Caveolin-1 Expression Patterns in Pleural Mesothelioma
Caveolin-1 Expression Patterns in Pleural Mesothelioma
With the aim of defining the role of CAV1 expression in MPM, we report herein the differences in immunohistochemical reactivity in the different MPM histotypes. We also describe specific expression patterns in neoplastic vs stromal MPM cells that have an influence on patient outcomes. CAV1 was previously described in MPM as an immunohistochemical marker for differentiating MPM from lung adenocarcinoma. However, to the best of our knowledge, this is the first time that the intratumoral CAV1 expression was assessed and its potential prognostic role investigated.
In our experience, CAV1 had variable immunohistochemical expression with a finely granular pattern, as described in epithelial tumor cells of various organs. When the neoplastic cells were considered, CAV1 expression in epithelial MPM had the lowest expression levels, with a mean HS value of 70, thus indicating that not all cases expressed the protein at detectable levels; in most of these cases, a focal distribution was detected. Interestingly, in biphasic MPM, an intermediate CAV1 expression level was detected, comparable in both components to those of the corresponding pure histotypes. It is well known that histotype is one of the most important prognostic factors in MPM. The increased CAV1 expression, according to low-grade epithelial to high-grade sarcomatous histotypes, is in line with the known correlation of increased CAV1 expression with a more aggressive behavior in various epithelial and nonepithelial malignancies, including non–small cell lung cancer, brain tumors, renal cell carcinoma, and, recently, gastric cancer. Furthermore, expression of CAV1 was described in several soft tissue and bone sarcomas as a marker associated with the degree of differentiation. In line with this observation, our series of sarcomatous MPMs were all positive (100%) with the highest N-CAV1 levels (mean, 215). We thus confirmed that this neoplasia exhibits a CAV1 gain along with dedifferentiation when the spectrum of mesothelial histotype, from the low-grade differentiated epithelial to the high-grade sarcomatous histotype, is considered Figure 6.
(Enlarge Image)
Figure 6.
Schematic hypothetical caveolin-1 (CAV1) modulation in the malignant mesothelial transdifferentiation process ranging from the low (epithelial) to high (sarcomatous) dedifferentiated histotype. S, stromal.
As a further confirmation, we tested three MPM cell lines, two (MPP89 and REN) derived from epithelial MPM and one (MSTO) from biphasic MPM pleural effusion (as stated by ATCC). The different CAV1 distribution according to the histotype observed in patient tissue samples was confirmed at both protein and mRNA levels in cell line models, thus also proving the reliability of the CAV1 antibody used.
A unique finding of our investigation was the detection of a peculiar CAV1 expression in stromal cells of a fraction of epithelial MPMs (67% of cases). S-CAV1 expression allowed the identification of a particular type of spindle-shaped stromal cells characterized by bland atypical morphology and embedded in a soft stroma; this S-CAV1 expression was absent in cases having a poorly cellular sclerohyaline stroma. Immunostaining revealed that S-CAV1–positive cells expressed diffuse SMA and vimentin, and coexpressed scattered mesothelial markers calretinin, cytokeratin, podoplanin, and WT1 in the same cells. This suggests that such cells could originate from mesothelial cells able to transdifferentiate into stromal activated cells with low proliferative potential. Double immunofluorescence staining seemed to confirm this hypothesis. It is not clear whether these cells represent early transforming neoplastic cells into a biphasic/sarcomatous phenotype, thus confirming the mesothelioma histotypes as a continuous transformation of the mesothelial cell. Although the low proliferative index may not support this notion, the clinical behavior of epithelial MPM having stromal CAV1 expression is more aggressive, overlapping that of biphasic MPM type. These findings lead to a question on the real nature of biphasic MPM compared with epithelial MPM with areas of "reactive" stroma. In particular, could MPM having a stroma enriched for CAV1-expressing cells represent a real, more aggressive variant, irrespective of these cases being classified as epithelial or biphasic histotypes?
An additional aim of our study was to investigate the potential prognostic role of CAV1 expression in MPM. Because prognosis in MPM is greatly affected by MPM histotype and type of treatment administered, we specifically grouped the cases to overcome biases related to these two factors. The sarcomatous MPM group (group C) was in fact used as control only, and specific subanalyses were conducted separately in the series of patients with MPM who were treated with the same chemotherapy regimen represented by platinum-pemetrexed (group A) and control cases with unknown treatment regimens (group B). No association with survival was detected for N-CAV1 expression of the epithelial MPM type in both groups. On the contrary, the detection of the S-CAV1 positivity identified a subgroup of epithelial MPMs characterized by a worse outcome compared with those devoid of stromal positivity, and with a survival outcome comparable with that of biphasic and sarcomatous histotypes. Together these findings led to the identification of a specific population of stromal cells in epithelial MPM subtype with peculiar morphologic and phenotypic characteristics and linked to prognostic information.
The exact mechanism of CAV1 action in tumor cells is incompletely understood and even more obscure is its role in stromal cells. Physiologically, CAV1 acts as a multifunctional scaffolding protein with multiple binding partners and is associated with cell surface caveolae in the regulation of lipid raft domains. But it also seems to be involved in cancer growth and progression, modulating tissue responses through architectural regulation of the microenvironment. Our findings are in line with prior evidence by Goetz and coworkers, showing an involvement of CAV1 in biochemical remodeling of the microenvironment that favors tumor invasion and metastases. These authors showed that stromal tissue associated with human carcinomas and melanoma metastases is enriched in CAV1-expressing carcinoma-associated fibroblasts (CAFs) that favor directional migration and invasiveness of carcinoma cells in vitro. In ovarian serous carcinomas, which share morphologic, biological, and histogenetic similarities with MPM, the presence of CAFs has been associated with a poor prognosis. However, the opposite association has been found in other tumor models such as breast cancer and melanoma. Furthermore, CAV1 was also recently described to promote progression of gastric cancer by upregulating EMT by crosstalk of signaling mechanisms. More specifically in MPM, SMA-positive tumor-associated fibroblast was previously described to promote mesothelioma cell progression in vitro and in vivo through cell recruitment and activation by the malignant cytokine network; however, the exact nature of these cells still remains undefined. Normal mesothelium is of a complex cell type, sharing epithelial and mesenchymal properties that are maintained in neoplastic conditions, and having potential EMT features. Based on this particular nature of the mesothelial cells, we think that they could acquire a (myo)fibroblast-like phenotype through an EMT-like process or, more appropriately, a mesothelial-to-mesenchymal transition process, as described in a recent study in peritoneal mesothelium. Yanez-Mo et al reported that mesothelial cells in the course of peritoneal dialysis could acquire a mesenchymal phenotype through a mesothelial-to-mesenchymal transition. Sandoval et al later showed that cancer-associated fibroblasts could originate from mesothelial cells, allowing carcinoma cells to invade the peritoneal compact zone. Furthermore, mesothelial cells with a mesenchymal phenotype acquire the capacity to synthesize inflammatory and angiogenic factors as well as ECM components. Thus these mesothelial cells contribute to the deterioration of the peritoneum during peritoneal injury. CAV1 expression was also detected in mesothelial cells reactive to Freund's adjuvant therapy injected into the peritoneal cavity of rats. This experimental model was used as an in vivo system to investigate the steps of the EMT mechanism: the possibility that mesothelial cells are not entirely differentiated and might undergo EMT was suggested, also based on a documented change of the distribution of caveolae during inflammation. These data support the idea that CAV1/caveolae might be involved in the regulation of signal transduction, leading to transdifferentiation of mesothelial cells and an increase of cytoplasmic caveolae internalization. Finally, upregulation of EMT-related transcription factors in MPM was recently described as playing a significant role in morphologic tumor features.
In our opinion, the presence of stromal CAV1-positive cells also observed in some pleuritis cases is not in contrast with our model; these findings expand the concept that mesothelial cells might transdifferentiate into a (myo) fibroblastic or mesenchymal phenotype (including CAV1 expression), but do not modify our data and interpretation of the peculiar function of these particular cells in the neoplastic setting.
In summary, CAV1 may represent a marker of differentiation and transition from epithelial to sarcomatous MPM, and an early marker of aggressive behavior in epithelial MPM, independent of the treatment, which needs further validation.
Discussion
With the aim of defining the role of CAV1 expression in MPM, we report herein the differences in immunohistochemical reactivity in the different MPM histotypes. We also describe specific expression patterns in neoplastic vs stromal MPM cells that have an influence on patient outcomes. CAV1 was previously described in MPM as an immunohistochemical marker for differentiating MPM from lung adenocarcinoma. However, to the best of our knowledge, this is the first time that the intratumoral CAV1 expression was assessed and its potential prognostic role investigated.
In our experience, CAV1 had variable immunohistochemical expression with a finely granular pattern, as described in epithelial tumor cells of various organs. When the neoplastic cells were considered, CAV1 expression in epithelial MPM had the lowest expression levels, with a mean HS value of 70, thus indicating that not all cases expressed the protein at detectable levels; in most of these cases, a focal distribution was detected. Interestingly, in biphasic MPM, an intermediate CAV1 expression level was detected, comparable in both components to those of the corresponding pure histotypes. It is well known that histotype is one of the most important prognostic factors in MPM. The increased CAV1 expression, according to low-grade epithelial to high-grade sarcomatous histotypes, is in line with the known correlation of increased CAV1 expression with a more aggressive behavior in various epithelial and nonepithelial malignancies, including non–small cell lung cancer, brain tumors, renal cell carcinoma, and, recently, gastric cancer. Furthermore, expression of CAV1 was described in several soft tissue and bone sarcomas as a marker associated with the degree of differentiation. In line with this observation, our series of sarcomatous MPMs were all positive (100%) with the highest N-CAV1 levels (mean, 215). We thus confirmed that this neoplasia exhibits a CAV1 gain along with dedifferentiation when the spectrum of mesothelial histotype, from the low-grade differentiated epithelial to the high-grade sarcomatous histotype, is considered Figure 6.
(Enlarge Image)
Figure 6.
Schematic hypothetical caveolin-1 (CAV1) modulation in the malignant mesothelial transdifferentiation process ranging from the low (epithelial) to high (sarcomatous) dedifferentiated histotype. S, stromal.
As a further confirmation, we tested three MPM cell lines, two (MPP89 and REN) derived from epithelial MPM and one (MSTO) from biphasic MPM pleural effusion (as stated by ATCC). The different CAV1 distribution according to the histotype observed in patient tissue samples was confirmed at both protein and mRNA levels in cell line models, thus also proving the reliability of the CAV1 antibody used.
A unique finding of our investigation was the detection of a peculiar CAV1 expression in stromal cells of a fraction of epithelial MPMs (67% of cases). S-CAV1 expression allowed the identification of a particular type of spindle-shaped stromal cells characterized by bland atypical morphology and embedded in a soft stroma; this S-CAV1 expression was absent in cases having a poorly cellular sclerohyaline stroma. Immunostaining revealed that S-CAV1–positive cells expressed diffuse SMA and vimentin, and coexpressed scattered mesothelial markers calretinin, cytokeratin, podoplanin, and WT1 in the same cells. This suggests that such cells could originate from mesothelial cells able to transdifferentiate into stromal activated cells with low proliferative potential. Double immunofluorescence staining seemed to confirm this hypothesis. It is not clear whether these cells represent early transforming neoplastic cells into a biphasic/sarcomatous phenotype, thus confirming the mesothelioma histotypes as a continuous transformation of the mesothelial cell. Although the low proliferative index may not support this notion, the clinical behavior of epithelial MPM having stromal CAV1 expression is more aggressive, overlapping that of biphasic MPM type. These findings lead to a question on the real nature of biphasic MPM compared with epithelial MPM with areas of "reactive" stroma. In particular, could MPM having a stroma enriched for CAV1-expressing cells represent a real, more aggressive variant, irrespective of these cases being classified as epithelial or biphasic histotypes?
An additional aim of our study was to investigate the potential prognostic role of CAV1 expression in MPM. Because prognosis in MPM is greatly affected by MPM histotype and type of treatment administered, we specifically grouped the cases to overcome biases related to these two factors. The sarcomatous MPM group (group C) was in fact used as control only, and specific subanalyses were conducted separately in the series of patients with MPM who were treated with the same chemotherapy regimen represented by platinum-pemetrexed (group A) and control cases with unknown treatment regimens (group B). No association with survival was detected for N-CAV1 expression of the epithelial MPM type in both groups. On the contrary, the detection of the S-CAV1 positivity identified a subgroup of epithelial MPMs characterized by a worse outcome compared with those devoid of stromal positivity, and with a survival outcome comparable with that of biphasic and sarcomatous histotypes. Together these findings led to the identification of a specific population of stromal cells in epithelial MPM subtype with peculiar morphologic and phenotypic characteristics and linked to prognostic information.
The exact mechanism of CAV1 action in tumor cells is incompletely understood and even more obscure is its role in stromal cells. Physiologically, CAV1 acts as a multifunctional scaffolding protein with multiple binding partners and is associated with cell surface caveolae in the regulation of lipid raft domains. But it also seems to be involved in cancer growth and progression, modulating tissue responses through architectural regulation of the microenvironment. Our findings are in line with prior evidence by Goetz and coworkers, showing an involvement of CAV1 in biochemical remodeling of the microenvironment that favors tumor invasion and metastases. These authors showed that stromal tissue associated with human carcinomas and melanoma metastases is enriched in CAV1-expressing carcinoma-associated fibroblasts (CAFs) that favor directional migration and invasiveness of carcinoma cells in vitro. In ovarian serous carcinomas, which share morphologic, biological, and histogenetic similarities with MPM, the presence of CAFs has been associated with a poor prognosis. However, the opposite association has been found in other tumor models such as breast cancer and melanoma. Furthermore, CAV1 was also recently described to promote progression of gastric cancer by upregulating EMT by crosstalk of signaling mechanisms. More specifically in MPM, SMA-positive tumor-associated fibroblast was previously described to promote mesothelioma cell progression in vitro and in vivo through cell recruitment and activation by the malignant cytokine network; however, the exact nature of these cells still remains undefined. Normal mesothelium is of a complex cell type, sharing epithelial and mesenchymal properties that are maintained in neoplastic conditions, and having potential EMT features. Based on this particular nature of the mesothelial cells, we think that they could acquire a (myo)fibroblast-like phenotype through an EMT-like process or, more appropriately, a mesothelial-to-mesenchymal transition process, as described in a recent study in peritoneal mesothelium. Yanez-Mo et al reported that mesothelial cells in the course of peritoneal dialysis could acquire a mesenchymal phenotype through a mesothelial-to-mesenchymal transition. Sandoval et al later showed that cancer-associated fibroblasts could originate from mesothelial cells, allowing carcinoma cells to invade the peritoneal compact zone. Furthermore, mesothelial cells with a mesenchymal phenotype acquire the capacity to synthesize inflammatory and angiogenic factors as well as ECM components. Thus these mesothelial cells contribute to the deterioration of the peritoneum during peritoneal injury. CAV1 expression was also detected in mesothelial cells reactive to Freund's adjuvant therapy injected into the peritoneal cavity of rats. This experimental model was used as an in vivo system to investigate the steps of the EMT mechanism: the possibility that mesothelial cells are not entirely differentiated and might undergo EMT was suggested, also based on a documented change of the distribution of caveolae during inflammation. These data support the idea that CAV1/caveolae might be involved in the regulation of signal transduction, leading to transdifferentiation of mesothelial cells and an increase of cytoplasmic caveolae internalization. Finally, upregulation of EMT-related transcription factors in MPM was recently described as playing a significant role in morphologic tumor features.
In our opinion, the presence of stromal CAV1-positive cells also observed in some pleuritis cases is not in contrast with our model; these findings expand the concept that mesothelial cells might transdifferentiate into a (myo) fibroblastic or mesenchymal phenotype (including CAV1 expression), but do not modify our data and interpretation of the peculiar function of these particular cells in the neoplastic setting.
In summary, CAV1 may represent a marker of differentiation and transition from epithelial to sarcomatous MPM, and an early marker of aggressive behavior in epithelial MPM, independent of the treatment, which needs further validation.