Activated caspase 3 and cleaved poly(ADP-ribose)polymerase in salivary epithelium suggest a pathogenetic mechanism for Sjögren's syndrome.

OBJECTIVE: Apoptosis is an organized energy-dependent process of cellular self-destruction carried out by proteolytic enzymes such as the caspases. These enzymes may play a role in epithelial cell apoptosis in Sjögren's syndrome (SS). A classical caspase substrate is poly(ADP-ribose)polymerase (PARP), a DNA repair enzyme. To elucidate the molecular mechanisms responsible for salivary gland dysfunction in SS, we studied the expression of caspase and PARP in SS salivary gland biopsies. METHODS: The presence of activated caspases (caspases 3 and 9) and cleaved PARP (85 kDa) in SS biopsies was demonstrated by immunohistochemistry using specific polyclonal antibodies. RESULTS: Initial studies performed with an antibody reagent that recognizes both active and inactive forms of caspase 3 identified this enzyme in SS salivary ductal and acinar cells. Activated caspase 3 and cleaved PARP were strongly expressed in ductal and acinar cells in SS salivary glands (13/15). Ductal and acinar cells from normal salivary glands (n=5) stained with less intensity compared with SS tissue. Staining for activated caspase 9 was negative in all samples. Likewise, infiltrating lymphocytes were negative for caspase 3, caspase 9 and cleaved PARP. CONCLUSION: This study shows that caspase 3 is important in the salivary dysfunction of SS, while caspase 9 appears not to be involved.

Fas-mediated apoptosis in a rat acinar cell line is dependent on caspase-1 activity.

Apoptosis plays an important role in the dysfunction of exocrine glands. Fas is a death-inducing receptor found on many types of cells including epithelial acinar cells. To elucidate the intracellular mechanism of Fas-mediated cell death in exocrine glands, an epithelial acinar cell line, SMG-C6, was studied. Caspase-1, -3, -8, and -9 activities were elevated in SMG-C6 cells after the induction of apoptosis by soluble Fas ligand (FasL). The activation of caspase-1 and -8 occurred prior to caspase-3 and -9 activation. The caspase-1 inhibitor, zYVAD-fmk, was effective in preventing cell death, whereas the caspase-3 and -8 inhibitors (ac-DEVD-CHO and ac-IETD-CHO, respectively) were not. zYVAD-fmk was able to inhibit caspase-3 activation indicating that caspase-1 is upstream to caspase-3. Furthermore, kinetic studies show that caspase-1 is an early event in the Fas apoptotic pathway. This study shows that caspase-1 participates in Fas-mediated apoptosis of epithelial cells by initiating the caspase cascade.

Elevated proapoptotic Bax and caspase 3 activation in the NOD.scid model of Sjögren's syndrome.

OBJECTIVE: Salivary gland epithelial cells in patients with Sjögren's syndrome (SS) and in NOD and NODscid mice express Fas and Fas ligand, and these cells die from apoptosis. To elucidate the intracellular molecular mechanisms responsible for this salivary gland epithelial cell apoptosis, expression of the Bcl-2 family of proteins (Bcl-2, Bcl-xL, Bax) and caspase (caspases 3 and 8) was studied in young (ages 8-10 weeks) and old (ages 17-28 weeks) NOD and NOD.scid mice. METHODS: Sections of frozen salivary gland tissue were obtained from NOD and NOD.scid mice and from the lip biopsy material of SS patients. Immunohistochemistry or Western blot analysis was performed to assess the apoptotic-associated proteins. RESULTS: Levels of Bax and caspase 3 were elevated in the epithelial cells of glands from old NOD mice, but not in those from young NOD mice. In contrast, epithelial cells from both young and old NOD.scid mice exhibited strong expression of Bax and caspase 3. Western blot analysis showed that the activated form of caspase 3 was increased 2-5-fold in the glands from old NOD, old NOD.scid, and young NOD.scid mice compared with those from young NOD mice. Caspase 3 was also significantly elevated (P < 0.01) in SS patients whose focus scores were grade 3 or 4. In the SS patients' biopsy tissue and in the mouse glands, cells with fragmented DNA were positive for caspase 3. CONCLUSION: These results demonstrate that salivary gland epithelial cells in NOD and NOD.scid mice overexpress the proapoptotic molecules Bax and caspase 3. Bax could be the gene responsible for initiation of caspase activation, epithelial cell destruction, and lymphocyte glandular localization in SS.

G-protein signaling abnormalities mediated by CD95 in salivary epithelial cells.

Salivary epithelial cells from patients with primary Sjögren's syndrome (SS) undergo Fas-mediated apoptosis. Bcl-2 and Bcl-xL are apoptosis suppressing oncogenes. Very little is known about the role of these oncogene molecules in salivary epithelial cells. To investigate the possible prevention of salivary glandular destruction in SS by Bcl-2 and Bcl-xL, stable transfectants expressing these molecules were made from HSY cells, a human salivary epithelial cell line. HSY cells were transfected with an expression vector for human Bcl-2 or Bcl-xL. Stable transfectants were selected and apoptosis was induced by anti-Fas antibody. Apoptosis was quantified by propidium iodide staining followed by flow cytometry. Caspase activity was detected by immunohistochemical analysis and enzyme cleavage of DEVD-AMC, a fluorescent substrate. Response to carbachol, a muscarinic receptor agonist, and EGF was measured by Ca2+ mobilization and influx. Fas-mediated apoptosis was significantly inhibited in Bcl-2 and Bcl-xL transfectants compared to wild-type and control transfectants (empty vector). Surprisingly, caspase activity was not inhibited in Bcl-2 and Bcl-xL transfectants. Activation of the Fas pathway in the Bcl-2 and Bcl-xL transfectants by antibody also inhibited carbachol and EGF responsiveness (i.e., Ca2+ mobilization and/or influx) by 50-60%. This Fas-mediated inhibition of cell activation was partially or completely restored by specific peptide interference of caspase enzyme activity. The prevention of Fas-mediated apoptosis by the overexpression of Bcl-2 and Bcl-xL in salivary gland epithelial cells results in injured cells expressing caspase activity and unable to respond normally to receptor agonists. Such damaged cells may exist in SS patients and could explain the severe dryness out of proportion to the actual number of apoptotic cells seen on salivary gland biopsy.

Murine cytomegalovirus immediate-early promoter directs astrocyte-specific expression in transgenic mice.

Murine cytomegalovirus (MCMV), which causes acute, latent, and persistent infection of the natural host, is used as an animal model of human cytomegalovirus (HCMV) infection. Transcription of MCMV immediate-early (IE) genes is required for expression of the early and late genes and is dependent on host cell transcription factors. Cell-type-specific expression activity of the MCMV IE promoter was analyzed in transgenic mice generated with the major IE (MIE) enhancer/promoter involving nucleotides -1343 to -6 (1338 bp) connected to the reporter gene lacZ. Distinct expression was observed in the brain, kidneys, stomach, and skeletal muscles. Weak expression was observed in a portion of the parenchymal cells of the salivary glands and pancreas, and expression was hardly detected in the lungs, intestine, or immune and hematopoietic organs such as the thymus, spleen, lymph nodes, and bone marrow. The spectrum of organs positive for expression was narrower than that of the HCMV MIE promoter-lacZ transgenic mice reported previously and showed a greater degree of cell-type specificity. Interestingly, astrocyte-specific expression of the transgene was observed in the brain and primary glial cultures from the transgenic mice by combination of beta-galactosidase (beta-Gal) expression and immunostaining for cell markers. However, the transgene was not expressed in neurons, oligodendroglia, microglia, or endothelial cells. Furthermore, the beta-Gal expression in glial cultures was stimulated significantly by MCMV infection or by addition of calcium ionophore. These observations indicated that expression activity of the MCMV IE promoter is strictly cell-type specific, especially astrocyte-specific in the brain. This specific pattern of activity is similar to that of natural HCMV infection in humans.

Murine cytomegalovirus induces apoptosis in non-infected cells of the developing mouse brain and blocks apoptosis in primary neuronal culture.

Cytomegalovirus (CMV) is the most common cause of congenital infection, resulting in birth defects such as microcephaly. In this study, we found that apoptosis is induced in the developing mouse brain infected with murine cytomegalovirus (MCMV) in an association with neuronal cell loss. With the combination of the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) technique and immunohistochemical staining, 3.8% of the TUNEL-positive cells were double-stained with the antibody to neuron-specific enolase, while none of the TUNEL-positive cells were stained with antibodies to the immediate early and early viral antigens of MCMV. Furthermore, distribution pattern of the TUNEL-positive cells was different from that of viral DNA-positive cells detected by the in situ DNA-DNA hybridization. More than 30% of the TUNEL-positive cells were double-stained with the F4/80 antibody specific for microglia/macrophages, which were sometimes swollen, presumably the consequence of engulfment of the neuronal apoptotic cells. In the primary neuronal cultures, MCMV infection inhibited the induction of apoptosis either by serum deprivation or by glutamate treatment. It was also confirmed by the double-staining method that apoptosis was not induced in the viral-infected neuronal cultures. These results suggest that MCMV infection induces apoptosis in non-infected neuronal cells, presumably by indirect mechanisms, and that apoptotic cells are engulfed by microglia/macrophages. The induction and blocking of neuronal apoptosis by viral infection may be important for morphological and functional brain disorders in the congenital CMV infection.

Differential expression of the immediate-early and early antigens in neuronal and glial cells of developing mouse brains infected with murine cytomegalovirus.

Brain disorders induced by congenital cytomegalovirus (CMV) infection may appear at a later time after birth as a consequence of persistent infection and/or the activation of a latent infection of the neural cells. We have analyzed the infection dynamics of the neural cells in the neonatal mouse brains infected with murine CMV (MCMV) in the late stage of gestation. First we prepared a rat monoclonal antibody to the major immediate-early (IE)-89K antigen and then used the antibody for comparison of the expression of early and late viral genes in the developing mouse brains. The cells expressing the IE-89K antigen were mostly localized in the ventricular and subventricular zones and were preferentially double stained with anti-glial fibrillary acidic protein and anti-nestin antibodies. In contrast, the cells expressing the early nuclear antigen, detected by the monoclonal antibody D5, were diffusely distributed in the cortex and the hippocampus and were mostly double labeled with anti-neuron-specific enolase antibody. In neonatal mouse brains infected congenitally with recombinant MCMV, which expressed lacZ as a late gene, the number of the early nuclear antigen-positive cells was much higher than that of the beta-galactosidase-expressing cells, the number of which was almost the same as that of the IE-89K antigen-positive cells. In addition, the distribution of viral DNA-rich cells detected by DNA-DNA hybridization was similar to that of the IE-89K antigen-positive cells. These results suggest that CMV may persistently infect neuronal cells, whereas lytic infection may preferentially occur in the glial cells in the developing brain.

Disordered migration and loss of virus-infected neuronal cells in developing mouse brains infected with murine cytomegalovirus.

Microcephaly is the most prominent symptom of the developmental brain abnormalities induced by congenital cytomegalovirus (CMV) infection. To investigate the effect of CMV infection on neuronal migration in developing brains, mouse embryos on one side of uteri received, on day 15.5 of gestation (E15.5), an injection of murine CMV (MCMV) into the cerebral ventricles, and the embryos on the other side of the uteri were injected with minimum essential medium (MEM). Labeling with 5-bromo-2-deoxyuridine (BrdU) was accomplished by intraperitoneal injection of BrdU 6 h later. Disturbance of the neuronal migration and loss of neurons were observed postnatally in the brains of MCMV-infected mice, which were identified by immunohistochemical staining of viral antigen. Double staining of BrdU-labeled and viral antigen-positive cells in brains on the 7th postnatal day showed that the migration of BrdU-single-labeled cells, mainly localized in cerebral layers II-III, mostly preceded that of the viral antigen-positive cells. However, about 7.5% of the cells observed were double-labeled, especially in the layers III-IV, and a few double-stained cells were markedly disturbed in migration. In the brains of offspring labeled with BrdU 72 h after infection with MCMV on E15.5, most of the double-stained cells were seen around the ventricular and subventricular zones. These findings suggest that a disturbance of neuronal migration in addition to neuronal loss may play a crucial role in the development of microcephaly in congenital CMV infection in humans.

A novel allelic variant of serum amyloid A, SAA1 gamma: genomic evidence, evolution, frequency, and implication as a risk factor for reactive systemic AA-amyloidosis.

Reactive systemic amyloidosis, also called AA-amyloidosis is a rare fatal complication of common chronic inflammatory diseases such as rheumatoid arthritis. It has been proposed that as yet undefined factors other than persistent elevation of serum level of the precursor protein, serum amyloid A (SAA), are also important for the development of AA-amyloidosis. In this work we show genomic evidence for a novel allelic variant of human SAA, SAA1 gamma, which we have recently identified at the protein level. The SAA1 gamma [Ala52(GCC), Ala57(GCG)] differed from SAA1 alpha [Val52(GTC), Ala57(GCG)] only at one base, indicating a single point mutation. On the other hand, SAA1 beta [Ala52(GCC), Val57(GTG)] had not only one, but additional differences in a nearby intron and this portion was identical to the SAA2 gene, suggesting a crossing-over between the SAA1 and SAA2 genes. Furthermore, we report that there was a significant difference in the observed numbers of SAA1 alleles between rheumatoid arthritis patients with AA-amyloidosis and the control population (chi 2(2) = 11.59, p = 0.003) with a higher frequency of gamma-allele in the AA-amyloid group (0.70 vs. 0.37). There was also a notable difference in the distribution of SAA1 genotypes (chi 5(2) = 14.63, p = 0.012) with an increased frequency of gamma/gamma-homozygotes in the AA-amyloid group (0.60 vs. 0.18). Thus our findings indicate that this novel allelic variant may be an important risk factor for the development of AA-amyloidosis.

Prolonged infection of mouse brain neurons with murine cytomegalovirus after pre- and perinatal infection.

The susceptibility of mice at different developmental stages to a relatively low titer of cell culture-passaged murine cytomegalovirus (MCMV) infection was compared in terms of the urinary excretion of MCMV examined by plaque assay and in terms of the distribution of viral infection, determined by immunohistochemistry, using antibodies specific to the early nuclear antigen of MCMV. Viral infection on day 8.5 of gestation (E8.5) into the conceptus and intraperitoneal infection on day 15.5 of gestation (E15.5), postnatal day 2 (P2), postnatal day 11 (P11), and 30 days after birth (P30), respectively, were performed. Embryonal and perinatal mice were more susceptible to MCMV in terms of urinary excretion of the virus and the presence of viral antigen-positive cells in the brain, lungs, and kidneys. In the embryonal and perinatal infection, the viral antigen-positive cells in the neurons of the cerebral cortex and hippocampus were retained late after birth, even though the positive cells in the lungs and kidneys had disappeared. In the mice infected on E8.5, small clusters of viral antigen-positive cells were detected only in the cortex and hippocampus late after birth, without the urinary excretion of virus. These results suggest that when mice are infected with MCMV at the embryonal and perinatal stages, elimination of the infected neurons is delayed compared with that of the other cells in the lungs and kidneys. These findings provide a model for the analysis of pathogenesis of the subclinical congenital CMV infection that manifested clinically late after birth in humans as brain disorders.