Phase I and pharmacokinetic study of continuous twice weekly intravenous administration of Cilengitide (EMD 121974), a novel inhibitor of the integrins alphavbeta3 and alphavbeta5 in patients with advanced solid tumours.

A single-agent dose escalating phase I and pharmacokinetic study with Cilengitide, an inhibitor of the integrins alphavbeta3 and alphavbeta5, was performed to determine its safety and toxicity. Cilengitide was administered as a one-hour infusion twice weekly without interruption to patients with histologically- or cytologically-confirmed metastatic solid tumours. Plasma pharmacokinetics were determined at days 1 and 15. 37 patients were enrolled into the study. Dose levels studied were 30, 60, 120, 180, 240, 400, 600, 850, 1200, and 1600 mg/m(2)/infusion. There was no dose-limiting toxicity (DLT). Pharmacokinetics were dose-independent and time-invariant. Apparent terminal half-life ranged from 3 to 5 h. At 120 mg/m(2)/infusion, peak plasma concentrations were attained that optimally inhibited tumour growth in preclinical models. Cilengitide can be safely administered using a continuous twice-weekly infusion regimen. As DLT was not reached, future trials should explore Cilengitide at different doses.

Surface antigens of human mesangial cells: impact of growth surface or IL-1alpha.

The interactions of mesangial cells (MC) with their environment are important events in glomerular physiology and pathology, yet a detailed characterization of the MC-surface antigens mediating these interactions is still lacking. In this study, a comparative phenotype analysis of primary human MC in culture using 191 monoclonal antibodies directed against 108 antigens was performed by flow-cytometry. The MC were grown on three different surfaces (human matrix, fibronectin, polystyrene) and cultured in the presence or absence of IL-1alpha. Seventy-one antibodies recognizing 35 different antigens (integrins: CD29, 49b, 49c, 49e, 51, 61; immunoglobulin gene family: CD54, 58, 90, 106, 146, 147, 166; growth factor receptors: CD105, 140b; apoptosis related: CD95; hemostatis related: CD141, 142; miscellaneous: CD44, 109, 138, 151, 157, 165, and 11 nonclustered antigens) reacted with mesangial cells. CD58, 109, 146, 147, 151, 157, 165, and 166 are reported for the first time to be present on human mesangial cells. In comparison to growth on polystyrene, CD44, 54, 95, 105, 109, 140b, 146, 147, 157, 165 and 166, were up-regulated on fibronectin, and CD44, 54, 90, 95, 105, 106, 109, 138, 140b, 141, 142, 146, 147, 151, 157, 165 and 166 were up-regulated on human matrix. The stimulation by IL-1alpha up-regulated CD44, 49e, 51, 54, 61, 106 on MC on polystyrene; CD49e, 51, 61, 106, 146, 165 on MC on fibronectin, and CD49e, 51, 54 on MC grown on human matrix. This analysis of surface antigen expression provides new information to enable a better understanding of the role of mesangial cells in glomerular pathophysiology.

Association of two MTHFR polymorphisms with total homocysteine plasma levels in dialysis patients.

The effect of the combined 5,10-methylenetetrahydrofolate reductase (MTHFR) 677C-->T and 1298A-->C genotype on total homocysteine (tHcy), folate, and vitamin B(12) plasma levels was investigated in 983 subjects, including 415 hemodialysis patients, 179 peritoneal dialysis patients, and 389 healthy individuals. Mean tHcy plasma concentrations were 27.2 +/- 15.8 micromol/L in hemodialysis patients, 25.4 +/- 19.1 micromol/L in peritoneal dialysis patients, and 8.9 +/- 3.5 micromol/L in healthy individuals. Hyperhomocysteinemia (tHcy > 15 micromol/L) was detected in 81.6% of patients and 2.6% of controls. Multiple stepwise regression analysis showed that the MTHFR 677C-->T/1298A-->C genotype (CC/AA, CC/AC, CC/CC, CT/AA, CT/AC, TT/AA), vitamin use, age, folate and vitamin B(12) plasma level were significant predictors of tHcy plasma levels. Analysis of variance showed that this effect of MTHFR genotypes on tHcy level was caused by significantly greater tHcy levels in 677TT/1298AA hemodialysis and peritoneal dialysis patients versus other genotypes. Compound heterozygous controls (677CT/1298AC genotype) had significantly greater tHcy levels compared with 677CC/1298AA controls. There was no major effect of MTHFR polymorphisms on folate and vitamin B(12) plasma concentrations. This study shows that the MTHFR 677TT/1298AA genotype, but not the 677CT/1298AC genotype, is a significant predictor of tHcy plasma levels in dialysis patients.

Human melanoma therapy in the SCID mouse: in vivo targeting and reactivation of melanoma-specific cytotoxic T cells by bi-specific antibody fragments.

The adoptive transfer of tumor-specific cytotoxic T cells (CTL) offers a promising perspective in cancer immunotherapy. However, the ex vivo-generated T lymphocytes are mostly IL-2-dependent. Here we explored the possibility of circumventing the requirement for IL-2, known for severe side effects in the patient, and of simultaneously targeting the CTL towards the tumor by the use of 2 bi-specific antibody fragments. As a model system, we used SCID mice bearing an s.c.-implanted human melanoma line (BLM-gp100) and in vitro-generated CTL specific for the gp100-derived immunogenic peptide YLEPGPVTA, which were injected i.v. with delay. To maintain the cytotoxic potential of the transferred CTL, 2 bi-specific antibody (biAb) fragments were generated which bound with one arm either CD3 or CD28, a combination known to support the activation of CTL. For targeting the CTL, both biAbs contained the F(ab') part of HD-Me13, an antibody recognizing p97, a non-immunogenic melanoma-associated surface molecule. In vitro and in vivo, the addition of the 2 biAbs increased the cytotoxic potential of the gp100-specific CTL and supported their clonal expansion in the absence of IL-2. Correspondingly, significantly higher numbers of CTL were recovered from melanoma-bearing SCID-mice that received the 2 biAb than from mice treated with the CTL only. In animals treated with CTL plus both biAbs, the primary tumor did not grow, and none of the mice developed metastases. Thus, this set of bi-specific antibody fragments was proved to target effector cells in the tumor-bearing host and to efficiently support in vivo clonal expansion and cytolytic activity of in vitro-generated CTL.

Blockade of metastasis formation by CD44-receptor globulin.

CD44 standard as well as variant isoforms have been frequently reported to be involved in the process of metastasis formation. Whereas in the rat system, but also in some human tumours, the variant exon v6 is of importance in the lymphatic spread of carcinomas, in human malignant melanoma CD44s and, possibly, CD44v10 appear to facilitate local invasion and haematogenous spread. This has been tested in the B16F10 murine melanoma model by treating B16F10-bearing C57BL/6 mice either with a CD44s-/ CD44v10-specific antibody, or with receptor globulins (Rg) containing the extracellular part of CD44s or CD44v10 linked to the constant region of the immunoglobulin kappa light chain. Prior characterization of the CD44s and CD44v10 Rg had shown that both Rgs bound to components of the extracellular matrix, CD44s in particular to hyaluronic acid. Immunohistological screening of organ sections from adult C57BL/6 mice revealed additional evidence for both Rgs binding to elements of the extracellular matrix, particularly in bone marrow, intestine and lung. In the absence of any further treatment, the CD44s Rg reduced the number of lung colonies by 70%, while application of the CD44v10 Rg resulted in 60% reduction. CD44-specific antibodies were equally efficient with regard to B16F10 settlement in the lung. However, only the CD44 Rgs prevented spread and settlement of melanoma cells in distant organs. The finding confirms the involvement of both CD44s and CD44v10 in melanoma progression, and is suggestive for the use of Rgs as therapeutic reagents.

Stable expression of a retrovirally transferred adhesion molecule in a human melanoma-specific cytotoxic T lymohocyte clone.

The adoptive transfer of in vitro generated tumor-specific cytotoxic T lymphocytes (CTL) is considered a promising perspective in cancer therapy. One possible drawback lies in the inappropriate homing of in vitro cultured lymphocytes, which could be circumvented by introducing the appropriate targeting molecules. Here we describe a protocol that allows a rapid and stable transfection of cytotoxic T cell clones. As a model system we used a CTL clone specific for the melanoma-associated antigen gp100 and a cDNA encoding for murine CD14 containing the variant exen v10 which is supposed to facilitate lymphocyte homing towards the skin. CD44v10 cDNA was ligated into the retroviral vector pMV-7, which was used to transfect the ecotropic GP-E-86 and the amphotropic PA317 cells. After several cycles of transduction to increase the viral titre, supernatants of the amphotropic PA317-CD44v10 line were used for transduction of CD44v10 into a human CTL clone. After three cycles of transduction at 12-h intervals, low but stable expression of CD44v10 was observed throughout the culture period of 10 weeks. The phenotype of the transduced CTL clone was unaltered and the cytotoxic potential was only slightly reduced as compared to the parental clone. The efficiency of stable transduction within a period of 1 week makes the protocol well suited for the in vivo transfer of transduced cells and, in the special case, should guarantee appropriate homing of the transduced CTL clone.

Variant cell lines selected for alterations in the function of the hyaluronan receptor CD44 show differences in glycosylation.

CD44 is a major cell surface receptor for the extracellular matrix glycosaminoglycan hyaluronan (HA). However, the ability of CD44 to bind ligand is strictly regulated. Three activation states of CD44 have been demonstrated: (a) inactive; (b) inducible (by certain CD44-specific mAb); and (c) constitutively active. Starting with two parental cell lines expressing CD44 in the inactive state, a pre-B cell (RAW 253) and a fibroblast (L cells), we used fluorescence-activated cell sorting with fluorescein-conjugated hyaluronan in the presence of inducing mAb to derive variant cell lines with CD44 in the inducible state. Constitutively active derivatives were isolated from the inducible variants by a further round of fluorescence-activated cell sorting in the absence of inducing antibody. However, constitutively active variants could not be isolated directly from parental cells expressing CD44 in the inactive state. These results suggest that two genetic events must occur to obtain an active CD44-HA receptor from an inactive receptor. Variant and parental cell-derived CD44 molecules exhibited differences in migration on sodium dodecyl sulfate-polyacrylamide gel electrophoresis that were partly attributable to differences in N-linked glycosylation. Furthermore, culture in tunicamycin for 2-3 d converted parental and inducible cell lines into cells showing constitutive CD44-mediated HA binding. Also, removal of cell surface glycosaminoglycan chains by culture of cells in p-nitrophenyl beta-D-xylopyranoside or treatment with chondroitinase ABC resulted in conversion of cells with an inactive CD44 receptor to an inducible state. These results indicate that carbohydrate side chains of CD44 and/or other molecules on the cell surface that interact with CD44 are potentially involved in regulating the HA-binding function of CD44 on the cell surface.

Transmembrane domain of CD44 is required for its detergent insolubility in fibroblasts.

The hyaluronan receptor CD44 is an abundant glycoprotein expressed on a variety of different cell types. In fibroblasts a significant portion of receptor molecules remain in the detergent-insoluble fraction after Triton X-100 extraction. Detergent insolubility of these CD44 molecules has been interpreted to reflect their association with the cytoskeleton. In this study we examined the structural features of CD44 required for its Triton X-100 insolubility in murine fibroblasts. We expressed in L cells the wild-type hematopoietic form of CD44, a mutant CD44 lacking the cytoplasmic domain, and two mutant CD44 molecules with substituted transmembrane domains. Immunofluorescence and cell surface iodination were performed and the detergent extraction profile of the transfected CD44 molecules was determined. No difference in detergent solubility was observed between wild-type and tailless mutant-transfected molecules. However, both CD44 mutants with a heterologous transmembrane domain, derived from either the CD3 zeta chain or CD45, were completely soluble in Triton X-100. These results demonstrate that the transmembrane region but not the cytoplasmic domain of CD44 is required for the detergent-insolubility in these cells. No obvious colocalization of CD44 and actin stress fibers was observed before or after treatment with cytochalasin D, and no change in the detergent extraction profile of wild-type and mutant CD44 molecules was effected by cytochalasin D. In equilibrium density sucrose gradients the Triton-insoluble CD44 component was found in the low density fractions, indicating an association with Triton X-100-insoluble lipids.(ABSTRACT TRUNCATED AT 250 WORDS)

Role of CD44 cytoplasmic domain in hyaluronan binding.

The hyaluronan (HA) binding activity of mutant CD44 constructs expressed in AKR1 T-lymphoma cells was evaluated by flow cytometry using fluorescein-conjugated HA (Fl-HA). Previous studies showed that wild-type hematopoietic CD44 bound Fl-HA when expressed in AKR1, but that truncated "tailless" CD44, lacking all but six amino acids of the cytoplasmic domain, did not bind. Here, we show that a disulfide-bonded dimer of CD44, formed by substituting the transmembrane region of CD3 zeta chain for that of CD44, binds Fl-HA, even when the cytoplasmic domain of the CD44 dimer is absent. We conclude that dimerization of CD44 abrogates the requirement for the cytoplasmic domain, suggesting that the cytoplasmic domain of CD44 may contribute to HA binding by promoting CD44 clustering. These results suggest that changes in the distribution of CD44 on the cell surface, induced by molecular interactions either from within the cell or from outside, may regulate its role as a receptor. Further studies sought to localize the region of the CD44 cytoplasmic domain contributing to HA binding by the construction of a series of cytoplasmic domain truncation mutants and internal deletion mutants. All of the mutant CD44 molecules bound Fl-HA similarly to wild-type CD44. Thus, it was not possible to assign the function mediating HA binding to a specific region of the cytoplasmic domain, suggesting either that multiple regions of the cytoplasmic domain can promote enhancement of HA binding, or that the role of the cytoplasmic domain in mediating this function does not require a specific amino acid sequence.

Hyaluronan binding function of CD44 is transiently activated on T cells during an in vivo immune response.

Though CD44 functions as a cell surface receptor for hyaluronan (HA) in some cell lines, most normal hematopoietic cells expressing CD44 do not bind HA. Certain CD44-specific monoclonal antibodies (mAbs) can rapidly induce CD44-mediated HA binding in normal murine T cells. This observation suggests that in vivo mechanisms may exist for activating the HA receptor function of CD44 on normal T cells. Here, it is shown that up to one third of splenic T cells are capable of CD44-mediated binding of fluorescein-conjugated HA (Fl-HA) during an in vivo allogeneic response. HA binding activity peaks at 7-8 d postinjection and declines rapidly. These rapid kinetics could be the result of transient activation of CD44 function and/or differentiation or expansion of short-lived population(s) that have constitutive HA-binding function. Both CD4 and CD8 T cells are included in the HA binding population which is strongly CD44 positive. After separation of HA-binding cells from nonbinding cells by cell sorting, it is shown that almost all cytotoxic effector cells are found in the HA-binding population. However, there is no evidence that CD44-mediated HA recognition is directly involved in the killing of target cells, since cytotoxicity could not be inhibited by CD44-specific mAbs that inhibit HA binding or by soluble HA. PCR amplification of cDNA reverse transcribed from RNA of sorted HA-binding cells indicated no evidence for CD44 isoforms other than the standard (hematopoietic) form. Though CD44 expression is known to be elevated upon T cell activation, and, as shown here, HA-binding function is induced in a portion of CD44-expressing T cells including cytotoxic effector cells, the role of CD44 and HA-recognition in immune responses is not known.

Mouse cathepsin D gene: molecular organization, characterization of the promoter, and chromosomal localization.

A cloned mouse genomic DNA fragment containing the gene encoding cathepsin D (Catd) encompasses 11 kb of genomic DNA and is composed of 9 exons. Using recombinant inbred strains, we localized the Catd gene on chromosome 4, tightly linked to the loci Mtv-13, Cyp4a, Ms15-1, and Pmv-19. The exon-intron organization of the Catd gene was shown to be very similar to that of its human counterpart. Presence of a CpG island, absence of a TATA box, and initiation of transcription at more than one site indicate that the Catd gene is a "housekeeping" gene. A 1.2-kb fragment containing the 5'-flanking region of the gene displayed promoter activity in BHK-21 cells. Comparison of the nucleotide sequences of mouse and human cathepsin D promoter regions revealed conservation of three potential regulatory elements: an E box, a GC box and a potential cAMP-responsive element. In contrast to the 5' region of human cathepsin D, the murine gene contains three CCAAT boxes but lacks any of the four AP2 binding sites found in the human gene.

Further evidence that preoptic anterior hypothalamic GABAergic neurons are part of the GnRH pulse and surge generator.

The in vivo release rates of GABA from the preoptic/anterior hypothalamic area (PO/AH) of ovariectomized rats were assessed by means of a focal perfusion cannula system. Seven days after surgery all animals received a sc silastic capsule implant containing either estradiol-17 beta (E2) or corn oil, and they were perfused 3 days later. Perfusate fractions were sampled at 5-min intervals and blood was collected every 10 min over a period of 5 h. In ovariectomized animals PO/AH GABA release was pulsatile without any diurnal rhythm. Prior to frequency analysis by means of the pulsar-programme, LH and GABA values were z-transformed. Significant LH peaks of all examined ovariectomized rats were superimposed and GABA data were arranged accordingly. It became evident that LH episodes are preceded by a significant reduction of preoptic anterior hypothalamic GABA release. The secretion patterns of GABA and LH were profoundly affected by E2 replacement. During early noon when LH levels were low, constantly elevated hypothalamic GABA release rates were observed in E2-substituted rats in comparison to ovariectomized rats. GABA release rates fell significantly during the E2-induced LH surge. Our previous demonstration of the existence of a large number of estrogen-respective GABAergic neurons in the PO/AH is suggestive of these neurons changing their activity in response to estrogen treatment. We conclude that these estrogen-respective GABAergic neurons are involved in the generation of GnRH pulses as well as in the generation of the so-called positive feedback effect of E2 on LH release.(ABSTRACT TRUNCATED AT 250 WORDS)

In vivo release rates of substance P in the preoptic/anterior hypothalamic area of ovariectomized and ovariectomized estrogen-primed rats: correlation with luteinizing hormone and prolactin levels.

Push-pull cannulae were implanted into the preoptic/anterior hypothalamic area (POA) of ovariectomized (OVX) rats. After recovery animals were treated with estradiol (E2) or corn oil and they were perfused 3 days later. Substance P (SP) concentrations were measured in 15 min perfusate fractions, blood samples were taken at similar intervals. SP concentration in POA perfusates were readily measurable. Following estrogen priming SP release increased significantly each afternoon prior to the estrogen induced prolactin and luteinizing hormone (LH) surges. No such increase of SP release was observed in OVX rats with constant LH and prolactin levels throughout the day. Mean SP rates in OVX rats were significantly higher in comparison to OVX estrogen-primed rats. These results indicate that SP may be involved in the feedback mechanisms of estrogen on prolactin and LH release. Authenticity of SP in POA perfusates was made probable by high-performance liquid chromatography (HPLC) where synthetic SP eluted at the same retention time as the signal measured in push-pull perfusates.