Immunohistochemical localization of CD44 variants 5 and 6 in human gastric mucosa and gastric cancer

The aim of this study was to compare the distribution of CD44 variants 5 and 6 in normal mucosa and gastric cancer and to determine their relationship with histoclinical Parameters of the disease. Experimental material included 112 paraffin blocks of various human gastric carcinomas. Both variants of CD44 were detected immunohistochemically with primary antibodies deriving from Bender Med. Systems. In normal gastric mucosa positive reactions with both antibodies were observed in surface epithelium, parietal cells, myocytes and vascular endothelia. They were also found in intestinal enterocytes, esophageal epithelium and myoepithelial. Additional reactivity with antibodies against variant 6 was observed in Paneth cells. In gastric cancer, variants 5 and 6 were demonstrated in 91% and 64% of cases respectively, without significant correlation with the tumor type. The occurrence of variant 6 correlated positively with tumor size (p = 0.081) and negatively with histological grading (p = 0.093). The relationship with metastases was insignificant.

Differential CD44 expression patterns in primary brain tumours and brain metastases

Splicing variants of CD44 (CD44v) are increasingly recognised as metastasis-promoting factors in rodent and some human cancers. However, the frequency for CD44v expression in human cancers and their metastases and the status of CD44v expression in low or non-metastatic tumours is still uncertain. To address this issue, we investigated CD44 expression patterns in brain metastases (BMTs) spread from more than ten organs and five types of primary brain tumours (PBTs) by Northern blot, reverse transcription-polymerase chain reaction (RT-PCR) and immunocytochemical analysis. The results demonstrated that all of the 56 PBTs examined express standard form of CD44 (CD44s) but none of them express CD44v. In contrast, 22 of 26 BMTs studied were found with CD44v expression. Our data thus present direct evidence of a general distribution of CD44 in BMTs but suggest that such expression is an extremely rare event in PBTs. Therefore, the presence or absence of CD44v expression may be related to high or low metastatic potential of human malignancies.

[The role of CD44 adhesion molecules]

CD44 is a transmembrane glycoprotein involved in the interaction between cells and extracellular matrix. CD44 is ubiquitously expressed on cells, and has been thought to be a cell adhesion molecule with proposed functions in extracellular matrix binding, cell migration and lymphocyte homing. In 1991, Stamenkovic et al. showed the existence of two forms of CD44: a hematopoietic (standard) and an epithelial form which was highly expressed by carcinomas. The alternative splicing of 10 variant exons encoding the membrane proximal portion of the external domain of CD44, in particular, results in many variant isoforms. These may play a critical role in malignant behavior and in determining organ specificity in metastasis.

Biotinylated hyaluronan as a probe for detection of binding proteins in cells and tissues

A convenient and reliable method for preparing and using biotinylated hyaluronan for detection of hyaluronan-binding proteins is described. The biotinylated hyaluronan can be used to detect binding proteins in transblots after electrophoresis or as a histological probe for localization of binding proteins in cultured cells and tissue sections.

[Expression of CD44 variant form in human renal cell carcinoma]

To examine whether renal cell carcinoma display altered CD44 expression we performed reverse transcriptase-polymerase chain reaction (RT-PCR) analysis of CD44 in the tissues resected from 19 patients with renal cell carcinoma and 6 renal cancer cell lines. To detect the CD44 variants, we utilized the RT-PCR Southern blot method reported by Matsumura et al. In 12 of 17 (70.6%) cases, about a 700 base pairs band was emphasized in cancerous tissues compared with normal kidney. Moreover, we found that this isoform is the CD44 variant sharing only exon v10. Examination by Northern blot analysis has revealed that all tumors express a higher level of CD44 variants sharing exon v10. Our findings suggest that this variant form plays some roles in renal cell carcinoma.

[Expression of CD44 in endometrial cancer and metastasis]

We have examined expression of CD44 in 17 normal endometria and 58 endometrial carcinomas by reverse transcription and polymerase chain reaction (RT-PCR) and by immunohistochemistry using antibodies to CD44. CD44 variants were expressed in 13 out of 17 (76.5%) normal endometria, whereas 11 out of 58 (19.0%) endometrial carcinomas. Furthermore, lymph-vascular space involvement of cancer cells was observed to be statistically significant in the CD44 negative group as opposed to the positive group. Immunohistochemical examination revealed that intense staining in normal endometria was seen on the gland cells at the basement membrane side. CD44 may have a role in the function of normal endometrium such as implantation, and CD44 might be related with metastasis of endometrial cancer cells through lymph-vascular space.

Up-regulation of CD44 expression by tumor necrosis factor-alpha is neutralized by interleukin-10 in Langerhans cells

CD44 is a principal cell-surface receptor for hyaluronate and is found on a wide variety of cells. CD44 plays an important role in lymphocyte homing, lymphohemopoiesis, and T-cell activation as well as in cell motility and migration. CD44 is expressed on the cell surface of epidermal Langerhans cells (LC), and is one of the candidates for molecules that are involved in the migratory capability of LC, but little is known about its regulatory properties. We examined the modulatory effects of tumor necrosis factor (TNF)-alpha and interleukin (IL)-10 on the CD44 expression in LC. We found 1) that TNF-alpha significantly up-regulated the expression of CD44 in a concentration-dependent manner, 2) that IL-10 down-regulated the expression of CD44 in a concentration-dependent manner, 3) that the effect of TNF-alpha or IL-10 was readily detectable as early as 24 h after the initiation of culture, and 4) that the simultaneous addition of TNF-alpha and IL-10 mutually neutralized the effect of each other. These data suggest that in the epidermal microenvironment the expression of CD44 in LC may be reciprocally regulated by TNF-alpha and IL-10, both of which are known to be produced by surrounding keratinocytes.

Expression of CD44 on rheumatoid synovial fluid lymphocytes

OBJECTIVES

To investigate the involvement of the adhesion molecule CD44 in the homing of lymphocytes to synovial tissue, by examining the density of expression and molecular mass of CD44 on rheumatoid synovial fluid lymphocytes.

METHODS

Twenty patients with rheumatoid arthritis were studied. Peripheral blood and synovial fluid lymphocytes were isolated by Ficoll-Hypaque sedimentation. CD44 expression was analysed by two colour flow cytometry of CD3 positive T lymphocytes with calculation of mean fluorescence intensity. Expression of activation markers M21C5, M2B3, interleukin (IL)-2 receptor and transferrin receptor was quantitated. In addition, CD44 molecular mass was examined by Western blot in six patients.

RESULTS

CD44 expression was markedly increased on synovial fluid T lymphocytes of rheumatoid patients relative to peripheral blood lymphocytes from the same individuals. CD44 molecular mass on peripheral blood mononuclear cells was 88 kDa, but that on synovial fluid lymphocytes was only 83 kDa. CD44 expression correlated significantly with expression of activation markers M21C5, M2B3, and the IL-2 receptor.

CONCLUSIONS

Alterations in density of expression or of the molecular mass of CD44 could contribute to local tissue injury, either directly by facilitating adhesion, or indirectly through effects on other adhesion molecules.

Two waves of gamma delta T cells expressing different V delta genes are recruited into schistosome-induced liver granulomas

Isolated granulomas provide a unique model to study T cells in the site of inflammation. Schistosoma mansoni-infected mice develop liver granulomas in response to schistosome egg deposition and the granulomas contain gamma delta T cells that appear to be activated (Pgp-1high and L-selectin low). Analysis of kinetics and TCR gene usage of granuloma gamma delta T cells revealed a limited TCR repertoire restricted to V gamma 1.1, V delta 4, and V delta 6 genes, suggesting that the occurrence of gamma delta T cells in the granuloma is influenced by TCR V gene usage. The V delta 4+, but not the V delta 6+, gamma delta T cells expressed CD69, a marker of recent activation. To determine if there was a preferred order of accumulation of the gamma delta T cells in granulomas, s.c. sponge grafts were implanted into schistosome-infected mice, schistosome eggs were injected into the grafts, and accumulated T cells were sequentially analyzed. The earliest gamma delta T cell immigrants expressed V delta 6 and later immigrants expressed V delta 4 V genes. Additional evidence for a role of TCR specificity in the accumulation of gamma delta T cells in granulomas is their absence from schistosome granulomas in TCR transgenic mice that express only a single MHC-specific gamma delta TCR. Finally, gamma delta T cell recruitment into the granulomas did not require beta 2-microglobulin, since gamma delta T cells were present in liver granulomas of beta 2-microglobulin gene-disrupted mice. The analysis of the influx of gamma delta T cells into schistosome-induced, liver granulomas and schistosome egg-containing sponges provides a model system to investigate the role, if any, of gamma delta T cells in schistosome infections.

Expression of CD44 and its clinical implication in diffuse-type and intestinal-type gastric adenocarcinomas

Knowing the differential expression of CD44 isoforms in intestinal- and diffuse-type gastric carcinomas, we used antibodies against the standard form of CD44 (CD44s) and the domain encoded by exon v6 (CD44v6) in 103 patients with primary gastric adenocarcinomas to explore the role of CD44 isoforms in metastases of both types of gastric cancer. Carcinomas of the intestinal type were more frequently CD44s and CD44v6 positive than carcinomas of the diffuse type (p = 0.034 for CD44s and p = 0.022 for CD44v6). The reactivity to these two antibodies did not correlate with histopathological and clinical prognostic factors in intestinal-type carcinoma. In contrast, expression of CD44v6 was associated with infiltrative tumor growth (p = 0.021), depth of invasion (p = 0.012), lymph node involvement (p = 0.005) and a higher incidence of distant metastasis (p = 0.069) in cancers of the diffuse type. CD44s-expressing diffuse-type tumors had a higher incidence of distant metastasis at presentation (p = 0.001), but expression of CD44s was not correlated with other clinicopathologic indices. For all cases, there was a nonsignificant association between CD44s expression and poor survival. Unexpectedly, there was also no significant difference in survival regarding expression of CD44v6 for all cases or the diffuse-type subset. This study showed the role of CD44v6 in invasion and metastases of diffuse-type gastric carcinoma and demonstrated the necessity of subclassifying tumor types when studying the clinical significance of CD44 in human cancers.

Abnormalities in intramyocardial arteries detected in cardiac transplant biopsy specimens and lack of correlation with abnormal intracoronary ultrasound or endothelial dysfunction in large epicardial coronary arteries

OBJECTIVES

We sought to determine whether abnormalities in small intramyocardial vessels could be detected on routine cardiac transplant biopsy specimens and whether these features correlate with intimal thickening by intracoronary ultrasound and endothelial dysfunction in large epicardial vessels.

BACKGROUND

Variability in clinical presentation of allograft vasculopathy suggests differential involvement of large and small vessels. Intracoronary ultrasound and endothelial function studies detect large-vessel abnormalities but may not reflect changes in small intramyocardial arteries. The latter could be detected in routine cardiac biopsy specimens by histologic and immunohistochemical studies.

METHODS

Thirty-nine cardiac transplant recipients underwent intracoronary ultrasound and acetylcholine studies 5 to 7 days after endomyocardial biopsy. Biopsy tissue was evaluated for coronary artery endothelial plumping and intimal thickening and increased immunostaining for fibronectin, tumor necrosis factor-alpha and receptor for hyaluronan-mediated motility. Large-vessel disease was assessed by calculating an average intimal index from intracoronary ultrasound of the left anterior descending coronary artery. Endothelial function was determined by quantitative coronary analysis after acetylcholine challenge.

RESULTS

Coronary arteries were found in the biopsy tissue of 30 (76%) of the 39 patients who formed the study group. Fourteen of 30 patients had abnormal histologic findings. Immunohistochemical analysis for fibronectin, possible in 20 of 30 patients, was positive in 14 (70%) of 20 and correlated with abnormal histologic findings (p = 0.01). Immunostaining was positive for tumor necrosis factor-alpha and receptor for hyaluronan-mediated motility in 12 (40%) and 13 (43%) of 30 patients, respectively. All patients had intimal thickening by intracoronary ultrasound, but intimal index did not correlate significantly with small-artery disease by histologic or immunohistochemical analysis. Large-vessel endothelial dysfunction in 13 patients (43%) did not correlate with either abnormal ultrasound findings or small-vessel disease.

CONCLUSIONS

Intramyocardial arteries are readily observed in biopsy specimens from cardiac transplant recipients and provide useful information about allograft vasculopathy. Lack of correlation between intramyocardial and epicardial vessel disease suggests discordant progression of allograft vasculopathy.

Expression of CD44 in prostate cancer cells: association with cell proliferation and invasive potential

High level expression of the cell surface adhesion molecule CD44 standard form and its splice variants, have been causally linked to tumor metastasis. In this study, we investigated the significance of CD44 expression in human prostatic carcinoma cells. Immunocytochemistry showed high level expression of CD44 in cells from a high grade prostate tumor, and two androgen-independent, invasive prostatic carcinoma lines, PC-3 and TSU-Pr1. Normal prostatic epithelial cells and LNCaP, a low metastatic, androgen sensitive cell line, expressed none to a very low level of CD44, although mRNA transcripts were detected in all cell lines. Immunoprecipitation detected two proteins of M(r) approximately 140 kDa and 210 kDa in PC-3, and predominantly the M(r) approximately 95 kDa protein in TSU Pr1, but none in LNCaP. Most importantly, a neutralizing antibody to CD44 inhibited cell proliferation and basement membrane invasive activity, suggesting a definitive role of CD44 in prostate tumor growth and metastasis.

Expression of CD44 splice variants in human cutaneous melanoma and melanoma cell lines is related to tumor progression and metastatic potential

Expression of CD44, particularly of certain splice variants, has been linked to tumor progression and metastasis formation in a number of different animal and human cancers. Because human cutaneous melanoma is among the most aggressive human cancers, we explored expression of CD44 isoforms (CD44v) in lesions of melanocytic tumor progression. In addition, by RT-PCR and FACS analysis we assessed CD44v RNA species and cell surface expression of CD44v in cultured melanocytes isolated from human foreskin and in a panel of 2 non-, 2 sporadically and 2 highly metastatic human melanoma cell lines. We observed that all melanocytic lesions examined showed strong uniform expression of standard CD44 (CD44s) epitopes. We did not detect CD44v6 expression in the melanocytic lesions. However, CD44 isoforms containing v5 or v10 were differentially expressed. V5 was expressed in 16%, 0%, 20%, 67% and 58% of common nevi, atypical nevi, early primary melanomas (< or = 1.5 mm), advanced primary melanomas (> 1.5 mm) and metastases, respectively, and hence was related to tumor progression. In contrast, CD44v10 was expressed in all common nevi, whereas part of the atypical nevi and most primary melanomas and metastases lacked v10. CD44v RNA patterns were closely similar in cultured melanocytes and all melanoma cell lines. Melanocytes expressed high levels of CD44s but no CD44v, whereas all melanoma cell lines expressed CD44v at the surface. Interestingly, expression of v5 was strongly increased in the highly metastatic cell lines. Our results suggest a role for CD44 variant domains, particularly v5 and v10, in human melanocytic tumor progression.

Prevalence and activation phenotype of Sendai virus-specific CD4+ T cells

The relative prevalence of Sendai virus-specific CD4+ T helper precursors (Thp) has been assessed by limiting dilution analysis (LDA). Within 10 days after intranasal inoculation of C57BL/6 mice, Thp prevalence increased from < 1 in 10(5) to approximately 1 in 200 in CD4+ T-cell-enriched spleen and mediastinal lymph node populations. These frequencies remained elevated relative to naive animals, with the majority of the Thp recovered at 2-3 months after infection being found in the spleen. The "memory" Thp express an "activated" L-selectin-low, CD44-high, alpha 4-integrin-high phenotype, comparable to that described previously for CD8+ cytotoxic T lymphocyte precursors (CTLp) specific for Sendai virus. Background effects for the IL-2-based Thp analysis are, however, less predictable than those found previously for the 51Cr release CTLp LDA, as the extent of non-virus-specific lymphokine production varies. However, provided the analysis is appropriately controlled, the approach does allow useful comparisons between phenotypically different subsets of antigen-specific CD4+ T cells.

A new phenotypic classification of bone marrow plasmacytosis

Here, we propose a new phenotypic classification of bone marrow plasmacytosis. By 2-color phenotypic analysis with FITC anti-CD38 and PE anti-CD19, -CD56, -VLA-5 or MPC-1 antibody, plasma cells are easily identified on the histogram, even though no more than 1% of plasma cells are found in the bone marrow. Hence, plasma cells are phenotypically classified into polyclonal (reactive) (CD19+CD56-) or monoclonal (neoplastic) plasma cells (mostly CD19-CD56+), and furthermore immature (VLA-5-MPC-1-), intermediate (VLA-5-MPC-1+) and mature plasma cells (VLA-5+MPC-1+). According to these findings, plasmacytosis in the bone marrow can be classified into polyclonal marrow plasmacytosis (POMP) and monoclonal marrow plasmacytosis (MOMP) states. The MOMP state is further subclassified into MOMP-1 and MOMP-2, MOMP-3 and MOMP-4; MOMP-1 is defined by co-existence of monoclonal plasma cells and polyclonal plasma cells, and MOMP-2 to MOMP-4 are dependent on increased proportions of VLA-5-MPC-1- immature myeloma (plasma) cells. We found that the cases of benign monoclonal gammopathy (BMG) according to the conventional classification were in the MOMP-1 state, and myelomas could be classified into the MOMP-2 to MOMP-4 state. Subclassification of the MOMP state may be useful in determining the prognosis of myelomas, where an increase in immature myeloma cells is reported to correlate well with their aggravation during the clinical courses. Therefore, this new phenotypic classification of bone marrow plasmacytosis (POMP and MOMP-1 to MOMP-4) will contribute to differential diagnosis and understanding of therapeutic responses and prognosis in myelomas.

Splice variants of CD44 in human cervical cancer stage IB to IIB

Aberrant expression of the cell adhesion molecule CD44 has been detected in human tumors and the expression of specific CD44 isoforms (splice variants) has been shown to be associated with metastasis and poor prognosis in human malignancies. We used three different variant exon sequence-specific murine monoclonal antibodies to epitopes encoded by exon v5, exon v6, or exon v7-v8 of human variant CD44 to study the expression of CD44 splice variants by immunohistochemistry in human cervical cancer. One-hundred five patients with surgically treated squamous cell carcinomas of the cervix stages IB to IIB were included in the study. CD44 splice variants CD44v5, CD44v6, and CD44v7-8 were detected in 70, 67, and 26%, respectively. Tumors expressing exon v6 had significantly more often metastasized to the pelvic nodes (58 vs 79%, P = 0.04). Expression of exon v6 was significantly correlated with a greater probability of vascular space invasion (73 vs 50%, P = 0.04) and a significantly lower rate of inflammatory stromal reaction (48 vs 78%, P = 0.004). Patients suffering from tumors expressing splice variant CD44v6 showed poorer overall survival (P = 0.03). In cases with negative pelvic lymph nodes we found a poorer prognosis when tumors expressed CD44v6 (P = 0.01) or CD44v7-8 (P = 0.02). Among the investigated CD44 splice variants expression of exon v6 is the most promising prognostic marker in surgically treated cervical cancer.

Lymph node (but not spleen) invasion by murine lymphoma is both CD44- and hyaluronate-dependent

Similar to activated T cells, LB T cell lymphoma expresses the CD44 cell surface Ag. In addition, the vast majority of LB cells also express the beta 2 (CD18) and alpha L (CD11a) chains of LFA-1 integrin. In view of the finding that anti-CD18 mAb blocked spleen, but not lymph node invasion by LB cells inoculated s.c. into BALB/c mice, we tested the ability of anti-CD44 mAb (IM 7.8.1) to block the infiltration of LB cells into the lymph nodes. We found that, as opposed to anti-CD18 mAb, anti-CD44 mAb, as well as its F(ab')2 or Fab fragment, prevented lymph node infiltration but had no effect on spleen invasion. This conclusion was based on histologic examination and [3H]thymidine incorporation into proliferating LB cells invading the lymphoid organs. Histologic analysis further demonstrated that LB cells invade the lymph node via the afferent lymphatics. The surface expression of CD44 molecules on LB cells was enhanced after PMA activation. PMA activation also enabled in vitro binding of the lymphoma to hyaluronic acid (HA), a known ligand of CD44. Because anti-CD44 mAb, its F(ab')2 or Fab fragment, and hyaluronidase blocked this binding, we also tested the ability of the enzyme to inhibit lymph node invasion by LB cells. We established through histologic examination and [3H]thymidine incorporation that hyaluronidase protected the lymph node, but not the spleen, from invasion by the lymphoma.

Increased expression of CD44v6 and CD44v3 in ulcerative colitis but not colonic Crohn's disease

Immune mechanisms, possibly involving cell-surface molecules such as CD44, have been invoked to explain the pathogenesis of inflammatory bowel disease. We used monoclonal antibodies against epitopes encoded within the variable region of CD44 to investigate CD44 isoform expression in colon, small intestine, and liver in patients with various intestinal disorders and in controls. Biopsy samples from patients with ulcerative colitis showed significantly increased epithelial expression of CD44 isoforms containing the v6 and v3 epitopes, detected with antibodies 2F10 and 3G5, respectively. CD44v6 was detected on colonic crypt epithelial cells in 23 of 25 ulcerative colitis samples compared with 3 of 18 colonic Crohn's disease samples (p = 3.0 x 10(-6); odds ratio 57.5 [95% CI 6.83-702]) and 3 of 52 controls (22 normal colon, 10 infective colitis, 2 radiation colitis, and 18 colonic Crohn's disease; p < 1 x 10(-8); odds ratio 199 [25.5-2294]). No significant expression of CD44v6, CD44v3, or CD44v8/9 was found in samples of normal proximal colon from 4 patients with distal ulcerative colitis, whereas samples from the affected area showed staining for CD44v6 and CD44v3. No expression of CD44 variants was found in 15 samples of normal small intestine, 11 small-bowel pouchitis, 8 coeliac disease, 3 small-bowel Crohn's disease, 6 normal liver, 6 primary biliary cirrhosis, or 9 primary sclerosing cholangitis. The high intensity of CD44v6 and v3 epitope expression on crypt epithelial cells in ulcerative colitis suggests that CD44 isoforms may have an important role in ulcerative colitis. Their detection could have diagnostic potential in differentiating ulcerative colitis from other forms of colonic inflammation including Crohn's disease.

Induction of CD44 expression by the Epstein-Barr virus latent membrane protein LMP1 is associated with lymphoma dissemination

Epstein-Barr Virus (EBV) is implicated in the pathogenesis of endemic Burkitt's lymphoma (BL), B-cell lymphomas occurring under immunosuppression, nasopharyngeal carcinoma and Hodgkin's disease. Two distinct patterns of latent EBV gene expression occur in EBV-associated lymphomas. BLs typically display expression of the nuclear antigen EBNAI only, whereas EBV-associated, non-Burkitt B-cell lymphomas express at least 9 latent viral genes (6 EBNAs and 3 latent membrane proteins), reminiscent of in vitro EBV-immortalized lymphoblastoid cell lines (LCL). BLs are characterized by local, extra-nodal growth, whereas EBV-associated B-cell lymphomas often disseminate to peripheral lymphoid tissue. We show here that BL cells forming local tumors after xenotransplantation into SCID mice disseminate to lymphoid tissue following introduction of the latent membrane protein I (LMP 1) gene. Introduction of LMP 1 into BL cells induced expression of CD44 on the cell surface, a molecule implicated in enhanced lymphoid tumor growth and dissemination. Introduction of CD44 into LMP 1-/CD44-BL cells was observed to confer the disseminated tumor growth pattern associated with LMP 1 expression. Taken together our results show that expression of LMP 1 may regulate expression of CD44 and play an important role in the behavior of EBV-based lymphomas.

CD44 isoform expression in primary and metastatic pancreatic adenocarcinoma

CD44 is the transmembrane adhesion molecule which binds hyaluronate. The gene encoding CD44 is found on chromosome 11p and comprises 20 exons. Differential splicing of the 10 extracellular juxtamembranous exons (v1-10) generates the major isoforms of CD44. The major CD44 isoform found on hematopoetic cells (CD44s) contains none of the variably expressed exons, while the major isoform expressed on epithelial cells [CD44(v8-10)] contains exons v8-10. Metastasis-specific isoforms of CD44 were first documented in a model of rat pancreatic adenocarcinoma [CD44(v4-7), CD44(v6-7)] and subsequently in other cancers. This study is the first characterization of CD44 isoforms in primary and metastatic human pancreatic adenocarcinomas. CD44 isoforms were analyzed in specimens of 15 primary and 6 metastatic pancreatic adenocarcinomas as well as in 6 specimens of control pancreata by two different methods. Radiolabeled reverse transcriptase-PCR coupled with 8% PAGE allowed analysis of the major isoforms of CD44, while Southern blot hybridization with [alpha-32P]dCTP-labeled probes permitted analysis for metastasis-specific CD44 isoforms containing CD44(v6) or CD44(v8-10). No differences in the expression of CD44(v8-10) and CD44s were found among the primary and metastatic pancreatic adenocarcinomas, and control specimens of pancreata. However, a novel CD44(v6) isoform was found in metastatic lesions and may represent the human homologue of the rat pancreatic adenocarcinoma metastasis-associated CD44 isoform.

Circulating allergen-reactive T cells from patients with atopic dermatitis and allergic contact dermatitis express the skin-selective homing receptor, the cutaneous lymphocyte-associated antigen

The cutaneous lymphocyte-associated antigen (CLA) is the major T cell ligand for the vascular adhesion molecule E-selectin, and it has been proposed to be involved in the selective targeting of memory T cells reactive with skin-associated Ag to cutaneous inflammatory sites. To further investigate the relation of CLA and cutaneous T cell responses, we analyzed the CLA phenotype of circulating memory T cells in patients with allergic contact dermatitis and atopic dermatitis (AD) alone vs in patients manifesting bronchopulmonary atopy (asthma with or without AD) and nonallergic individuals. Significant T cell proliferative responses to Ni, a contact allergen, and to the house dust mite (HDM), an allergen to which sensitization is often observed in AD and/or asthma, was noted only in allergic and atopic individuals, respectively. When the minor circulating CLA+CD3+CD45RO+ subset was separated from the major CLA-CD3+CD45RO+ subpopulation in Ni-sensitive subjects, the Ni-dependent memory T cell response was largely confined to the CLA+ subset. A similar restriction of the T cell proliferative response to the CLA+ memory subset was observed for HDM in patients with AD alone. In HDM-sensitive patients with asthma with or without AD, however, the CLA- subset exhibited a strong antigen-dependent proliferation, in contrast to patients with AD alone, whose CLA- subset proliferated very weakly to HDM. In asthma with or without AD, the HDM-dependent proliferation slightly predominated in the CLA- when compared to the CLA+ subset. The functional linkage between CLA expression and disease-associated T cell effector function in AD was also demonstrated by the finding that the circulating CLA+ T cell subset in AD patients, but not nonatopic controls, selectively showed both evidence of prior activation (human histocompatibility antigen-DR expression) and spontaneous production of interleukin 4 but not interferon-gamma. Taken together, these observations demonstrate the correlation of CLA expression on circulating memory T cells and disease-associated memory T cell responses in cutaneous hypersensitivity, and they suggest the existence of mechanisms capable of sorting particular T cell Ag specificities and lymphokine patterns into homing receptor-defined memory subsets.

Alternative RNA splicing of the hyaluronic acid receptor CD44 in the normal human brain and in brain tumors

The cell-surface receptor for hyaluronic acid, CD44, is expressed by both normal and malignant cells. Numerous CD44 isoforms have recently been identified that are derived by alternative ribonucleic acid splicing. The expression of some CD44 isoforms has been shown to be involved in tumor progression and metastatic spread in a rat carcinoma model and in human carcinomas. In the present study, CD44 isoform expression was evaluated by reverse transcriptase-polymerase chain reaction (PCR) analysis in frozen sections derived from three samples of normal brain tissue and from 40 brain tumors, including samples of glioblastoma multiforme, anaplastic astrocytoma, low-grade astrocytoma, cerebral primitive neuroectodermal tumor, medulloblastoma, metastatic colon carcinoma, and metastatic melanoma. Normal brain tissue adjacent to the tumors was also examined in 14 of 18 glioblastomas. In all normal brain and tumor samples, with the exception of metastases from colon carcinoma, PCR analysis demonstrated one prominent product that corresponded to the CD44H hematopoietic form of CD44. Metastases from colon carcinoma demonstrated two prominent PCR amplification products corresponding to CD44H and CD44R1. These results suggest that CD44H is the predominant isoform of this protein in normal human brain tissue and in human neuroectodermal tumors of varying degrees of malignancy. The ability of CD44H to mediate tumor cell motility and invasiveness (in contrast to CD44R1) suggests that the CD44 alternative splicing pattern of neuroectoderm-derived tumors may enhance their local biological aggressiveness and intracerebral spread. The lack of expression of larger molecular weight CD44 variants by primary brain tumors may also partially explain why these tumors rarely metastasize to distant sites.

Differential expression of adhesion molecules in acute leukemia

Interactions between hematopoietic cells and the stromal microenvironment are mediated by membrane-bound adhesion molecules. As the expression patterns of these molecules may alter the adhesive qualities of leukemic blasts, leukemic samples were investigated for the expression of beta 1-, beta 2-, beta 3-integrins, CD44, the three selectins and several members of the immunoglobulin family. CD44 (167/169), LFA-3 (158/169), the beta 1-integrins VLA-4 (120/123) and VLA-5 (45/51) and the beta 2-integrin LFA-1 (149/157) were found on > 70% of blasts in most cases of leukemias. Other molecules were restricted to specific differentiation stages and lineage. The beta 2-integrins Mac-1 (CD11b/CD18) and gp 150,95 (CD11c/CD18) were preferentially expressed on M4 and M5 subtypes, and NCAM (CD56) was only found on a subset of acute myeloid leukemias (17/113). Unexpectedly, the beta 1-integrins VLA-1 (1/51), VLA-2 (18/123), VLA-3 (5/43), VLA-6 (15/29) and the E-selectin (2/47) were expressed on > 70% blasts on a subset of leukemias of varied phenotype. These molecules were absent on normal CD34+ bone marrow precursors. The simultaneous analysis generally revealed a higher percentage of positive blasts in the blood than in bone marrow. Our observations therefore suggest that in leukemia these antigens are displayed on a non-adherent population that is defective and is unable to convert to an adherent, functionally active conformational state.