The prognostic impact of Akt isoforms, PI3K and PTEN related to female steroid hormone receptors in soft tissue sarcomas

Background The PI3K/Akt pathway is involved in cellular survival pathways by inhibiting apoptotic processes and stimulating cell growth and proliferation. Its negative prognostic value has been proven in many types of cancer. In soft tissue sarcomas, the expression profiles of the PI3K/Akt pathway components are poorly defined and their significance uncertain. We aimed to investigate the prognostic impact of Akt (Akt1) phosphorylated at threonine308 and serine473, Akt2, Akt3, PI3K and PTEN, alone and in coexpression with ER and PgR in non-gastrointestinal stromal tumor soft tissue sarcomas (non-GIST STSs). Patients and methods Tumor samples and clinical data from 249 patients with non-GIST STS were obtained, and tissue microarrays (TMAs) were constructed. Immunohistochemistry (IHC) was used to evaluate marker expression in tumor cells. Results In univariate analyses, the expression levels of p-Akt Thr308 (P = 0.002), Akt2 (P = 0.008) and PI3K (P < 0.001) were significant prognostic factors. In the multivariate analysis, high PI3K expression was an independent negative prognosticator (HR = 1.5, 95% CI = 1.0-2.2, P = 0.042) in addition to advanced age, tumor depth, high malignancy grade, metastasis at diagnosis, surgery and positive resection margins. p-Akt Thr308 expression had strong unfavorable effect in men only (P = 0.009). In contrast, p-Akt Ser473 expression had strong unfavorable impact in women (P = 0.023). PgR-/p-Akt Ser473+ phenotype tended to have less favorable impact in women (P = 0.087), but was the most favorable one in men (P = 0.010). Conclusion Expression of PI3K was significantly associated with aggressive behavior and shorter DSS in non-GIST STSs. The site of Akt phosphorylation seems to have gender-dependent impact on survival in STS patients.


Background
Soft tissue sarcomas (STS) are malignant tumors arising from extraskeletal connective tissues. They are heterogeneous neoplasms, consisting of more than 50 subtypes, and comprise less than 1% of adult malignancies [1,2]. Approximately 50% of the STS patients will succumb to their disease because of metastasis or local progression [3]. The prognostic factors determining tumor evolution and ultimately patients' fate include tumor grade, size, location, depth, histological entity, positive resection margins and presence of local relapse [4][5][6][7][8][9][10]. In addition, an array of recurrent gene aberrations are found to be prognostic and predictive biomarkers in STSs [11][12][13].
Akt is a serine/threonine protein kinase that exists in three possible isoforms, including Akt1, Akt2, and Akt3. Akt can be activated by phosphorylation at threonine 308 or at serine 473 for Akt1 or homologous sites for Akt2 and Akt3 by phosphatases which along with Akt isoforms, belong to the phosphoinositide 3-kinase (PI3K)/ Akt pathway. The PI3K/Akt pathway has been linked to an extraordinarily diverse group of cellular functions, including cell growth, proliferation, differentiation, motility, survival, intracellular trafficking and angiogenesis [14]. Both PI3K and Akt isoforms have been implicated as major players in many types of cancer [15][16][17].
The PI3K/Akt pathway seems to be more often deregulated in cancer than any other pathway [18]. However, in the literature there is disagreement regarding the prognostic impact of Akt expression. While the majority of studies agree that Akt expression overtly indicates a poor prognosis [19][20][21], there are several studies showing the opposite effect [22,23]. Expressions of PI3K/Akt pathway components have rarely been investigated in STSs and there are almost no studies devoted to their prognostic value [24].
Different physiological function of the Akt family kinases implies that the expression of its isoforms may also have different prognostic impact in cancer. The significance of this variation for the survival of the STS patients is not well investigated and it is not clear whether the site of phosphorylation and the pattern of expression can play prognostic roles.
In previous studies, we have shown the prognostic value of female steroid hormone receptors in STSs, both alone and in the coexpression with TGF-β and fascin [25,26]. Such prognostic impact is not surprising, since both ER and PgR regulate growth and cell differentiation upon ligand-dependent and ligand-independent activation and are in essence growth factors. In this context we wanted to explore the correlations between female hormone receptors and the members of PI3K/Akt signaling pathway. To our knowledge, these correlations have not been described previously.
In this study, we investigate the prognostic impact of all isoforms of Akt (phosphorylated at threonine 308 and Akt phosphorylated at serine 473 , non-phosphorylated Akt2, and total Akt3), PI3K, PTEN, ER and PgR in 249 non-GIST STS patients. GIST cases were excluded from the study since patients with this subtype of sarcoma receive a specific treatment regimen which resulted in significantly better survival.

Patients and clinical samples
Primary tumor tissue from anonymized patients diagnosed with non-GIST STS at the University Hospital of North Norway (UNN) 1973-2006 and The Hospitals of Arkhangelsk region, Russia, were used in this retrospective study. In total, 496 patients were registered from the hospital databases. Of these, 247 patients were excluded due to missing clinical data (n = 86) or inadequate material for histological examination (n = 161). Thus, 249 STS patients with full clinical records and adequate paraffin-embedded tissue blocks were eligible.
This report includes follow-up data as of September 2009. The median follow-up was 38 (range 0.1 -392) months. Formalin-fixed and paraffin-embedded tumor specimens were obtained from the archives of the Departments of Pathology at UNN and the Arkhangelsk hospitals. The tumors were graded according to the French Fèdèration Nationales des Centres de Lutte Contre le Cancer (FNCLCC) [27].

Microarray construction
All sarcomas were histologically reviewed by two trained pathologists (S.S. and A.V.) and the most representative areas of viable tumor cells (neoplastic cells) were carefully selected and marked on the hematoxylin and eosin (H&E)-stained slides and sampled for the tissue microarray blocks (TMAs). The TMAs were assembled using a tissue-arraying instrument (Beecher Instruments, Silver Springs, MD). The Detailed methodology has been previously reported [28]. Briefly, we used a 0.6 mm diameter stylet, and the study specimens were routinely sampled with two replicate core samples (different areas) of neoplastic tissue. To include all core samples, 12 tissue array blocks were constructed. Multiple 4-μm sections were cut with a Micron microtome (HM355S) and stained using specific antibodies for immunohistochemistry (IHC) analyses.

Immunohistochemistry (IHC)
The applied antibodies were subjected to in-house validation by the manufacturer for IHC analysis on paraffin-embedded material. The applied antibodies had been subjected to in-house validation by the manufacturer for IHC analysis on paraffin-embedded material. The antibodies used in the study were as follows: Phospho-Akt (Ser473) ( Sections were deparaffinised with xylene and rehydrated with ethanol. Antigen retrieval was performed by placing the specimen in 0.01mol/l citrate buffer at pH 6.0 and exposed to two repeated microwave heatings of 10 minutes at 450W. The DAKO EnVision + System-HRP (DAB) kit was used as endogen peroxidase blocking. Primary antibodies were incubated overnight at 4°C (except PI3K, for 32 minutes at room temperature). The DAKO EnVision+ System-HRP (DAB) kit was used to visualize the antigens for all stains. This yielded a brown reaction product at the site of the target antigen. As negative staining controls, the primary antibodies were replaced with the primary antibody diluent. Finally, all slides were counterstained with hematoxylin to visualize the nuclei. For each antibody, including negative controls, all TMA staining were performed in one single experiment. The immunohistochemical staining for ER and PgR was performed as described earlier [25].

Scoring of IHC
The ARIOL imaging system (Genetix, San Jose, CA) was used to scan the slides with immunohistochemically stained TMAs. The specimens were scanned at a low resolution (1.25×) and high resolution (20×) using Olympus BX 61 microscope with an automated platform (Prior). The slides were loaded in the automated slide loader (Applied Imaging SL 50). Representative and viable tissue sections were scored manually on computer screen, semiquantitatively for cytoplasmic staining for PI3K/Akt pathway components and for nuclear staining for ER and PgR. The dominant staining intensity in neoplastic cells was scored subjectively as: 0 = negative; 1 = weak; 2 = intermediate; 3 = strong ( Figure 1). For ER and PgR, the modified All Red scoring system [25] was used. All samples were anonymized and independently scored by two pathologists (A.V. and S.S.). In cases where score difference was equal to or exceeding 2, the slides were re-examined and a consensus was reached by the observers. When assessing a score for a given core, the observers were blinded to the scores of the other variables and to outcome. Mean score for duplicate cores from each individual was calculated.

Statistical methods
All statistical analyses were done using the statistical package SPSS (Chicago, IL), version 16. The IHC scores from each observer were compared for interobserver reliability by use of a two-way random effect model with absolute agreement definition. The intraclass correlation coefficient (reliability coefficient) was obtained from these results. The Chi-square test and Fishers Exact test were used to examine the association between molecular marker expression and various clinicopathological parameters. Univariate analyses were done by using the Kaplan-Meier method, and statistical significance between survival curves was assessed by the log rank test. Disease-specific survival (DSS) was determined from the date of histological confirmed STS diagnosis to the time of STS death. To assess the independent value of different pretreatment variables on survival, in the presence of other variables, multivariate analysis was performed using the Cox proportional hazards model. Only variables with value 0.10 or less from the univariate analysis were entered into the Cox regression analysis. The significance level used in both univariate and multivariate analyses was P < 0.05, but in the post hoc subgroup analysis the significance level was moved from P = 0.05 to P = 0.01 due to risk of false positivity.

Ethical clearance
The National Cancer Data Inspection Board and The Regional Committee for Research Ethics approved the study. The Regional Committee approved that written consent from the patients for their information to be stored in the hospital database and used for research was not needed because most of the material was more than 20 years old and most of the patients are now dead. The material was collected from our approved biobank for paraffin-embedded material and slides. All material was anonymously collected. The data were analyzed anonymously.

Expression pattern and correlations with clinicopathological variables
In the immunohistochemical analyses, we used antibodies against all Akt isoforms, including Akt phosphorylated at Ser 473 and at Thr 308 , non-phosphorylated Akt2 and total (both phosphorylated and non-phosphorylated) Akt3. Besides, we investigated expression of total PI3K and PTEN. The p-Akt Ser 473 , p-Akt Thr 308 , Akt2, Akt3, PI3K and PTEN showed expression in the cytoplasm or both in the cytoplasm and in the nuclei of tumor cells in the majority of cases, while pure nuclear staining was demonstrated in a smaller proportion of the tumors, varying from 7% of all immunohistochemically positive tumors for PTEN to 19% for p-Akt Thr 308 and Akt3.
In order to find out whether subcellular location of proteins belonging to the Akt/PI3K signaling pathway has impact on survival, we performed a series of univariate analyses to compare the impact of their expression in nucleus, cytoplasm or both. Nuclear expression of p-Akt Thr 308 expression showed a significantly favorable prognosis (P = 0.029), compared to cytoplasmic and  especially mixed cytoplasmic and nuclear expression, Figure 2, D. The other factors did not show any significant prognostic differences in the subcellular location. Subgroup analysis based on clinical variables revealed that high expression of both p-Akt Thr 308 (P = 0.006) and Akt3 (P = 0.001) were adverse prognostic indicators for STSs located to extremities and for tumors larger than 5 cm in largest dimension (P = 0.001 for both markers). Interestingly, high expression of p-Akt Thr 308 was a negative prognostic factor particularly for men (P = 0.009 vs. P = 0.064 for women). In contrast, p-Akt Ser 473 , which appeared to be a negative prognosticator exclusively for female patients (P = 0.023 vs. P = 0.868 for men), Table 2.

Multivariate Cox proportional hazards analyses
The results of the multivariate analysis are presented in Table 3. Advanced age of the patient (P = 0.038), deep site (P = 0.018), high malignancy grade (p < 0.001), metastasis at time of diagnosis (P = 0.010), lack of surgery (P = 0.031), non-free resection margins (p < 0.001), and PI3K expression by tumor cells (P = 0.042) were significant independent negative prognostic indicators of DSS.

Co-expression of activated Akt and PI3K with female steroid hormone receptors
The co-expression profiles of both types of activated Akt and PI3K with female steroid hormone receptors, in the group as a whole and stratified into gender were tested as shown in Table 4. The co-expression phenotypes PgR +/p-Akt Thr 308 + among men (P = 0.023, HR = 2.4, 95% CI = 1.1-5.2), ER-/PI3K+ both in whole cohort (P = 0.005, HR = 2.0, 95% CI = 1.2-3.2) and among women (P = 0.014, HR = 2.4, 95% CI = 1.2-4.8), as well as PgR-/PI3K+ (P = 0.007, HR = 1.9, 95% CI = 1.2-3.0) and PgR+/PI3K+ (P = 0.014, HR = 1.9, 95% CI = 1.1-3.2) in the whole cohort of patients were significant independent negative prognostic factors. Interestingly, both steroid hormone receptors and Akt phosphorylation site seem to have opposite prognostic impact depending on the gender. This was further proved by the co-expression of these factors. Indeed, PgR-/p-Akt Ser 473 + phenotype tended to have an unfavorable impact in women (P = 0.087) but was favorable in men (P = 0.010). Co-expression of ER and p-Akt Ser 473 showed similar results, with significantly adverse influence of -/+ profile on DSS among female patients (P = 0.006). There was no significant difference among the four possible profiles in men, but the -/+ curve demonstrated the best survival rate, Figure 2E and 2F.

Discussion
In this large-scale retrospective study we have investigated the prognostic impact of a set of biomarkers belonging to the Akt-PI3K signaling pathway in non-GIST STS patients, both separately and in relation to gender. Further, we have also elucidated the coexpression of these markers and the female hormone receptors ER and PgR. These proteins participate in a diversity of processes in physiological and pathological conditions, especially in cancer development and progression [14]. p-Akt Thr 308 , Akt2 and PI3K showed significant unfavorable influence on survival of the whole cohort of patients in univariate analyses and, in addition, high expression of PI3K was a significant independent negative prognostic factor. p-Akt Thr 308 expression had a strong unfavorable impact among men, but was not significant in women. p-Akt Ser 473 expression had strong adverse impact in women but was not significant in men or in the whole cohort. PgR-/p-Akt Ser 473 + phenotype showed less favorable impact in women, but was the most favorable one in men. To our knowledge, this is the first prognostic evaluation of these biomarkers in non-GIST STSs.
In a healthy organism, Akt1 is a key signaling protein in the cellular pathways that result in skeletal muscle hypertrophy, and general tissue growth [30]. Akt can be phosphorylated by its two activating kinases, phosphoinositide dependent kinase 1 (PDK1) -at threonine 308 , and mammalian target of rapamycin complex 2 (mTORC2), previously putatively named PDK2, -at serine 473 . Both mTORC2 and PDK1 are products of the PI3K pathway. Activated Akt can activate or deactivate its multiple substrates, including mammalian target of rapamycin (mTOR), bcl-2 family member BAD, transcription factor forkhead homolog 1 in rhabdomyosarcoma (FKHR), Mdm2 protein, glycogen synthase kinase 3 (GSK3) and many others, via its kinase activity [31,32].  Akt1 is involved in cellular survival pathways by inhibiting apoptotic processes. Since it thereby promotes cell survival, Akt1 has been regarded as a major factor in many types of cancer [15][16][17]. The majority of studies agree that high expression of Akt by tumor cells indicates a poor prognosis [19][20][21]. However, in a recent study by Baba et al., phosphorylated Akt expression was reported to have a favorable impact on DSS in 717 colorectal cancer patients [22]. Similar results were obtained by Mori et al. in a study devoted to Akt expression in endometrial carcinoma [23]. This discrepancy can probably be explained by the site of Akt phosphorylation. Both studies utilized antibodies against p-Akt Ser 473 , while the articles describing negative influence of Akt are based on p-Akt Thr 308 expression [20,33,34]. Al-Saad et al. [33] has recently compared the prognostic impact of Akt phosphorylated on both sites and demonstrated that expression of p-Akt Thr 308 , unlike p-Akt Ser 473 , negatively influenced prognosis in patients with non-small cell lung cancer.
For the whole cohort we also found that p-Akt Thr 308 expression was associated with a shorter STS survival in univariate analyses, while p-Akt Ser 473 expression had no significant value. However, calculated separately for each gender, high expression of p-Akt Thr 308 was a negative prognostic factor particularly for men, in contrast to p-Akt Ser 473 , which appeared to be a negative prognosticator exclusively for female patients. This prompted us to further investigate this phenomenon by studying of co-expression profiles of both types of activated Akt with female steroid hormone receptors. In our previous works we have shown that ER and PgR expression possess variable prognostic significance depending of gender both per se [25] and in co-expression with TGF-β and fascin [26]. ERβ was shown to activate PI3K/Akt signalling pathway [35]. Tsai et al. demonstrated an activation of Akt by estrogen in ER negative breast cancer cell culture [36]. In the present study, the prognostic diversity of these factors in men and women was enhanced in the co-expression profiles: male patients with STSs expressing simultaneously p-Akt Thr 308 and PgR had statistically significant minimal survival rate. For women, the ER-/p-Akt Ser 473 + expression profile was the most unfavorable phenotype.
Taking into consideration a possible distortion of the results by gender-related sarcomas (i.e. leiomyosarcoma in uterus) we have attempted to exclude these sarcomas and recalculate all analyses. There were no considerable differences in the results by exclusion of gender-related sarcomas comparing to those obtained for whole cohort (data not shown).
Akt2 is an important molecule in the insulin signaling pathway, but in Akt1 deficient mice it is also proved to substitute, at least partly, the role of Akt1 in growth and proliferation [37]. We found Akt2 expression to be associated with significantly shorter DSS in univariate analysis. This might be explained by the extra-endocrine function of Akt2. The role of Akt3 is less clear, it Abbreviations: ER, estrogen receptor; PgR, progesterone receptor; NR, not reached; NS, not significant; * overall significance as a prognostic factor appears to be predominantly expressed in the central nervous system [29]. In this study, we failed to demonstrate any association of Akt3 with the survival of STS patients. PI3K is, via PDK1 and mTORC2 dependent activation, an upstream regulator of all Akt isoforms, and plays an important role in the PI3K/Akt pathway. Its high expression has been implicated as an adverse prognostic factor in many types of cancer [38][39][40]. In STS, we observed that PI3K expression was an independent significant indicator of shorter DSS. Not surprisingly, the co-expression of PI3K with both ER and PgR showed multiple independent negative impacts on survival in STS patients with the phenotypes ER-/PI3K+ in women and PgR+/PI3K+ in men being the least favorable.
The tumor suppressor gene PTEN negatively regulates the PI3K/Akt signaling pathway. It is a proapoptotic and antineoplastic factor and shown to be a favorable prognosticator in cancer patients [41]. In our study, we failed to find any statistical difference in survival between patients having PTEN-positive and PTEN-negative STSs.
In our material, all investigated factors showed three distinct patterns of expression; nuclear, cytoplasmic and combined nuclear and cytoplasmic. Little is known about the prognostic value of such subcellular stratification. Le Page et al. reported that nuclear Akt-1 and Akt-2 expression were significantly correlated with favorable outcome in 63 prostate cancer patients, while cytoplasmic Akt-1 expression was correlated with a higher risk of postoperative prostate-specific antigen (PSA) recurrence and shorter PSA recurrence interval [42]. In the present study, we were able to find such dependence only for nuclear p-Akt Thr 308 expression, which proved to be prognostically favorable compared to cytoplasmic and especially mixed cytoplasmic and nuclear location.

Conclusion
We have characterized the occurrence and distribution of several proteins belonging to PI3K/Akt signaling pathway in STS patients with respect to tumor aggressiveness and DSS. Our findings are largely in agreement with the results of a number of studies that have investigated the roles of these markers in other, especially epithelial, tumors. Nevertheless, the diverse prognostic values depending on the site of Akt phosphorylation and on the co-expression with female steroid hormones have not been described earlier.
Our findings may help to identify subgroups of patients with aggressive tumors requiring adjuvant therapy which, due to relationship of the PI3K/Akt pathway components with female steroid hormone receptor proteins, could include specific endocrine treatment. Moreover, since the investigated biomarkers belong to the family of serine-threonine kinases, which are comprehended as "drugable" [18], they may represent molecular targets for personalized, small-molecule targeted therapy. This currently is a hotspot of oncological research, and it was shown effect of such Akt targeted agents on several subtypes of sarcomas in vitro [43,44].