Variant V6 Research Articles

In vitro expression of soluble and cell surface-associated CD44 on endometrial cells from women with and without endometriosis.

Endometriosis is one of the most common benign gynaecological diseases, and attachment of retrogradely shed viable endometrial cells is considered to be important in its development. CD44 is a multifunctional adhesion molecule that undergoes alternative splicing, giving rise to different isoforms. The expression of cell surface-associated CD44 std, v4, v5, v6 and v10 variants before and after cytokine treatment was investigated in endometrial cultures derived from 10 endometriosis patients and 22 women without the disease using immunocytochemistry. The immunoreactivity of soluble CD44 std, v5 and v6 variants was measured in culture medium using an enzyme immunoassay kit. We report on the presence of soluble CD44 in endometrial culture supernatants. In particular, circulating CD44 standard form levels were significantly higher than levels of splice variants. We also found that both epithelial and stromal cells express surface-associated CD44 molecules in a distinct pattern and that this expression is not modulated by tumour necrosis factor (TNF)-alpha or/and interleukin 1 (IL-1) alpha/beta. Finally, cell surface-associated as well as soluble CD44 expression was similar in the two groups. Our results indicate that endometrial cells can serve as a source of circulating CD44, but a direct role in the pathogenesis of endometriosis is rather improbable.

Expression of CD44 splice variant v10 in Hodgkin's disease is associated with aggressive behaviour and high risk of relapse

Expression of CD44 isoforms has been shown to correlate with the progression and prognosis of some malignant tumours. The aim of this study was to investigate the expression of CD44 standard (CD44s) and CD44 splice variants (CD44v) v5, v6, and v10 in lymph node specimens from patients with nodular sclerosing Hodgkin's disease (NSHD), with or without initial bone marrow involvement and with or without relapse. Specimens were studied by immunohistochemistry to determine CD44s and CD44v in Hodgkin- and Reed–Sternberg (HRS) cells. For validation of the immunohistochemical detection of CD44v10 in paraffin-embedded samples, selected cases were analysed in parallel immunohistochemically using fresh frozen material and by reverse transcription-polymerase chain reaction (RT-PCR). There was high expression of CD44 isoforms containing the variant exon v10 selectively in HRS cells of patients with relapse within 2–3 years or with initial bone marrow involvement. In patients without relapse, however, no or only very few HRS cells were positive. These differences were statistically highly significant (p≤ 0001), whereas evaluation of CD44s, CD44v5, and v6 expression revealed no marked differences. It is concluded that evaluation of CD44v10 expression could serve as a new prognostic marker in NSHD. These results are considered to be of sufficient importance to initiate a large multi-institutional study for confirmation; furthermore, they might suggest causal involvement of CD44v10 in the progression of NSHD. Copyright © 1998 John Wiley & Sons, Ltd.

Expression of CD44 splice variant v10 in Hodgkin's disease is associated with aggressive behaviour and high risk of relapse.

Expression of CD44 isoforms has been shown to correlate with the progression and prognosis of some malignant tumours. The aim of this study was to investigate the expression of CD44 standard (CD44s) and CD44 splice variants (CD44v) v5, v6, and v10 in lymph node specimens from patients with nodular sclerosing Hodgkin's disease (NSHD), with or without initial bone marrow involvement and with or without relapse. Specimens were studied by immunohistochemistry to determine CD44s and CD44v in Hodgkin- and Reed-Sternberg (HRS) cells. For validation of the immunohistochemical of detection of CD44v10 in paraffin-embedded samples, selected cases were analysed in parallel immunohistochemically using fresh frozen material and by reverse transcription-polymerase chain reaction (RT-PCR). There was high expression of CD44 isoforms containing the variant exon v10 selectively in HRS cells of patients with relapse within 2-3 years or with initial bone marrow involvement. In patients without relapse, however, no or only very few HRS cells were positive. These differences were statistically highly significant (p < or = 0.001), whereas evaluation of CD44s, CD44v5, and v6 expression revealed no marked differences. It is concluded that evaluation of CD44v10 expression could serve as a new prognostic marker in NSHD. These results are considered to be of sufficient importance to initiate a large multi-institutional study for confirmation; furthermore, they might suggest causal involvement of CD44v10 in the progression of NSHD.

CD44 splice variant expression in normal and malignant uterine cervical epithelium

CD44 expression was investigated immunohistochemically on paraffin sections obtained from 88 uterine cervical cancers and 31 normal cervices, using monoclonal antibodies against CD44 variant epitopes v4, v5, v6, v7/8, v9 and the standard form of the CD44 protein. Normal epithelium showed expression of all CD44 splice variants, at least in traces, and it was located predominantly in basal and parabasal cells. In cervical carcinomas CD44 expression was widely heterogeneous. CD44 variant v9 staining was moderate or strong in 86% of the tumors, and it was significantly correlated with CD44 v6 staining. Also a significant correlation between expression of CD44 v4 and v6 occurred. Highly significant correlations between CD44 expression of variants v4 and v6 and tumor stage as well as patients age were found. In addition, these variants were more frequently expressed in squamous cell carcinomas than in adenocarcinomas. However, in contrast to the recently reported data, we were not able to confirm the hypothesis that CD44 v6 represents a prognostic indicator in cervical cancer. In FIGO stages III and IV, patients with CD44 variant v4 positive tumors had a significantly longer disease-free and overall survival than patients with CD44 variant v4 negative tumors. In conclusion, our data indicate that CD44 v6 tissue expression cannot be considered as a prognostic factor in cervical cancer. Regarding the unexpected outcome of patients with CD44 v4 positive tumors, further investigations are needed to elucidate the exact clinical value of this variant isoform.

CD44v3 and v6 variant isoform expression correlates with poor prognosis in early-stage vulvar cancer.

Expression of alternatively spliced CD44 isoforms has been reported to correlate with poor prognosis in human squamous cell cancers, i.e. squamous cell cancer of the lung and cervix. The aim of this study was to evaluate whether CD44 isoform expression is a prognostic factor in early-stage squamous cell cancer of the vulva. Seventy cases of squamous cell carcinoma of the vulva International Federation of Gynaecology and Obstetrics (FIGO) stage I were examined immunohistochemically for expression of CD44 isoforms. We used four different variant exon sequence-specific murine monoclonal antibodies to epitopes encoded by exons v3, v5, v6 and v7-8 of human variant CD44. The correlation of CD44 expression with histological grade and disease-free and overall survival was investigated. CD44 isoforms CD44v3, CD44v5, CD44v6 and CD44v7-8 were detected in 28% (20/70), 47% (33/70), 33% (23/70) and 17% (12/70) of the tumour samples respectively. Patients suffering from tumours expressing CD44v6 had a poorer relapse-free (log-rank test, P = 0.02) and overall survival (log-rank test, P = 0.03). Likewise, patients suffering from tumours expressing CD44v3 had a poorer relapse-free (log-rank test, P = 0.04) and overall survival (log-rank test, P = 0.01). Expression of CD44v5 and CD44v7-8 did not compromise the patients' outcome. Histological grade did not correlate with CD44 isoform expression. Immunohistochemically detected expression of CD44 isoforms containing variant exon v6 or v3 is correlated with a poor relapse-free and overall survival in FIGO stage I vulvar cancer patients.

Expression of CD44 variant proteins in adenocarcinoma of Barrett's esophagus and its relation to prognosis.

None of the commonly used staging criteria accurately determine the prognosis of a patient with adenocarcinoma of Barrett's esophagus. The authors therefore assessed the expression pattern and prognostic impact of CD44 standard and CD44 isoforms CD44v4, v5,v6,v7, and v10 in adenocarcinoma of Barrett's esophagus. Specimens from 41 patients with adenocarcinoma of Barrett's esophagus who underwent esophageal resection were embedded in paraffin and studied immunohistochemically to determine the expression of CD44 splice variants. Histomorphologic parameters and survival time were not known at the time of the investigation. Correlations between favorable clinical or histomorphologic parameters and CD44s or any of the split variants could not be established. Down-regulation of CD44s and the split variant v10 was significantly correlated with pT classification. Furthermore, down-regulation of CD44v10 and up-regulation of CD44v7 were significantly correlated with ploidy. There was a significant correlation between CD44s and split variants in tumorous and nontumorous tissue from the same patient. Down-regulation of CD44s and CD44v4 had a significant influence on prognosis in that it was associated with shortened life expectancy. Multivariate analysis revealed that the expression of CD44v4 was an independent factor in prognosis. The results obtained for this small patient sample suggest that CD44v4 is a new independent prognostic parameter for adenocarcinoma of Barrett's esophagus that can be determined preoperatively by biopsy. It may therefore be helpful in planning therapy by allowing the identification of patients who may benefit from esophageal resection as well as those who are at high risk for morbidity and mortality even when the tumor is otherwise resectable. Further studies of larger patient samples are required to validate the results of the current study.

Expression and Characterization of Canine Cytochrome P450 2D15

CYP2D15 is the canine ortholog of human CYP2D6, the human CYP2D isoform involved in the metabolism of drugs such as antiarhythmics, adrenoceptor antagonists, and tricyclic antidepressants. Similar to human, canine CYP2D15 is expressed in the liver, with detectable levels in several other tissues. Three different CYP2D15 cDNA clones were obtained by RT-PCR from dog liver RNA. Two clones corresponded to variant full-length CYP2D15 cDNAs (termed CYP2D15 WT2 and CYP2D15 V1); the third was identified as a splicing variant missing exon 3 (termed CYP2D15 V2). Recombinant baculoviruses were constructed containing full-length cDNAs and used to express CYP2D15 WT2 and CYP2D15 V1 inSpodoptera frugiperda(Sf9) cells with expression levels of up to 0.14 nmol/mg cell protein. As with human CYP2D6, the recombinant CYP2D15 enzymes exhibited bufuralol 1′-hydroxylaseand dextromethorphanO-demethylase activities whencoexpressed with rabbit NADPH:P450 oxidoreductase. For bufuralol 1′-hydroxylase, apparentKmvalues were 4.9, 3.7, and 2.5 μM and theVmaxvalues were 0.14, 0.034, and 0.60 nmol/min/mg protein for dog liver microsomes, CYP2D15 WT2, and the variant CYP2D15 V1, respectively. For dextromethorphanO-demethylase, apparentKmvalues were 0.6, 0.6, and 2.0 μM and theVmaxvalues were 0.18, 0.034, and 0.057 nmol/min/mg protein for dog liver microsomes, CYP2D15 WT2, and the variant CYP2D15 V1, respectively. The human CYP2D6-specific inhibitor quinidine and the rat CYP2D1-specific inhibitor quinine were both shown to be inhibitors of bufuralol 1′-hydroxylase activity for dog liver microsomes, CYP2D15 WT2, and the CYP2D15 V1 variant with nearly equal potency. Thus, the dog expresses a CYP2D ortholog possessing enzymatic activities similar to human CYP2D6, but is affected by the inhibitors quinine and quinidine in a manner closer to that of rat CYP2D1.

Predominant Expression of CD44 Splice Variant v10 in Malignant and Reactive Human Skin Lymphocytes

The remarkable functional diversity of the cell surface receptor CD44 may be due to expression of multiple variant isoforms generated by alternative splicing of variant exons. Functional and correlative data implicate a role of CD44 variant isoforms in adhesion dependent processes such as lymphocyte recirculation and tumor progression and metastasis. We have analyzed 25 primary cutaneous lymphomas and 35 reactive lymphoid cell skin infiltrates or T cell-mediated skin diseases for the expression of CD44 variant isoforms. Irrespective of histologic typing, staging, and grading, cutaneous lymphomas as well as nonmalignant skin-infiltrating CD3+ CD4+ and CD8+ T and CD19+ B lymphocytes exhibited a strong expression of CD44v10 and a moderate expression of CD44v3 as determined by immunohistochemistry, immunofluorescence microscopy, and mRNA analysis. Expression of v5, v6, v7, and v9-containing CD44 variant isoforms was not detected. Furthermore, flow cytometry revealed expression of CD44v10 on a significant proportion of peripheral blood lymphocytes from Sézary's syndrome patients and a remarkable co-expression with cutaneous lymphocyte antigen. These results indicate a distinct CD44 variant isoform expression pattern associated with skin-homing lymphocytes different to lymphatic cells at noncutaneous sites. This differential expression pattern of CD44 variant isoforms may contribute to the development of lymphocyte skin infiltrates and/or the unique biologic behavior of cutaneous lymphomas.

Expression of CD44 variant exons by primary and metastatic oral squamous carcinomas.

Abnormal CD44 expression in many neoplasms correlates with behaviour, but reports on its role in oral squamous carcinoma are contradictory. CD44 expression was characterised in a closely matched series of oral carcinomas with and without metastases in both frozen and formalin-fixed tissue and correlated with behaviour and histological grading parameters. Eleven primary oral squamous carcinomas without metastases and nine primary carcinomas with 19 matched metastases were stained immunocytochemically for CD44H and products of variant exons v3, v4/5, v6 and v9. Patterns of staining in frozen and formalin-fixed tissue were correlated with invasive front grading and behaviour using exact inferential statistics. Most primary carcinomas stained for all exons tested but some showed loss of expression of v4/5. Loss of expression was more marked in metastases, but there was no correlation between expression and behaviour or grade. Stromal surfaces of epithelial cells often expressed variant exon products reflecting loss of polarity. This, together with selective loss of v4 and v5 in primary carcinomas and their more frequent loss in metastases, suggests that CD44 may play a role in metastasis of some oral squamous carcinomas.

Activation-Dependent Modulation of Hyaluronate-Receptor Expression and of Hyaluronate-Avidity by Human Monocytes

During inflammation, activated monocytes (Mo) migrate into tissues where they interact with extracellular matrix components such as hyaluronate (HA), produced in high amounts at inflammatory sites. We determined whether Mo that had invaded sites of cutaneous inflammation bind HA and express the putative HA receptors CD44 isoforms, ICAM-1, or receptor for hyaluronate-mediated motility (RHAMM). In cutaneous inflammation, activated infiltrating Mo displayed high HA avidity and expressed epitopes encoded by CD44s, CD44 variant exons v3, v4, v5, v6, v7, and v9, and ICAM-1, but not RHAMM. We further investigated how activation affects the avidity of Mo for HA and which receptors were responsible for such binding. Mo freshly purified from human peripheral blood bound little HA and expressed CD44s but no epitopes encoded by CD44v exons, ICAM-1, or RHAMM. During short-term tissue culture, Mo upregulated their HA avidity and expression of ICAM-1, CD44s, and epitopes encoded by CD44v, all of which were further augmented by IFN-gamma or lipopolysaccharide, whereas RHAMM was not detectable. Thus in vitro activated Mo resembled Mo that had migrated to inflammatory sites in vivo. Lipolysaccharide or IFN-gamma-induced HA binding was inhibited by more than 90% with monoclonal antibodies directed against N-terminal HA binding domains of CD44s, but not by monoclonal antibodies against CD44v epitopes or ICAM-1. In conclusion, we show that upon in vitro or in vivo activation, Mo enhance their capacity to bind HA. This is critically dependent upon the expression ofCD44s epitopes. Regulated CD44-HA interactions may be important for the ability of Mo to migrate into and within sites of inflammation and for Mo effector functions.

Immunhistochemischer Nachweis von CD44v-Proteinen bei präneoplastischen und neoplastischen Veränderungen der Vulva

Purpose : The cell surface glycoproteins of the CD44 family are divided into standard CD44 (CD44 s) and splice variants of CD44 (CD44v). In normal human tissues CD44v expression is mainly restricted to certain epithelia. Its expression could also be demonstrated in human gynaecological malignancies by immunohistochemistry and semi-quantitative reverse transcription PCR. In this study we investigated the expression of CD44v epitopes encoded by the variant exons v5, v6 and v3 to v10 in preneoplastic and neoplastic lesions of the vulva. Method and material: We investigated the expression of CD44v in biopsies of 30 patients with VIN and 11 patients with vulval carcinoma by immunohistochemical APAAP staining. In addition HPV-DNA was tested using PCR on 10 μm thick paraffin sections. We examined samples for HPV 6 and 11 as well as for 16 and 18. Results: In normal vulval keratinocytes CD44v5, CD44v6 and CD44v3-v10 is present in the membranes of the lower two thirds of the epithelium. The same distribution pattern of CD44v is seen in mild and moderate intraepithelial neoplasia of the vulva (VIN I/II). In severe intraepithelial neoplasia of the vulva (VIN III) we observed alterations in the CD44v expression in comparison to normal vulval tissue. The detection of CD44v isoforms is reduced. In squamous cell carcinoma of the vulva CD44v5, CD44v6 and CD44v3-v10 is present in the membranes of carcinoma cells in non-keratinizing areas. In poor differentiated areas of squamous cell carcinoma we found a loss of CD44v expression. CD44v expression is independent of age, histological subtype of VIN III or vulval carcinoma and detection of HPV-DNA type 6/11, 16 and 18. Conclusion: According to our first immunohistochemical results, CD44v expression is not correlated to poor clinical outcome of patients with preneoplastic and neoplastic diseases of the vulva.

The prognostic value of CD44 isoform expression in endometrial cancer.

Isoforms of the transmembrane glycoprotein CD44 have been implicated in tumour cell adhesion, tumour differentiation and metastatic spread in various human malignancies. We investigated the expression of CD44 isoforms containing variant exons v3, v5, v6 and v7-8 in 156 human endometrium cancer specimens by means of immunohistochemistry. CD44 isoforms CD44v3, CD44v5, CD44v6 and CD44v7-8 were detected in 26% (41 out of 156), 31% (48 out of 156), 22% (35 out of 156) and 15% (23 out of 156) of the tumour samples respectively. The expression of CD44 isoforms CD44v3, CD44v5 and CD44v7-8 showed no prognostic impact. In the univariate analysis, the expression of CD44v6 showed an association with shortened overall survival (log-rank test, P = 0.06). Multivariate analysis correcting for the confounding variable histological grading revealed CD44v6 not to be a prognostic factor in endometrial cancer (log-rank test, P = 0.06). Comparing the expression of CD44 isoforms CD44v3, CD44v5, CD44v6 and CD44v7-8 in 45 specimens of normal endometrial tissue, we found an up-regulation of all investigated CD44 isoforms in the secretory phase compared with the proliferative phase of the menstrual cycle. Our data indicate that the expression of CD44 isoforms, while obviously playing a role in the functional changes of normal endometrium, is not an adverse predictive factor in endometrial cancer.

CD44 Isoforms, Including the CD44 V3 Variant, Are Expressed on Endothelium, Suggesting a Role for CD44 in the Immobilization of Growth Factors and the Regulation of the Local Immune Response

Endothelium plays a central role in the regulation of site and inflammation specific leukocyte migration. Some of the mediators involved in leukocyte migration, such as chemokines, can bind to heparan sulfate on the endothelium resulting in immobilization near their sites of production. Because CD44 variants expressing V3 have been shown to carry heparan sulfate side chains and to interact through these side chains with heparan sulfate binding growth factors, we investigated the expression of CD44 variants on endothelium. We found a strong expression of V5, V7-8 and V10 CD44 variants and a weaker expression of V3 and V6 CD44 variants on endothelium by using immuno-histochemistry and by FACS analysis. Expression of CD44 V3 variants was confirmed at both the protein and mRNA levels by Western blotting and by reverse transcriptase-PCR respectively. Expression of CD44 variants was unaffected by IL-1β, IL-8, TNFα, IFNγ or IL-4 treatment, indicating either constitutive expression of these variants or involvement of other cytokines in their regulation.

CD44 and variants in melanocytic skin neoplasms.

Expression of cell surface molecules that mediate cell-matrix and cell-cell interactions largely contributes to the ability of melanoma cells to migrate and spread beyond the primary site of the tumor. CD44, the principal cell-surface receptor for hyaluronate, and its numerous splice variants have been reported to play a crucial role in invasion and the metastatic process of different human neoplasms, including primary malignant melanoma (PMM). The aim of this study was to clarify which isoforms of CD44 (standard CD44 and CD44 variants) are distributed in PMM with a vertical tumor thickness of >1.4 mm. Staining of CD44 standard (CD44s) and splice variants was further examined for diagnostic and prognostic relevance in a panel of melanocytic skin lesions. Ten cases of PMM with Breslow >1.4 mm were analysed by immunohistochemistry using monoclonal antibodies specific for CD44s and the splice variants v3, v5, v6, v7, v7-8, and v10. In addition, using anti-CD44s, v5, and v6 antibodies, 55 melanocytic lesions, including dermal nevi (n=12), Clark nevi (dysplastic nevi) (CN; n=11), melanoma in situ (Mis; n=8), PMM (n=18), and cutaneous metastasis of malignant melanoma (cMMM; n=6) were assessed. Staining intensities were scored visually and evaluated by means of a staining index. In ten cases of PMM with a Breslow index >1.4 mm positive staining was ascertained for CD44s, v5 and for v6 in three cases. No staining was found for v3, v7, v7-8, and v10. Examination of CD44s, v5, and v6 in 55 melanocytic skin lesions revealed a high index for CD44s in all specimens and a weak staining of v5 in Mis; dermal nevi and CN did not stain for v5. However, in PMM and cMMM we found v5 to be strongly positive. The isoform v6 showed a variable index only in PMM, but without connection to established prognostic criteria. We conclude that CD44s and splice variants can not be regarded as indicators for tumor progression in malignant melanomas. However, v5 may potentially serve as a diagnostic marker for melanocytic skin lesions.

Expression of CD44 and Variant Proteins in Human Colorectal Cancer and Its Relevance for Prognosis

Background: CD44 is a cell adhesion molecule often expressed in the form of various splice variants. The role of standard CD44 isoform (CD44s) and its variants in colorectal carcinogenesis is partly conflicting. Therefore, we compared the expression of CD44s (hermes-3) and its splice variants (v3 and v6) with traditional prognostic factors in 194 colorectal cancer patients treated at Kuopio University Hospital and followed up for a mean of 14 years. Methods: Formalin-fixed, paraffin-embedded tissue sections from 194 patients with colorectal carcinoma were examined immunohistochemically to detect the expression of different forms of CD44. The hypothesis that CD44s, CD44v3, and CD44v6 expression intensities and distribution in cancer cells correlated with survival was tested with the log-rank test, hazard ratios, and their confidence intervals. Results: In high-grade tumours CD44s and CD44v6 expression intensities and CD44s percentages were stronger than in low-grade tumours. CD44v6, CD44v3, and CD44s expression intensities in tumour epithelium were also stronger in Dukes C and D tumours than in A and B tumours. In the univariate survival analysis patients with strong CD44s, CD44v3, and CD44v6 expression intensities in tumour epithelium had lower cancer-related survival than the patients who had weak CD44s, CD44v3, and CD44v6 expression intensities. Recurrence-free survival was also shorter in patients with intense signals for CD44v3 and CD44v6 in tumour epithelium. In the multivariate analysis the CD44v6 expression intensity in tumour epithelium predicted independently both cancer-related and recurrence-free survival in T1-4N0-3M0 and T1-3N0M0 cases. In addition, the CD44v3 expression intensity in tumour epithelium was a significant predictor of RFS in T1-3N0M0 cases. Conclusions: These results strongly suggest that the CD44 splice variants v6 and v3 have prognostic significance in colorectal cancer.

Soluble CD44 v5 and v6 in serum of patients with breast cancer. Correlation with expression of CD44 v5 and v6 variants in primary tumors and location of distant metastasis.

Primary breast cancers were shown to overexpress CD44 v5 and v6 at the plasma membrane. However, the clinical significance of this overexpression remains unclear. Overexpression of CD44 v5 and v6 in primary breast cancers was found to correlate with metastasis and poor prognosis by some investigators, yet this correlation could not be confirmed by others using different antibodies. In this study the influence of metastatic disease, the site of metastasis, and the amount of CD44 v5 and v6 expression in the primary tumor on serum levels of the soluble forms of CD44 v5 and v6 (sCD44 v5 and v6) in breast cancer patients was investigated. Soluble CD44 v5 and v6 serum levels were measured by enzyme linked immuno sorbent assay in a group of breast cancer patients who developed metastases in various organs and in another group of patients with single organ metastasis. For control, sCD44 v5 and v6 levels were measured in breast cancer patients who remained free of metastasis and in healthy blood donors. Expression of plasma membrane bound CD44 v5 and v6 in the primary tumors of the patients with metastasis in various organs was correlated to sCD44 v5 and v6 levels in serum. Furthermore the size of sCD44 v6 was analyzed by immunoblot using a monoclonal antibody directed against CD44 v6. When metastases were detected, sCD44 v5 and v6 serum levels were increased as compared to levels measured one month after tumor surgery in patients free of metastases (p= 0.0025 and p=0.0004). Six of 19 and 6 of 20 patients had sCD44 v5 and v6 serum levels above a cut-off level of 85 and 275 ng/mL, respectively. In these cases expression of CD44 v5 and v6 in the primary cancers was also elevated. Low sCD44 v5 and v6 serum levels were associated with weak expression of CD44 v5 and v6 in the respective primary cancers. As shown by statistical analysis of sCD44 v5 and v6 levels in 57 patients with single organ metastases, elevated sCD44 v6 levels but not sCD44 v5 levels were associated with metastases in liver or bone (p=0.0025). Immunoblot analysis of soluble CD44 proteins in serum revealed two CD44 v6 specific signals of approximately 120 and 170 kDa. Increased sCD44 v5 and v6 serum levels in patients with breast cancer were influenced by the amount of CD44 v5 and v6 expression in the primary tumor by the site of metastasis. Elevated sCD44 v6 serum levels were preferentially found in patients with metastases in liver or bone.

CD44 variant isoforms are essential for the function of epidermal Langerhans cells and dendritic cells.

Upon antigen encounter epidermal Langerhans cells (LC) and dendritic cells (DC) emigrate from peripheral organs and invade lymph nodes through the afferent lymphatic vessels and then assemble in the paracortical T cell zone and present antigen to T lymphocytes. Part of this process is mimicked by metastasizing tumor cells. Since splice variants of CD44 promote metastasis to lymph nodes we explored the expression of CD44 proteins on migrating LC and DC. We show that following antigen contact, LC and DC upregulate pan CD44 epitopes and epitopes encoded by variant exons v4, v5, v6 and v9. Antibodies against CD44 epitopes arrest LC in the epidermis, prevent the binding of activated LC and DC to the T cell zones of lymph nodes, and severely inhibit their capacity to induce a delayed type hypersensitivity reaction to a skin hapten in vivo. Our results demonstrate that CD44 splice variant expression is obligatory for the migration and function of LC and DC.

Expression and prognostic value of the CD44 splicing variants v5 and v6 in gastric cancer.

In the present study, the expression and prognostic role of the CD44 splicing variants v5 and v6 were immunohistochemically investigated in 418 curatively resected gastric carcinomas. CD44v5 was expressed in 65.3 per cent (n = 273) and CD44v6 in 77.0 per cent (n = 322) of the tumours. Whereas the expression of CD44v5 was correlated with advanced pT categories, with lymph node involvement, and with the presence of blood and lymphatic vessel invasion, such a correlation could not be found for the variant v6. As shown by univariate analysis, patients with CD44v5-positive tumours had a significantly shorter overall survival than patients with CD44v5-negative tumours (P = 0.049). In contrast, expression of CD44v6 had no impact on prognosis (P = 0.574). In a multivariate analysis including the prognostic parameters pT category and pN category, as well as blood and lymphatic vessel invasion, the prognostic impact of CD44v5 expression could not, however, be maintained. Although in the present study the expression of CD44v5 was correlated with a more aggressive tumour type, these data suggest that neither CD44v5 nor CD44v6 can predict survival in patients with gastric cancer, nor is their expression a suitable tool for identifying subgroups of patients who may be at higher risk.

CD44 variant expression in cutaneous T-cell lymphoma.

Expression of the lymphocyte homing receptor CD44 and its splice variants have been linked to tumour dissemination and poor prognosis in non-Hodgkin's lymphoma. Specifically, the in vitro expression of variant exon V6 confers metastatic potential in rat pancreatic carcinoma cell lines. In this study, we investigated the expression of CD44 splice variants in cutaneous T-cell lymphomas, including patients with mycosis fungoides (MF), Sezary syndrome (SS), large-cell anaplastic lymphoma (LCAL) and HTLV1-associated cutaneous lymphoma. In addition, 4 involved lymph nodes from 2 patients with MF and 1 patient with SS were examined. Inflammatory dermatoses, lichen planus and psoriasis, and normal skin were also studied. Immunohistochemistry was performed using a panel of monoclonal antibodies, including those with specificity for CD44H (standard isoform) and variant exons V3, V6 and V8-9. Normal epidermal keratinocytes were consistently CD44H and CD44 V3, V6 and V8-9 positive. In all the different clinicopathological subtypes and stages of cutaneous T-cell lymphomas, including involved lymph nodes, tumour cells consistently expressed CD44H, but were CD44 V3 and V6 negative. CD44 V8-9 was expressed on a majority of tumour cells in 2/5 LCAL and on occasional tumour cells in 2/5 LCAL. Occasional V8-9 positive tumour cells were also identified in 6/13 MF, 1/4 SS and 3/4 HTLV1. In 2/3 lymph node samples from 2 patients with tumour-stage MF, CD44 V8-9 expression was found on a small percentage of atypical mononuclear cells. Scattered V8-9 positive dermal mononuclear cells were present in sections of lichen planus and psoriasis. We have found no evidence to suggest that the metastasis-associated CD44 variant exon (V6) is expressed in cutaneous T-cell lymphoma, or that CD44H expression is associated with an adverse prognostic group. It is not clear whether the strong expression of CD44 V8-9 in 2 patients with CD30 positive LCAL reflects activation status or metastatic potential.

Expression of CD44 variant exons by normal oral epithelia

Altered expression of CD44H an CD44 splice variants is associated with metastasis in several malignancies but its analysis requires knowledge of the normal pattern of expression which is tissue specific. There are considerable regional variations in epithelial differentiation within the mouth, including differences in cell surface glycoproteins. The aim of this study was to determine whether there are regional variations in the expression of CD44 variants in oral epithelia. Frozen and formalin-fixed paraffin-embedded sections of lip vermilion border, buccal mucosa, dorsum and ventrum of tongue, floor of mouth, gingiva and hard palate were stained immunohistochemically for CD44H and the produce of variant exons v3, v4/5, v6 and v9. Paraffin sections were subjected to microwave antigen retrieval. All epithelia stained strongly for all variants in basal, suprabasal and prickle cells and cornified and surface layers and the basal surface of basal cells were negative. Patterns of staining were identical in frozen and formalin-fixed tissue. Despite the structural heterogeneity within oral epithelium, no regional variation was detected.