PA200 is upregulated in the sciatic nerves of S63del mice

The levels of mRNA of PSME4, the gene that encodes the protein PA200, are increased by ∼50% in the sciatic nerves of S63del mice (Fig 1A). This magnitude of increase is comparable to that observed previously in transcriptomic studies on sciatic nerves of WT and S63del mice (D’Antonio et al, 2013). The levels of PA200 protein are ∼fourfold greater in S63del than in WT, as seen by Western blot analysis of sciatic nerve lysates (Fig 1B).

PA200 contains a bipartite nuclear localization sequence (NLS) and was detected as a nuclear protein in HeLa cells in the first report that described PA200 as an activator of the 20S proteasome (Ustrell et al, 2002). To evaluate PA200’s localization in sciatic nerves, we performed subcellular fractionation in buffer conditions in which 26S proteasomes remain assembled and functional (VerPlank et al, 2020). Unexpectedly, PA200 was located predominantly in the cytosolic fraction of sciatic nerves of both WT and S63del mice (Fig 1C). Schwann cells constitute ∼90% of the dry mass of sciatic nerves (Rutkowski et al, 1992). Therefore, PA200 is located predominantly in the cytosolic fraction of Schwann cells. In sciatic nerves from S63del mice, PA200 protein was increased in both the cytosolic and nuclear fractions, but most noticeably in the cytosolic fraction (Fig 1C).

Lysates of sciatic nerves from PA200−/− mice (Khor et al, 2006) were included as controls in the above experiments. PCR analysis of genomic DNA from these mice showed biallelic deletion of PSME4 (Fig S1) and, accordingly, PSME4 mRNA was undetectable via qRT-PCR in the sciatic nerves of PA200−/− mice (Fig 1A). In WT and S63del mice, Western blot analysis with antibodies against PA200 detected two bands: one with the expected molecular weight of ∼200 kD and another with a molecular weight of ∼150 kD (Fig 1B and C). In PA200−/− lysates, the 200 kD band was absent, whereas the 150 kD band was present. Notably, this 150 kD band was detected more readily than the 200 kD band (Fig 1B and C) and was located primarily in the nuclear fraction of WT and S63del sciatic nerves (Fig 1C). This nonspecific band detected by some commercially available antibodies (Welk et al, 2019) may be the reason PA200 has been repeatedly reported to be an exclusively nuclear protein.

PA200-bound proteasomes are increased in the sciatic nerves of S63del mice

To test whether the increase in the levels of PSME4 mRNA and PA200 protein led to an increase in PA200-bound proteasomes, we first immunoprecipitated proteasomes from the sciatic nerve lysates of WT and S63del mice with an antibody against the 20S proteasome subunit α1 and then immunoblotted for PA200. In S63del, more PA200 protein was detected in the proteasome immunoprecipitate than in WT (Fig 2A).

Centrifugation of the sciatic nerve lysates through a 10–50% glycerol gradient followed by measurements of proteasomal peptidase activity and Western blot analysis showed that PA200 protein is found only in fractions that contained the 20S proteasome subunit α1 and proteasomal chymotrypsin-like activity in both WT (fractions 1–4) and S63del (fractions 1–7) (Fig 2B). PA200 protein was not found in fractions that did not contain a 20S proteasome subunit or peptidase activity (fractions 10–19), and little to no PA200 was found in the unbound fraction of the proteasome immunoprecipitate (Fig 2A), strongly suggesting that the majority of PA200 protein in the sciatic nerves of WT and S63del mice is associated with proteasomes.

There are three types of PA200-bound proteasomes: a 20S with PA200 on one end (PA200-20S), a 20S with PA200 on both ends (PA200₂-20S), and a 20S with PA200 on one end and a 19S on the other (PA200-20S-19S) (Qian et al, 2013). To identify which types of PA200-bound proteasomes are present in the sciatic nerves and increased in S63del, we separated sciatic nerve lysates by native PAGE and performed Western blot for PA200 and the 20S proteasome subunit α4. The most prominent band detected with an antibody against PA200 that was also absent in sciatic nerve lysates from PA200−/− mice co-migrated with the 20S (Fig 2C), indicating a 20S proteasome with PA200 on one end (PA200-20S). These PA200-20S proteasomes were the most abundant type of PA200-bound proteasomes in sciatic nerves, and their levels were much greater in S63del than in WT (Fig 2C). Faint higher molecular weight bands were detected in S63del that were absent from the lysates of PA200−/− mice (Fig 2C). These bands migrated at approximately the same size as the 26S proteasome (19S-20S) and are most likely PA200₂-20S and PA200-20S-19S proteasomes. However, based on band intensities, these must represent a small fraction of the population of PA200-bound proteasomes in Schwann cells.

PA200-20S proteasomes were found predominantly in the cytosolic fractions of sciatic nerves from both WT and S63del mice (Fig 2D). In S63del, PA200-20S proteasomes were increased in both the cytosolic and nuclear fractions (Fig 2D), but most noticeably in the cytosolic fraction. The levels of PA200-20S proteasomes and their subcellular locations thus seem to resemble those of PA200 protein (Fig 1C).

We previously reported that the amount of assembled 26S proteasomes is increased in the sciatic nerve lysates of S63del mice (VerPlank et al, 2018), and this was also seen here in lysates (Fig 2C), and in cytosolic and nuclear fractions (Fig 2D). In nuclear fractions, 26S and 20S proteasomes were readily detected in S63del, but not in WT (Fig 2D). This finding was unexpected and suggests an increase in nuclear proteasomes in S63del. However, it should be noted that our fractionation method does not separate the ER from the nucleus. Therefore, the proteasomes detected in the nuclear fractions generated from the sciatic nerves of S63del mice may not be in the nucleus, but instead associated with the ER; where the mutant MPZ accumulates (Sidoli et al, 2016). Whatever the reasons for this unexpected finding, in S63del sciatic nerve lysates, there is a clear increase in the amount of PA200 and PA200-bound proteasomes, in both the nuclear and cytosolic fractions.

The knockout of PA200 in S63del mice increases protein degradation by proteasomes

We hypothesized that the increase in PA200 and PA200-bound proteasomes in the sciatic nerves of S63del mice was a compensatory response to maintain proteasomal protein degradation when there is a decrease in the 26S proteasome’s ability to degrade proteins. To test this hypothesis genetically, we generated S63del mice that lacked PA200 (S63del//PA200−/−) and examined protein degradation and proteasome function in the sciatic nerves.

To measure protein degradation, we used a well-validated pulse-chase approach in which newly synthesized proteins are labeled with [³H]tyrosine in the sciatic nerves ex vivo (VerPlank et al, 2018). In sciatic nerve explant cultures, this technique primarily measures protein turnover in Schwann cells because they constitute ∼90% of the dry weight of peripheral nerves (Rutkowski et al, 1992) and perform the majority of protein synthesis ex vivo because of the absence of neuronal cell bodies.

With a short pulse of 20 min with [³H]tyrosine, followed immediately by measurements of radiolabeled TCA-soluble peptides in the chase media, we first investigated the degradation of proteins which are degraded rapidly after synthesis. These proteins have short half-lives, include misfolded, regulatory (e.g., transcription factors, rate-limiting enzymes), and damaged proteins, and are degraded almost exclusively by the 26S proteasome in a ubiquitin-dependent manner (Sha et al, 2018b). The rate of degradation of these short-lived proteins was similar in sciatic nerves explanted from PA200−/− and WT mice, and reduced in the sciatic nerves explanted from S63del mice (Fig 3A) (VerPlank et al, 2018). In sciatic nerves from S63del//PA200−/− mice, the rate of degradation of short-lived proteins was similar to that measured in S63del (Fig 3A). Thus, in WT or S63del mice, PA200 does not participate in the degradation of these short-lived proteins, which compose a very small fraction of cell protein content.

To measure the degradation of long-lived proteins, which constitute the vast majority of cell proteins, the pulse with [³H]tyrosine was extended to 18 h and we included a chase period of 2 h to exclude the degradation of short-lived proteins from the analysis. The sciatic nerves explanted from PA200−/− mice showed a slight (∼20%), but consistent decrease in the rate of degradation of these long-lived proteins (Fig 3B and C). This finding was unexpected. PA200 has been shown to participate in the proteasomal degradation of specific proteins (e.g., acetylated histones) (Qian et al, 2013; Mandemaker et al, 2018), but this is the first evidence that it contributes to protein degradation generally. In S63del, the rate of degradation of long-lived proteins was ∼50% less than that of WT (Fig 3B and C), as seen previously (VerPlank et al, 2018). Surprisingly, the sciatic nerves explanted from S63del//PA200−/− mice exhibited a rate of degradation that was ∼twofold greater than that of S63del (Fig 3B and C).

Long-lived proteins, unlike short-lived proteins, are degraded by both the proteasome and the autophagy-lysosome system. To determine which was responsible for the observed changes in degradation in the absence of PA200, the assay was repeated in the presence of an inhibitor of the proteasome (Bortezomib), to measure protein degradation by the lysosome, or an inhibitor of lysosome acidification (chloroquine), to measure degradation by the proteasome (Sha et al, 2018b). Proteasome degradation was decreased by ∼25% in the sciatic nerves explanted from PA200−/− mice (Fig 3D) and ∼50% from S63del (Fig 3D) (VerPlank et al, 2018). In S63del//PA200−/−, the rate of proteasomal degradation of long-lived proteins was, again, ∼twofold greater than that measured in S63del (Fig 3D). Protein degradation by the lysosome was similar across all four genotypes (Fig 3D). Thus, the decreased degradation of long-lived proteins in PA200−/− and S63del occurred via the proteasome, as did the increased degradation in S63del//PA200−/−.

S63del//PA200−/− mice have greater proteasome peptidase activity and more assembled, active 26S proteasomes than S63del

Because the knockout of PA200 in WT and S63del mice altered proteasomal degradation, we assayed in sciatic nerve lysates the proteasomal hydrolysis of peptides specific to its chymotrypsin-like (LLVY-amc) and caspase-like (nLPnLD-amc) active sites. In PA200−/−, the hydrolysis of both peptides was slightly less than WT (Fig 4A and B), as would be expected with the observed decreases in proteasomal protein degradation (Fig 3). The levels of proteasome subunits in the lysates from WT and PA200−/− mice were comparable (Fig 4C). Thus, the slight decrease in proteasome peptidase activity was not because of a decreased number of proteasomes, but most likely because of the absence of PA200 and its activation of the 20S proteasome.

In S63del mice, the hydrolysis of these peptides by the proteasome in sciatic nerve lysates occurred at similar rates as in WT (Fig 4A and B) and (VerPlank et al, 2018, 2022). The lysates from S63del mice contained higher levels of proteasome subunits and assembled 26S proteasomes than those of wild-type littermates (Figs 2C and D and 4B and C) and (VerPlank et al, 2018, 2022). In S63del, the decreased degradation of proteins by the proteasome induces the upregulation of proteasome subunits (VerPlank et al, 2018), as is seen in many mammalian cultured cell lines in response to pharmacological inhibition of the proteasome (Sha & Goldberg, 2014). This increase in proteasome number is likely the reason total proteasome peptidase activity in the lysates is similar in S63del and WT, even though in S63del, the specific activity of the proteasome is decreased (VerPlank et al, 2018).

Lysates from sciatic nerves of S63del//PA200−/− mice exhibited proteasomal chymotrypsin-like and caspase-like activity that was 50% greater than that of S63del (Fig 4A and B). Thus, the knockout of PA200 in WT and S63del seems to have opposite effects on proteasome peptidase activity and protein degradation in the sciatic nerves–in WT, an expected slight decrease, and in S63del, an unexpected increase.

To better understand how the knockout of PA200 paradoxically increased proteasome activity and protein degradation in S63del mice, we first investigated the levels of 26S proteasome subunits. A greater number of 26S proteasomes in S63del//PA200−/− than in S63del could increase both protein degradation and proteasomal peptidase activity. Western blot analysis for several 26S proteasome subunits showed lower levels in S63del//PA200−/− sciatic nerve lysates than in S63del (Fig 4C), similar to those seen in WT and PA200−/− (Fig 4C). This return of proteasome subunits to WT levels is likely because of the lessening of the proteotoxic stress and proteasome functional impairment that caused their upregulation, as seen previously when proteasomal activity and protein degradation were increased in S63del mice by a 3-wk treatment with the PDE5 inhibitor sildenafil (VerPlank et al, 2022).

Next, we evaluated the levels of assembled 26S proteasomes in the sciatic nerve lysates by native PAGE followed by Western blot analysis for the 20S subunit α1. S63del//PA200−/− had more 26S proteasomes, especially those that are doubly capped (19S₂-20S), than S63del, where levels of assembled 26S proteasomes are already greater than that seen in WT (Figs 2C and 4D). An increase in assembled 26S proteasomes was also seen in the sciatic nerve lysates of PA200−/− mice (Figs 2C and D and 4D), but not to the extent of that seen in S63del//PA200−/− (Fig 4D). In both PA200−/− and S63del//PA200−/−, this increase occurred without a corresponding increase in the levels of proteasome subunits (Fig 4C), suggesting a change in stability or assembly of the 26S proteasome. Perhaps the absence of PA200 leaves ends of the 20S proteasome unoccupied and available for binding by 19S (see the Discussion section).

Not all 26S proteasomes in cells are in an active conformation that is competent to degrade proteins (Asano et al, 2015; Albert et al, 2020). To assess whether the greater number of assembled 26S proteasomes in S63del//PA200−/− sciatic nerve lysates was increasing the amount of active proteasomes, we used an irreversible inhibitor of the proteolytic active sites of the proteasome that is conjugated to the fluorescent molecule Bodipy (BodipyFL-Ahx₃LeuVS) (Gan et al, 2019). Fluorescent-labeling of the proteasome active sites, which can be detected after separation of the lysates by SDS–PAGE, thus indicates active proteasomes. Sciatic nerve lysates from WT and S63del mice were incubated with BodipyFL-Ahx₃LeuVS for 30, 60, and 120 min. Bodipy-labeling of the active proteasome subunits increased linearly over time (Fig 4E), and was blocked by co-incubation with the proteasome inhibitor Carfilzomib (Fig 4E). S63del sciatic nerve lysates showed greater labeling at all time points than WT (Fig 4E), indicating a greater amount of active proteasome in S63del. This further shows proteasome dysfunction in S63del because an increased number of active proteasomes (Fig 4E) produced a similar amount of peptide hydrolysis (Fig 4A and B) as WT, which had a lesser number of active proteasomes (Fig 4E).

Lysates of sciatic nerves from S63del//PA200−/− mice showed even greater bodipy-labeling of proteasome proteolytic subunits than S63del (Fig 4F). Analysis of the S63del and S63del//PA200−/− samples by Native PAGE showed increased bodipy-labeling in S63del//PA200−/− of 26S proteasomes and no change in labeling of 20S proteasomes (Fig 4G). Thus, S63del//PA200−/− sciatic nerves have a greater amount of assembled, active 26S proteasomes than S63del. This increased level of active proteasomes, seen in S63del//PA200−/− but not in PA200−/−, is likely responsible for the increased proteasomal degradation of proteins and hydrolysis of peptides in the sciatic nerves of S63del//PA200−/− mice.

S63del//PA200−/− sciatic nerves have less proteotoxic stress than S63del

The knockout of PA200 in S63del increased proteasomal peptidase activity (Fig 4) and proteasomal degradation of proteins (Fig 3) in sciatic nerves. We’ve shown previously that increasing proteasome function in S63del by pharmacologically raising cGMP and stimulating phosphorylation of 26S proteasomes by PKG reduces the biochemical indicators of the proteotoxicity seen in S63del neuropathy (VerPlank et al, 2022). We therefore examined whether the increased 26S proteasome activity seen here in S63del by the knockout of PA200 also reduced proteotoxicity.

Polyubiquitinated proteins accumulate in the Schwann cells in the sciatic nerves of S63del mice (Fig 5A) and (VerPlank et al, 2018) because of their reduced degradation by the 26S proteasome (Fig 3) and (VerPlank et al, 2018). In lysates of sciatic nerves of S63del//PA200−/− mice, the levels of polyubiquitinated proteins, included those linked through Lysine48 which target proteins for degradation by the 26S proteasome, were less than in S63del–similar to levels seen in WT (Fig 5A). This decrease in polyubiquitinated proteins in S63del//PA200−/− is most likely because of their increased degradation by the greater number of active 26S proteasomes (Fig 4D, F, and G).

The expression of the mutant MPZ causes a canonical UPR in the Schwann cells of S63del mice (Pennuto et al, 2008) and consequently, the levels of p-PERK, p-eif2α, and BiP are greater in sciatic nerve lysates of S63del mice than in WT littermates (Fig 5B) and (Pennuto et al, 2008). In S63del//PA200−/− sciatic nerve lysates, the levels of these three markers of the UPR are decreased (Fig 5B). The reduction in S63del//PA200−/− of polyubiquitinated proteins, p-PERK, p-eif2α, and BiP indicates a lessening of the proteotoxicity in Schwann cells caused by the expression of mutant MPZ.

The deletion of PA200 rescues the morphological and functional deficits seen in the peripheral nerves of S63del mice

S63del mice, like human patients with CMT1B, have thinner myelin sheaths around the axons of peripheral nerves (Fig 6A) (Wrabetz et al, 2006). To determine whether in S63del//PA200−/− mice the increased proteasomal protein degradation and reduced proteotoxic stress ameliorated this deficiency in myelination, the amount of myelin per axon was measured via electron microscopy of thin sections of sciatic nerves from WT, PA200−/−, S63del, and S63del//PA200−/− mice. Myelin thickness is quantified as g-ratio, which is defined as the axon diameter divided by the fiber diameter. The higher the g-ratio, the thinner the myelin sheath relative to the axon diameter. In S63del, the average g-ratio is higher than in WT (Fig 6B), indicating hypomyelination. S63del//PA200−/− mice had an average g-ratio that was lower than that of S63del and comparable to that of WT (Fig 6B). This return in myelin thickness to WT levels in S63del//PA200−/− sciatic nerves was seen across all axon diameters (Fig 6C and D). Western blot analysis of the sciatic nerve lysates for the myelin proteins PMP22 and MBP showed less of these proteins in S63del than in WT (Fig 6E), and in S63del//PA200−/−, similar levels to WT (Fig 6E), mirroring the changes in myelin thickness (Fig 6A–D), as expected. The percentage of myelinated axons per diameter was unchanged across all four genotypes (Fig 6F). PA200−/− mice exhibited an average g-ratio and levels of myelin proteins that were comparable to that of WT (Fig 6B and E).

S63del mice model another aspect of the pathology seen in human patients with CMT1B: amyelinated fibers (red arrow in Fig 6A) (Wrabetz et al, 2006). These axons meet the requirements to be myelinated–are ensheathed by a single Schwann cell and have a diameter greater than 1 μm–but are not myelinated. Very few amyelinated fibers were present in the sciatic nerves of WT and PA200−/− mice (<1% of all myelinated axons) (Fig 6G), but more were found in S63del (∼4% of all myelinated axons) (Fig 6G). The sciatic nerves of S63del//PA200−/− mice had less amyelinated fibers than S63del (∼1% of all myelinated fibers) (Fig 6G).

The return of myelin thickness in S63del//PA200−/− mice to WT levels and the reduction in amyelinated fibers prompted us to investigate whether the function of the nerves was also restored. S63del mice, like patients with CMT1B, have reduced motor function because of slowing of nerve conduction velocities, prolongation of distal latencies, and prolongation of F-wave latencies that are caused by uniform hypomyelination (Fig 7A–F). In S63del//PA200−/− mice at 1 mo of age, the age at which all above analyses were performed, nerve conduction velocities (Fig 7A), distal latencies (Fig 7B), and F-wave latencies (Fig 7C) were indistinguishable from WT, indicating a restoration of the nerve conduction defects seen in S63del. At 6 mo of age, these same measures of electrical conduction in S63del//PA200−/− mice remained quicker than in S63del and comparable to WT (Fig 7D–F). Thus, the restored function in the peripheral nervous system of S63del//PA200−/− mice, which is dependent on myelin thickness, is seen at 1 mo of age and maintained at 6 mo of age.

Animal models

All experiments involving animals followed protocols approved by the Institutional Animal Care and Use Committee of University at Buffalo and Roswell Park Cancer Institute. Mice were housed under specific pathogen-free conditions at 70°F, 50% room humidity, 12-h light/12-h dark cycle and received ad libitum access to water and food. All mice were in a congenic C57BL/6 background and both sexes were used equally. S63del (Wrabetz et al, 2006) and PA200−/− (Khor et al, 2006) transgenic mice have been described previously. Genotyping of transgenic mice was performed by PCR using QuickLoad PCR mix (New England Biolabs) on tail genomic DNA extracted with phenol chloroform. The following three primers were used in a single reaction to detect the PA200wt allele (380 bp) and the PA200null allele (500 bp): PA200wt-AS: 5′-GTT GTT TGT TAG TTG TCA GGC TC-3′; Common-S: 5′-CCA CCA TCT AGG TTA AAG GT-3′; PA200null-AS: 5′-CCG CTC GAG GGC AGT ACA GTC TTA CT-3′. All experiments for Figs 1–6 were performed on sciatic nerves dissected from mice at ages between p28 and p32. The EMG experiments in Fig 7A–C were performed on mice at ages between p28 and p32. The EMG experiments in Fig 7D–F were performed on mice that were 6 mo old.

RNA isolation and qRT-PCR

Nerves were first pulverized in liquid nitrogen. TRIzol (Thermo Fisher Scientific) was added (400 μl). The mixture was then passed several times through a 26-gauge needle. Chloroform (80 μl) was added, and the mixture was centrifuged at 12,000g for 15 min at 4°C. The aqueous phase (180 μl) was isolated, combined with 200 μl of isopropanol, 38 μl 3M sodium acetate pH 5.5, and 1 μl of glycogen (20 mg/ml), and then incubated for 1 h at −80°C. The precipitated RNA was collected by centrifugation at 12,000g for 10 min at 4°C and then washed four times by the addition of 1 ml 75% ethanol followed by centrifugation at 7,600g for 5 min at 4°C. The pellets were resuspended in 10 μl DEPC water and the concentration of RNA was measured on a NanoDrop 2000c (Thermo Fisher Scientific). The RNA (0.7 μg) was reverse transcribed to cDNA using the SuperScript III First Strand Synthesis System (Invitrogen), according to the manufacturer’s instructions.

TaqMan qRT-PCR was performed on a Bio-Rad CFX96/384 real-time PCR machine with the following probes: Psme4, Mm01193108_m1 and GAPDH, Hs99999915_g1. Each reaction had 10 μl of 2X Taqman Universal PCR Master Mix, 1 μl of probe, 8 μl of DEPC water, and 1 μl of cDNA. Data were analyzed using the threshold cycle (Ct) and 2(^−ΔΔCt) methods.

Lysis for immunoblotting

Sciatic nerves were flash frozen and pulverized with mortar and pestle in liquid nitrogen. The pulverized tissues were suspended in CHAPS lysis buffer (25 mM HEPES-KOH, 150 mM NaCl, 1 mM EDTA, and 0.3% CHAPS) containing the following: 10 mM N-Ethylmaleimide, 1 mM NaF, 1 mM Na₃VO₄, 1 mM PMSF, and 1 μM Bortezomib. The crude lysates were rotated for 15 min at 4°C, sonicated (three times, 10 s on–30 s off, 25% power, Bandelin HD 2200) and centrifuged at 14,000g for 10 min at 4°C. Protein in the supernatant was quantified with the detergent-compatible Bradford Assay (Life Technologies). Lysates were separated by Bis–Tris SDS–PAGE (Genscript) and transferred to either nitrocellulose (Protran; VWR) or PVDF (Immobilin FL; EMD Millipore) membranes. Immunoblotting was performed using the following antibodies per manufacturer’s instructions: rb anti-ubiquitin (P4D1) (sc-8017), ms anti-PSMD1 (Rpn2) (sc-271775), ms anti-PSMB5 (β5) (sc-393931); rb anti-PSMA6 (α1) (A303-809), rb anti-PSMD2 (Rpn1) (A303-854) (Bethyl Laboratories); rb anti-PSMA7 (α4) (15219-1-AP), rb anti-PSMD11 (Rpn6) (14786-1-AP) (ProteinTech); ms anti-eIF2α (2103), rb anti-phospho eIF2α (3398), rb anti-K48-ubiquitin (8081), rb anti-PERK (3192), rb anti-p-PERK (3179) (Cell Signaling); rb anti-GRP78 (BiP) (NBP1-06274), rb anti-Beta-Tubulin (NB600-936) (Novus); rb anti-PA200 (PA1-1961), ms anti-LaminB1 (33–2000) (Invitrogen); rb anti-GAPDH (G9545), rb anti-PMP22 (SAB4502217) (Sigma-Aldrich); ms anti-MBP (836504) (BioLegend). When detecting ubiquitin, the nitrocellulose membranes were autoclaved after transfer. Visualization was performed with ECL, ECL Prime (GE Healthcare), or Odyssey CLx infrared imaging system (LiCor), and quantification was performed with ImageJ (National Institutes of Health) or ImageStudio (LiCor) software.

Generation of cytosolic and nuclear fractions from sciatic nerve

Cytoplasmic and nuclear fractions were produced as previously described (VerPlank et al, 2020), with minor changes to the compositions of the buffers. Briefly, nerves were pulverized in liquid nitrogen and suspended in STM buffer (50 mM Tris–HCl pH 7.5, 250 mM sucrose, 5 mM MgCl₂, 1 mM ATP, 1 mM DTT, 1 mM PMSF, and 1 mM NaF). The crude lysate was centrifuged at 800g for 15 min at 4°C. The supernatant (cytosolic fraction) was transferred to a new tube and the pellet (nuclear fraction) was resuspended in STM buffer, vortexed, sonicated, and centrifuged at 500g for 15 min at 4°C. The pellet was washed in STM buffer, collected by centrifugation at 1,000g for 10 min at 4°C, and then resuspended in NET buffer (25 mM HEPES-KOH pH 7.5, 150 mM NaCl, 5 mM MgCl₂, 10% glycerol, 1 mM ATP, 1 mM DTT, 0.1% NP-40, 1 mM PMSF, and 1 mM NaF). The crude nuclear fraction was incubated for 30 min on ice, sonicated, and then centrifuged at 9,000g for 30 min at 4°C. The supernatant was collected as the nuclear fraction. The cytosolic fraction was centrifuged at 10,000g for 10 min at 4°C to remove the mitochondria. NaCl was added to the cytosolic fractions to a final concentration of 150 mM, to be equimolar with the nuclear fractions. All fractions were normalized to the amount of total protein by detergent-compatible Bradford assay.

Immunoprecipitation of proteasomes from sciatic nerve lysates

Sciatic nerves were flash frozen, pulverized in liquid nitrogen, suspended in APB buffer (25 mM HEPES-KOH pH 7.5, 150 mM NaCl, 5 mM MgCl₂, 1 mM ATP, 1 mM DTT, and 10% glycerol) plus 1 μM Bortezomib (Enzo), sonicated (three times, 10 s on–30 s off, 25% power, Bandelin HD 2200), and centrifuged at 14,000g for 10 min at 4°C. Protein concentration of the clarified lysates was determined by Bradford assay and equal amounts of total protein per genotype were pre-cleared with Protein A/G Plus agarose beads (Santa Cruz Biotechnology) for 30 min at 4°C. Rabbit anti PSMA6 (Bethyl Laboratories) was added to the supernatant at a final concentration of 2 μg/ml and the samples rotated end over end for 2 h at 4°C. Protein A/G Plus agarose beads, washed 3x in APB buffer, were added and the mixture was rotated end over end for 1 h at 4°C. The beads were then collected by centrifugation and washed 3X in APB buffer for 5 min each time. The immunoprecipitates were eluted by boiling for 5 min in 2X Laemmli sample buffer (120 mM Tris–HCl pH 6.8, 4% SDS, 200 mM DTT, 0.002% xylene cyanol).

Morphological analysis of sciatic nerves by electron microscopy

Sciatic nerves were prepared as previously described (VerPlank et al, 2022). Ultrastructural images of thin sections of sciatic nerves were acquired on a FEI Tecnai G2 Spirit BioTWIN electron microscope. These images were used for the quantification of g-ratio and axon distribution. At least 15 images and 65 axons were quantified per mouse. The analysis of amyelinated fibers was performed using Image J software on semithin images of sciatic nerves. Seven hundred and fifty fibers were counted per mouse.

Electrophysiology

Mice were anesthetized with 2,2,2 tribromoethanol (Avertin; Sigma-Aldrich) in H₂O and placed under a heating lamp to avoid hypothermia. Nerve conduction velocities, distal latencies, and F-wave latencies were obtained with steel monopolar needle electrodes. To stimulate at distal, proximal, and spinal sites along the nerve, a pair of stimulating electrodes was inserted subcutaneously near the nerve at the ankle; a second pair at the sciatic notch, and a third over the dorsum of the spine.

Ex vivo sciatic nerve protein degradation measurements

The degradation rate of short- and long-lived proteins in sciatic nerves ex vivo was performed as described previously (VerPlank et al, 2018). Briefly, segments of equal length were dissected from the right and left sciatic nerves of each animal. To measure the degradation of long-lived proteins, the segments were incubated for 18 h at 37°C in 5% CO₂ in DMEM containing 10% FBS and 100 ng/ml nerve growth factor (culture media) and 5 μCi/ml L-[3,5³H]Tyrosine (Perkin Elmer). The segments were then incubated for 2 h in chase media (culture media containing 2 mM cold tyrosine) to exclude short-lived proteins from the analysis. At the end of the 2-h chase, one nerve segment per animal was transferred to fresh chase media. The other was collected and the TCA-insoluble radioactivity (i.e., radiolabeled proteins) was measured per sample on a Tri-Carb 2800TR scintillation counter (Perkin Elmer). Aliquots of media were taken at indicated times and the TCA-soluble radioactivity (i.e., radiolabeled amino acids and small peptides) was measured. Proteolysis was expressed as the amount of TCA-soluble radioactivity released over a time as a percentage of input TCA-insoluble radioactivity. To measure the degradation of long-lived proteins by the UPS or the autophagy–lysosome system, 5 μM bortezomib or 150 μM chloroquine was included in the chase media.

To measure the degradation of short-lived proteins, the following changes to the above protocol were made: nerve segments were incubated for 4 h in culture media at 37°C in 5% C02 before the pulse with 5 μCi/ml L-[3,5³H]Tyrosine, the time of pulse was reduced to 30 min, and after the pulse the nerve segments were washed 3x with chase media.

Native gel electrophoresis

For analysis of sciatic nerve lysates in Figs 1 and 2, sciatic nerves were pulverized in liquid nitrogen, suspended in APB, lysed by sonication, and centrifuged at 14,000g for 10 min at 4°C. 4% acrylamide resolving gels were prepared (90 mM Tris–HCl pH 7.5, 90 mM Boric Acid, 0.5 mM EDTA, 5 mM MgCl₂, 1 mM DTT, 1 mM ATP, 10% glycerol, 0.1% APS, 0.01% TEMED) and then overlaid with 2.5% acrylamide stacking gels (90 mM Tris–HCl pH 7.5, 90 mM Boric Acid, 0.5 mM EDTA, 5 mM MgCl₂, 1 mM ATP, 0.1% APS, 0.01% TEMED). Equal volumes of lysates that were normalized by protein concentration were added and electrophoresis was performed in a TBE-based buffer (90 mM Tris–HCl pH 7.5, 90 mM Boric Acid, 0.5 mM EDTA, 5 mM MgCl₂, 1 mM DTT, 1 mM ATP) at 145 V for 3–4 h at 4°C. The gels were then soaked for 15 min in a buffer pre-chilled to 4°C containing 25 mM Tris, 190 mM glycine, and 0.1% SDS and then transferred overnight at 15 V at 4°C to PVDF membranes in transfer buffer (25 mM Tris and 190 mM glycine) containing 1% MeOH.

For analysis of assembled proteasomes following subcellular fractionation, 3.5% acrylamide gels were prepared (90 mM Tris–HCl pH 7.5, 90 mM Boric Acid, 0.5 mM EDTA, 10% glycerol, 0.1% APS, 0.01% TEMED) without a stacking gel. Cytosolic and nuclear fractions were loaded into the gels, and electrophoresis was performed in a TBE-based buffer (90 mM Tris–HCl pH 7.5, 90 mM Boric Acid, 0.5 mM EDTA, 5 mM MgCl₂, 1 mM DTT, 1 mM ATP) at 145 V for 3–4 h at 4°C. Western analysis was performed as described above.

In Fig 4, sciatic nerves were pulverized in liquid nitrogen, suspended in APB containing 0.1% NP-40, rotated end over end for 15 min at 4°C, sonicated (three times, 10 s on—30 s off, 25% power, Bandelin HD 2200), and centrifuged at 14,000g for 10 min at 4°C. Lysates were loaded into 3–8% tris-acetate gels (Thermo Fisher Scientific) and run at 145 V for 3–4 h at 4°C in a running buffer containing: 25 mM Tris, 192 mM Glycine, 5 mM MgCl2, 1 mM ATP, 0.5 mM DTT. The gels were transferred to PVDF overnight at 4°C in the following transfer buffer: 48 mM Tris, 39 mM Glycine (pH 9.0), 0.03% SDS. After transfer, the PVDF membranes were incubated in a standard transfer buffer (25 mM Tris,190 mM glycine) with 20% MeOH for 20 min and then washed in TBS before blocking.

Measuring peptidase activity of the proteasomes in sciatic nerve lysate

Proteasome peptidase activity in sciatic nerve lysates was measured as described previously (VerPlank et al, 2018). For chymotrypsin-like peptidase activity, 20 μM Suc-LLVY-amc (Enzo) was used, and for caspase-like peptidase activity, 20 μM AC-nLPnLD-amc (Enzo) was used. Proteasome-specific cleavage was calculated by subtracting the overall rate of peptide hydrolysis in the lysates by the rate of hydrolysis measured in the presence of 5 μM Carfilzomib (Cayman Chemical Company).

Detecting levels of active proteasomes with Me4BodipyFL-Ahx3LeuVS

Sciatic nerves were lysed as described above for analysis by native gel electrophoresis in Fig 4. The following 3X reaction mixture was prepared on ice with and without 3 μM Carfilzomib: 150 mM Tris–HCl pH 7.5, 120 mM KCl, 15 mM MgCl2, 3 mM ATP, 3 mM DTT, 0.15 mg/ml BSA, and 0.75 μM Me4BodipyFL-Ahx3LeuVS. The mixture was added to the lysates to a final concentration of 1X and the reaction took place at 37°C, shaking at 225 rpm, for the indicated amounts of time and was terminated by the addition of Laemmli buffer and boiling for 5 min. SDS–PAGE was performed, and the gels were imaged on a VersaDoc (Bio-Rad) on the FITC setting.

For analysis by native PAGE, the experiment was performed as above, but the samples were not denatured by the addition of Laemmli buffer and boiling. Instead, the samples were loaded in 3–8% tris-acetate gels and electrophoresis was performed as described above in Fig 4.

Glycerol gradient centrifugation

20–50% glycerol gradients were prepared in a volume of 2 ml in a base buffer of 25 mM HEPES-KOH pH 7.4, 5 mM MgCl₂, 1 mM DTT, and 1 mM ATP. Sciatic nerve lysates were prepared as described above for the immunoprecipitation of proteasomes and layered on top of the gradients. Centrifugation was performed at 55,000g for 3 h at 4°C in a SW41i rotor in a Beckman Optima TL centrifuge. The gradients were then displaced upwards by Fluorinert and collected as 19 fractions of equal volume.