EMP2 is a novel therapeutic target for endometrial cancer stem cells.

Previous studies have suggested that overexpression of the oncogenic protein epithelial membrane protein-2 (EMP2) correlates with endometrial carcinoma progression and ultimately poor survival from disease. To understand the role of EMP2 in the etiology of disease, gene analysis was performed to show transcripts that are reciprocally regulated by EMP2 levels. In particular, EMP2 expression correlates with and helps regulate the expression of several cancer stem cell associated markers including aldehyde dehydrogenase 1 (ALDH1). ALDH expression significantly promotes tumor initiation and correlates with the levels of EMP2 expression in both patient samples and tumor cell lines. As therapy against cancer stem cells in endometrial cancer is lacking, the ability of anti-EMP2 IgG1 therapy to reduce primary and secondary tumor formation using xenograft HEC1A models was determined. Anti-EMP2 IgG1 reduced the expression and activity of ALDH and correspondingly reduced both primary and secondary tumor load. Our results collectively suggest that anti-EMP2 therapy may be a novel method of reducing endometrial cancer stem cells.

EMP2 regulates angiogenesis in endometrial cancer cells through induction of VEGF.

Understanding tumor-induced angiogenesis is a challenging problem with important consequences for the diagnosis and treatment of cancer. In this study, we define a novel function for epithelial membrane protein-2 (EMP2) in the control of angiogenesis. EMP2 functions as an oncogene in endometrial cancer, and its expression has been linked to decreased survival. Using endometrial cancer xenografts, modulation of EMP2 expression resulted in profound changes to the tumor microvasculature. Under hypoxic conditions, upregulation of EMP2 promoted vascular endothelial growth factors (VEGF) expression through a HIF-1α-dependent pathway and resulted in successful capillary-like tube formation. In contrast, reduction of EMP2 correlated with reduced HIF-1α and VEGF expression with the net consequence of poorly vascularized tumors in vivo. We have previously shown that targeting of EMP2 using diabodies in endometrial cancer resulted in a reduction of tumor load, and since then we have constructed a fully human EMP2 IgG1. Treatment of endometrial cancer cells with EMP2-IgG1 reduced tumor load with a significant improvement in survival. These results support the role of EMP2 in the control of the tumor microenvironment and confirm the cytotoxic effects observed by EMP2 treatment in vivo.

Can we develop biomarkers that predict response of cancer patients to immunotherapy?

PRIMARY OBJECTIVE: The primary objective is to delineate the potential utility of cancer biomarkers that correlate and predict response to immunotherapy in cancer patients who are refractory to conventional therapeutics. Unlike significant development of biomarkers that predict response to chemotherapy, very few biomarkers have been developed to predict the response to immunotherapy. MAIN OUTCOMES AND RESULTS: This article describes briefly the importance of characterizing and validating biomarkers for immunotherapy. A few examples have been provided, such as the transcription factor NF-kappaB, the transcription repressor Yin-Yang 1 (YY1), the pro-apoptotic gene product (Smac/DIABLO) and the circulating Fas and Fas ligand. These biomarkers have been determined to be of prognostic significance in different cancers. CONCLUSIONS: Immunotherapy is considered as an alternative therapy in the treatment of cancer patients who are refractory to chemotherapy/radiation/hormonal therapies. Cross-resistance to apoptosis develops between cancer cells that are resistant to conventional therapeutics and immunotherapy. Therefore, it is important to develop biomarkers that will determine patient response to immunotherapy.

Epithelial membrane protein 2, a 4-transmembrane protein that suppresses B-cell lymphoma tumorigenicity.

A murine homologue of the epithelial membrane protein 2 (EMP2) gene was identified in a search for genes associated with B-cell lymphoma tumorigenicity by using suppression subtractive hybridization. Expression of EMP2 messenger RNA in primary mouse tissues was limited to certain epithelial cell types and the peritoneal lymphoid compartment. EMP2 was expressed in the poorly tumorigenic DAC B-lymphoma cell line but was significantly down-regulated in a subline selected for in vivo tumor formation in Balb/c mice. Recombinant restoration of EMP2 expression in the subline suppressed its tumorigenicity, suggesting that loss of EMP2 was a causal factor in the malignant phenotype. Recombinant overexpression of EMP2 was studied in B lymphoma and NIH3T3 cells. EMP2 in both cell types induced cell death on serum deprivation. EMP2-induced cell death correlated with the expression level of EMP2 protein and was prevented by caspase inhibitors Z-VAD and Z-DEVD. These findings for the first time describe an apoptotic effect of a GAS3 family gene in lymphocytes. They also suggest that EMP2 may influence B-lymphoma tumorigenicity through a functional tumor suppressor phenotype. (Blood. 2001;97:3890-3895)

Mapping the B cell superantigen binding site for HIV-1 gp120 on a V(H)3 Ig.

The emerging class of B cell superantigens includes HIV-1 gp120, which binds to many members of the V(H)3 Ig gene family. The present study addresses the structural features of V(H)3 antibodies conferring gp120 binding activity using a panel of recombinant full-length and Fab Ig proteins. Binding activity was fully conferred by the Fab portion of the Ig molecule. The V(H) region was the major determinant of binding; diverse light chains were permissive for gp120 binding. A series of recombinant V(H)3-V(H)1 chimeric molecules was created to analyze the contribution of different subregions of V(H)3 to gp120 binding. Hypervariable loop 1 (H1) substitution alone caused a 10-fold reduction in binding activity. The framework subregions (FR1, FR2 and FR3) and H2 also influenced binding, since substitutions of various combinations of these subregions conferred 10- to 100-fold binding reductions. We conclude that gp120 binding occurs through a non-conventional interaction involving multiple discontinuously arrayed residues spanning the V(H), and including roles in gp120 contact and favorable conformation of the V(H).

Monocytic differentiation of HL-60 promyelocytic leukemia cells correlates with the induction of Bcl-xL.

Treatment of human promyelocytic leukemia HL-60 cells with phorbol esters ultimately induces the differentiation of these cells along the monocyte/macrophage lineage, whereas treatment with retinoic acid or DMSO induces granulocytic/neutrophillic differentiation. In this study, we demonstrate the disparate fates of HL-60 cells treated with the phorbol ester 12,13-phorbol dibutyric acid (PDBu) or DMSO. After DMSO treatment, HL-60 cells eventually died via apoptosis, whereas the viability of PDBu-treated cells was not affected during the same interval. The levels of the apoptosis effector proteins Bak and Bad were enhanced, whereas there was a slight down-regulation of the apoptosis suppressor protein Bcl-2 after treatment of the cells with PDBu and DMSO. Treatment with DMSO resulted in the elevation of the apoptosis effector Bax, whereas treatment with PDBu did not significantly alter the levels of this protein. However, treatment of HL-60 cells with PDBu induced the rapid expression of the apoptosis suppressor protein Bcl-xL, whereas the expression of this protein remained unaltered in DMSO-treated cells. The generality of this finding was confirmed by the induction of Bcl-xL in human myeloid U-937 cells, human peripheral blood monocytes exposed to phorbol ester, and mouse thioglycollate-activated and resident peritoneal macrophages. PDBu-treated HL-60 cells remained viable for 7 days and thereafter began to die via apoptosis, with a concomitant down-regulation of Bcl-xL. In conclusion, we propose that Bcl-xL expression is associated with differentiation and survival of hematopoietic cells along the monocyte/macrophage lineage.

HIV-1 gp120: a novel viral B cell superantigen.

The envelope glycoprotein of the human immunodeficiency virus (HIV-1), gp120, has recently been characterized as a novel immunoglobulin superantigen (Ig-SAg) [1,2]. Analogous to the interaction of SAgs with T cells, gp120 binds to an unusually large proportion of immunoglobulins (lgs) from HIV-uninfected individuals; most, if not all of these Igs are members of the VH3 family [3]. Functionally, gp120 preferentially stimulates VH3 B cells in vitro. This stimulation correlates with an in vivo VH3 activation during HIV infection. Curiously, this initial activation is followed by a subsequent depletion of VH3-expressing B cells as individuals progress to AIDS. In this article we will review our current understanding of the superantigenic properties of HIV gp120. Specifically we will focus on structural aspects of the binding interaction. on the ontological development of these superantigen-binding antibodies, and on potential roles that this unconventional Ig-pathogen interaction might play in the pathogenesis of HIV-induced disease.

Growth factor responses and protooncogene expression of murine mesothelial cell lines derived from asbestos-induced mesotheliomas.

Repeated intraperitoneal injections of crocidolite asbestos fibers induced diffuse malignant mesotheliomas in mice. A series of mesothelial cell lines was isolated from mice at different stages in the development of these tumors. The cell lines isolated from mice with mesotheliomas recapitulated their growth pattern in vivo and were tumorigenic when reinjected into syngeneic mice. Similar to human mesothelial cells, growth of the murine cell lines was stimulated by epidermal growth factor. Reactive mesothelial cells and mesotheliomas expressed the receptor for this growth factor. Crocidolite asbestos fibers have been reported to induce sustained expression of the c-fos and c-jun protooncogenes in rat pleural mesothelial cells in vitro (Heintz et al, Proc. Natl. Acad. Sci. USA 90: 3299-303, 1993). Human malignant mesotheliomas have been shown to express c-fos in situ (Ramael et al, Histol. Histopathol. 10: 639-643, 1995). Two of the cell lines derived from highly invasive murine mesotheliomas overexpressed c-fos and c-jun. This murine model recapitulates the histopathology, growth factor responses, and protooncogene expression of human malignant mesotheliomas.

Human immunodeficiency virus-1 (HIV-1) gp120 superantigen-binding serum antibodies. A host factor in homosexual HIV-1 transmission.

HIV-1 gp120 is an immunoglobulin superantigen which can bind to preimmune serum Ig. We hypothesize that levels of such preimmune antibodies vary in the population and might affect host resistance or susceptibility to viral transmission. This study tests two predictions: (a) levels of preimmune anti-gpl20 Igs are a polymorphic trait; and, (b) these levels are correlated with resistance or susceptibility to HIV-1 transmission. The first prediction was confirmed in a longitudinal study of a low-risk seronegative population. In this group, levels of both endogenous anti-gpl20 IgM and IgG varied widely, but were characteristic and stable for each individual. The second prediction was addressed in a study of participants of the Multicenter AIDS Cohort Study, in which men "susceptible" and "resistant" to HIV infection were identified based on numbers of sexual partners and eventual seroconversion. Specimens consisted of archival sera obtained > 2 yr before seroconversion. Men in the susceptible population (low-risk seroconverters) were distinguished by low levels of anti-gpl20 IgG. We conclude that the level of preimmune anti-gpl20 IgG is a polymorphic population trait, and low levels are a potentially specific and significant factor in homosexual transmission of HIV infection.

Mapping the Ig superantigen-binding site of HIV-1 gp120.

The envelope glycoprotein, gp120, of HIV-1 has recently been identified as a member of the new family of Ig superantigens (Ig-SAg). This classification is based on the selective binding of gp120 to an unusually high proportion of endogenous, nonimmune Ig, and the selective activation of nonimmune B cells by gp120 in vitro. Many, if not all of the nonimmune Ig that bind to gp120 are members of the VH3 Ig gene family. The aim of this study was to determine the epitope on gp120 that was responsible for its Ig-SAg binding activity. To do this, we utilized a panel of 30 peptides derived from gp160 in a competition-binding assay. For five Igs that were tested, as well as for polyclonal serum IgM, two overlapping peptides (each 20 amino acids in length) were identified that were potent inhibitors of gp120 binding. Similarly, the 10 amino acid overlap region of these two peptides had inhibitory activity. Thus, this decamer sequence represented the optimal Ig-SAg epitope or mimotope. The amino acid residue at position 1 of the decamer, and to a lesser extent at position 10, was critical for peptide binding. In addition to this decamer peptide, other peptides that shared modest sequence homology were also selectively inhibitory for specific Ig samples. These findings provide the first definition of an Ig-SAg ligand at the peptide level and will facilitate further structural and biologic characterization of this new class of pathogenic Ags.

A novel octamer regulatory element in the VH11 leader exon of B-1 cells.

B-1 cells (CD5 B cells) represent an initial fetal wave of B cell lymphopoiesis. B-1 cells have fundamental properties that are unique from conventional B cells, including a restricted Ab repertoire. We investigated the mechanism for the overrepresentation of one such Ig H chain variable-region gene, VH11, by murine B-1 cells. We postulated that a cis-regulatory element contributed to the use of VH11. We observed that the DNA encoding the leader peptide of VH11 was atypically A/T rich and thus was a candidate for nuclear protein binding. By electrophoretic mobility shift analysis, we found that the VH11 leader DNA specifically bound to three protein complexes present in the nucleus of the B-1 cell line AJ9. Of these bands, one was ubiquitous for all cells examined (lymphoid and nonlymphoid); another band was present only in B cells, and the third band was specific for B-1 cells that expressed VH11 or VH12. In addition to its binding properties, the VH11 leader sequence also displayed modest tissue-specific enhancer activity. By DNA footprint analysis, all three protein complexes were found to bind to an octamer motif embedded within the VH11 leader DNA. To identify the octamer-binding proteins, a panel of octamer-specific Abs was used. We found that the ubiquitous band was Oct-1, and the B cell-specific band was Oct-2. The B-1 cell-specific nuclear binding protein was neither Oct-1 nor Oct-2, but may be a novel POU domain protein. We hypothesize that the VH11 leader octamer site may target this gene for preferential rearrangement and/or expression and therefore would be a contributing factor in the increased use of this gene by B-1 cells.

Immunoglobulin VH3 gene products: natural ligands for HIV gp120.

Infection with human immunodeficiency virus-type 1 (HIV-1) depletes T cells expressing CD4 and B cells expressing immunoglobulin (Ig) VH3 gene products. A subpopulation of normal B cells from non-HIV-infected individuals was shown to bind to HIV gp120 by means of membrane Ig; most of these B cells expressed VH3 family Ig. Serum VH3 IgM from uninfected individuals also avidly bound gp120. Finally, gp120 selectively induced Ig secretion by VH3 B cells, indicating that the binding of gp120 functionally activated these cells. These results indicate that naturally occurring VH3 Ig is a second ligand for gp120 and a candidate superantigen for VH3 B cells.

Cytotoxicity of long and short crocidolite asbestos fibers in vitro and in vivo.

Fiber length and diameter are important factors in the pathogenicity of asbestos. We examined the relative toxicity of long and short crocidolite asbestos fibers in vitro and in vivo. Both long and short crocidolite asbestos fibers were toxic to elicited macrophages in vitro. Similar to native crocidolite asbestos, long and short fibers stimulated the release of reactive oxygen metabolites from elicited macrophages in vitro. We evaluated whether in vitro cytotoxicity was dependent on the production of reactive oxygen metabolites. In the presence of the reactive oxygen metabolite scavenging enzymes, superoxide dismutase or catalase, the toxicity of long and short crocidolite fibers to macrophages was prevented. Furthermore, macrophages were not killed when either long or short fibers were soaked in the iron chelator, deferoxamine. Native, long, and short crocidolite fibers also caused depolarization of the mitochondrial membrane potential prior to cell death. In vivo, a single i.p. injection of long crocidolite fibers stimulates an intense inflammatory reaction, release of reactive oxygen metabolites near sites of fiber deposition, and cell death. In contrast, these events were minimal after a single injection of short fibers due to the removal of fibers from the peritoneal cavity. After five daily injections of short fibers, however, fibers were present on the surface of the mesothelium and provoked an inflammatory response. Cell death was observed on the surface of the mesothelium. Reactive oxygen metabolites were also produced near accumulations of short fibers. Our results suggest that both long and short crocidolite asbestos fibers are toxic to macrophages in vitro via an oxidant and iron-dependent mechanism. In vivo, short fibers are cytotoxic when the clearance of these fibers is prevented.

Altered calcium homeostasis in irreversibly injured P388D1 macrophages.

Sequestration of calcium by mitochondria is an important mechanism to maintain normal intracellular calcium homeostasis. Anoxic or toxic damage to these organelles has been postulated to disrupt intracellular calcium compartmentalization, leading to cell death. The authors examined the potential relationship between mitochondrial dysfunction, altered calcium homeostasis, and irreversible injury in a model system of silica-induced toxicity to P388D1 cells. Exposure to toxic silica particles, but not to nontoxic latex heads, disrupted mitochondrial membrane potential, increased membrane-associated calcium, elevated free cytosolic calcium, and killed 50% to 60% of the cell population after 6 to 8 hours. To test whether disruption of the mitochondrial membrane potential was sufficient to cause irreversible injury, P388D1 cells were exposed to either the proton ionophore, carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) or to the mitochondrial inhibitor, antimycin A. Over 90% of the treated cells showed depolarization of the mitochondrial membrane as indicated by the fluorescent probe rhodamine 123. Carbonyl cyanide p-trifluoromethoxyphenylbydrazone also caused an elevation in free cytosolic calcium as monitored by fura-2. However, even after 6 hours of exposure to these proton ionophores or mitochondrial inhibitors, P388D1 cells did not show increased chlorotetracycline (CTC)-induced fluorescence or loss of viability. P388D1 cells exposed to silica have been shown previously to lose 80% of their adenosine triphosphate (ATP) content. The effect of reduced ATP levels on intracellular calcium homeostasis and viability was assessed by exposing P338D1 cells to FCCP in the presence of sodium azide and 2-deoxyglucose, which reduced ATP content by more than 90%. Under these conditions, none of the cells were killed, and only 5.5% showed increased CTC-induced fluorescence after 6 hours. These data indicate that disruption of the mitochondrial membrane potential, even in combination with reduced ATP content, is not sufficient to kill P388D1 cells.

Evaluation of the causal relationship between crocidolite asbestos-induced lipid peroxidation and toxicity to macrophages.

In vitro, crocidolite asbestos toxicity to macrophages is mediated by the production of reactive oxygen metabolites. We examined whether exposure of macrophages to crocidolite asbestos induced lipid peroxidation as measured by the thiobarbituric acid assay. When elicited mouse peritoneal macrophages were exposed to crocidolite, a dose- and time-dependent increase in lipid peroxidation breakdown products accompanied cell death. Superoxide dismutase plus catalase or deferoxamine prevented both lipid peroxidation and loss of viability caused by crocidolite. We tested whether crocidolite-induced lipid peroxidation was causally responsible for cell death. Macrophages were not killed by crocidolite when incubated with 10 mM 3-aminobenzamide. The level of thiobarbituric acid-reactive material was the same, however, for cells incubated with crocidolite in the presence or absence of 3-aminobenzamide. When macrophagaes were pretreated for 24 h with 25 microM vitamin E and then incubated with crocidolite, no thiobarbituric acid-reactive products were detected. Vitamin E, however, did not prevent crocidolite cytotoxicity. These results suggest that exposure of macrophages to crocidolite asbestos produces lipid peroxidation as measured by thiobarbituric acid-reactive products. This reaction, however, is not directly responsible for irreversible injury in this model system.

Acute injury and regeneration of the mesothelium in response to asbestos fibers.

The mesothelium is a target of the toxic and carcinogenic effects of asbestos fibers. Fibers greater than 8 mu in length and less than 0.25 mu in diameter have been found to be highly tumorigenic in rodents, while shorter asbestos fibers or spherical mineral particles have not been shown to produce mesotheliomas. For investigation of early mesothelial reactions associated with the development of mesotheliomas, C57BL/6 mice were given intraperitoneal injections of 200 micrograms of short or long crocidolite asbestos fibers, toxic silica particles, or nontoxic titanium dioxide particles. At intervals between 3 hours and 21 days after a single injection, the mesothelial surface of the diaphragm was examined by stereomicroscopy, scanning electron microscopy, and autoradiography. Within 6 hours after injection of asbestos fibers, mesothelial cells in the lacunar regions of the diaphragm retracted opening stomata 10.7 +/- 2.3 mu in diameter leading to the submesothelial lymphatic plexus. Short asbestos fibers (90.6% less than or equal to 2 mu in length), silica, or titanium dioxide particles (less than or equal to 5 mu in diameter) were cleared through these stomata without provoking an inflammatory reaction or mesothelial injury. In contrast, long asbestos fibers (60.3% greater than or equal to 2 mu in length) were trapped at the lymphatic stomata in the lacunar regions on the peritoneal surface of the diaphragm. At these sites, an intense inflammatory reaction developed with accumulation of activated macrophages and a 5.5-fold increase in albumin recovered in the peritoneal lavage fluid after 3 days. As early as 12 hours after injection of long asbestos fibers, the adjacent mesothelial cells were unable to exclude trypan blue and lost their surface microvilli, developed blebs, and detached. Recovery of lactate dehydrogenase activity in the peritoneal lavage fluid was increased 5.8-fold after 3 days and returned to normal levels after 14 days. Regenerating mesothelial cells appeared at the periphery of asbestos fiber clusters 3 days after injection. Maximal incorporation of 3H-thymidine by mesothelial cells occurred after 7 days, followed by partial restoration of the mesothelial lining after 14-21 days. As late as 6 months after a single injection of crocidolite asbestos fibers, clusters of fibers remained in the lacunar regions, partially covered by mesothelium but surrounded by macrophages and regenerating mesothelial cells. The anatomic distribution and size of lymphatic stomata on the peritoneal surface of the diaphragm account for the selective accumulation of long asbestos fibers in these regions.(ABSTRACT TRUNCATED AT 400 WORDS)

Role of reactive oxygen metabolites in crocidolite asbestos toxicity to mouse macrophages.

Crocidolite asbestos is toxic to macrophages in vitro. We hypothesize that this toxicity is mediated by the generation of reactive oxygen metabolites. Elicited mouse peritoneal macrophages were found to release reactive oxygen metabolites upon incubation with crocidolite asbestos in vitro. Crocidolite toxicity to both primary cultures of mouse peritoneal macrophages and P388D1 cells, a mouse macrophage-like cell line, could be prevented by a hypoxic environment or by addition of the reactive oxygen metabolite scavengers, superoxide dismutase and catalase. In addition, if crocidolite fibers were presoaked with the iron chelator deferoxamine, no macrophage death occurred. In an attempt to mimic crocidolite-induced cytotoxicity, P388D1 cells or primary elicited macrophages were exposed to the nontoxic mineral particle titanium dioxide in the presence and absence of ferric chloride. Titanium dioxide was only lethal when ferric chloride was added. This toxicity was prevented by superoxide dismutase, catalase, or deferoxamine. These results suggest that crocidolite-induced injury to macrophages depends on the formation of reactive oxygen metabolites. Iron present in crocidolite fibers may catalyze the production of hydroxyl radical from superoxide anion and hydrogen peroxide generated during phagocytosis. These highly reactive hydroxyl radicals are postulated to mediate lethal cell injury.

Reference intervals for amylase isoenzymes in serum and plasma of infants and children.

Measurement of pancreatic (P) and salivary-like (S) amylase isoenzyme activity in serum of adults is useful as an indirect indicator of pancreatic and salivary gland exocrine dysfunction. To extend the use of this assay to the pediatric population, we measured amylase isoenzymes in 546 serum and plasma samples and defined normal reference intervals for the P and S isoenzymes as a function of age in newborns, infants, and children. The mean activity of P isoenzyme in newborns is 3% of that of adults, begins to increase at seven to eight months, and reaches adult values by five years. The mean activity of S isoenzyme in serum is 32% of the adult mean at birth, begins to increase by three to four months, and reaches adult values by 19 months; children five to 12 years old have slightly higher values than adults. These changes with age underscore the importance of the use of age-matched reference intervals when serum amylase isoenzyme activities are measured as diagnostic indicators.