A) Immunofluorescence analysis for P0 or KDEL on teased fibers shows that P0 staining is augmented in the ER of myelinating Schwann cells from P0S63del (arrows), as compared to wt (arrowheads), sciatic nerves. B-C) BiP and CHOP mRNA levels were measured in P28 sciatic nerves by quantitative RT-PCR and normalized to 18S rRNA, wt = 1. L, low, H, high expressor S63del transgenes; OE, P0 wildtype overexpressor; tun, tunicamycin; error bars, s.e.m. for n = individual nerves from 3-5 animals. * p < 0.001, ** p < 0.05 relative to wt by Student's t-test. D) Western analysis for BiP and CHOP was performed on freshly dissected nerves or wt nerve segments cultured overnight with tunicmycin or sham media. Note that CHOP, but not BiP, mRNA and protein levels parallel the ratio of P0S63del (arrowhead) to P0wt (arrow), and the severity of neuropathy (Wrabetz et al., 2006) in S63del mice. Calnexin (CNX) levels do not change. BiP and CHOP are induced upon tunicamycin treatment of wt nerve (wt/Tun +). β-Tubulin (Tub) provides a control for loading. E) DOC1, 6 and 4 mRNA levels are induced by tunicamycin treatment of NIH3T3 cells (note the logarithmic scale), but only DOC1 and 6 are induced in S63del nerves. DOC4 is not induced even by tunicamycin treatment of nerve.

A) IRE-1 catalyzes excision of a 26 nt fragment from XBP-1 mRNA. Semi-quantitative RT-PCR for XBP-1 unspliced (XBP-1u), spliced (XBP-1s), normalized to GAPDH, shows that the spliced XBP-1 is enriched in S63del and tunicamycin-treated nerves. P0wt overexpressor (OE) nerves are included as expression-matched controls for S63del low (L) and high (H). Error bars, s.e.m. for n = 6 individual nerves, *p < 0.05 relative to wt by Student's t-test. B) Levels of eIF2α, eIF2α–P (phosphorylated), and ATF4, were measured by Western analysis. Tubulin (Tub) shows equal loading. C) ATF3 mRNA was measured by quantitative RT-PCR and normalized to phosphoglycerate kinase 1(PGK1) mRNA. ATF3 is induced in a dose-dependent fashion in S63del mice and in tunicamycin-treated livers. Error bars, s.e.m., n = 6, 5, 5, 3, and 3 nerves for wt, S63del-L, S63del-H, OE-L, OE-H, respectively and 3 livers. ** p < 0.01, * p < 0.05 relative to wt by Student's t-test. D) Immunostaining reveals CHOP in elongated, DAPI positive, Schwann cell nuclei in longitudinal sections of S63del nerves at P28 (b–b"), but is undetectable in normal nerves (a–a") Many CHOP-positive Schwann cells contain integral myelin sheaths (c–c"'). Bar = 15 μm in a–b"; 13 μm in c–c"'. E) Western analysis for ATF6 reveals dose-dependent cleavage of the 90kD precursor protein (arrowhead) to the 50kD product (arrow) only in S63del, but not S63C, P0 overexpressor (OE) or wt nerves. Numbers indicate relative molecular weights (M_r).

A) Western analysis was performed for P0 under reducing (DTT +) or non-reducing conditions (DTT −) on peripheral nerve lysates from wt, S63del/P0−/− (S63del) or S63C/P0−/− (S63C) mice. Therefore, in mutant mice, only mutant P0 is present. Under non-reducing conditions, wt and mutant P0 show a 3-4 kD shift as expected for the formation of one disulfide bond. Note that non-reduced S63del produces a smear (arrowhead), that might represent protein misfolding. The majority of P0S63del recognized in this analysis is unglycosylated (Wrabetz et al., 2006) and migrates at a lower M_r. In contrast, S63C produces discrete bands with the same mobility as P0wt. B) P0 β-strand C is characterized by an alternating hydrophilic (black) and hydrophobic (red) pattern of residues. In contrast to S63C, deletion of either S63 or F64, or the mutant S63F/T65F and S63F are predicted to alter the orientation of 4, 3, 2 and 1 hydrophobic residues, respectively. The correlation between number of altered (alt) hydrophobic residues and ER retention (Ret +) in transfected cells is shown. COS-7 cells were transfected with constructs encoding various P0wt– or P0mutant–DsRed fusion proteins and analyzed for expression of UPR markers (C) or for trafficking of P0 (D). BiP mRNA was measured by quantitative RT-PCR and normalized to 18S rRNA for transfection efficiency. XBP-1 and GAPDH were measured by semiquantitative RT-PCR. Note that BiP mRNA levels and XBP-1 splicing (XBP-1s, arrow) are enriched in cells expressing either P0S63del or P0F64del mutants, as well as in cells treated with tunicamycin (a representative experiment (of 3) in shown). D) Confocal microscopy of trafficking of P0-DsRed fusion proteins (P0, red) and immunostaining for calnexin (CNX, green) is shown in overlay. P0wt, P0S63C, and P0S63F are delivered to the cell membrane, whereas P0S63del and P0F64del are completely retained in the ER (yellow signal). P0S63F/T65F was mostly retained in the ER, although some was delivered to the plasma membrane in some cells. Bar = 20 μm.

P0 (A), BiP (B) or CHOP (C) mRNA levels in sciatic nerves were measured by quantitative RT-PCR and normalized to 18S rRNA in wt (gray bars) or S63del (black bars) mice. The level of P0 rises in both wt and S63del nerves at P7, with a peak near P28. Only in S63del, but not in wt nerves, BiP rises in parallel, and CHOP rises by P7 to a plateau level that then persists. For P0 and BiP, one representative experiment (animal) of three is shown; error bars, s.e.m.; n = 3. For CHOP the average of 7 experiments (animals) is shown (anchored to P28wt and then expressed relative to P1wt = 1; error bars, s.e.m.; n = 7; ** p < 0.05 relative to wt of the same age by Student's t-test.

Segments of wt or S63del-L sciatic nerves were snap-frozen either immediately after dissection (t = 0h, white bars) or after incubation ex vivo for 20 hours with (t = 20h TUN, black bars) or without (t = 20h, gray bars) tunicamycin, and analyzed for UPR markers. Levels of mRNA were measured by quantitative RT-PCR (A-B,D) and normalized to PGK1 mRNA. BiP and CHOP mRNAs are induced in S63del mice at t = 0h and remain induced with tunicamycin treatment for twenty hours after axotomy. Instead induction is rapidly reversed without tunicamycin, in parallel with marked reduction of the band containing P0S63del by Western analysis (C, arrowhead), due to the general reduction of P0 mRNA after axotomy (D). (S63del −/−) represents P0 −/− mice transgenic for MpzS63del. β-Tubulin (Tub) is a control for protein loading. Error bars, s.e.m., n = 3 animals.

A) Rotarod analysis of motor function showed that four month old S63del mice remain for 200 seconds less as compared to wt and Chop-null mice. Instead, S63del/Chop-null mice perform like wt mice. B) Semithin sections stained with toluidine blue are shown from P180 wt, S63del, Chop-null and S63del/Chop-null sciatic nerves. S63del nerves manifest hypomyelination and demyelinated fibers and onion bulbs. Although hypomyelination is not ameliorated in S63del/Chop-null nerves, the number of demyelinating fibers is significantly reduced (Table). Examples of pathological fibers from S63del or S63del/Chop-null nerves are shown magnified below. Arrows indicate demyelinated fibers, stars indicate onion bulbs; and arrowheads indicates remyelinating fibers. Bar = 15 μm in upper panels and 5 μm in the magnified panels below. C) S63del transgene expression was analyzed in rescued nerves by semi-quantitative RT-PCR. Digestion of the amplimer with Dpn II distinguishes wt (326 bp) from transgenic cDNA (244 and 82 bp fragments). D-H) BiP, CHOP and DOC1,6 and 4 expression was analyzed by quantitative RT-PCR and normalized to 18S rRNA. Note that transgene expression and BiP activation remain in S63del/Chop-null nerves, whereas CHOP is undetectable and DOC1 falls below wt levels. DOC6 and 4 are not induced in S63del-L nerves.

A) TUNEL analysis in longitudinal sections of wt and mutant sciatic nerves at P28 (grey bars), P42 (white bars) and P180 (black bars) reveals that cell death is significantly increased in S63del nerves as compared to wt or other mutant nerves. TUNEL-positive nuclei were usually elongated suggesting they belonged to Schwann cells (data not shown). Ablation of Chop reduces cell death to wt levels in S63del/Chop-null nerves. Note that cell death in S63del nerves at P180 is nearly double that of P28, whereas BiP and Chop expression peak at or before P28 (Figure 4). Error bars, s.e.m., for P28 and P180, n = 27, 5, 14, 5, 6, or 6 animals in wt, Chop-null, S63del, S63del/Chop-null, S63C, or OE (P0 wildtype overexpressor), respectively; for P42 wt and S63del n=3. *p < 0.001 by Student's t-test relative to wt or S63del/Chop-null. B) Cell number per 40x field was estimated by counting DAPI-stained nuclei in longitudinal sections of sciatic nerves. Chop ablation does not alter the nuclei/field in S63del nerves at either P28 or P180. p < *0.002 or ^&0.05 by Student's t-test relative to wt for n= 9-11 sections (2-3 sections from each of 3-5 animals). C) Cell proliferation was estimated by BrdU incorporation in P28 (grey bars) and P180 (black bars) sciatic nerves from wt and mutant mice. Cell proliferation is increased in S63del mice at P28 and P180 and in S63del/Chop-null mice at P28. Error bars, s.e.m., n = 5, 3, 4, 3, wt, Chop-null, S63del, S63del/Chop-null, respectively. D) Proteolytic cleavage of caspase-12 was analyzed by Western analysis of sciatic nerve lysates from P42 wt, S63del, S63C and P0 OE or P28 wt, Chop-null, S63del or S63del/Chop-null animals. The amount of the 42 kD activated form of caspase-12 (arrowhead) is strongly increased in S63del nerves and remains after Chop ablation. A representative experiment (of 3) is shown. Numbers indicate M_r.