PMP22-C3 Tg (hetero) mice [B6.Cg-Tg(PMP22)C3Fbas/J] were purchased from Jackson Laboratory (Bar Harbor, ME, USA) with permission for breeding and genetical modifying. They were bred in Axcelead Drug Discovery Partners, Inc. (ADDP, Kanagawa, Japan) and male PMP22-C3 mice and age-matched wild-type (WT) littermates were used. Humanized mice [hPMP22-C3 Tg/mPmp22 KO (ho/ho); homo-humanized, hPMP22-C3 Tg/mPmp22 KO (he/ho); hetero-humanized] were generated by crossing the PMP22-C3 mouse line with mPmp22 KO mouse line that lacks the full length (31 kbp) mPmp22 gene using the CRISPR/Cas9 method and supplied by ADDP. In detail, hPMP22-C3 Tg /mPmp22 KO (he/he) mice were initially generated by crossing hPMP22-C3 Tg (hetero) mice with mPmp22 KO (homo) mice using In Vitro Fertilization method. Then hPMP22-C3 Tg /mPmp22 KO (he/he) mice were crossed with hPMP22-C3 Tg /mPmp22 KO (wt/he) mice to generate hPMP22-C3 Tg /mPmp22 KO (he/ho) mice. Consequently, hPMP22-C3 Tg /mPmp22 KO (he/ho) mice were bred to generate hPMP22-C3 Tg /mPmp22 KO (ho/ho) mice. The ratio of these mouse lines was broadly consistent with Mendel’s Laws. Genotyping analysis using PCR method (50 °C for 2 min, then 95 °C for 20 s, following 40 cycles of 95 °C, 1 s and 60 °C, 20 s) was conducted to identify the genotypes of each mouse. The following primer sets were used: mPmp22 (Forward: 5′-GCGTCCTCTTTCATGAATCAAAA, Reverse: 5′-TCTACCCCTAACCTCTGGTTCCT, Probe: 5′-CATTCACGAATTGCC; Thermo Fisher Scientific, Waltham, MA, USA), hPMP22 (Forward: 5′-GTGCTGCGGCCATCTACAC, Reverse: 5′-GTAGGCGAAACCGTAGGAGTAATC, Probe: 5′-CCGGAGTGGCATCT; Thermo Fisher Scientific). The phenotypic characteristics of these two humanized mice were stable across generations. As the control group for humanized models, age-matched C57BL/6 J Jcl mice (WT), the background mouse line of PMP22-C3 mice and mPmp22 KO mice were purchased from CLEA Japan, Inc. (Tokyo, Japan). The mice were housed in groups with unrestricted access to food and water and maintained under a 12-h light/dark cycle (lights on at 7:00 a.m.) at a temperature of 22 ± 1 °C. To study the natural history of NCV and CMAP, mice were subjected to continuous NCV and CMAP recording at 6, 10, 14, and 18 weeks of age from the same individuals (n = 8–10/group). The mice were then sacrificed, and histopathological analysis and evaluation of mRNA expression were performed. Behavioral assessments and electrical impedance myography (EIM) measurements were performed at 17 and 18 weeks of age, respectively (n = 8–10/group). Another group of 14-week-old humanized and PMP22-C3 mice was used for the PMP22 protein assay (n = 4/group). The overall experimental schedule is shown in Figure 1. All experimental procedures in this study were conducted in accordance with animal care protocols approved by the Institutional Animal Care and Use Committee of Shonan Health Innovation Park (protocol number: AU-00031034 and AU-00031278), in compliance with Act on Welfare and Management of Animals (Act No. 105 of October 1, 1973), harmonized with the American Association for Accreditation of Laboratory Animal Care (AAALAC) International.

Mice were anesthetized via intraperitoneal injection of a combination of anesthetics (medetomidine, 0.3 mg/kg; midazolam, 4 mg/kg; and butorphanol, 5 mg/kg) and then euthanized by exsanguination. The left sciatic nerve was removed and homogenized using a FastPrep-24 5G homogenizer (MP Biochemicals LLC, Irvine, CA, USA) in Lysing Matrix I tubes (6918–100, MP Biochemicals LLC) filled with QIAzol Lysis Reagent (79,306, Qiagen, Hilden, Germany). Total RNA was extracted using the RNeasy Plus Universal Mini Kit (73,404, Qiagen) according to the manufacturer’s instructions. Total RNA (0.1 μg) was used for complementary DNA synthesis using SuperScript IV VILO Master Mix (11,756,500, Thermo Fisher Scientific). RT-qPCR was conducted using QuantStudio (Thermo Fisher Scientific) with TaqPath qPCR Master Mix, CG (A15297; Thermo Fisher Scientific). Relative copy numbers of mPmp22 and hPMP22 were calculated. The mRNA expression levels of myelin-related genes were normalized to those of Gldn, a Schwann cell-specific gene, and compared with each WT group. The following primer sets were used: mPmp22 (Forward: 5′-CTTCCAAATCCTTGCTGGTCTG, Reverse: 5′-GGATGTAGGCGAAGCCATAGG, Probe: 5′-ATGCCACTCACTGTGCCTCACTGTGTAGAT; Sigma-Aldrich, Saint Louis, USA), hPMP22 (Forward: 5′-CTTGCTGGTCTGTGCGTGAT, Reverse: 5′-ACCGTAGGAGTAATCCGAGTTGAG, Probe: 5′-CATCTACACGGTGAGGCACCCGG; Sigma-Aldrich), Sc5d (Mm. PT.58.33540072; IDT, San Jose, CA, USA), Mpz (Mm. PT.58.5771188; IDT), Pou3f1 (Mm. PT.58.33607006. g; IDT), Gldn (Mm. PT.58.43925599; IDT).

After euthanasia by exsanguination, the left sciatic nerve was removed. 2.5% homogenate of sciatic nerve was homogenized by Precellys Evoluation (Beritin, France) with lysis buffer including 50 mM Tris pH 7.5, 4% sodium dodecyl sulfate, 150 mM NaCl, Complete (04693132001, Roche, Switzerland), and PhosSTOP (04906837001, Roche). After incubation at room temperature (20–25 °C) for 30 min, the samples were subjected to ultrasonication at 40 kHz for 10 min and then incubated at 95 °C for 10 min using an Eppendorf ThermoMixer C (Eppendorf, Germany). After cooling to room temperature, the samples were centrifuged at 21,600 × g for 10 min at room temperature. The protein concentration in the supernatant was measured using a bicinchoninic acid assay. Sciatic nerve (7.5 μg) was mixed with 1 M phosphoric acid buffer (10 μL), 5 M NaCl (10 μL), and cold acetone (0.5 mL) at −30 °C for 2 h. After centrifugation (21,600 × g, 4 °C, 20 min), the supernatant was completely removed and the precipitant was kept at room temperature for 15 min. The precipitant was dissolved in 8 M urea (50 μL), 0.5 M Tris(2-carboxyethyl) phosphine (5 μL), and 0.5 M iodoacetamide (5 μL) and incubated for 15 min at room temperature. This was followed by sonication for 10 min and a subsequent incubation for 10 min at 37 °C. Next, 800 μL of 0.1 M Tris–HCl (pH 8.0) was added, and proteolytic digestion was performed using 12.5 μg of chymotrypsin (Promega, WI, USA) overnight at 25 °C. The resulting sample was mixed with 4% phosphoric acid (200 μL) and 100 mg/mL of internal standard peptides (10 μL), then purified via solid-phase extraction using Oasis MCX cartridges (Waters, MA, USA). The final eluted solution was dried by nitrogen gas. The residue was reconstituted in 200 μL of a formic acid/acetonitrile/purified water (1,40,960, v/v/v), and an aliquot of 0.2 μL was injected into an EASY-nLC 1,200 system/Q Exactive HF-X Orbitrap mass spectrometer (Thermo Fisher Scientific). The signature peptides were monitored in parallel reaction monitoring mode. The detailed analytical conditions are listed in Supplementary Tables S1, S2.

NCV and CMAP were recorded from the right sciatic nerve of WT and humanized mice using a Natus UltraPro S100 (Natus Medical Inc., Middleton, WI, USA). Disposable monopolar needle electrodes (TECA^® Elite, 902-DMF25-TP (28G), Natus Medical Inc.) were used for stimulating and recording. Mice were anesthetized with 1.5–2.0% isoflurane on a heating pad to maintain body temperature. To record the NCV and CMAP, a sensing recording electrode was inserted into the tibialis anterior muscle, and the reference electrode was inserted into the ipsilateral plantar muscle. For stimulation, the stimulating cathodes were subcutaneously placed at the sciatic nerve, 4 mm (distal) and 16 mm (proximal) from the sensing recording electrode. The anode was inserted into the contralateral plantar muscle. Square-wave pulses of 0.1 ms duration were delivered to the distal and proximal sites at a 0.5 Hz frequency, and each CMAP was measured at supramaximal stimulation. The NCV was calculated from the difference in peak latencies between the distal and proximal CMAP responses and the distance (12 mm) between the stimulating electrodes. The CMAP evoked by proximal stimulation was used to analyze the CMAP amplitude.

The right sciatic nerve was immersed in 4% paraformaldehyde phosphate buffer solution (163–20,145, FUJIFILM Wako Pure Chemical Corp., Osaka, Japan) for 5–7 days at 4 °C. The fixative was replaced with 0.01 mol/L phosphate-buffered saline and stored at 4 °C until the histopathological analysis. For morphometric assessment of axons and myelin, the sciatic nerve was embedded in resin and sectioned transversely (0.5 μm thick), and stained with toluidine blue (104,172, Merck, Darmstadt, Germany). Digital images were obtained using a digital slide scanner (NanoZoomer S60; Hamamatsu Photonics, Shizuoka, Japan), and the inner and outer areas/diameters of the axon and myelin were subsequently analyzed using HALO image analysis software (Indica Labs, Albuquerque, NM, USA). The g-ratio of each axon was calculated as a ratio between axon outer diameter and inner diameter.

Blood samples were collected from the jugular vein of the mouse, which was gently restrained by hand, at 9 and 14 weeks of age, and from the abdominal vena cava of the mouse anesthetized with the combination anesthetic described above at 18 weeks of age. Approximately 200 μL of whole blood was collected using a 1 mL syringe with a 27-guage needle coated with 0.1 mol/L EDTA-2Na (344–09255, Dojindo Laboratories, Kumamoto, Japan) and kept on ice. After the centrifugation at 17,400 × g at 4 °C for 1 min, 40 μL of plasma was collected into a Protein LoBind^® Tube (0030108434, Eppendorf), then chilled on dry ice and stored at −80 °C until analysis. NfL levels in plasma were measured using an HD-X analyzer (Quanterix Corporation, Massachusetts, USA) with a Simoa NF-Light™ V2 Advantage Kit (104,073, Quanterix) according to the manufacturer’s instructions.

EIM measurements were performed as previously described (24). Briefly, mice were anesthetized with 1.5–2.0% isoflurane and placed on a heating pad to maintain their body temperature. The left hindlimb was shaved and pretreated with skin preparation gel (Nuprep^®; Weaver and Company, Aurora, CO, USA) to reduce impedance, and positioned at a 45° angle. After application of an electrode cream (SignaCreme^®; Parker Labs, Fairfield, NJ, USA), a four-electrode array with a weight of 20 g was placed over the gastrocnemius (GC) muscle in the longitudinal direction and then, a weak electrical current with multi-frequencies was applied. The EIM parameters (reactance, resistance, and phase) were obtained using an impedance spectroscopy system (mScan-V^™; Myolex Inc., Boston, MA, USA). Measurements were performed three times and averaged within the 1–10,000 kHz range. Values from 10 to 1,000 kHz were used for the analysis to avoid artifacts.

The hindlimbs were removed from each mouse and immersed in 4% paraformaldehyde phosphate buffer solution (163–20,145; FUJIFILM Wako Pure Chemical Corp., Osaka, Japan). After 24 h, the fixative was replaced with 0.01 mol/L phosphate-buffered saline and stored at 4 °C. To assess muscle pathology, GC muscle tissues were assessed using laminin immunohistochemical staining as previously described (24). Digital images were obtained using a digital slide scanner (NanoZoomer S60; Hamamatsu Photonics, Shizuoka, Japan), and the cross-sectional area (CSA) and muscle fiber count were subsequently analyzed using HALO image analysis software (Indica Labs).

Statistical analyses were performed using the SAS software (version 9.4; SAS Institute Japan, Tokyo, Japan). For multiple comparison for PMP22 mRNA and protein levels, one-way ANOVA with a post hoc Bonferroni’s test was utilized. To compare two groups genetic differences between humanized or PMP22-C3 and WT mice, data were assessed using Student’s t-test for homogeneous data or Mann–Whitney U test for nonhomogeneous data. For the analysis of repeated measures data of NAV, CMAP, and plasma NfL, two-way Repeated Measures ANOVA with a post hoc Bonferroni’s test was performed. Statistical significance was set at p < 0.05. Data were reported as the mean ± standard error of the mean (SEM).

To confirm the genetic characteristics of humanized mice, we investigated the mRNA expression of mPmp22 and hPMP22 in the sciatic nerves of WT, humanized, and PMP22-C3 mice at 18 weeks of age using RT-qPCR. mPmp22 was not detected in hetero-humanized or homo-humanized mice, whereas WT and PMP22-C3 mice showed similar levels of intrinsic mPmp22 expression (Figure 2A). The expression level of hPMP22 in hetero-humanized mice was similar to that in PMP22-C3 mice. In contrast, homo-humanized mice showed higher levels of hPMP22 expression than PMP22-C3 or hetero-humanized mice (2.5-fold increase compared to PMP22-C3 mice). Consistent with the mRNA expression of mPmp22 and hPMP22, hetero- and homo-humanized mice showed no expression of mPmp22 protein, and homo-humanized mice exhibited a higher level of hPMP22 protein than PMP22-C3 mice or hetero-humanized mice (2.3-fold increase compared to PMP22-C3 mice) (Figure 2B). These data confirmed that humanized mice had no mPmp22 gene or protein and that gene dosage-dependent hPMP22 mRNA and protein expression was observed in the sciatic nerve.

Transcriptional changes in several myelin-related genes, such as Sc5d, Mpz, and Pou3f1, were investigated in the sciatic nerves of humanized and PMP22-C3 mice because changes in the transcript levels of these genes have been reported in the CMT1A rodent model (19). The mRNA expression of these genes in the sciatic nerve was assessed using RT-qPCR at 18 weeks of age. In homo-humanized mice, the transcript levels of Sc5d and Mpz were decreased, and Pou3f1 was increased relative to those in WT mice (Sc5d: 0.46-fold reduction, Mpz: 0.58-fold reduction, Pou3f1: 3.2-fold increase; p < 0.005) (Figure 3A), consistent with those in PMP22-C3 mice (Figure 3C) and a previous report (19). In contrast, hetero-humanized mice showed transcript levels of these genes comparable to those in WT mice (Figure 3B). These data suggest that the mRNA expression levels of myelin-related genes changed depending on PMP22 levels in the sciatic nerve of humanized mice.

Patients with CMT1A and animal models of this disease exhibit electrophysiological impairment of peripheral nerves due to demyelination (9, 15, 21, 22). To investigate functional changes in the peripheral nerves, NCV and CMAP were repeatedly measured in humanized and PMP22-C3 mice at 6, 10, 14, and 18 weeks of age. Individual values of NCV and CMAP are shown in Supplementary Table S3. Robustly slower NCV (6.8, 7.1, 8.5, and 9.7% of WT mice at 6, 10, 14, and 18 weeks of age, respectively; p < 0.01 or 0.001) and reduced CMAP amplitude (21.7, 24.0, 20.4, and 32.3% of WT mice at 6, 10, 14, and 18 weeks of age, respectively; p < 0.001) were observed in homo-humanized mice compared to WT mice (Figures 4A,D), reflecting demyelination in the sciatic nerve and indicating an abnormality in the lower motor unit of this model. These results were comparable to those observed in the PMP22-C3 mice (Figures 4C,F). In contrast, hetero-humanized mice showed no differences from WT mice in NCV (102.2, 107.4, 105.5, and 99.7% of WT mice at 6, 10, 14, and 18 weeks of age, respectively) or CMAP (103.0, 99.6, 105.2, and 106.0% of WT mice at 6, 10, 14, and 18 weeks of age, respectively) at all ages (Figures 4B,E). These findings suggest that homo-humanized mice show peripheral nerve dysfunction similar to that in PMP22-C3 mice, which depends on the total amount of PMP22 expression.

As homo-humanized and PMP22-C3 mice showed changes in myelin-related genes and demyelination-related nerve dysfunction in the sciatic nerve, histopathological analyses of the sciatic nerve were performed using these mice at 18 weeks of age. In homo-humanized mice, many demyelinated axons were observed (Figure 5A, middle panel, arrowhead), and quantitative analysis revealed a smaller myelin area (58.3% of WT mice, p < 0.001) than in WT mice (Figure 5B). To further investigate the characteristics of demyelination in homo-humanized mice, the g-ratio of each axon was calculated and analyzed by axon size (Figure 5C). Homo-humanized mice exhibited higher g-ratio, especially in larger axons, consistent with previous reports for PMP22-C3 mice (25, 26). In addition, the axonal inner (54.4% of WT mice, p < 0.001) and outer areas (56.8% of WT mice, p < 0.001) were significantly smaller in homo-humanized mice than in WT mice (Figure 5B), and the distribution of axon size in homo-humanized mice shifted toward smaller compared to that in WT mice (Figure 5D), indicating axonal atrophy in the sciatic nerve of this model. These morphological changes were comparable to those observed in the PMP22-C3 mice (Figure 5A, right panel, E–G).

NfL is considered to reflect axonal damage (27), and patients with CMT1A show increased levels of NfL in the blood (28). As axonal atrophy was observed in the sciatic nerves of homo-humanized and PMP22-C3 mice at 18 weeks of age, we investigated plasma NfL levels in these mice at 18 weeks of age and two other earlier points (9 and 14 weeks of age). Individual values of plasma NfL are shown in Supplementary Table S4. Homo-humanized mice showed significantly higher plasma NfL levels than those in WT mice at all ages (8.3-, 5.3-, and 4.2-fold increases at 9, 14, and 18 weeks of age, respectively; p < 0.001) (Figure 6A), as well as PMP22-C3 mice (Figure 6C), implying that axonal atrophy occurs earlier than 18 weeks of age. In contrast, hetero-humanized mice showed no statistically significant difference from WT mice (Figure 6B).

Motor disabilities such as muscle weakness and impaired balance are common features of patients with CMT1A (29, 30) and mouse models (19, 23, 31, 32). To evaluate whether peripheral nerve dysfunction affects motor ability in the humanized mouse model, we performed three behavioral assessments (grip strength, beam walking, and rotarod tests) at 17 weeks of age.

Homo-humanized mice showed a weaker grip strength in both forelimb (WT: 163.6 ± 3.4 g, homo-humanized: 130.6 ± 3.2 g, p < 0.001) and hindlimb (WT: 58.8 ± 2.4 g, homo-humanized: 14.7 ± 2.4 g, p < 0.001) than those in WT mice (Figure 7A), while hetero-humanized mice showed a similar level of grip strength as WT mice in both forelimb (WT: 164.3 ± 4.9 g, hetero-humanized: 175.0 ± 3.3 g) and hindlimb (WT: 64.4 ± 2.5 g, hetero-humanized: 70.2 ± 2.3 g) (Figure 7B).

In the beam walking test that assesses the motor balance and gait, WT mice typically crossed the beam in approximately 12 s with less than one slip, while homo-humanized mice took longer to cross the beam (WT: 12.6 ± 1.2 s, homo-humanized: 24.7 ± 1.9 s, p < 0.001) (Figure 7D) with many slips (WT: 0.40 ± 0.16, homo-humanized: 11.3 ± 1.8, p < 0.001) (Figure 7G), indicating the abnormal motor balance and gait. In contrast, no difference was observed in crossing time (WT: 11.2 ± 1.1 s, hetero-humanized: 10.6 ± 1.1 s) (Figure 7E) or slip counts (WT: 0.38 ± 0.18, hetero-humanized: 0.63 ± 0.26) (Figure 7H) between WT mice and hetero-humanized mice.

The motor coordination was tested using a rotarod. Homo-humanized mice dropped off the rotating rod much earlier than WT mice (WT: 193.3 ± 10.5 s, homo-humanized: 61.1 ± 9.0 s, p < 0.001) (Figure 7J). In contrast, hetero-humanized mice showed a latency comparable to that of WT mice (WT: 199.8 ± 14.3 s, hetero-humanized: 187.1 ± 11.2 s) (Figure 7K).

Similar to homo-humanized mice, PMP22-C3 mice showed weaker grip strength in both forelimbs and hindlimbs, longer crossing time and slipping on the beam, and shorter latency to fall off the rod (Figures 7C,F,I,L). These behavioral findings suggested that homo-humanized mice exhibited motor disabilities similar to those of PMP22-C3 mice, whereas hetero-humanized mice maintained normal motor function. This is consistent with the morphological and functional peripheral nerve characteristics observed in all three mouse lines.

We recently reported that PMP22-C3 mice exhibited muscle weakness and alterations in EIM parameters in the GC muscle innervated by tibial nerve branched off from the sciatic nerve (24). To further confirm that muscle weakness can be detected as changes in muscle composition, EIM data were obtained from the GC muscles of WT, homo-humanized, and hetero-humanized mice at 18 weeks of age. Individual values of EIM parameters are shown in Supplementary Table S5. Homo-humanized mice showed obvious alterations in EIM parameters (Figure 8A), which were consistent with EIM changes in PMP22-C3 mice (24). Statistical comparison using EIM parameters at 50 kHz revealed significant decreases in reactance (79.1% of WT mice, p < 0.05) and phase (69.0% of WT mice, p < 0.001) and an increase in resistance (119.0% of WT mice, p < 0.001) in homo-humanized mice than those in WT mice (Figure 8B). In contrast, hetero-humanized mice showed similar values for all parameters as WT mice (Figure 8C), and no statistical difference in the 50 kHz parameters was detected (Figure 8D). As decreased EIM reactance and phase or increased EIM resistance indicate a reduction in muscle fiber size or muscle mass, respectively (33–35), homo-humanized mice have skeletal muscular deterioration, whereas hetero-humanized mice maintain a healthy muscle state.

The histopathology of the GC muscle of homo-humanized mice was investigated at 18 weeks of age to identify morphological changes in muscle fibers underlying EIM alterations. A higher number of thin muscle fibers were observed in homo-humanized mice than in WT mice (Figure 9A). In WT mice, the CSA histogram of individual muscle fibers showed a wider distribution pattern with a peak at approximately 1,300 μm², while homo-humanized mice shifted the distribution to a right-skewed pattern with a peak at the smallest muscle fiber group in the GC muscle (Figure 9B). This distribution change in muscle fibers led to a significant decrease in the average CSA of the GC muscle of homo-humanized mice compared to that in WT mice (65.0% of WT mice, p < 0.001) (Figure 9C). These data indicate muscle fiber atrophy in the GC muscles of homo-humanized mice, resulting in alterations in EIM parameters.