOR expression and PTEN inactivation. We did not find any significant association between these two parameters, suggesting that these events arise independent from each other. However, PTEN-deleted cancers with loss of pSer2448-mTOR expression had a significantly worse outcome as compared to PTEN-deleted tumors with pSer2448-mTOR expression. Importantly, this effect was not seen in cancers with normal PTEN copy numbers and our data show that loss of pSer2448-mTOR expression confers an additional adverse prognostic effect to PTEN inactivated tumors. This effect can be explained by simultaneous activation of two co-operative oncogenic events, namely PTEN loss-driven activation of AKT signaling and mTOR pathway. This further suggests that expression of pSer2448-mTOR is a marker for restrained mTOR activity, whereas loss of Ser2448 phosphorylation indicates full mTOR pathway activation. In line with our findings, Sircar et al. and Dai et al. failed to show associations between PTEN deletion27 or mTOR protein downregulation18 and pSer2448-mTOR expression. Our findings are further supported by functional studies, showing that mTOR Ser2448 is not phosphorylated via PTEN/PI3K/AKT signaling pathway as suggested previously,9 but by the downstream target of mTOR, that is S6 kinase, which controls mTOR activity in a negative feedback manner.28, 29 The associations between loss of pSer2448-mTOR and adverse features of prostate cancer in our study, also including increased cell proliferation, are consistent with such a negative feedback mechanism. Taken together, these findings strongly suggest that serine 2448 phosphorylation of mTOR is a marker for impaired mTOR signaling rather than for its oncogenic activation.

Loss of pSer2448-mTOR expression was strongly linked to ERG-fusion-negative cancers in our study. We used both IHC and FISH analysis to detect ERG alterations. The identical results obtained with FISH corroborate our ERG IHC findings and exclude the possibility that the significant association between loss of pSer2448-mTOR and ERG-fusion-negative prostate cancers might have been biased by general fixation-related variations of immunoreactivity of the arrayed tissues. Although the differences of pSer2448-mTOR expression between ERG-fusion-negative and -positive cancers were only minimal in absolute numbers, significant associations between loss of pSer2448-mTOR expression and prostate cancer phenotype and prognosis were limited to the subset of ERG-fusion-positive cancers. These findings suggest a specific role of mTOR signaling for the biology of ERG-fusion-positive prostate cancers. Several other molecular alterations are linked to the ERG status, including deletions at 3p14 and of PTEN to ERG-fusion positive and deletions of 5q21 and 6q15 to ERG-fusion-negative cancers30–32. These alterations define molecularly distinct subgroups of ERG-fusion-positive or -negative prostate cancers, some of which are linked to poor prognosis23, 30–32. Our study adds mTOR to this growing list of ERG-fusion-linked prognosticators. The cumulative prognostic impact of these genes seems to be comparable in ERG-fusion-positive and ERG-fusion-negative cancers as these subgroups do not differ overall in terms of clinical outcome as reported by our group previously.21

AR signaling is a prerequisite for the development of ERG fusions,33, 34 and activated mTOR is a negative regulator of AR signaling.35, 36 Consequently, tumors harboring activated mTOR are expected to have less ERG fusions owing to impairment of AR signaling. In fact, ERG-fusion-negative tumors were frequently characterized by loss of pSer2448-mTOR expression, indicating activated mTOR and impaired AR signaling. The strong association between reduced AR expression and loss of pSer2448-mTOR staining in our tumor set directly supports this assumption.

At present, clinical trials are undertaken in several cancer types to estimate the efficacy of mTOR inhibition using inhibitors of mTORC1 or mTORC1 and mTORC2.37, 38 The results of our study suggest that the small (4%) but clinically highly relevant fraction of prostate cancers characterized by loss of pSer2448-mTOR expression, ERG fusion and PTEN deletion might optimally benefit from such therapies (Fig. 4). In addition, combination of mTOR therapy with AR inhibitors might be recommended to avoid loss of AR signaling control owing to mTOR inhibition. In line with this notion, it has been shown that rapamycin monotherapy has little effect in prostate cancer, whereas parallel application of AR inhibitors has provided better results.35 This is also consistent with the results of an in vitro study, showing that rapamycin has growth inhibitory effects only in AR-independent C42 cells (an AR-independent clone derived from LNCaP) but not in the androgen-dependent LNCaP cells.39

thumbnail image

Figure 4. Therapeutic implication of the association of loss of pSer2448-mTOR expression with ERG fusion and PTEN deletion. The subset of 4% ERG-fusion-positive (ERG+)/PTEN deleted/pSer2448-mTOR-negative prostate cancers that might optimally benefit from anti-mTOR treatment is highlighted.

Download figure to PowerPoint

In summary, the results of our study suggest that pSer2448-mTOR expression is a marker for physiologically balanced or even impaired mTOR signaling activity in prostate cancer. Loss of pSer2448-mTOR expression, together with ERG fusion and PTEN deletion, characterizes a molecularly distinct subgroup of tumors with particular aggressive features and early PSA recurrence that might optimally benefit from anti-mTOR therapies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank Christina Koop, Julia Schumann, Sünje Seekamp, and Inge Brandt for excellent technical assistance.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
IJC_27768_sm_SuppFig1.tif14861KSupporting Information Figure 1.
IJC_27768_sm_SuppFig2.tif11823KSupporting Information Figure 2.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.