Peripherin aggregates in DRG neurons of peripherin transgenic mice. Cryostat sections from L4-L5 DRG from peripherin transgenic mice were labeled by indirect immunofluorescence with antibody to peripherin or neurofilaments (NF-H; SMI32). The panel labeled Peripherin and NF-H Double is an overlay of the images obtained with antibodies recognizing peripherin (red) and NF-H (green), and shows peripherin aggregates either alone or together with neurofilaments (arrows). Bar, 75 μm.

Overexpression of peripherin induces apoptosis of DRG neurons. (A) TUNEL assays were performed on WT and Per dissociated spinal cord cultures. Cultures were then labeled by indirect immunofluorescence with peripherin antibody so that TUNEL-positive DRG neurons could be correlated with the presence of peripherin aggregates. TUNEL-positive DRG neurons were shrunken and had condensed chromatin and an intact plasma membrane, consistent with the changes associated with apoptosis. (B) Dissociated spinal cord cultures were labeled by indirect immunofluorescence with antibody recognizing activated caspase-3 (C-3). A number of DRG neurons in Per cultures were labeled with C-3. No labeling of WT DRG neurons was observed. (C) Electron microscopic evaluation of apoptotic Per DRG neurons showed that there was severe disruption of mitochondrial cristae (arrowheads). (D) Comparison of specific apoptosis of Per DRG neurons in dissociated spinal cord cultures from three separate litters. The percentage of TUNEL-positive DRG neurons in WT and Per dissociated spinal cord cultures was calculated after 14 d in vitro. Specific apoptosis was calculated as the percentage of TUNEL-positive Per DRGs, minus the percentage of TUNEL-positive WT DRGs/100, minus the percentage of TUNEL-positive WT DRGs. Specific apoptosis of Per DRG neurons was usually between 30 and 40%. The results shown are from three litters and are typical of our findings. Bars: (A) 20 μM; (B) 15 μM; (C) 0.4 μM.

Apoptosis of Per DRG neurons in dissociated spinal cord cultures is precluded by TNF-α–neutralizing antibody. (A) Indirect immunofluorescence labeling with Mac-2 antibody (American Type Culture Collection) showing the abundance of activated microglia in dissociated spinal cord cultures. 30 μm. (B) The effects of TNF-α–neutralizing antibody on the percentage of TUNEL-positive DRG neurons in WT and Per dissociated spinal cord cultures was compared. Quadruplet cultures from each embryo were prepared and 5 or 10 μg/ml of TNF-α antibody added to two of the cultures, the other two acting as control. The chart and conjoining table show that inclusion of TNF-α antibody within the culture medium significantly reduced the number of TUNEL-positive DRG neurons in Per cultures (P < 0.001*** by two-way analysis of variance). Addition of 10 μg/ml TNF-α antibody to the culture medium enhanced the protective effect to levels similar to results found for WT. There was no significant effect of TNF-α–neutralizing antibody on WT DRG neuronal viability (unpublished data) Bar, 50 μM.

Intranuclear microinjection of the peripherin gene induces death of cultured motor neurons. (A) Mammalian expression plasmids encoding either NF-L or peripherin were microinjected into the nuclei of motor neurons in dissociated spinal cord cultures. Cultures were then labeled by indirect immunofluorescence with antibody recognizing NF-L (NR4) or peripherin (MAB1527). The results in this figure show the organization of NF-L or peripherin in microinjected motor neurons after 3 d of expression. Overexpression of NF-L led to intense immunofluorescence labeling of motor neurons, characterized by the formation of coiling loops in the perikaryon. In contrast, overexpressed peripherin did not integrate normally into the existing cytoplasmic network, but instead formed punctate aggregates that were also clearly apparent in neurites. (B) Double immunofluorescence labeling with antibodies to peripherin (MAB1527) and NF-L (AB1983) of the neurites from a motor neuron that had been microinjected with peripherin, showing the disruption, by peripherin, of the endogenous neurofilament network (arrows). (C) Motor neurons in dissociated spinal cord cultures were microinjected with expression plasmids encoding (a) peripherin; (b) NF-L; (c) NF-L and peripherin; or (d) expression plasmid alone (pRcCMV; control). The number of viable motor neurons was counted each day for 7 d, and the results from each microinjection experiment were compared. The chart shows that peripherin was extremely toxic to motor neurons. In contrast, there was no significant effect of NF-L overexpression on viability (Student's t test; P < 0.05). Indeed, comicroinjection of NF-L with peripherin offered protection from the toxic effects observed with peripherin microinjection alone. Bars: (A) 25 μm; (B) 15 μm.

Peripherin aggregates in Per transgenic DRG neurons. (A) Comparison of peripherin labeling of dissociated spinal cord cultures derived from WT and Per transgenic embryos. DRG neurons were the predominant cell type labeled by peripherin antibody in these cultures. Peripherin in DRG neurons in WT cultures displayed a normal cytoplasmic distribution, whereas DRG neurons in Per cultures formed aggregates that displayed a variety of phenotypes, including distinct spherical aggregates, punctate-like aggregates, and amorphous aggregates that filled the perikaryon (arrows). (B) Double immunofluorescence labeling of DRG neurons from WT and Per cultures with peripherin and neurofilament (N52, specific for NF-H) antibodies showed the colocalization of neurofilaments with the peripherin aggregates. (C) Electron microscopic examination of aggregates showed that they were composed of 10-nm filaments, representative of intermediate filaments (*), excluded microtubules, and contained entrapped organelles, particularly mitochondria (arrow). (D) To estimate the increase in peripherin expression in Per DRG neurons relative to WT, pure DRG neuronal cultures from WT and Per transgenic embryos were prepared, and samples of each compared by immunoblotting with antibody to peripherin. Labeling with actin antibody was used to control for protein loadings. The immunoblot shows the findings from cultures derived from five different embryos and is representative of all cultures analyzed. The increase in peripherin expression in Per DRG neurons was estimated using NIH image software and found to be approximately sixfold. Bars: (A) 25 μM; (B) 20 μM; (C) 75 μm.

Apoptosis of Per DRG neurons requires CNS cell culture environment. (A) It was observed that localized groupings of Per DRG neurons within dissociated spinal cord cultures were largely TUNEL negative despite containing peripherin aggregates. Double labeling of such a grouping with TUNEL and with peripherin antibody shows that all the DRG neurons in the field have peripherin aggregates, but that only a few are TUNEL positive and have condensed chromatin (arrows). (B) Pure DRG neuronal cultures were prepared from both WT and Per transgenic embryos, and the number of TUNEL-positive neurons was compared with the findings from dissociated spinal cord cultures (spinal culture). The chart and conjoining table show that there was no significant increase in the number of TUNEL-positive DRG neurons in pure Per DRG cultures compared with WT DRG neurons (DRG culture). To investigate the effect of the cell culture environment on Per DRG viability, pure DRG neurons were grown for 4 d on glass coverslips and then placed in inserts in plates that had (Coculture) or had not (coculture control) been seeded with dissociated spinal cord cultures. Our findings show that when pure cultures of Per DRG neurons were cocultured with dissociated spinal cord cultures, there was a significant increase in apoptosis compared with control P < 0.001*** (by two-way analysis of variance). (C) A comparison of TUNEL labeling of pure cultures of Per DRGs, cocultured in the absence (coculture control) or presence of dissociated spinal cord cells (coculture). The arrows indicate DRG neurons. Note the increase in number of TUNEL-positive DRG neurons and their shrunken appearance in spinal cord coculture compared with control DRG neurons (arrows). Bars: (A) 50 μM; (C) 50 μM.