Reduction in membranous immunohistochemical staining for the intracellular domain of epithelial cell adhesion molecule correlates with poor patient outcome in primary colorectal adenocarcinoma.

BACKGROUND: Epithelial cell adhesion molecule (epcam) is a multifunctional transmembrane glycoprotein expressed on both normal epithelium and epithelial neoplasms such as gastric, breast, and renal carcinomas. Recent studies have proposed that the proteolytic cleavage of the intracellular domain of epcam (epcam-icd) can trigger signalling cascades leading to aggressive tumour behavior. The expression profile of epcam-icd has not been elucidated for primary colorectal carcinoma. In the present study, we examined epcam-icd immunohistochemical staining in a large cohort of patients with primary colorectal adenocarcinoma and assessed its performance as a potential prognostic marker. METHODS: Immunohistochemical staining for epcam-icd was assessed on tissue microarrays consisting of 137 primary colorectal adenocarcinoma samples. Intensity of staining for each core was scored by 3 independent pathologists. The membranous epcam-icd staining score was calculated as a weighted average from 3 core samples per tumour. Univariate analysis of the average scores and clinical outcome measures was performed. RESULTS: The level of membranous epcam-icd staining was positively associated with well-differentiated tumours (p = 0.01); low preoperative carcinoembryonic antigen (p = 0.001); and several measures of survival, including 2-year (p = 0.02) and 5-year survival (p = 0.05), and length of time post-diagnosis (p = 0.03). A number of other variables-including stage, grade, and lymph node status-showed correlations with epcam staining and markers of poor outcome, but did not reach statistical significance. CONCLUSIONS: Low membranous epcam-icd staining might be a useful marker to identify tumours with aggressive clinical behavior and potential poor prognosis and might help to select candidates who could potentially benefit from treatment targeting epcam.

RET ligand-induced internalization and its consequences for downstream signaling.

RET is a receptor tyrosine kinase (RTK) with roles in cell growth, differentiation and survival. Ligand-induced activation of RET results in stimulation of multiple signal transduction pathways, including the MAP kinase/Erk and PI3 kinase/Akt pathways. However, the mechanisms governing receptor internalization and signal down- regulation have not been explored. As other RTKs are internalized through the clathrin-coated pit pathway in a ligand-dependant manner, we have investigated whether RET is internalized through a similar process. Using a highly sensitive fluorescence resonance energy transfer (FRET)-based assay, we have shown that RET is internalized from the plasma membrane in a ligand-dependant manner that requires RET kinase activity as well as the GTPase activity of the clathrin-coated vesicle scission protein dynamin 2. Further, we have demonstrated that RET colocalizes with Rab5a, a marker of clathrin-coated vesicles and early endosomes, after internalization. Finally, we demonstrated that RET internalization is required for complete activation of Erk1/2, but not for activation of Akt signaling. Our data suggest that ligand-induced internalization of RET not only plays an overall role in downregulation and termination of signaling, but also functions to traffic RET to subcellular locations where it can fully activate certain downstream signaling pathways.

Cloning and characterization of the human GFRA2 locus and investigation of the gene in Hirschsprung disease.

The glial-cell-line-derived neurotrophic factor (GDNF) family receptors alpha (GFRalpha) are cell surface bound glycoproteins that mediate interactions of the GDNF ligand family with the RET receptor. These interactions are crucial to the development of the kidney and some peripheral nerve lineages. In humans, mutations of RET or RET ligands are associated with the congenital abnormality Hirschsprung disease (HSCR) in which nerves and ganglia of the hind gut are absent. As the GFRalpha family are required for normal activation of the RET receptor, they are also candidates for a role in HSCR. The GFRA2 gene, which is required for the development of the myenteric nerve plexus, is an excellent candidate gene for HSCR. In this study, we cloned the human GFRA2 locus, characterized the gene structure, and compared it with other GFRA family members. We further investigated the GFRA2 gene for mutations in a panel of HSCR patients. GFRA2 has nine coding exons that are similar in size and organization to those of other GFRA family genes. We identified six sequence variants of GFRA2, four of which did not affect the amino acid sequence of the GFRalpha-2 protein. Two further changes that resulted in amino acid substitutions were found in exon 9 and were predicted to lie in the amino acid sequence encoding the glycosylphosphatidylinositol-linkage signal of GFRalpha-2. There was no difference in frequency of any of the sequence variants between control and HSCR populations. Our data indicate that members of the GFRA gene family are closely related in intron/exon structure and in sequence. We have not detected any correlation between sequence variants of GFRA2 and the HSCR phenotype.

Over-representation of a germline variant in the gene encoding RET co-receptor GFRalpha-1 but not GFRalpha-2 or GFRalpha-3 in cases with sporadic medullary thyroid carcinoma.

In contrast to the hereditary form of medullary thyroid carcinoma (MTC), little is known about the etiology of sporadic MTC. Somatic gain-of-function mutations in the RET proto-oncogene, encoding a receptor tyrosine kinase, are found in an average of 40% of sporadic MTC. We analysed 31 sporadic MTC for somatic and germline variants in GFRA1, GFRA2 and GFRA3 which encode the co-receptors of RET. Although there were no somatic mutations in any of the three genes, a sequence variant (-193C>G) in the 5'-UTR of GFRA1 was found in 15% of cases. Three patients were heterozygous (het); another three patients homozygous (hom) for the G variant. The G allele was not observed in 31 race-matched normal controls. Hence, the relative frequency of this variant in sporadic MTC cases and controls differed significantly (P<0.05). Since this variant lies in the 5' UTR, likely at the transcriptional start site, we analysed for differential expression of GFRalpha-1 at the transcript and protein levels. At the mRNA level, GFRA1 was over-expressed in tumors harboring the rare variant (P=0.06). The presence of the G polymorphic allele seemed to be associated with increased expression by immunostaining for GFRalpha-1. Interestingly, cytoplasmic staining was stronger in intensity for het patients and nuclear staining predominant in hom cases. In conclusion, mutation analysis of GFRA1, GFRA2 and GFRA3 revealed over-representation of a rare variant in GFRA1 (-193C>G) in the germline of sporadic MTC cases. Our data suggest that the mechanism is related to over-expression of GFRalpha-1 and differential subcellular compartmentalization but the precise mechanism as to how it acts as a low penetrance susceptibility allele for the development of sporadic MTC remains to be elucidated.

Reconsideration of biallelic inactivation of the VHL tumour suppressor gene in hemangioblastomas of the central nervous system.

OBJECTIVES: Cerebellar haemangioblastoma occurs sporadically or as a component tumour of autosomal dominant von Hippel-Lindau disease. Biallelic inactivation of the VHL tumour suppressor gene, which is located on chromosome 3p, has been shown to be involved in the pathogenesis of both tumour entities. Mechanisms of VHL inactivation are intragenic mutations, mitotic recombination events, and hypermethylation of the promoter region. The systematic and complete examination of these genetic and epigenetic phenomena in large series of von Hippel-Lindau disease related and sporadic hemangioblastomas has, thus far, not been performed. METHODS: In the largest series to date, 29 von Hippel-Lindau disease associated and 13 sporadic haemangioblastomas were investigated for all suggested inactivating mechanisms of the VHL gene using single strand conformational polymorphism (SSCP), loss of heterozygosity (LOH), and methylation analyses. Additionally, corresponding blood samples of all patients were screened for VHL germline mutations by SSCP and Southern blotting. RESULTS: Germline mutations were identified in 94% of patients with von Hippel-Lindau disease and their tumours and 62% of these hemangioblastomas showed LOH of chromosome 3p. Of the 13 sporadic tumours, 23% showed a single somatic mutation of the VHL gene that was not present in the germline. 3p LOH was identified in 50% of informative sporadic tumours. No von Hippel-Lindau disease related or sporadic tumour demonstrated VHL promoter hypermethylation. CONCLUSIONS: For most von Hippel-Lindau disease related haemangioblastomas, the inactivation or loss of both alleles of the VHL gene, as predicted by the Knudson two hit theory, is required. However, in a subset of tumours including most sporadic haemangioblastomas, the genetic pathways involved in tumorigenesis have yet to be defined and may represent alterations of a different pathway or pathways.

Conservation of RET proto-oncogene splicing variants and implications for RET isoform function.

The RET proto-oncogene encodes a receptor tyrosine kinase required for development of the kidney and neural crest-derived cell types. Alternative splicing of the 3' exons of human RET results in three protein isoforms with distinct C-termini: RET9, RET51, and RET43. These RET isoforms show differential binding to downstream adapter molecules, suggesting they may have distinct signaling functions. We have characterized Ret 3' sequences in mouse and investigated alternative splicing of this region. We found that the organization of Ret 3' sequences is very similar to human RET. The mouse locus also has alternatively spliced C-terminal coding regions, and the sequences corresponding to RET9 and RET51 are highly conserved in both position and sequence with the human locus. Further, we compared the predicted C-terminal amino acids of RET9 and RET51 in seven vertebrate species, and found that they are well conserved. We have identified sequence encoding a putative ret43 isoform in mouse, however the predicted amino acid sequence showed low homology to human RET43. Our data suggest that RET isoforms are evolutionarily highly conserved over a broad range of species, which may indicate that each isoform has a distinct role in normal RET function.

Multiple endocrine neoplasia type 2 and RET: from neoplasia to neurogenesis.

Multiple endocrine neoplasia type 2 (MEN 2) is an inherited cancer syndrome characterised by medullary thyroid carcinoma (MTC), with or without phaeochromocytoma and hyperparathyroidism. MEN 2 is unusual among cancer syndromes as it is caused by activation of a cellular oncogene, RET. Germline mutations in the gene encoding the RET receptor tyrosine kinase are found in the vast majority of MEN 2 patients and somatic RET mutations are found in a subset of sporadic MTC. Further, there are strong associations of RET mutation genotype and disease phenotype in MEN 2 which have led to predictions of tissue specific requirements and sensitivities to RET activity. Our ability to identify genetically, with high accuracy, subjects with MEN 2 has revolutionised our ability to diagnose, predict, and manage this disease. In the past few years, studies of RET and its normal ligand and downstream interactions and the signalling pathways it activates have clarified our understanding of the roles played by RET in normal cell survival, proliferation, and differentiation, as well as in disease. Here, we review the current knowledge of the normal functions of RET and the effects of mutations of this gene in tumorigenesis and in normal development.

Sp1 and Sp3 transactivate the RET proto-oncogene promoter.

The RET proto-oncogene plays an important role in the initiation and progression of tumors derived from the neural crest. The cis-regulatory elements responsible for RET basal promoter activity have not been identified. To characterize these elements, a RET promoter DNA fragment (-453 to +227bp) was fused to a luciferase reporter and introduced into TT, a neural crest-derived cell line. Sequential 5' deletions of the promoter revealed that optimal expression of the RET promoter in TT cells required only 70bp of sequence upstream of the transcription start site, and contains two Sp1 binding sites. DNase I footprinting, electrophoretic mobility shift analysis (EMSA), and supershift assays revealed that this region binds both Sp1 and its related protein, Sp3. Additionally, RET basal promoter activity was abrogated by removal of these Sp1/Sp3 binding sites. The proximal two GC boxes were sufficient to allow transactivation of the RET promoter in Drosophila SL2 cells. Sp3 expression in these cells caused an additional activation of the promoter. These results demonstrate that the transactivation of the RET promoter within a neural crest-derived cell line is dependent on Sp1 and Sp3.

Characterisation of the human GFRalpha-3 locus and investigation of the gene in Hirschsprung disease.

BACKGROUND: The GDNF family receptor alpha (GFRalpha) proteins are extracellular cell surface bound molecules that act as adapters in binding of the GDNF family of soluble neurotrophic factors to the RET receptor. These molecules are essential for development of many neural crest derived cell types and the kidney. Mutations in RET and in two members of the GDNF ligand family are associated with Hirschsprung disease (HSCR), a congenital absence of the enteric ganglia. Members of the GFRalpha family are also candidates for HSCR mutations. One such gene is GFRalpha-3, which is expressed in the peripheral nervous system and developing nerves. OBJECTIVE: We have characterised the structure of the human GFRalpha-3 locus and investigated the gene for sequence variants in a panel of HSCR patients. METHODS: Long range PCR or subcloning of PAC clones was used to investigate GFRalpha-3 intron-exon boundaries. A combination of single strand conformation polymorphism (SSCP) analysis and direct sequencing was used to investigate GFRalpha-3 sequence variants. RESULTS: GFRalpha-3 spans eight coding exons and has a gene structure and organisation similar to that of GFRalpha-1. We identified three polymorphic variants in GFRalpha-3 in a normal control population, a subset of which also occurred in HSCR patients. We did not detect any sequence variants within the coding sequence of GFRalpha-3. We found a base substitution in the 5' UTR of GFRalpha-3, 15 base pairs upstream of the translation start site. A second substitution was identified in intron 4 (IVS4-30G>A) between the splice branch site and the splice acceptor site. The final variant was a 2 base pair insertion within the splice donor consensus sequence of exon 7 (IVS7+4ins GG). CONCLUSIONS: We did not detect any correlation between variants of GFRalpha-3 and the HSCR phenotype. Our data suggest that mutations of this gene are not a cause of HSCR.

Immunohistochemical evidence of loss of PTEN expression in primary ductal adenocarcinomas of the breast.

Germline mutations in PTEN, encoding a dual-specificity phosphatase on 10q23.3, cause Cowden syndrome (CS), which is characterized by a high risk of breast and thyroid cancers. Loss of heterozygosity of 10q22-24 markers and somatic PTEN mutations have been found to a greater or lesser extent in a variety of sporadic component and noncomponent cancers of CS. Among several series of sporadic breast carcinomas, the frequency of loss of flanking markers around PTEN is approximately 30 to 40%, and the somatic intragenic PTEN mutation frequency is <5%. In this study, we analyzed PTEN expression in 33 sporadic primary breast carcinoma samples using immunohistochemistry and correlated this to structural studies at the molecular level. Normal mammary tissue had a distinctive pattern of expression: myoepithelial cells uniformly showed strong PTEN expression. The PTEN protein level in mammary epithelial cells was variable. Ductal hyperplasia with and without atypia exhibited higher PTEN protein levels than normal mammary epithelial cells. Among the 33 carcinoma samples, 5 (15%) were immunohistochemically PTEN-negative; 6 (18%) had reduced staining, and the rest were PTEN-positive. In the PTEN-positive tumors as well as in normal epithelium, the protein was localized in the cytoplasm and in the nucleus (or nuclear membrane). Among the immunostain negative group, all had hemizygous PTEN deletion but no structural alteration of the remaining allele. Thus, in these cases, an epigenetic phenomenon such as hypermethylation, -ecreased protein synthesis or increased protein degradation may be involved. In the cases with reduced staining, 5 of 6 had hemizygous PTEN deletion and 1 did not have any structural abnormality. Finally, clinicopathological features were analyzed against PTEN protein expression. Three of the 5 PTEN immunostain-negative carcinomas were also both estrogen and progesterone receptor-negative, whereas only 5 of 22 of the PTEN-positive group were double receptor-negative. The significance of this last observation requires further study.

Mutation and deletion analysis of GFR alpha-1, encoding the co-receptor for the GDNF/RET complex, in human brain tumours.

Glial cell line-derived neurotrophic factor (GDNF) plays a key role in the control of vertebrate neuron survival and differentiation in both the central and peripheral nervous systems. GDNF preferentially binds to GFRalpha-1 which then interacts with the receptor tyrosine kinase RET. We investigated a panel of 36 independent cases of mainly advanced sporadic brain tumours for the presence of mutations in GDNF and GFRalpha-1. No mutations were found in the coding region of GDNF. We identified six previously described GFRalpha-1 polymorphisms, two of which lead to an amino acid change. In 15 of 36 brain tumours, all polymorphic variants appeared to be homozygous. Of these 15 tumours, one also had a rare, apparently homozygous, sequence variant at codon 361. Because of the rarity of the combination of homozygous sequence variants, analysis for hemizygous deletion was pursued in the 15 samples and loss of heterozygosity was found in 11 tumours. Our data suggest that intragenic point mutations of GDNF or GFRalpha-1 are not a common aetiologic event in brain tumours. However, either deletion of GFRalpha-1 and/or nearby genes may contribute to the pathogenesis of these tumours.

Investigation of germline GFR alpha-1 mutations in Hirschsprung disease.

Inactivating mutations of the RET proto-oncogene and of one of its soluble ligand molecules, glial cell line derived neurotrophic factor (GDNF), have been found in a subset of patients with Hirschsprung disease (HSCR). However, the majority of HSCR mutations remain unidentified. As normal RET function requires a multicomponent ligand complex for activation, other members of the RET ligand complex are primary candidates for these mutations. We investigated the presence of mutations in another member of the RET signalling complex, GDNF family receptor alpha-1 (GFR alpha-1), in a panel of 269 independent cases of HSCR. We identified 10 polymorphisms at the GFR alpha-1 locus. Surprisingly, however, we did not identify any sequence variants in our HSCR population that were not also present in a normal control population. Our data suggest that mutations of the GFR alpha-1 gene are not a common aetiological event in HSCR.

Analysis of the 10q23 chromosomal region and the PTEN gene in human sporadic breast carcinoma.

We examined a panel of sporadic breast carcinomas for loss of heterozygosity (LOH) in a 10-cM interval on chromosome 10 known to encompass the PTEN gene. We detected allele loss in 27 of 70 breast tumour DNAs. Fifteen of these showed loss limited to a subregion of the area studied. The most commonly deleted region was flanked by D10S215 and D10S541 and encompasses the PTEN locus. We used a combination of denaturing gradient gel electrophoresis and single-strand conformation polymorphism analyses to investigate the presence of PTEN mutations in tumours with LOH in this region. We did not detect mutations of PTEN in any of these tumours. Our data show that, in sporadic breast carcinoma, loss of heterozygosity of the PTEN locus is frequent, but mutation of PTEN is not. These results are consistent with loss of another unidentified tumour suppressor in this region in sporadic breast carcinoma.

Mutation of the RET proto-oncogene is correlated with RET immunostaining in subpopulations of cells in sporadic medullary thyroid carcinoma.

Mutations in the RET proto-oncogene, which encodes a receptor tyrosine kinase, are associated with the pathogenesis of medullary thyroid carcinoma (MTC). Somatic mutations in RET, predominantly at codon 918, and very rarely at codon 883, have been found in a proportion of sporadic MTC. We have previously shown that approximately 80% of sporadic MTCs had at least one subpopulation with a somatic RET mutation. Uneven distribution of somatic mutation within a single tumor or among metastases from a single individual was notable. In the present study, we sought to correlate RET expression, as demonstrated by RET immunohistochemistry, with mutation status in sporadic MTC for each tumor. Seventy evaluable subpopulations, belonging to 28 unrelated sporadic cases, comprising primary MTC and metastases, were immunostained with two different polyclonal antibodies raised against the C-terminus of RET. The regional presence of codon 918 or 883 seemed to coincide with increased RET immunopositivity in at least 62 of 70 (89%, P < 0.000001) tumor subpopulations. The reasons for this concordance are not entirely clear but could be related to either RNA or protein stability. Preliminary studies have suggested that the presence of somatic codon 918 mutation in MTC has a prognostic significance. If these preliminary results prove true, then given our data, we can further explore the feasibility of RET immunocytochemistry as a rapid assessment for the presence of somatic codon 918 for molecular diagnostic and prognostic purposes.

Analysis of PTEN and the 10q23 region in primary prostate carcinomas.

Deletions involving chromosome 10q23 occur frequently in prostatic carcinomas. Recently, a novel tumour suppressor gene, PTEN, mapping to this interval, has been identified. Mutation or deletion of PTEN has been observed in a proportion of prostate cancer cell lines; however, primary prostate carcinomas have not been studied. We have investigated the involvement of PTEN in primary prostatic adenocarcinomas using a panel of 51 matched normal and prostate tumour DNAs. We first determined the proportion of tumours with allele loss at loci in 10q23 which span the region containing the PTEN gene. Our results show that LOH involving 10q23 is common in primary prostate carcinomas. Twenty-five of 51 (49%) tumours showed loss of heterozygosity (LOH) over the region spanning the PTEN locus. We next directly analysed the PTEN gene for mutations of the coding region using single strand conformation polymorphism (SSCP) and sequence analyses. Of those tumours with LOH, only a single tumour was found to carry a missense mutation in PTEN. No mutations in PTEN were identified in tumours without LOH. Our results suggest either that mutation of PTEN is a late event in prostate tumorigenesis, or that another tumour suppressor gene important in prostate cancer may lie close to PTEN in 10q23.

Investigation of the genes for RET and its ligand complex, GDNF/GFR alpha-I, in small cell lung carcinoma.

RET is a receptor tyrosine kinase expressed in neuroendocrine cells and in tumors of these cell types. RET activation may be mediated by a ligand complex comprising glial cell line-derived neurotrophic factor (GDNF) and GDNF family receptor alpha-1 (GFR alpha-1). Activating RET mutations are found in the inherited cancer syndrome multiple endocrine neoplasia type 2 and in a subset of the related sporadic tumors, medullary thyroid carcinoma and pheochromocytoma, both being derived from neuroendocrine tissues. In one small study, mutations were identified in another tumor with neuroendocrine features, small cell lung carcinoma (SCLC). To determine whether RET mutations contribute to the pathogenesis of SCLC, we examined a panel of 54 SCLC cell lines. No mutations were identified in RET exons 10, 11, and 13-16, regions previously implicated in SCLC or other neuroendocrine tumors. We further examined the expression pattern of RET and the genes encoding the components of its ligand complex GDNF and GFR alpha-1, in 21 SCLC lines by using RT-PCR. Although we found no consistent pattern of expression for these three genes, RET was expressed in 57% of SCLC lines. Thus, although RET mutations appear unlikely to be an important step in the tumorigenesis of SCLC, the frequent expression of this gene suggests that RET may have a mitogenic role in a subset of SCLC cell lines.

Genomic structure and chromosomal localization of the human GDNFR-alpha gene.

GDNFR-alpha is a glycosyl-phosphotidylinositol-linked receptor for glial cell line-derived neurotrophic factor (GDNF). GDNF binds to GDNFR-alpha and this complex, in turn, is believed to interact with the RET receptor tyrosine kinase to effect downstream signalling. GDNFR-alpha belongs to a novel gene family without strong homology to known genes. Thus, little information has been available to help predict genomic structure or location of this gene. In this study, the genomic organization of human GDNFR-alpha was delineated through a combination of PAC clone characterization, long distance PCR and sequence analyses. Exon-intron boundaries were defined by comparing the size and sequence of the genomic PCR products to those predicted by the cDNA sequence. The human GDNFR-alpha gene comprises 9 exons. GDNFR-alpha PAC clones were used for FISH analysis to map this gene to 10q26.

Expression of RET 3' splicing variants during human kidney development.

The mature mammalian kidney arises through a series of reciprocal inductive interactions between two different cell groups, the ureteric bud epithelium and the metanephric mesenchyme. The RET receptor tyrosine kinase is required for induction and development of the metanephric kidney. Differential splicing at the 3' end of RET results in transcripts encoding three isoforms that differ with respect to their C-terminal 9 (RET9), 51 (RET51) or 43 (RET43) amino acids. In vitro assays have identified differences in the abilities of the RET9 and RET51 isoforms to induce differentiation suggesting functional differences between these proteins. We examined the relative expression levels of the three RET 3' splicing variants in developing human kidney using semi-quantitative RT-PCR. We observed consistent expression of the RET9 and RET43 variants in kidney samples spanning 7.5 through 24 weeks gestation. At early gestational ages (7.5-8.5 weeks), RET51 expression was very low (+/-5%) compared to RET9; however, a rapid seven fold increase in expression was detected by 9 weeks. Our data suggest that RET51 may contribute to differentiation-related events occurring after 8.5 weeks gestation rather than to induction of the human kidney.

Germline dinucleotide mutation in codon 883 of the RET proto-oncogene in multiple endocrine neoplasia type 2B without codon 918 mutation.

The autosomal dominant multiple endocrine neoplasia type 2 syndromes (MEN 2) comprise three clinically distinct entities, MEN 2A, familial medullary thyroid carcinoma and MEN 2B, which share a common clinical feature: medullary thyroid carcinoma (MTC). MEN 2B is considered to have the most aggressive form of MTC. Therefore, early detection of MEN 2B in order to prevent potentially lethal MTC is important. More than 95% of all MEN 2B cases are caused by germline mutation at codon 918 (M918T) in exon 16 of the RET proto-oncogene. In this study, we demonstrate the presence of germline codon 883 mutation (A883F) in 2 of 3 unrelated MEN 2B cases without codon 918 mutation. Our data demonstrate a novel etiologic event which may have roles in predisposition to MEN 2B when present in the germline and in the pathogenesis of sporadic MTC when somatic.