The leukocyte common antigen-related protein tyrosine phosphatase receptor regulates regenerative neurite outgrowth in vivo

Abstract

Drosophila and leech models of nervous system development demonstrate that protein tyrosine phosphatase (PTP) receptors regulate developmental neurite outgrowth. Whether PTP receptors regulate neurite outgrowth in adult systems or in regenerative states remains unknown. The leukocyte common antigen-related (LAR) receptor is known to be present in rodent dorsal root ganglion (DRG) neurons; therefore, the well established model of postcrush sciatic nerve regeneration was used to test the hypothesis that LAR is required for neurite outgrowth in the adult mammalian nervous system. In uninjured sciatic nerves, no differences in nerve morphology and sensory function were detected between wild-type and LAR-deficient littermate transgenic mice. Sciatic nerve crush resulted in increased LAR protein expression in DRG neurons. In addition, nerve injury led to an increase in the proportion of LAR protein isoforms known to have increased binding affinity to neurite-promoting laminin-nidogen complexes. Two weeks after nerve crush, morphological analysis of distal nerve segments in LAR-deficient transgenic mice demonstrated significantly decreased densities of myelinated fibers, decreased axonal areas, and increased myelin/axon area ratios compared with littermate controls. Electron microscopy analysis revealed a significant twofold reduction in the density of regenerating unmyelinated fibers in LAR-/- nerves distal to the crush site. Sensory testing at the 2 week time point revealed a corresponding 3 mm lag in the proximal-to-distal progression of functioning sensory fibers along the distal nerve segment. These studies introduce PTP receptors as a major new gene family regulating regenerative neurite outgrowth in vivo in the adult mammalian system.

Fig. 1.

LAR protein in DRG after nerve crush. DRG frozen sections obtained from uninjured mice (A) and 2 weeks after nerve crush (B) were immunostained using monoclonal antibody directed against the N terminal of LAR ∼150 kDa ectodomain. Staining patterns from both uninjured and postcrush tissue appear similar and are consistent with previous localizations of LAR to neurons and neural processes. Scale bar, 0.1 mm.C–F, Protein extracts were prepared from lumbar DRG (L4–L6) pooled from eight normal uninjured mice and from eight mice 2 weeks after bilateral nerve crush. Western blot analysis of DRG extracts was performed using the same LAR N-terminal antibody used for tissue immunostaining. Blots were reprobed using antibody directed against the LASE-c insert present in the LAR ∼150 kDa extracellular domain, followed by antibody to β-actin to control for protein loading. E, The ratio of LAR to β-actin signal increased by 66% (n = 6; Western analyses; mean ± SE; *p < 0.05; Mann–Whitney rank sum test). F, In the same samples, the ratio of LASE-c to LAR signal decreased by 38% (n = 6; Western analyses; mean ± SE; *p < 0.05; Mann–Whitney rank sum test).

Fig. 2.

Uninjured sciatic nerve morphology and function in LAR+/+ versus LAR−/− mice. a, Protein extracts were prepared from lumbar DRG (L4–L6) pooled from four LAR+/+ and four LAR−/− uninjured mice. Western blot analysis of DRG extracts was performed using monoclonal antibody directed against the N terminal of the LAR ∼150 kDa ectodomain. Blots were reprobed using antibody directed against β-actin to control for protein loading. Only trace levels of LAR protein were detected in DRG tissue derived from LAR−/− mice. b, Visual examination of 1-μm-thick semithin cross-sections of uninjured sciatic nerve stained with paraphenylenediamine revealed no apparent differences in nerve structure between LAR+/+ and LAR−/− mice. Scale bar, 10 μm.c–e, Quantitative morphological analysis performed on nerves derived from three mice of each genotype showed no detectable differences in myelinated fiber density or in axonal and myelin areas. Mean ± SE. f, Analysis of axonal areas demonstrated no significant difference in size distributions between LAR+/+ and LAR−/− nerves (p = 0.214; Kolmogorov–Smirnov test). g, Pain threshold testing using the hot plate test revealed no detectable differences in temperature reaction latency between LAR+/+ (n = 11 mice) and LAR−/− (n = 8 mice) mice. Mean ± SE.

Fig. 3.

Gross morphology of sciatic nerve 2 weeks after crush. Two weeks after sciatic nerve crush, 1-μm-thick semithin paraphenylenediamine-stained cross-sections obtained from LAR+/+ and LAR−/− mice at the crush site (0 mm) and at 2 mm increments distal to the crush site were examined. Gross inspection suggested decreased density of myelinated fibers in LAR−/− sections at the 2 and 4 mm points. In proximal LAR−/− sections (0 and 2 mm), axonal areas and myelin areas also appeared to be reduced. Scale bar, 10 μm.

Fig. 4.

Sciatic nerve fiber density, axonal area, and myelin area analyses 2 weeks after crush. Two weeks after nerve crush, cross-sections were obtained from LAR+/+ (open circles) and LAR−/− (filled circles) nerves at points 2 mm proximal to the crush site (−2 mm), at the crush site (0 mm), and at 2 mm increments distal to the crush site. Six nerves of each genotype were examined. The number of myelinated fibers per area, axonal areas, and myelin areas were measured in two to four randomly selected fields per section. A, The mean number of myelinated fibers per field was calculated and expressed as the number of fibers per 10⁴ μm². The mean fiber density derived from a given nerve segment and genotype was calculated. Mean ± SE are shown. At the 2 and 4 mm points, significant decreases in fiber density in LAR−/− compared with LAR+/+ sections were found (Mann–Whitney rank sum test). In addition, in LAR+/+ nerves, fiber density at the 2 mm point was increased significantly compared with that at the crush site (p = 0.008) and the −2 mm proximal site (p = 0.049). B, Axonal areas in sections from LAR−/− mice were significantly reduced at the crush site and in all distal sections. Reductions were particularly large at the crush site and at the 2 mm site (*p < 0.001; Mann–Whitney rank sum test). C, Myelin areas in LAR−/− mice were significantly increased at the −2 mm proximal site and significantly decreased at the crush site and all distal sites. This decrease was particularly evident at the crush site and at the 2 mm site (*p < 0.001; Mann–Whitney rank sum test). D, In all sections except the 8 mm distal point, myelin/axon area ratios were increased in LAR−/− mice (*p < 0.001; Mann–Whitney rank sum test).

Fig. 5.

Histogram analysis of axonal areas. Data collected for axonal area analysis was plotted in histogram format over the indicated area ranges. Hatched bars, LAR+/+ values;black bars, LAR−/− values. For eachpanel, the inset shows a cumulative percentage distribution plot of axonal areas. Analysis of axonal area distributions using the Kolmogorov–Smirnov test demonstrates a highly significant loss in the proportion of larger fibers in LAR−/− nerves at the crush site (0 mm) and the 2 mm distal site.

Fig. 6.

Electron microscopy analysis of axonal sprouting. Electron micrographs of LAR+/+ (A) and LAR−/− (B) transverse nerve sections 4 mm distal to the crush site demonstrate typical profiles of unmyelinated fibers (indicated by the asterisks). Characteristic profiles have relatively low electron density, contain abundant microtubule profiles, and are surrounded by a relatively electron-dense axolemma (Jeng and Coggeshall, 1984; Longo et al., 1986; Gibbels, 1989). LAR−/− sections had lower numbers of unmyelinated fibers per field.C, Blinded counts demonstrated a significant decrease in the number of unmyelinated fibers per area in sections 2 mm (*p < 0.05) and 4 mm (**p < 0.000001) distal to the crush site. Mann–Whitney rank sum test; mean ± SE; n = 90 fields counted per genotype. Scale bars, 1 μm.

Fig. 7.

Functional testing of sensory fiber regeneration in normal and LAR-deficient mice. The pinch test was performed 2 weeks after nerve crush using LAR+/+ and LAR−/− littermate mice in a blinded manner. Results were combined from two independent experiments using littermates derived from different heterozygous crosses. The maximum distal distances beyond the crush site at which nerve pinch responses were detected were as follows: LAR+/+, 7.08 ± 0.14 mm (n = 18 nerves); LAR−/−, 3.50 ± 0.17 mm (n = 10 nerves). p < 0.0001; two-tailed Student's t test; mean ± SE.