Limits...
A distinct molecular profile associated with mucinous epithelial ovarian cancer.

Heinzelmann-Schwarz VA, Gardiner-Garden M, Henshall SM, Scurry JP, Scolyer RA, Smith AN, Bali A, Vanden Bergh P, Baron-Hay S, Scott C, Fink D, Hacker NF, Sutherland RL, O'Brien PM - Br. J. Cancer (2006)

Bottom Line: The results show that MOC express a genetic profile that both differs and overlaps with other subtypes of epithelial ovarian cancer.In particular, galectin 4 (LGALS4) was highly and specifically expressed in MOC, but expressed at lower levels in benign mucinous cysts and borderline (atypical proliferative) tumours, supporting a malignant progression model of MOC.Hence LGALS4 may have application as an early and differential diagnostic marker of MOC.

View Article: PubMed Central - PubMed

Affiliation: Cancer Research Program, Garvan Institute of Medical Research, Darlinghurst, NSW 2010, Australia.

ABSTRACT
Mucinous epithelial ovarian cancers (MOC) are clinically and morphologically distinct from the other histological subtypes of ovarian cancer. To determine the genetic basis of MOC and to identify potential tumour markers, gene expression profiling of 49 primary ovarian cancers of different histological subtypes was performed using a customised oligonucleotide microarray containing >59 000 probesets. The results show that MOC express a genetic profile that both differs and overlaps with other subtypes of epithelial ovarian cancer. Concordant with its histological phenotype, MOC express genes characteristic of mucinous carcinomas of varying epithelial origin, including intestinal carcinomas. Differences in gene expression between MOC and other histological subtypes of ovarian cancer were confirmed by RT-PCR and/or immunohistochemistry. In particular, galectin 4 (LGALS4) was highly and specifically expressed in MOC, but expressed at lower levels in benign mucinous cysts and borderline (atypical proliferative) tumours, supporting a malignant progression model of MOC. Hence LGALS4 may have application as an early and differential diagnostic marker of MOC.

Show MeSH

Related in: MedlinePlus

(A) Principal components analysis based on 500 genes with the most variable signal intensities (based on variance) separates the histological subtypes of EOC. MOC (n=3) are circled; (B) Hierarchical clustering and heat map of differentially expressed genes (n=167 upregulated and n=18 down-regulated) in MOC compared to serous and endometrioid ovarian cancers. Clustering was performed on all transcript profiled samples (n=3 MOC; n=4 mucinous borderline tumours; n=8 endometrioid ovarian cancers (endo); n=3 serous borderline tumours; n=31 serous ovarian cancers (unlabelled columns); and four normal ovaries) as described in the Materials and Methods. Expression levels are colour coded with red, green and black corresponding to an increase, a decrease, and no change in gene expression, respectively.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2361366&req=5

fig1: (A) Principal components analysis based on 500 genes with the most variable signal intensities (based on variance) separates the histological subtypes of EOC. MOC (n=3) are circled; (B) Hierarchical clustering and heat map of differentially expressed genes (n=167 upregulated and n=18 down-regulated) in MOC compared to serous and endometrioid ovarian cancers. Clustering was performed on all transcript profiled samples (n=3 MOC; n=4 mucinous borderline tumours; n=8 endometrioid ovarian cancers (endo); n=3 serous borderline tumours; n=31 serous ovarian cancers (unlabelled columns); and four normal ovaries) as described in the Materials and Methods. Expression levels are colour coded with red, green and black corresponding to an increase, a decrease, and no change in gene expression, respectively.

Mentions: Principal components analysis on the top 500 most variable genes identified by transcript profiling showed that MOC can be clearly distinguished from the other subtypes of ovarian cancer by their expression profile, and cluster more closely to endometrioid ovarian cancer than to serous carcinomas (Figure 1A), as previously observed (Schwartz et al, 2002; Hart, 2005). Using a penalised t-statistic, we identified 167 probesets with higher expression in MOC compared to serous and endometrioid ovarian cancers (P-value adjusted for multiple testing <0.01) (Table 2 and Supplementary Data), and 18 probesets whose expression was lower in MOC compared to the other cancers (Supplementary Data). Hierarchical clustering illustrated that these genes can clearly separate MOC from the other subtypes of ovarian cancer, and shows that in most cases mucinous borderline tumours cluster closely with MOC (Figure 1B). Genes identified as having low expression in MOC compared to the other subtypes had similar expression levels in normal (whole) ovaries (Figure 1B), and their identification here may reflect their high expression in serous/endometrioid ovarian cancers rather than reduced expression in MOC.


A distinct molecular profile associated with mucinous epithelial ovarian cancer.

Heinzelmann-Schwarz VA, Gardiner-Garden M, Henshall SM, Scurry JP, Scolyer RA, Smith AN, Bali A, Vanden Bergh P, Baron-Hay S, Scott C, Fink D, Hacker NF, Sutherland RL, O'Brien PM - Br. J. Cancer (2006)

(A) Principal components analysis based on 500 genes with the most variable signal intensities (based on variance) separates the histological subtypes of EOC. MOC (n=3) are circled; (B) Hierarchical clustering and heat map of differentially expressed genes (n=167 upregulated and n=18 down-regulated) in MOC compared to serous and endometrioid ovarian cancers. Clustering was performed on all transcript profiled samples (n=3 MOC; n=4 mucinous borderline tumours; n=8 endometrioid ovarian cancers (endo); n=3 serous borderline tumours; n=31 serous ovarian cancers (unlabelled columns); and four normal ovaries) as described in the Materials and Methods. Expression levels are colour coded with red, green and black corresponding to an increase, a decrease, and no change in gene expression, respectively.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2361366&req=5

fig1: (A) Principal components analysis based on 500 genes with the most variable signal intensities (based on variance) separates the histological subtypes of EOC. MOC (n=3) are circled; (B) Hierarchical clustering and heat map of differentially expressed genes (n=167 upregulated and n=18 down-regulated) in MOC compared to serous and endometrioid ovarian cancers. Clustering was performed on all transcript profiled samples (n=3 MOC; n=4 mucinous borderline tumours; n=8 endometrioid ovarian cancers (endo); n=3 serous borderline tumours; n=31 serous ovarian cancers (unlabelled columns); and four normal ovaries) as described in the Materials and Methods. Expression levels are colour coded with red, green and black corresponding to an increase, a decrease, and no change in gene expression, respectively.
Mentions: Principal components analysis on the top 500 most variable genes identified by transcript profiling showed that MOC can be clearly distinguished from the other subtypes of ovarian cancer by their expression profile, and cluster more closely to endometrioid ovarian cancer than to serous carcinomas (Figure 1A), as previously observed (Schwartz et al, 2002; Hart, 2005). Using a penalised t-statistic, we identified 167 probesets with higher expression in MOC compared to serous and endometrioid ovarian cancers (P-value adjusted for multiple testing <0.01) (Table 2 and Supplementary Data), and 18 probesets whose expression was lower in MOC compared to the other cancers (Supplementary Data). Hierarchical clustering illustrated that these genes can clearly separate MOC from the other subtypes of ovarian cancer, and shows that in most cases mucinous borderline tumours cluster closely with MOC (Figure 1B). Genes identified as having low expression in MOC compared to the other subtypes had similar expression levels in normal (whole) ovaries (Figure 1B), and their identification here may reflect their high expression in serous/endometrioid ovarian cancers rather than reduced expression in MOC.

Bottom Line: The results show that MOC express a genetic profile that both differs and overlaps with other subtypes of epithelial ovarian cancer.In particular, galectin 4 (LGALS4) was highly and specifically expressed in MOC, but expressed at lower levels in benign mucinous cysts and borderline (atypical proliferative) tumours, supporting a malignant progression model of MOC.Hence LGALS4 may have application as an early and differential diagnostic marker of MOC.

View Article: PubMed Central - PubMed

Affiliation: Cancer Research Program, Garvan Institute of Medical Research, Darlinghurst, NSW 2010, Australia.

ABSTRACT
Mucinous epithelial ovarian cancers (MOC) are clinically and morphologically distinct from the other histological subtypes of ovarian cancer. To determine the genetic basis of MOC and to identify potential tumour markers, gene expression profiling of 49 primary ovarian cancers of different histological subtypes was performed using a customised oligonucleotide microarray containing >59 000 probesets. The results show that MOC express a genetic profile that both differs and overlaps with other subtypes of epithelial ovarian cancer. Concordant with its histological phenotype, MOC express genes characteristic of mucinous carcinomas of varying epithelial origin, including intestinal carcinomas. Differences in gene expression between MOC and other histological subtypes of ovarian cancer were confirmed by RT-PCR and/or immunohistochemistry. In particular, galectin 4 (LGALS4) was highly and specifically expressed in MOC, but expressed at lower levels in benign mucinous cysts and borderline (atypical proliferative) tumours, supporting a malignant progression model of MOC. Hence LGALS4 may have application as an early and differential diagnostic marker of MOC.

Show MeSH
Related in: MedlinePlus