Temporal manipulation of KCC3 expression in juvenile or adult mice suggests irreversible developmental deficit in hereditary motor sensory neuropathy with agenesis of the corpus callosum

Abstract

Hereditary motor sensory neuropathy (HMSN/ACC) with agenesis of the corpus callosum (ACC) has been documented in the French-derived populations of Charlevoix and Saguenay/Lac St. Jean in Quebec, Canada, as well as a few sporadic families throughout the world. HMSN/ACC occurs because of loss-of-function mutations in the potassium-chloride cotransporter 3 (KCC3). In HMSN/ACC, motor deficits occur early in infancy with rapid and continual deterioration of motor and sensory fibers into juvenile and adulthood. Genetic work in mice has demonstrated that the disease is caused by loss of KCC3 function in neurons and particularly parvalbumin (PV)-expressing neurons. Currently, there are no treatments or cures for HMSN/ACC other than pain management. As genetic counseling in Quebec has increased as a preventative strategy, most individuals with HSMN/ACC are now adults. The onset of the disease is unknown. In particular, it is unknown if the disease starts early during development and whether it can be reversed by restoring KCC3 function. In this study, we used two separate mouse models that when combined to the PV-Cre^ERT2 tamoxifen-inducible system allowed us to 1) disrupt KCC3 expression in adulthood or juvenile periods; and 2) reintroduce KCC3 expression in mice that first develop with a nonfunctional cotransporter. We show that disrupting or reintroducing KCC3 in the adult mouse has no effect on locomotor behavior, indicating that expression of KCC3 is critical during embryonic development and/or the perinatal period and that once the disease has started, reexpressing a functional cotransporter fails to change the course of HMSN/ACC.

Keywords: Andermann Syndrome; HSMN/ACC; KCC3 mouse model; parvalbumin-CreERT2; tamoxifen.

Figure 1.

Generation of the K-Cl cotransporter-3 (KCC3)-rescue mouse model. A: structure of the Slc12a6 gene showing exons 5-11, the location of 2 SphI restriction sites flanking an 8-kb fragment, and the position of a fragment used as Southern blot probe. B: structure of the gene after recombination with 5′- and 3′-homology arms, a duplicated exon 7 flanked by loxP sites, a PGK-driven neomycin-resistance gene cassette flanked by frt sites, and a 3rd SphI restriction site generating a 4-kb fragment. C: Southern blot analysis of embryonic stem (ES) cell clones showing the 8-kb SphI fragment and an additional 4-kb SphI fragment in clone 2A10. D: PCR genotyping of a mouse litter of KCC3-rescue mice carrying the PV-Cre^ERT2 allele. Sizes indicated in parentheses are expected fragment sizes. Crossing PV-Cre^ERT2 positive x KCC3-rescue heterozygote with PV-Cre^ERT2 negative x KCC3-rescue heterozygote mice led to 2 KCC3-rescue homozygote mice, with one of them carrying the PV-Cre^ERT2 allele. PV, parvalbumin; mut, mutant; w.t., wild type.

Figure 2.

Timeline of tamoxifen and locomotor experiments. Tamoxifen was administered at 2 different time points in 2 separate cohorts (n = 7 per cohort) of mice. After a baseline determination of locomotor activity at postnatal day 55 (P55) (adult mouse), tamoxifen was administered at P60. Two locomotor activity measurements followed the tamoxifen administration 10 days later (+10) or 30 days later (+30). In the 2nd cohort, tamoxifen was given by gavage at P20 and the behavior was assessed 50 (+50) and 80 days (+80) postadministration.

Figure 3.

Locomotor behavior in mice with induced deletion of K-Cl cotransporter-3 (KCC3) in parvalbumin (PV)-positive neurons in adulthood. A: breeding scheme utilized to produce tamoxifen-inducible knockout mice and their controls. B: mice treated with tamoxifen at postnatal day 60 (P60) were first tested on the accelerated rotarod at P55 and then retested at P70. Mice treated with tamoxifen at P20 were tested with rotarod assay at P70. No significant differences in accelerated rotarod performance were observed post-tamoxifen administration (n = 7 mice/cohort, P = 0.5, one-way ANOVA).

Figure 4.

Balance beam performance and pain perception in mice with induced deletion of K-Cl cotransporter-3 (KCC3) in parvalbumin (PV)-positive neurons. A: juvenile mice (n = 7) treated with tamoxifen (tamox) at postnatal day 20 (P20) displayed a shorter time interval to cross a 12-mm beam. In contrast, mice treated at P60 displayed similar behavior than mice before treatment. B: the neuro-scores, while slightly higher, are not statistically different between groups (P = 0.6). C and D: on the 6-mm beam, mice treated at P20 with tamoxifen displayed faster times in traversing beam (P = 0.02). The neuro-scores, while slightly higher, are not statistically different between groups (P = 0.7). E: hot plate assay in PV-Cre^ERT2 x KCC3-flox mice treated with tamoxifen (18.75 mg/mL) at P20 or P60. When treated at P20 and tested at P75, they displayed some increased sensitivity to the heat-evoked noxious stimulus (n = 7 mice, *P = 0.04, one-way ANOVA).

Figure 5.

Immunofluorescence analysis of dorsal root ganglions (DRGs) isolated from PV-Cre^ERT2 x KCC3-flox mice pre- and post-tamoxifen treatment. PV, parvalbumin; KCC3, K-Cl cotransporter-3. A–D: DRGs (n= 2 mice/ganglia, 8 sections) with no tamoxifen treatment displayed parvalbumin (A, red) and KCC3 (B, green) immunoreactivity. C: DAPI (blue) is used to stain nuclei. D: overlay of A–C gives an orange color. E–H: DRGs isolated from mice treated with tamoxifen. E and F: while the parvalbumin staining is visible (E, red), there is no expression of KCC3 (F, absence of green). G and H: DAPI (G, blue) and overlay (H, red) are also shown for the tamoxifen condition. DRGs were extracted from L1–L5 segments. They consisted of small, medium, and large ganglia. Scale bars = 20 μm.

Figure 6.

Poor rotarod performance of PV-Cre^ERT2 x KCC3-stop mice. PV, parvalbumin; KCC3, K-Cl cotransporter-3. A: complex breeding scheme utilized to produce the tamoxifen-induced rescue mice and controls. Heterozygous KCC3^+/r carrying or not the PV-Cre^ERT2 transgene were crossed to obtain 6 separate genotypes: controls, heterozygous, and homozygous KCC3-rescue with or without the PV-Cre^ERT2 allele. B: note that mice performed slightly better on day 3 than day 1, indicating an ability to learn. There are no error bars, as statistics could not be done on this cohort: n = 2 mice for pretamoxifen condition (black circles), and n = 1 mouse for post-tamoxifen condition (red squares). Inset: photographs of control and mutant mice.

Figure 7.

Immunofluorescence analysis of dorsal root ganglions (DRGs) isolated from PV-Cre^ERT2 x KCC3-rescue mice treated or not treated with tamoxifen. PV, parvalbumin; KCC3, K-Cl cotransporter-3. A–D: DRG neurons from untreated mice (n = 2 mice/ganglia, 8 sections) display parvalbumin expression (A, red) but no KCC3 expression (B, green). C: DAPI (blue) is used to stain nuclei. D: overlay of A–C gives red color. E–H: DRG neurons isolated from mice treated with tamoxifen for 3 days display parvalbumin expression (E, red) and KCC3 expression (F, green). G: DAPI (blue) is used to stain nuclei. H: overlay of E–G gives yellow color indicating overlapping expression of parvalbumin and cotransporter. DRGs were extracted from L1–L5 segments. They consisted of small, medium, and large ganglia. Scale bars = 20 μm.