Localisation of RA175 (Cadm1), a cell adhesion molecule of the immunoglobulin superfamily, in the mouse testis, and analysis of male infertility in the RA175-deficient mouse.

RA175, a member of the immunoglobulin superfamily, plays an important role in cell adhesion, and RA175 gene-deficient mice (RA175(-/-) ) show oligoastheno-teratozoospermia. To understand the function of RA175, location in the testis and the morphological features of its spermatogenic cells in RA175(-/-) mice were investigated. Immunohistochemical studies revealed that RA175 immunoreactivity was observed on the cell surface of the spermatogenic cells at specific stages. A strong reaction was detected from type A spermatogonia to pachytene spermatocytes at stage IV and from step 6 to step 16 spermatids during spermatogenesis. From pachytene spermatocytes at stage VI to step 4 spermatids, the reaction was not detected by the enzyme-labelled antibody method and was faintly detected by the indirect immunofluorescence method. Abnormal vacuoles in the seminiferous epithelium, showing exfoliation of germ cells, and ultrastructural abnormality of the elongate spermatids were revealed in the RA175(-/-) testes. Other members of the immunoglobulin superfamily such as basigin, nectin-2 and nectin-3, which have an important role in spermatogenesis, were immunohistochemically detected in the RA175(-/-) testis. These observations indicate a unique expression pattern of RA175 in the testis and provide clues regarding the mechanism of male infertility in the testis.

Autism spectrum disorder is related to endoplasmic reticulum stress induced by mutations in the synaptic cell adhesion molecule, CADM1.

Autism spectrum disorder (ASD) is a neurodevelopmental disorder with an unknown molecular pathogenesis. A recent molecular focus has been the mutated neuroligin 3, neuroligin 3(R451C), in gain-of-function studies and for its role in induced impairment of synaptic function, but endoplasmic reticulum (ER) stress induced by mutated molecules also deserves investigation. We previously found two missense mutations, H246N and Y251S, in the gene-encoding synaptic cell adhesion molecule-1 (CADM1) in ASD patients, including cleavage of the mutated CADM1 and its intracellular accumulation. In this study, we found that the mutated CADM1 showed slightly reduced homophilic interactions in vitro but that most of its interactions persist. The mutated CADM1 also showed morphological abnormalities, including shorter dendrites, and impaired synaptogenesis in neurons. Wild-type CADM1 was partly localized to the ER of C2C5 cells, whereas mutated CADM1 mainly accumulated in the ER despite different sensitivities toward 4-phenyl butyric acid with chemical chaperone activity and rapamycin with promotion activity for degradation of the aggregated protein. Modeling analysis suggested a direct relationship between the mutations and the conformation alteration. Both mutated CADM1 and neuroligin 3(R451C) induced upregulation of C/EBP-homologous protein (CHOP), an ER stress marker, suggesting that in addition to the trafficking impairment, this CHOP upregulation may also be involved in ASD pathogenesis.

Caspase-12 compensates for lack of caspase-2 and caspase-3 in female germ cells.

Previously, we analyzed mice lacking either caspase-2 or caspase-3 and documented a role for caspase-2 in developmental and chemotherapy-induced apoptosis of oocytes. Those data also revealed dispensability of caspase-3, although we found this caspase critical for ovarian granulosa cell death. Because of the mutual interdependence of germ cells and granulosa cells, herein we generated caspase-2 and -3 double-mutant (DKO) mice to evaluate how these two caspases functionally relate to each other in orchestrating oocyte apoptosis. No difference was observed in the rate of spontaneous oocyte apoptosis between DKO and wildtype (WT) females. In contrast, the oocytes from DKO females were more susceptible to apoptosis induced by DNA damaging agents, compared with oocytes from WT females. This increased sensitivity to death of DKO oocytes appears to be a specific response to DNA damage, and it was associated with a compensatory upregulation of caspase-12. Interestingly, DKO oocytes were more resistant to apoptosis induced by methotrexate (MTX) than WT oocytes. These results revealed that in female germ cells, insults that directly interfere with their metabolic status (e.g. MTX) require caspase-2 and caspase-3 as obligatory executioners of the ensuing cell death cascade. However, when DNA damage is involved, and in the absence of caspase-2 and -3, caspase-12 becomes upregulated and mediates apoptosis in oocytes.

ER stress (PERK/eIF2alpha phosphorylation) mediates the polyglutamine-induced LC3 conversion, an essential step for autophagy formation.

Expanded polyglutamine 72 repeat (polyQ72) aggregates induce endoplasmic reticulum (ER) stress-mediated cell death with caspase-12 activation and vesicular formation (autophagy). We examined this relationship and the molecular mechanism of autophagy formation. Rapamycin, a stimulator of autophagy, inhibited the polyQ72-induced cell death with caspase-12 activation. PolyQ72, but not polyQ11, stimulated Atg5-Atg12-Atg16 complex-dependent microtubule-associated protein 1 (MAP1) light chain 3 (LC3) conversion from LC3-I to -II, which plays a key role in autophagy. The eucaryotic translation initiation factor 2 alpha (eIF2alpha) A/A mutation, a knock-in to replace a phosphorylatable Ser51 with Ala51, and dominant-negative PERK inhibited polyQ72-induced LC3 conversion. PolyQ72 as well as ER stress stimulators upregulated Atg12 mRNA and proteins via eIF2alpha phosphorylation. Furthermore, Atg5 deficiency as well as the eIF2alpha A/A mutation increased the number of cells showing polyQ72 aggregates and polyQ72-induced caspase-12 activation. Thus, autophagy formation is a cellular defense mechanism against polyQ72-induced ER-stress-mediated cell death by degrading polyQ72 aggregates, with PERK/eIF2alpha phosphorylation being involved in polyQ72-induced LC3 conversion.

ER stress induces caspase-8 activation, stimulating cytochrome c release and caspase-9 activation.

Excess ER stress induces caspase-12 activation and/or cytochrome c release, causing caspase-9 activation. Little is known about their relationship during ER stress-mediated cell death. Upon ER stress, P19 embryonal carcinoma (EC) cells showed activation of various caspases, including caspase-3, caspase-8, caspase-9, and caspase-12, and extensive DNA fragmentation. We examined the relationship between ER stress-mediated cytochrome c/caspase-9 and caspase-12 activation by using caspase-9- and caspase-8-deficient mouse embryonic fibroblasts and a P19 EC cell clone [P19-36/12 (-) cells] lacking expression of caspase-12. Caspase-9 and caspase-8 deficiency inhibited and delayed the onset of DNA fragmentation but did not inhibit caspase-12 processing induced by ER stress. P19-36/12 (-) cells underwent apoptosis upon ER stress, with cytochrome c release and caspase-8 and caspase-9 activation. The dominant negative form of FADD and z-VAD-fmk inhibited caspase-8, caspase-9, Bid processing, cytochrome c release, and DNA fragmentation induced by ER stress, suggesting that caspase-8 and caspase-9 are the main caspases involved in ER stress-mediated apoptosis of P19-36/12 (-) cells. Caspase-8 deficiency also inhibited the cytochrome c release induced by ER stress. Thus, in parallel with the caspase-12 activation, ER stress triggers caspase-8 activation, resulting in cytochrome c/caspase-9 activation via Bid processing.

Expression of RA175 mRNA, a new member of the immunoglobulin superfamily, in developing mouse brain.

RA175, a new member of the immunoglobulin superfamily, is highly expressed during neuronal differentiation of P19 embryonal carcinoma cells. In situ hybridization showed that RA175 mRNA was detected in the developing nervous system, as well as the epithelium of the various non-neuronal tissues of mouse embryo. In contrast with the epithelia of the non-neuronal tissues, RA175 mRNA was not co-expressed with Sonic hedgehog mRNA and Patched mRNA during brain morphogenesis. RA175 mRNA was highly expressed in the anterior horn cells and the peripheral nervous system at embryonic day (E) 11.5 and in the central nervous system at E14.5-E18.5, but its expression decreased after birth and was undetectable in the adult mouse brain.

Induction of bud formation of embryonic mouse tracheal epithelium by fibroblast growth factor plus transferrin in mesenchyme-free culture.

Embryonic mouse tracheal epithelium, which branches in an epithelial-mesenchymal recombination culture with bronchial mesenchyme, was cultured under mesenchyme-free conditions. When embedded in a basement-membrane-like matrix and cultured in a serum-free medium supplemented with fibroblast growth factor 1 (FGF1), the tracheal epithelium did not branch, whereas the bronchial epithelium underwent branching morphogenesis. When the medium was enriched with transferrin (Tf), bud formation was induced in the tracheal epithelium and some buds branched secondarily. FGF7 and FGF10, in cooperation with Tf, induced tracheal bud formation to the same extent as FGF1, although the optimum concentrations differed. A bromodeoxyuridine-labeling study comparing cultures with and without Tf showed no Tf-specific amplification of cell proliferation. A whole-mount in situ hybridization study of the expression of Bmp4 and Shh genes in explants of mesenchyme-free culture revealed that both genes were ubiquitously expressed and that expression did not correlate with bud formation. This expression pattern was different from the distally localized expression pattern observed in normal lung rudiments and in extratracheal buds induced by the recombined bronchial mesenchyme. These results suggest that both bronchial and tracheal bud formations were initiated without localized exposure of the epithelium to FGFs and were not accompanied by localized expression of Bmp4 and Shh in the epithelium.

Keratin attenuates tumor necrosis factor-induced cytotoxicity through association with TRADD.

Keratin 8 and 18 (K8/18) are the major components of intermediate filament (IF) proteins of simple or single-layered epithelia. Recent data show that normal and malignant epithelial cells deficient in K8/18 are nearly 100 times more sensitive to tumor necrosis factor (TNF)-induced cell death. We have now identified human TNF receptor type 1 (TNFR1)-associated death domain protein (TRADD) to be the K18-interacting protein. Among IF proteins tested in two-hybrid systems, TRADD specifically bound K18 and K14, type I (acidic) keratins. The COOH-terminal region of TRADD interacted with the coil Ia of the rod domain of K18. Endogenous TRADD coimmunoprecipitated with K18, and colocalized with K8/18 filaments in human mammary epithelial cells. Overexpression of the NH2 terminus (amino acids 1-270) of K18 containing the TRADD-binding domain as well as overexpression of K8/18 in SW13 cells, which are devoid of keratins, rendered the cells more resistant to killing by TNF. We also showed that overexpressed NH2 termini of K18 and K8/18 were associated with endogenous TRADD in SW13 cells, resulting in the inhibition of caspase-8 activation. These results indicate that K18 may sequester TRADD to attenuate interactions between TRADD and activated TNFR1 and moderate TNF-induced apoptosis in simple epithelial cells.

Caspase activity is involved in, but is dispensable for, early motoneuron death in the chick embryo cervical spinal cord.

We examined the role of caspases in the early programmed cell death (PCD) of motoneurons (MNs) in the chick embryo cervical cord between embryonic day (E) 4 and E5. An increase in caspase-3-like activity in MNs was observed at E4.5. Treatment with an inhibitor of caspase-3-like activity, Ac-DEVD-CHO, for 12 h blocked this increase and revealed that caspase-3-like activity is mainly responsible for DNA fragmentation and the nuclear changes during PCD but not for degenerative changes in the cytoplasm. When a more broad-spectrum caspase inhibitor was used (bocaspartyl (OMe)-fluoromethyl ketone, BAF), the appearance of degenerative changes in the cytoplasm was delayed by at least 12 h. However, following treatment with either Ac-DEVD-CHO or BAF for 24 h, the number of surviving healthy MNs did not differ from controls, indicating a normal occurrence of PCD despite the inhibition of caspases. These results suggest that caspase cascades that occur upstream of and are independent of the activation of caspase-3-like activity are responsible for the degenerative changes in the cytoplasm of dying cervical MNs. These data also suggest that, although one function of caspases may be to facilitate the kinetics of PCD, caspases are nonetheless dispensable for at least some forms of normal neuronal PCD in vivo.

Discovery of a novel compound: insight into mechanisms for acrylamide-induced axonopathy and colchicine-induced apoptotic neuronal cell death.

The exposure of humans and experimental animals to certain industrial toxins such as acrylamide is known to cause nerve damage classified as axonopathy, but the mechanisms involved are poorly understood. Here we show that acrylamide induces morphological changes and tyrosine phosphorylation of focal adhesion kinase (FAK) and proline-rich tyrosine kinase 2 (Pyk2), a member of the FAK subfamily, in human differentiating neuroblastoma SH-SY5Y cells. Furthermore, we identified a novel molecule designated 'compound-1' that inhibits the morphological and biochemical events. Daily oral administrations of the compound also effectively alleviated behavioral deficits in animals elicited by acrylamide in inclined plane testing, landing foot spread testing and rota-rod performance testing. The compound also effectively inhibited the biological and biochemical responses caused by another axonopathy inducer, colchicine, including tyrosine phosphorylation of Pyk2, formation of an 85-kDa poly(ADP-ribose)polymerase (PARP) fragment and apoptosis-associated induction of the NAPOR gene as well as neuronal cell death. Our findings not only provide insight into FAK and Pyk2 functions in neuronal cells, but may also be important in the development of therapeutic agents for peripheral neuropathy and neurodegeneration.

Massive muscle cell degeneration in the early stage of merosin-deficient congenital muscular dystrophy.

Primary merosin-deficient congenital muscular dystrophy (CMD) is a severe form of congenital muscular disorder which is caused by mutations in the laminin alpha2 chain gene (LAMA2). The disease is characterized by marked dystrophic changes in skeletal muscles during early infancy, while little is known about the pathological process of the muscle fiber degeneration. Here, we report the immunohistochemical analysis of skeletal muscle in ten patients with primary merosin-deficient CMD using a panel of molecular markers for skeletal muscle proteins, cellular necrosis, and apoptosis. In the youngest patient (a 52 day old baby), prominent massive muscle cell degeneration occurred in association with the deposition of the C5-9 complement membrane attack complex (MAC). Most of the MAC-positive muscle fibers showed a severely deranged immunoreaction to dystrophin, dystroglycans, and other sarcolemmal proteins. In addition, we found scattered positive signals for apoptosis. Similar but milder changes were also observed in six other patients younger than 1 year. In the patients older than 3 years, muscle fibers positive for MAC and apoptotic signals were barely detectable. These findings imply that massive muscle fiber degeneration occurs in the very early stage of merosin-deficient CMD and may contribute to the severe dystrophic changes in muscle from early infancy.

Caspase-3-deficiency induces hyperplasia of supporting cells and degeneration of sensory cells resulting in the hearing loss.

Caspase-3 is one of the cystein proteases that play essential roles in programmed cell death. As such, brain development is profoundly affected by caspase-3-deficiency, resulting in hyperplasia and abnormal cell organization (Kuida et al., Nature 1996;384:368-372). In the present study, we used caspase-3 (-/-) mice to show that caspase-3 deficiency results in severe hearing loss, hyperplasia of supporting cells and degeneration of sensory hair cells. The greater epithelial ridge, a remnant of the primordial organ of Corti, persists throughout all of the turns of cochlea in 2-week-old caspase-3 (-/-) mice, which indicates that the morphology of the cochlea is immature. The number of border cells, that develop from the greater epithelial ridge and are one of the supporting cells of the inner hair cell, increase significantly in both 2- and 5-week-old caspase-3 (-/-) mice. On the other hand, abnormal fused stereocilia can be seen in both 2- and 5-week-old caspase-3 (-/-) mice, and disarrangement and loss of sensory hair cells are observed in 5-week-old caspase-3 (-/-) mice. Taken together, both hyperplasia and degeneration occur simultaneously in the inner ear of the caspase-3 (-/-) mice, suggesting that caspase-3-dependent apoptosis is necessary for the development and formation of a properly functioning auditory system in mammals.

FTY720, a novel immunosuppressive agent, induces apoptosis in human glioma cells.

FTY720, a metabolite from Isaria sinclairii, has been developed to be a potent immunosuppressive drug with induction of apoptosis in T cells and several cell lines. We investigated whether FTY720 induces apoptosis in human glioma cell lines, since they are relatively resistant to multiple apoptotic stimuli. In human glioma cells including T98G, FTY720 induced apoptosiswith ED50 between 1 to 10 microg/ml, while etoposidedid not induce apoptosis at the same doses. Among the caspase family proteases, mainly caspase-6 was activated during the apoptosis by FTY720 but not etoposide. In addition, FTY720 caused tyrosine dephosphorylation of FAK and did not activate a FAK-PI3-kinase survival pathway. This was confirmed also by the observation that orthovanadate prevented FTY720-induced dephosphorylation of FAK and inhibited FTY720-induced cell death. We assumed that FTY720 induced FAK dephosphorylation and cut off the FAK-PI3-kinase pathway resulting in the induction of apoptosis via caspase-6 activation in these glioma cells.

Suppression of hair follicle development inhibits induction of sonic hedgehog, patched, and patched-2 in hair germs in mice.

Embryonic induction of hair follicles is a fascinating model of localized morphogenesis from a simple homogeneous epithelial cell sheet. Accumulating evidence indicates that Sonic hedgehog (Shh) signaling plays a central role in hair follicle formation. We quantitated the expression levels of Shh and its receptor genes, Patched (Ptc) and Patched-2 (Ptch2), in two distinct experimental systems in which the development of hair follicles was suppressed. Shh, Ptc, and Ptch2 were induced about six- to tenfold in normal embryonic hair germs in vivo as well as in developing skin tissue maintained in organ culture. This induction was almost completely inhibited both in the developing skin tissue of ICR mice cultured with 30ng/ml epidermal growth factor and in embryos of Tabby mutant mice (a model of hypohidrotic ectodermal dysplasia) at 14.5-15.5 days postcoitus. Expression of Shh, Ptc and Ptch2 was induced in the Tabby embryos at 16.5 days postcoitus, indicating that Shh signaling may be involved in the formation not only of the well-studied guard hair but also of the awl hair. The potential of the two biological systems for studying molecular mechanisms in hair follicle formation, particularly at an early phase including Shh signaling, is discussed.

Caspases that are activated during generation of nuclear polyglutamine aggregates are necessary for DNA fragmentation but not sufficient for cell death.

Truncated polypeptides containing expanded polyglutamine (polyQ) stretches tend to form cytoplasmic or nuclear aggregates in cultured cells, leading to cell death. Although it has been shown recently that caspase-8 coaggregates with polyQ and is activated during polyQ-mediated cell death, little is known of the location and timing of caspase-8 activation by nuclear polyQ aggregates. Also, the relationship between nuclear polyQ aggregate-mediated cell death and activation of other caspases is unclear. In P19 embryonal carcinoma (EC) cells, which can be made to differentiate into neuronal cells, polyQ72 repeats preferentially aggregate in the nucleus. Nuclear aggregates of polyQ72 induced P19 EC cell death, with a high frequency of cells exhibiting morphology characteristic of apoptosis (i.e., roundness, cell shrinkage, chromatin condensation) and DNA fragmentation. In the present study, we used antisera that specifically recognized the active forms of caspase-8, -3, and -9 but not their proforms, and showed that only caspase-8 and -3 were activated during the generation of polyQ72 aggregates in P19 EC cell nuclei. Furthermore, we showed that the caspase inhibitor z-VAD-fmk inhibited DNA fragmentation, but only partially inhibited the appearance of apoptotic morphology. Thus, caspase activation, including caspase-8 and -3, is necessary for polyQ-mediated DNA fragmentation but not sufficient for polyQ-mediated cell death in P19 EC cells.

Deficiency in caspase-9 or caspase-3 induces compensatory caspase activation.

Dysregulation of apoptosis contributes to the pathogenesis of many human diseases. As effectors of the apoptotic machinery, caspases are considered potential therapeutic targets. Using an established in vivo model of Fas-mediated apoptosis, we demonstrate here that elimination of certain caspases was compensated in vivo by the activation of other caspases. Hepatocyte apoptosis and mouse death induced by the Fas agonistic antibody Jo2 required proapoptotic Bcl-2 family member Bid and used a Bid-mediated mitochondrial pathway of caspase activation; deficiency in caspases essential for this pathway, caspase-9 or caspase-3, unexpectedly resulted in rapid activation of alternate caspases after injection of Jo2, and therefore failed to protect mice against Jo2 toxicity. Moreover, both ultraviolet and gamma irradiation, two established inducers of the mitochondrial caspase-activation pathway, also elicited compensatory activation of caspases in cultured caspase-3(-/-) hepatocytes, indicating that the compensatory caspase activation was mediated through the mitochondria. Our findings provide direct experimental evidence for compensatory pathways of caspase activation. This issue should therefore be considered in developing caspase inhibitors for therapeutic applications.

Detection of caspase-9 activation in the cell death of the Bcl-x-deficient mouse embryo nervous system by cleavage sites-directed antisera.

Caspases, which play crucial roles during apoptosis, are activated from their inactive proforms in a sequential cascade of cleavage by other members of the caspase family. Caspase-9 is autoprocessed by the Apaf-1/cytochrome c pathway and acts at an early point in this cascade, whereas Bcl-xL, an antiapoptotic member of the Bcl-2 family, prevents activation of caspases in vitro. Little is known, however, about the relation between caspase-9 and Bcl-xL during development of the mammalian nervous system. We used antisera against two cleavage sites in mouse caspase-9 that recognize only the activated form of mouse caspase-9, and we examined immunohistochemically the activation of mouse caspase-9 in the nervous system of Bcl-x-deficient mouse embryos. Mouse caspase-9 is processed at both D(353) and D(368), but it is processed preferentially at D(368) during apoptosis of cultured cells induced by various stimuli and in the nervous system of Bcl-x-deficient mouse embryos. We show that Bcl-xL protects against caspase-9- and/or caspase-3-dependent apoptosis in the caudal portion of the ventral hindbrain, anterior horn cells, and dorsal root ganglia neurons of the normal mouse embryos and against caspase-9/caspase-3-independent apoptosis in the dorsal region of the nervous system including the dorsal spinal cord. Furthermore, we demonstrate that Bcl-xL blocks cytochrome c release from mitochondria, causing activation of caspase-9 in anterior horn cells and dorsal root ganglia neurons in mouse embryos at embryonic day 11.5.

Clinicopathologic study of mass-screened neuroblastoma with special emphasis on untreated observed cases: a possible histologic clue to tumor regression.

Spontaneous regression and maturation of neuroblastoma (NB) are well documented and occur frequently in infants, including those detected by mass screening. To seek histologic clues for regression/maturation in mass-screened NB, clinicopathologic features of 12 tumors that were resected after 2 to 18 months of untreated observation were reviewed. Unobserved screened and age-matched unscreened patients were also studied. To evaluate the possible important role of apoptosis, apoptotic cells were detected by in situ deoxyribonucleic acid (DNA) nick end labeling and immunohistochemical stain for activated caspase-3. Nests with a varying degree of reduced cellularity ("less cellular" and "hypocellular" nests) were common in patients younger than 18 months of age, and were rare in older patients. Two characteristic cells, which have not been focused previously, were frequent, especially in the hypocellular nests. One showed amorphic eosinophilic cytoplasm with pyknotic nuclei and the other contained plump cytoplasm with well-maintained nuclei. These cells were also observed in 89% of the unobserved screened NBs and 79% of the age-matched unscreened patients with good outcome, whereas they could not be confirmed in any of the age-matched unscreened NBs with poor outcome. The amorphic and plump cells were negative for activated caspase-3 and in situ DNA nick end labeling. From these results, the authors hypothesize that these cells most likely represent a degenerative process, in either a state before the activation of caspase-3 or a caspase-independent form of cell death. The presence of less cellular and hypocellular nests with amorphic/plump cells may serve as one of the important clues in predicting tumor prognosis.

A novel congenital myopathy with apoptotic changes.

We report on a female child with congenital myopathy with delayed developmental milestones and mental retardation. The most striking pathological finding was the presence of many condensed to fragmented myonuclei. DNA fragmentation was confirmed by the TUNEL method and supported by the ultrastructural characteristics of apoptotic nuclear changes. We also demonstrated immunohistochemically the activation of caspase-3 and caspase-9. This appears to be the first reported case of congenital myopathy with apoptotic process.

Localization of active form of caspase-8 in mouse L929 cells induced by TNF treatment and polyglutamine aggregates.

The relation between activation of caspase-8 and polyglutamine aggregates has been focused. We prepared an antiserum (anti-m8D387) that recognizes the active form but not the proform of mouse caspase-8. We used immunostaining with anti-m8D387 antiserum to compare the localizations of activated mcaspase-8 in L929 (clone 1422) cells induced by TNF and polyglutamine aggregates. Anti-m8D387 was positive throughout cytoplasm of the TUNEL-positive cells induced by TNF treatment, whereas the anti-m8D387 reactivity was not positive throughout cytoplasm of the cells expressing polyglutamine but was restricted to polyglutamine aggregates. In contrast with TNF-treated cells, cells expressing anti-m8D387-positive cytoplasmic polyglutamine aggregates did not undergo TUNEL-positive apoptosis. Thus activated caspase-8 associated with polyglutamine aggregates alone was not sufficient to induce TUNEL-positive apoptosis of L929 (clone 1422) cells. The distribution of activated caspase-8 associated with polyglutamine aggregates may be essential for the polyglutamine-mediated cell death or downstream of caspase-8 may be different in the TNF-treated cells and cells expressing polyglutamine.