• Open Access

Coexpression of aPKCλ/ι and IL-6 in prostate cancer tissue correlates with biochemical recurrence

Authors


To whom correspondence should be addressed. E-mail: ynagas@med.yokohama-cu.ac.jp

Abstract

Atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6) have been implicated in prostate cancer progression, the mechanisms of which have been demonstrated both in vitro and in vivo. However, the clinical significance of the correlation between the expressions of these factors remains to be clarified. In the present study, we report a significant correlation between aPKCλ/ι and IL-6 proteins in prostate cancer tissue by immunohistochemical staining. We evaluated the association of both proteins by analyzing clinicopathological parameters using chi-square test, Kaplan–Meier with log–rank test, and a Cox proportional hazard regression model in univariate and multivariate analyses. The results again showed that the expression of aPKCλ/ι and IL-6 correlates in prostate cancer tissue (< 0.001). Atypical protein kinase C λ/ι was also found to correlate with the Gleason score (< 0.001) and with biochemical recurrence after prostatectomy (= 0.02). Furthermore, aPKCλ/ι correlated with biochemical recurrence in a Kaplan–Meier and log–rank test (= 0.01) and Cox analysis (= 0.02 in the univariate analysis, = 0.02 in the multivariate analysis). The coexpression of aPKCλ/ι and IL-6 also correlated with biochemical recurrence by Kaplan–Meier and log–rank test (= 0.005) and Cox analysis (= 0.01 in the univariate analysis, = 0.03 in the multivariate analysis). These results indicate a strong correlation between aPKCλ/ι and IL-6 in prostate tumors, and that the aPKCλ/ι–IL-6 axis is a reliable prognostic factor for the biochemical recurrence of this cancer. (Cancer Sci 2011; 102: 1576–1581)

Prostate cancer is the second leading cause of cancer death in the USA,(1) with the number of patients also increasing in Japan.(2) A serious concern in treating this disease is the development of castration-resistant prostate cancer (CRPC). This involves mutation or amplification of the androgen receptor (AR) gene, resulting in hypersensitivity to low levels of androgen, or the acquisition of a dependence on other hormones, causing resistance to anti-androgen treatments.(3) In addition, many signaling pathways and molecules are considered to be involved in the process of CRPC development.(4) However, there is currently no effective strategy for the prognosis and treatment of CRPC patients.(5,6)

Atypical protein kinase C λ/ι (aPKCλ/ι) is implicated in cell growth and maintenance, and in cell polarity.(7–11) It is overexpressed in several cancers, such as non-small cell lung cancer, ovarian cancer, glioma, colon cancer, breast cancer, gastric cancer, and prostate cancer.(12–19) Atypical protein kinase C λ/ι has also been suggested to be a prognostic factor.(13–19) However, the mechanism by which the overexpression of aPKCλ/ι causes cancer progression remains largely unclear.(20–22) We previously reported a clinical correlation between aPKCλ/ι mRNA expression and recurrence-free survival in prostate cancer.(12) In addition, studies of cultured prostate cancer cells have revealed that the overexpression of aPKCλ/ι results in interleukin (IL)-6 secretion, thus forming an aPKC–IL-6 autocrine loop through the transcriptional activation of the IL-6 gene by aPKCλ/ι.(12) Interleukin-6 is one of the cytokines implicated in cancer progression.(23) For example, it has been reported that the IL-6 protein level is elevated in CRPC patients and is involved in the activation of AR in an androgen-independent manner.(24–27) Although these observations suggest the potential to target the aPKCλ/ι–IL-6 axis as a therapeutic strategy, the significance of the correlation between prostate cancer progression and IL-6 and aPKCλ/ι remains to be evaluated.

In our current study, we found a strong, positive correlation between the aPKCλ/ι–IL-6 axis and prostate cancer progression, and further demonstrated a correlation between biochemical recurrence after prostatectomy and the expression of these two factors. These results support our hypothesis that aPKCλ/ι and IL-6 play an important role in prostate cancer progression.

Materials and Methods

Patients and tissue samples.  Prostate cancer tissues were obtained during radical prostatectomies performed at the Chigasaki Municipal Hospital (Kanagawa, Japan). The clinicopathological data for these samples are summarized in Table 1 and were obtained from archival clinical records. One patient received neo-adjuvant therapy prior to the radical prostatectomy. Six patients received adjuvant therapy before biochemical recurrence. No patients received radiation therapy before biochemical recurrence. All patients were followed up from the date of radical prostatectomy until the date of the last follow up. We divided these cases into two groups according to median age and the preoperative serum prostatic specific antigen (PSA) level (71 years and 12.5 ng/mL, respectively). The postoperative serum PSA levels were assayed every 2–3 months. The occurrence of three consecutive elevations of the PSA level by more than 0.20 ng/mL was defined as biochemical recurrence.(28) The patients were periodically followed up until biochemical recurrence was recorded.

Table 1.   Clinicopathological characteristics of examined prostate cancer samples
 n (%)
  1. PSA, prostatic specific antigen; UICC, International Union against Cancer.

Age (years)
 ≤7123 (53.5%)
 >7120 (46.5%)
 Median (range)71.0 (60.0–80.0)
Preoperative PSA (ng/mL)
 ≤12.522 (51.2%)
 >12.521 (48.8%)
 Median (range)12.5 (4.9–72.0)
UICC stage
 pT2a-b18 (41.9%)
 pT3a-b25 (58.1%)
Lymph node involvement
 pN041 (95.3%)
 pN12 (4.7%)
Gleason score
 ≤728 (65.1%)
 ≥815 (34.9%)
Margin status
 −22 (51.2%
 +21 (48.8%)
Biochemical failure (months)
 −29 (67.4%)
 +14 (32.6%)
 Median (range)14.8 (0–39.0)
Observation time (months)
 Median (range)54 (4.0–109.0)
Neo-adjuvant/adjuvant therapy
 None36 (83.7%)
 Neo-adjuvant1 (2.3%)
 Adjuvant6 (14.0%)

The sampled prostate tissues were immediately fixed in 20% formalin, embedded in paraffin, and subjected to routine pathological diagnosis. The pathological stage was determined according to the International Union against Cancer (UICC).(29) This study was approved by the Institutional Ethical Committee of Chigasaki Municipal Hospital. Recorded informed consent was obtained from each patient enrolled in the study.

Antibodies, tissue microarray, and immunohistochemical staining.  Mouse anti-human aPKCι (clone 23) and mouse antihuman IL-6 antibodies (NCL-L-IL6) were purchased from BD Biosciences (San Jose, CA, USA) and Leica Biosystems (Newcastle, UK), respectively. Tissue microarray (TMA) blocks were constructed as described previously.(30) Briefly, cores (3 mm in diameter) were collected from archival paraffin-embedded tissue specimens using a manual tissue microarray technique (Azumaya, Tokyo, Japan). Tissue microarray blocks, each containing 24 cores, were subsequently established. The core sizes were chosen considering the heterogeneity of prostate cancer tissues.

For immunohistochemical staining, 4 μm-thick paraffin sections of the TMA samples were used. Immunohistochemical staining for aPKCλ/ι was performed, and the results were scored according to a previous method(16) as follows: 0, no staining; 1+, weak to normal intensity compared with normal prostate cells; 2+, moderate intensity; and 3+, strong intensity. Scores of 0 and 1+ were defined as negative (aPKCλ/ι[−]). Scores of 2+ and 3+ were defined as positive (aPKCλ/ι[+]).

For IL-6 immunohistochemical staining, the sections were autoclaved in citrate buffer at 121°C for 15 min for antigen retrieval, and then incubated with anti-IL-6 antibody (1:50 dilution in Signal Stain antibody diluent; Cell Signaling Technology, Danvers, MA, USA) at 4°C overnight. Labeled antigens were visualized with a Histo Fine kit (Nichirei, Tokyo, Japan), followed by a 3,3′-diaminobenzidine plus reaction (Dako Cytomation, Glostrup, Denmark). Finally, the sections were counterstained with hematoxylin. Low- and high-intensity signals were scored based on the predominant pattern in each case. Immunohistochemical staining for IL-6 was scored as follows; 0, no staining; 1+, weaker positivity than reactive lymphocytes in tissue specimens; 2+, comparable positivity with reactive lymphocytes; and 3+, stronger positivity than reactive lymphocytes. A score of 0 was defined as negative (IL-6[−]) and all other scores (1+ ≤) were defined as positive (IL-6[+]).

Statistics.  All statistical analyses were performed using a software for statistic analysis PASW statistics 18 (SPSS, Chicago, IL, USA). Atypical protein kinase C λ/ι and IL-6 expressions and clinicopathological characteristics were compared using chi-square test or the Fisher’s exact test. For analysis of the correlation between aPKCλ/ι and IL-6 expressions, clinicopathological characteristics, and biochemical recurrence, Kaplan–Meier and log–rank tests were utilized. Each parameter and biochemical recurrence was further analyzed using a Cox promotional hazard regression model in univariate and multivariate analyses. The hazard ratio and 95% confidence interval (95%) are also shown in the results.  0.05 was considered to be statistically significant.

Results

Expression of aPKCλ/ι protein in prostate cancer tissue.  We prepared 43 prostate cancer tissue samples for the TMA (Table 1) and immunohistochemical analyses for aPKCλ/ι. As shown in Table 2, the positive expression of aPKCλ/ι (aPKCλ/ι[+], 2+ and 3+ cases) was noted in 44.2% (19/43 cases) of the cases, and negative results (aPKCλ/ι[−], 0 and 1+ cases) were obtained in 55.8% (24/43 cases) of the patient samples. The representative results and immunohistochemical scoring in Figure 1(a–d) show that aPKCλ/ι is localized in the cytoplasm of prostatic epithelial cells, but not in stromal cells, which is consistent with the findings of our previous study.(12)

Table 2.   Summary of immunostaining results for atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6)
Total (= 43)0 (%)1+ (%)2+ (%)3+ (%)
aPKCλ/ι expression3 (7.0)21 (48.8)12 (27.9)7 (16.3)
IL-6 expression22 (51.2)15 (34.9)4 (9.3)2 (4.6)
Figure 1.

 Representative images of atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6) immunohistochemical staining results. Positive and negative staining of aPKCλ/ι (a–d) and IL-6 (e–h). Bars: 100 μm.

Expression of IL-6 protein in prostate cancer tissue.  The same 43 samples used for the above analysis of aPKCλ/ι expression were used for the analysis of IL-6 expression. The results also demonstrated positive IL-6 expression in prostate cancer tissue (Fig. 1e–h). The results were positive in 48.8% (score 1+ ≤, IL-6[+], 21/43 cases) and negative in 51.2% (score 0, IL-6[−], 22/43 cases) of the patient samples, as shown in Table 2. The IL-6 protein was also found to be localized in the cytoplasm of prostate cancer cells, as reported in previous studies.(31,32) Interleukin-6 expression was not detected in the non-neoplastic cells in our samples.

Correlation between aPKCλ/ι and IL-6 expression, and the clinicopathological features of prostate cancer tissues.  We next analyzed the association between aPKCλ/ι and IL-6 expression, and the clinical characteristics of our prostate cancer specimens by chi-square test or Fisher’s exact test (Table 3). Atypical protein kinase C λ/ι expression showed a significant positive correlation with IL-6 expression (< 0.001). In terms of the relationship between the clinicopathological features of our prostate cancer cohort and aPKCλ/ι, there were some associations found using the Gleason score (< 0.001), including margin status (= 0.03) and biochemical recurrence (= 0.02). Figure 2(a–d) reveals the association between the Gleason grade and aPKCλ/ι staining. However, none of the other clinicopathological features showed an association with the aPKCλ/ι protein level.

Table 3.   Association between atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6) expressions and clinicopathological features of prostate cancer patients
 aPKCλ/ι expressionIL-6 expression
+P-value+P-value
  1. *Fisher’s exact test. Bold values indicate statistical significance. PSA, prostatic specific antigen; UICC, International Union against Cancer.

Age (years)
 ≤7112110.7613100.55
 >71128911
UICC stage
 pT2a-b1170.761260.12
 pT3a-b13121015
Lymph node involvement
 pN023181.00*21200.74*
 pN11111
Gleason score
 ≤7217<0.001*18100.03*
 ≥8312411
Margin status
 −1660.031480.13
 +813813
Preoperative PSA (ng/mL)
 ≤12.51480.3612100.76
 >12.510111011
Biochemical recurrence
 −2090.02*18110.06*
 +410410
IL-6 expression
 −184<0.001*   
 +615  
Figure 2.

 Representative images of atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6) immunohistochemical staining results for a specific Gleason grade. Positive staining of aPKCλ/ι in Gleason grade 3 (a), 4 (b), 5 (c), and non-neoplastic (d) prostate tissue. Positive staining of IL-6 in Gleason grade 3 (e), 4 (f), 5 (g), and non-neoplastic (h) prostate tissue. Bars: 100 μm.

We further investigated the association between IL-6 expression and the clinicopathological features of our prostate cancer samples. Figure 2(e–h) reveals the association between the Gleason grade and IL-6 staining. The Gleason score showed a correlation with IL-6 expression (Table 3, Gleason score, = 0.03), but no other clinicopathological features did. To investigate the association between biochemical recurrence and the expressions of aPKCλ/ι and IL-6 in more detail, Kaplan–Meier and log–rank tests were utilized. As expected, aPKCλ/ι expression was found to correlate with biochemical recurrence (Fig. 3a, = 0.01). Further, IL-6 expression showed a correlation with biochemical recurrence (Fig. 3b, = 0.05). Among the clinical characteristics of our patient cohort, the identified association was with patient age (Fig. 3c, = 0.01) and margin status (Fig. 3d, = 0.001).

Figure 3.

 Analysis of the association between atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6), age, margin status and biochemical recurrence by Kaplan–Meier and log–rank testing. Positive and negative staining of aPKCλ/ι (a) and IL-6 (b) analyzed using Kaplan–Meier and log–rank tests. Positive cases are denoted aPKCλ/ι(+) and IL-6(+), and negative cases are indicated by aPKCλ/ι(−) and IL-6(−), respectively. (c) Age, divided into two groups by the median (71 years), was analyzed using Kaplan–Meier and log–rank tests. (d) Positive and negative margin statuses were analyzed using Kaplan–Meier and log–rank tests. Positive margin status is indicated as Margin(+), and a negative margin status is indicated as Margin(−).  0.05 indicates statistical significance. (a) P = 0.01, (b) P = 0.05, (c) P = 0.01 and (d) P = 0.001.

By univariate and multivariate analyses, we found that aPKCλ/ι was a prognostic factor (Table 4, = 0.02 in the univariate analysis, = 0.02 in the multivariate analysis). Patient age, classified by the median value of our samples (71 years old), and the involvement of the surgical margins were also shown to be prognostic factors (Table 4).

Table 4.   Relative recurrence-free survival probability determined by univariate and multivariate analyses
 UnivariateMultivariate
HR95% CIP-valueHR95% CIPHR95% CIP-value
  1. Bold values indicate statistical significance. 95% CI, 95% confidence interval; aPKCλ/ι, atypical protein kinase C λ/ι; HR, Hazard ratio; IL-6, interleukin-6; PSA, prostatic specific antigen; UICC, International Union against Cancer.

aPKCλ/ι expression
 −3.931.23–12.560.024.591.28–16.410.02   
 +
IL-6 expression
 −3.020.94–9.640.06      
 +
aPKCλ/ι and IL-6 expression
 aPKCλ/ι/IL-6(±) and aPKCλ/ι(+)/IL-6(–)4.081.36–12.230.01   3.801.17–12.30.03
 aPKCλ/ι(+)/IL-6(+)
UICC stage
 pT2a-b1.990.62–6.360.25      
 pT3a-b
Margin status
 −7.711.72–34.530.0084.821.04–22.380.054.951.07–22.930.04
 +
Lymph node
 −0.050.00–994.730.54      
 +
Gleason score
 ≤71.470.51–4.230.48      
 ≥8
Age (years)
 ≤714.131.29–13.290.025.801.59–21.160.0084.811.37–16.90.02
 >71
Preoperative PSA (ng/mL)
 ≤12.51.870.63–5.590.26      
 >12.5

Atypical protein kinase C λ/ι and IL-6 coexpression (aPKCλ/ι[+]/IL-6[+]) correlates with biochemical recurrence in prostate cancer.  Finally, we investigated whether the coexpression of aPKCλ/ι and IL-6 correlated with biochemical recurrence in our prostate cancer patients. Cancer tissues positively expressing both aPKCλ/ι and IL-6 (aPKCλ/ι[+]/IL-6(+)) showed a statistically-significant correlation when compared with aPKCλ/ι(+)/IL-6(−) and PKCλ/ι(−)/IL-6(±) cases (Fig. 4, = 0.005). Table 4 further shows the results of the univariate and multivariate analyses. The aPKCλ/ι(+)/IL-6(+) group showed statistical significance (= 0.01 in the univariate analysis, = 0.03 in the multivariate analysis) in addition to the median age and margin status.

Figure 4.

 Analysis of the association between atypical protein kinase C λ/ι (aPKCλ/ι) and interleukin-6 (IL-6) coexpression on biochemical recurrence by both Kaplan–Meier and log–rank testing. The coexpression of aPKCλ/ι and IL-6 was analyzed using the Kaplan–Meier and log–rank tests.  0.05 indicates statistical significance. Double-positive expression cases are denoted as aPKCλ/ι(+)/IL-6(+). Other cases are denoted aPKCλ/ι(−)/IL-6(±) and aPKCλ/ι(+)/IL-6(−). P = 0.005.

Discussion

We previously reported that a clinical correlation exists between aPKCλ/ι mRNA expression and recurrence-free survival in prostate cancer patients.(12) Further, we also found a molecular link between aPKCλ/ι and IL-6 in prostate cancer cell lines, where aPKCλ/ι upregulates IL-6 gene transcription to overproduce IL-6 and promote proliferation.(12) Interleukin-6 has been implicated in prostate cancer progression in a number of previous studies, particularly in CRPC.(23,25,33,34) These observations led us to hypothesize that the aPKCλ/ι–IL-6 axis” might play a role in prostate cancer progression.(12) From a clinical viewpoint, the combination of aPKCλ/ι and IL-6 provides a potential novel strategy for the prognosis and treatment of CRPC.

In our current study, we show that the expression of the aPKCλ/ι protein correlates with biochemical recurrence, as well as the aPKCλ/ι mRNA level in prostate cancer. Importantly, the aPKCλ/ι protein level in individual prostate cancer tissue samples, but not in the control paired tissue samples obtained from the same patient, also correlated with biochemical recurrence after a prostatectomy. Furthermore, non-neoplastic samples from our cohort were negative for these markers by immunostaining. Thus, aPKCλ/ι is expected to be a useful predictive marker for biochemical recurrence in prostate cancer.

In terms of correlations with the clinical parameters of the samples analyzed in our present study, median age and margin status were found to correlate with biochemical recurrence. This is not unexpected because age and margin status have been suggested in previous reports to be a predictive indicator of biochemical recurrence.(35,36) In contrast, other clinical factors, such as UICC stage, Gleason score, and preoperative PSA, were found not to be associated with biochemical recurrence in our present study. Nevertheless, we did find an association between aPKCλ/ι expression and biochemical recurrence. These results support the notion that aPKCλ/ι is a feasible prognostic factor for the biochemical recurrence of prostate cancer.

Previous studies have reported a relationship between biochemical recurrence, CRPC, and the serum IL-6 level.(23–25,37) We also confirmed a correlation between IL-6 expression and biochemical recurrence using Kaplan–Meier and log–rank testing in our present study. Furthermore, IL-6 was found to correlate with aPKCλ/ι expression in our current analysis. Based on these results, the aPKCλ/ι(+)/IL-6(+) group (34.9%) was also found to strongly correlate with biochemical recurrence. However, there was no statistically-significant difference found between biochemical recurrence in the aPKCλ/ι(+)/IL-6(+) and aPKCλ/ι(+)/IL-6(−) (9.3%) groups when we examined which was more useful as a prognostic indicator (data not shown, = 0.35). Atypical protein kinase C λ/ι(+)/IL-6(−) cases might show a transitional process of IL-6 overexpression by aPKCλ/ι-dependent IL-6 gene expression.(12) In addition, NF-κB and AP-1 control other molecules, such as interleukin-8 (IL-8).(38) Thus, in the aPKCλ/ι(+)/IL-6(−) group, other molecules induced through nuclear factor κB (NF-κB) and activator protein 1 (AP-1) by aPKCλ/ι overexpression might possibly be involved. The significance of aPKCλ/ι(+)/IL-6(+) as a prognostic factor might become clear through sustained follow up.

Our present findings provide strong evidence in support of the importance of the aPKCλ/ι–IL-6 axis in prostate cancer progression. Thus, the expression of aPKCλ/ι(+)/IL-6(+) could be a good prognostic indicator of biochemical recurrence in this disease. Furthermore, the aPKCλ/ι–IL-6 axis might also be a viable future molecular target for the treatment of prostate cancer.

Acknowledgments

We would like to thank Ms Rie Shimizu, Yoko Nakamura, Takako Yamaki, Yoshiko Imano, Tamiyo Taniguchi, Shizuko Kobayashi, Masumi Tanaka, Kayano Moriyama, Chie Kondo, Sayuri Kanaya, and Ayako Ishiyama for technical and secretarial assistance. This research was supported by a Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan for Scientific Research (to K Akimoto, Y Nagashima, Y Kubota, and S Ohno), a Grant from the Yokohama Foundation for Medical Research Promotion (to Y Nagashima), and a Grant for Strategic Research Promotion from Yokohama City University (to Y Nagashima).

Disclosure

The authors have no conflict of interest.

Ancillary