Abstract

Familial dysautonomia (FD) is an autosomal recessive neurodegenerative disease that affects the development and survival of sensory and autonomic neurons. FD is caused by an mRNA splicing mutation in intron 20 of the IKBKAP gene that results in a tissue-specific skipping of exon 20 and a corresponding reduction of the inhibitor of kappaB kinase complex-associated protein (IKAP), also known as Elongator complex protein 1. To date, several promising therapeutic candidates for FD have been identified that target the underlying mRNA splicing defect, and increase functional IKAP protein. Despite these remarkable advances in drug discovery for FD, we lacked a phenotypic mouse model in which we could manipulate IKBKAP mRNA splicing to evaluate potential efficacy. We have, therefore, engineered a new mouse model that, for the first time, will permit to evaluate the phenotypic effects of splicing modulators and provide a crucial platform for preclinical testing of new therapies. This new mouse model, TgFD9; IkbkapΔ20/flox was created by introducing the complete human IKBKAP transgene with the major FD splice mutation (TgFD9) into a mouse that expresses extremely low levels of endogenous Ikbkap (IkbkapΔ20/flox). The TgFD9; IkbkapΔ20/flox mouse recapitulates many phenotypic features of the human disease, including reduced growth rate, reduced number of fungiform papillae, spinal abnormalities, and sensory and sympathetic impairments, and recreates the same tissue-specific mis-splicing defect seen in FD patients. This is the first mouse model that can be used to evaluate in vivo the therapeutic effect of increasing IKAP levels by correcting the underlying FD splicing defect.

Introduction

Familial dysautonomia (FD, MIM 223900), also known as Riley–Day syndrome or hereditary sensory and autonomic neuropathy type III, is a rare, fatal, congenital sensory and autonomic neuropathy caused by a ‘leaky’ mRNA splicing defect that results in reduced levels of the inhibitor of kappaB kinase complex-associated protein (IKAP), also known as Elongator complex protein 1 (ELP1). The major clinical features of FD are all due to a striking progressive depletion of sensory and autonomic neurons (1–5). Common symptoms include gastrointestinal dysfunction, gastroesophageal reflux, vomiting crises, recurrent pneumonia, seizures, gait abnormalities, kyphoscoliosis, postural hypotension, hypertension crises, absence of fungiform papillae on the tongue, decreased deep-tendon reflexes, defective lacrimation and impaired pain and temperature perception (2,6–11). Despite advances in patient care, the disorder is inevitably fatal, with only 50% of patients reaching 40 years of age (12). FD is an autosomal recessive disorder with a high carrier frequency in the Ashkenazi Jewish population (13,14). The major mutation is an intronic non-coding point mutation that causes a T to C transition at base pair six of IKBKAP intron 20, which leads to variable tissue-specific skipping of exon 20 in the IKBKAP transcript and reduced levels of IKAP/ELP1 protein. All FD patients possess at least one copy of the IVS20+6T>C mutation; 99.5% of patients are homozygous for the major mutation, and five patients are compound heterozygotes for two missense mutations (15,16). We have previously shown that despite the fact that FD is a recessive disease, homozygous mutant cells express both wild-type (WT) and mutant (MU) IKBKAP mRNA, and are therefore capable of producing full length functional IKAP protein (16). Thus, the mutation weakens but does not completely inactivate the 5′ splice site of exon 20. We have also reported that the relative amount of MU and WT IKBKAP transcripts varies between tissues, with the lowest levels of WT IKBKAP mRNA and IKAP protein in tissues from the central and peripheral nervous system (17).

IKAP/ELP1 is a 150-kDa protein encoded by the IKBKAP gene. Although several functions of IKAP/ELP1 have been reported, currently it is widely acknowledged that IKAP/ELP1 is the scaffolding member of the six-subunit human Elongator complex (18,19). The Elongator complex is an evolutionary highly conserved protein complex that participates in distinct cellular processes, including transcriptional elongation, acetylation of cytoskeletal α-tubulin and tRNA modification. In the nucleus, the Elongator complex is required for efficient RNA polymerase II transcriptional elongation of a subset of genes through the acetylation of Histone H3 (19–22). However, most Elongator is found in the cytoplasm, where it contributes to intracellular trafficking and cell migration through acetylation of α-tubulin (23,24) and participates in tRNA modification. Currently, it is still not clear whether Elongator has distinct functions or whether its primary role in tRNA modification leads to multiple downstream events (24–28). Cytosolic IKAP/ELP1 has also been implicated in the remodeling of actin cytoskeletal organization, which is necessary for neuron differentiation, migration and target tissue innervation (29–31). Recent in vivo studies highlighted that the loss of IKAP/ELP1 leads to neuronal cell death as a consequence of failed tissue innervation rather than abnormal neuronal migration (32–35).

The mouse Ikap protein shows 80% amino acid identity with human IKAP (36). The consensus donor splice site of intron 20, which is mutated in FD, is also conserved in the mouse (36). However, attempts to generate a mouse model for FD by introducing the major splicing mutation into the mouse genome have been unsuccessful. As part of our efforts to understand and recapitulate the FD phenotype, we have recently generated a mouse model of FD that expresses very low levels of Ikap protein by introducing two distinct mutations into the mouse Ikbkap locus, IkbkapΔ20/flox (37). These mutations result in extremely reduced WT Ikbkap mRNA production from the flox allele, and expression of an Ikbkap mRNA lacking exon 20 derived from the Δ20 allele. Characterization of these mice shows that while they recapitulate a large number of FD phenotypic features through global reduction of endogenous mouse Ikap, they are not a precise model of the human molecular defect since the reduction of endogenous mouse Ikap expression is not due to mutations affecting splicing (37). Moreover, a very large fraction of them (95%) die perinatally, making it difficult to test therapeutic strategies. Our goal, therefore, was to develop a mouse that modeled both the disease and the tissue-specific IKAP reduction resulting from disrupted mRNA splicing in order to evaluate the potential efficacy of splicing modulator compounds, antisense oligonucleotides (ASOs) or combination therapies.

Through an NINDS sponsored Neurodegeneration Drug Screening Consortium, we identified kinetin (6-furfurylaminopurine) as a potent splicing enhancer for IKBKAP (38,39). After several studies highlighting the remarkable efficacy of kinetin to improve splicing in patient cells (39,40), we showed that kinetin has a potential therapeutic value since it has an extremely safe ADME profile and can modify IKBKAP splicing in vivo in both a transgenic mouse model and humans (41,42). Additional molecules identified as promising therapeutic drug candidates for FD include phosphatidylserine (43,44) and rectifier of aberrant splicing (RECTAS) (45). Despite these remarkable advances in the development of potential FD therapeutics, we lacked an animal model in which to evaluate these compounds for disease-relevant efficacy. Here, we describe the generation and characterization of a new mouse model of FD that models both the molecular and phenotypic aspects of the human disease. This advance is an important step in the pipeline as we work to bring novel targeted therapies to FD patients.

Results

Generating a new phenotypic mouse model for FD

Despite remarkable advances in the development of potential therapeutics for FD and the generation of mouse models that recapitulate the disease phenotype (37), attempts to generate an animal model that models simultaneously the mRNA splicing defect and the progressive FD phenotype have previously been unsuccessful.

Functional conservation of IKAP protein between human and mouse has been demonstrated by introduction of the human IKBKAP transgene into the knockout Ikbkap mouse, Ikbkap−/− (46). The human WT IKBKAP transgene completely rescued embryonic lethality in the Ikbkap−/− mice and resulted in phenotypically normal mice that expressed only human IKAP (46). Conversely, the introduction of the human IKBKAP transgene carrying the FD major splice mutation TgFD9 did not rescue embryonic lethality in the Ikbkap−/− mice although it prolonged embryonic survival of Ikbkap−/− embryos (data not shown). Therefore, we reasoned that introducing the human IKBKAP transgene with the major FD splice mutation (TgFD9) into the severe phenotypic model IkbkapΔ20/flox would yield a mouse with a ‘milder’ FD phenotype that could be modified using small molecules or ASOs that improve human IKBKAP splicing and increase IKAP protein. To generate a mouse model with these characteristics, we initially intercrossed the previously generated TgFD9 transgenic mouse line with mice heterozygous for the Ikbkapflox allele (Ikbkapflox/+) (Supplementary Material, Fig. S1). F1 TgFD9; Ikbkapflox/+ progeny were then crossed with heterozygous IkbkapΔ20 mice (IkbkapΔ20/+) (37) (Supplementary Material, Fig. S1). TgFD9; IkbkapΔ20/flox mice are born at mendelian ratio and unlike IkbkapΔ20/flox mice, which show perinatal lethality rates close to 95% (37), about 60% of TgFD9; IkbkapΔ20/flox mice survive postnatally (Fig. 1A). The TgFD9; IkbkapΔ20/flox mouse carries one Ikbkap allele in which exon 20 is deleted (IkbkapΔ20 allele), one Ikbkap allele with loxP sites flanking exon 20 (hypomorphic Ikbkapflox allele) and a human transgene containing the complete IKBKAP gene into which we inserted the major IVS20+6T>C splice mutation (Fig. 1B). To quantify the expression of both mouse Ikbkap and human IKBKAP, we performed qRTPCR analysis using primers specific for the mouse and human gene, respectively. TgFD9; IkbkapΔ20/flox mice expressed remarkably low levels of endogenous WT Ikbkap compared with TgFD9; Ikbkap+/+, WT, IkbkapΔ20/+ and Ikbkapflox/+ mice (Fig. 1C and D) while expressing the predicted level of human WT IKBKAP (Fig. 1E and F).

Figure 1.

Generation of TgFD9; IkbkapΔ20/flox mouse model. (A) Seven-week-old TgFD9; IkbkapΔ20/flox (arrow) and control littermate were photographed side by side. (B) Detection of Ikbkap alleles and human TgFD9 IKBKAP transgene in TgFD9; IkbkapΔ20/flox mouse by PCR on genomic DNA. All PCR products were sequenced to confirm identity. (C and D) qRTPCR of endogenous mouse WT Ikbkap transcripts in brain (C) and liver (D) of TgFD9; Ikbkap+/+ (n = 3), WT (n = 3), IkbkapΔ20/+ (n = 3), Ikbkapflox/+ (n = 3) relative to TgFD9; IkbkapΔ20/flox (n = 3) mice. TgFD9; IkbkapΔ20/flox mice expressed significantly low levels of endogenous WT Ikbkap compared with TgFD9; Ikbkap+/+, WT, IkbkapΔ20/+ and Ikbkapflox/+ mice, P < 0.0001. (E and F) Quantification of human WT IKBKAP transcripts in brain (E) and liver (F) of TgFD9; Ikbkap+/+ (n = 3), WT (n = 3), IkbkapΔ20/+ (n = 3), Ikbkapflox/+ (n = 3) relative to TgFD9; IkbkapΔ20/flox (n = 3) mice. The expression of the human IKBKAP is not significantly different between TgFD9; IkbkapΔ20/flox and TgFD9; Ikbkap+/+ mice. Data are expressed as mean ± SD.

Figure 1.

Generation of TgFD9; IkbkapΔ20/flox mouse model. (A) Seven-week-old TgFD9; IkbkapΔ20/flox (arrow) and control littermate were photographed side by side. (B) Detection of Ikbkap alleles and human TgFD9 IKBKAP transgene in TgFD9; IkbkapΔ20/flox mouse by PCR on genomic DNA. All PCR products were sequenced to confirm identity. (C and D) qRTPCR of endogenous mouse WT Ikbkap transcripts in brain (C) and liver (D) of TgFD9; Ikbkap+/+ (n = 3), WT (n = 3), IkbkapΔ20/+ (n = 3), Ikbkapflox/+ (n = 3) relative to TgFD9; IkbkapΔ20/flox (n = 3) mice. TgFD9; IkbkapΔ20/flox mice expressed significantly low levels of endogenous WT Ikbkap compared with TgFD9; Ikbkap+/+, WT, IkbkapΔ20/+ and Ikbkapflox/+ mice, P < 0.0001. (E and F) Quantification of human WT IKBKAP transcripts in brain (E) and liver (F) of TgFD9; Ikbkap+/+ (n = 3), WT (n = 3), IkbkapΔ20/+ (n = 3), Ikbkapflox/+ (n = 3) relative to TgFD9; IkbkapΔ20/flox (n = 3) mice. The expression of the human IKBKAP is not significantly different between TgFD9; IkbkapΔ20/flox and TgFD9; Ikbkap+/+ mice. Data are expressed as mean ± SD.

TgFD9; IkbkapΔ20/flox mouse recapitulates both disease phenotype and tissue-specific mis-splicing

The expression of the human FD IKBKAP transgene in TgFD9; IkbkapΔ20/flox mouse attenuates the severe phenotype of IkbkapΔ20/flox mouse (37), but does not result in complete rescue (Fig. 2A). At birth, TgFD9; IkbkapΔ20/flox mice appeared grossly normal but were significantly smaller than their WT or heterozygous littermates. After weaning (P28), TgFD9; IkbkapΔ20/flox mice displayed excessive grooming and neurological manifestations including ataxic gait and hindlimb clasping on tail suspension that became more severe over time (Supplementary Material, Fig. S2). These initial observations suggested that TgFD9; IkbkapΔ20/flox mice showed an intermediate phenotype between control mice and the more compromised IkbkapΔ20/flox mice (Fig. 2A). To further characterize this new model, we assessed the extent of several FD phenotypes. Similar to FD patients, TgFD9; IkbkapΔ20/flox mice showed reduced growth rate (Fig. 2B) (47) and skeletal abnormalities (Fig. 2C–G) (48). To quantify the magnitude of spinal deformities, we measured the Cobb angle in 6-month-old control and TgFD9; IkbkapΔ20/flox mice (Fig. 2C and E). This angle is formed by the intersection of two lines plotted at the end-vertebrae of the curve deformity. While the average Cobb angle in WT mice in the coronal plan is 61.7 ± 8.7, the Cobb angle average in TgFD9; IkbkapΔ20/flox mice is 81.9 ± 5.9 (Fig. 2G). Another hallmark of the FD phenotype is reduced numbers or absence of taste-bud papillae on the tip of the tongue, which is already evident in infant FD patients (10,49). To determine whether TgFD9; IkbkapΔ20/flox mice showed reduction in the number of taste-bud papillae, we performed methylene blue staining of mutant and control tongues at 4 weeks, 3 months, 6 months and 12 months of age. As shown in Figure 2H, at all ages TgFD9; IkbkapΔ20/flox mice had significantly fewer fungiform papillae compared with control littermates. Histological analysis revealed that although fungiform papillae of mutant tongues possessed taste buds, they were smaller with abnormal morphology (Fig. 2I and J) and less sensory innervation (Fig. 2K and L). To quantitatively assess the status of the aged TgFD9; IkbkapΔ20/flox mice, we examined the body lengths, body weights, and weights of brains and internal organs and compared them with the previous FD mouse models, IkbkapΔ20/flox and Ikbkapflox/flox, at 18 months of age (Table 1). The results presented in Table 1 show that the TgFD9; IkbkapΔ20/flox model is less compromised than the IkbkapΔ20/flox and Ikbkapflox/flox models (37), but is still significantly smaller (∼90% the length of their control littermates) and thinner (∼70% the weight of their control littermates) than controls. Interestingly, TgFD9; IkbkapΔ20/flox mice have very little fat content, as is typical in FD patients. Despite progressive deteriorating health, the lifespan of TgFD9; IkbkapΔ20/flox mice did not differ from control littermates. This new mouse model thus proves that increasing the amount of functional IKAP protein by introducing human protein expressed from the transgene can improve the disease phenotype, an important observation when considering therapies aimed at increasing IKAP protein.

Table 1.

Body length, body weight and organ weight of male mice at 18 months of age

 Controls (n = 7) IkbkapΔΔ20/flox (n = 3) Ikbkapflox/flox (n = 4) TgFD9; IkbkapΔ20/flox (n = 3) 
Body weight (g) 42.4 ± 3.9 21.3 ± 1.2*** 24.7 ± 1.6** 27.6 ± 2.3** 
Body length (cm) 10.7 ± 0.3 9.4 ± 0.1** 9.9 ± 0.1* 9.8 ± 0.2* 
Epididymal fat (mg) 1306.6 ± 82.2 0.0 ± 0.0*** 70.0 ± 29.4*** 283.3 ± 23.6*** 
Brain (mg) 551.7 ± 9.0 382.2 ± 11.8*** 384.8 ± 5.8*** 474.4.0 ± 23.6* 
Heart (mg) 274.9 ± 2.6 150.8 ± 14.9*** 177.6 ± 16.8** 223.5 ± 19.0* 
Kidney (mg) 314.3 ± 59.8 192.6 ± 10.6* 273.6 ± 87.0 275.4 ± 87.6 
Liver (mg) 1908.3 ± 187.8 1262.3 ± 52.5* 1363.1 ± 58.6* 1780.7 ± 87.5 
Testes (mg) 102 ± 2.0 78 ± 1.3*** 84.5 ± 5.4*** 90.1 ± 2.5*** 
 Controls (n = 7) IkbkapΔΔ20/flox (n = 3) Ikbkapflox/flox (n = 4) TgFD9; IkbkapΔ20/flox (n = 3) 
Body weight (g) 42.4 ± 3.9 21.3 ± 1.2*** 24.7 ± 1.6** 27.6 ± 2.3** 
Body length (cm) 10.7 ± 0.3 9.4 ± 0.1** 9.9 ± 0.1* 9.8 ± 0.2* 
Epididymal fat (mg) 1306.6 ± 82.2 0.0 ± 0.0*** 70.0 ± 29.4*** 283.3 ± 23.6*** 
Brain (mg) 551.7 ± 9.0 382.2 ± 11.8*** 384.8 ± 5.8*** 474.4.0 ± 23.6* 
Heart (mg) 274.9 ± 2.6 150.8 ± 14.9*** 177.6 ± 16.8** 223.5 ± 19.0* 
Kidney (mg) 314.3 ± 59.8 192.6 ± 10.6* 273.6 ± 87.0 275.4 ± 87.6 
Liver (mg) 1908.3 ± 187.8 1262.3 ± 52.5* 1363.1 ± 58.6* 1780.7 ± 87.5 
Testes (mg) 102 ± 2.0 78 ± 1.3*** 84.5 ± 5.4*** 90.1 ± 2.5*** 

Statistical analyses were performed to compare IkbkapΔ20/flox, Ikbkapflox/flox, TgFD9; IkbkapΔ20/flox mice with WT mice.

Data are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2.

FD phenotypic features in TgFD9; IkbkapΔ20/flox mouse model. (A) Two-month-old control littermate (left), TgFD9; IkbkapΔ20/flox (arrow) and IkbkapΔ20/flox (right) mice were photographed side by side. (B) Postnatal growth curves of TgFD9; IkbkapΔ20/flox (red, n = 6) and control (blue, n = 8) male mice. (CG) Spinal deformity in TgFD9; IkbkapΔ20/flox mouse. Cobb angle measured in the coronal CT scan in control (C) and TgFD9; IkbkapΔ20/flox (E) mice. Representative three-dimensional CT images of the whole spine of 6-month-old control (D) and TgFD9; IkbkapΔ20/flox (F) mice. Bar chart showing the severity of the kyphosis in TgFD9; IkbkapΔ20/flox (grey, n = 6) and control (black, n = 6) 6-month-old mice (G). (HL) Analysis of taste-bud papillae on the tongue from TgFD9; IkbkapΔ20/flox mouse. Counts of fungiform papillae using methylene blue staining in control (black, n = 6) and TgFD9; IkbkapΔ20/flox (grey, n = 6) tongues at 4 weeks, 3 months, 6 months and 12 months of age (H). Histological examination of tongue fungiform papillae. H&E staining of tongue coronal section in P23 control (I) and TgFD9; IkbkapΔ20/flox (J) littermates. Transverse paraffin sections of P23 control (K) and TgFD9; IkbkapΔ20/flox (L) tongues were immunostained with anti-CGRP antibody. Note the large number of CGRP-positive axons in WT fungiform papillae in contrast to their near absence in TgFD9; IkbkapΔ20/flox fungiform papillae. Data are expressed as mean ± SD. **P < 0.01, ***P < 0.001.

Figure 2.

FD phenotypic features in TgFD9; IkbkapΔ20/flox mouse model. (A) Two-month-old control littermate (left), TgFD9; IkbkapΔ20/flox (arrow) and IkbkapΔ20/flox (right) mice were photographed side by side. (B) Postnatal growth curves of TgFD9; IkbkapΔ20/flox (red, n = 6) and control (blue, n = 8) male mice. (CG) Spinal deformity in TgFD9; IkbkapΔ20/flox mouse. Cobb angle measured in the coronal CT scan in control (C) and TgFD9; IkbkapΔ20/flox (E) mice. Representative three-dimensional CT images of the whole spine of 6-month-old control (D) and TgFD9; IkbkapΔ20/flox (F) mice. Bar chart showing the severity of the kyphosis in TgFD9; IkbkapΔ20/flox (grey, n = 6) and control (black, n = 6) 6-month-old mice (G). (HL) Analysis of taste-bud papillae on the tongue from TgFD9; IkbkapΔ20/flox mouse. Counts of fungiform papillae using methylene blue staining in control (black, n = 6) and TgFD9; IkbkapΔ20/flox (grey, n = 6) tongues at 4 weeks, 3 months, 6 months and 12 months of age (H). Histological examination of tongue fungiform papillae. H&E staining of tongue coronal section in P23 control (I) and TgFD9; IkbkapΔ20/flox (J) littermates. Transverse paraffin sections of P23 control (K) and TgFD9; IkbkapΔ20/flox (L) tongues were immunostained with anti-CGRP antibody. Note the large number of CGRP-positive axons in WT fungiform papillae in contrast to their near absence in TgFD9; IkbkapΔ20/flox fungiform papillae. Data are expressed as mean ± SD. **P < 0.01, ***P < 0.001.

The evaluation of human IKBKAP splicing in the new TgFD9; IkbkapΔ20/flox mouse showed that the same tissue-specific mis-splicing pattern observed in FD patients is seen in the mice, suggesting conservation of the factors governing tissue-specificity. We performed RT-PCR analysis on total RNA extracted from brain, dorsal root ganglia (DRG), liver, heart, lung and kidney from TgFD9; IkbkapΔ20/flox and control mice at 4 weeks of age, 3 months and 6 months. Remarkably, at all ages the TgFD9; IkbkapΔ20/flox mouse showed the same tissue-specific mis-splicing pattern observed in FD patient tissues. The splicing efficiency of the human FD IKBKAP gene is significantly decreased in brain and DRG when compared with the other tissues (Fig. 3A and B), confirming our previous observations in the TgFD transgenic mouse models (50).

Figure 3.

Tissue-specific mis-splicing in TgFD9; IkbkapΔ20/flox mouse model. (A) Splicing analysis of human IKBKAP transcripts in brain, DRG, lung, liver, heart and kidney from 4-week-old, 3 and 6-month-old TgFD9; IkbkapΔ20/flox mice. (B) Quantification of the percent of exon 20 inclusion in brain, DRG, lung, liver, heart and kidney from TgFD9; IkbkapΔ20/flox mice. Data are expressed as mean ± SD.

Figure 3.

Tissue-specific mis-splicing in TgFD9; IkbkapΔ20/flox mouse model. (A) Splicing analysis of human IKBKAP transcripts in brain, DRG, lung, liver, heart and kidney from 4-week-old, 3 and 6-month-old TgFD9; IkbkapΔ20/flox mice. (B) Quantification of the percent of exon 20 inclusion in brain, DRG, lung, liver, heart and kidney from TgFD9; IkbkapΔ20/flox mice. Data are expressed as mean ± SD.

Sensory and sympathetic deficits in TgFD9; IkbkapΔ20/flox adult mice

In FD patients, the fetal development and postnatal maintenance of DRG neurons is highly compromised, resulting in DRG of grossly reduced size and significantly reduced neuronal numbers (5,51). In addition, in adult FD patients, the mean volume of superior cervical ganglia (SCG) is reduced to 34% of normal, with concomitant decrease in neuronal numbers (52). To determine if our new mouse model recapitulates these neuropathological abnormalities, we examined SCG (Fig. 4A–C) and DRG (Fig. 4D–F) of aged mice. Histological analyses of SCG revealed that a large fraction of surviving neurons in TgFD9; IkbkapΔ20/flox were smaller than in controls and vacuolated (Fig. 4A and B). In TgFD9; IkbkapΔ20/flox mice, the average volume of SCG was ∼50% that of control littermates by 18 months of age (Fig. 4C). Neuronal counts at 18 months of age revealed a striking reduction in neuronal numbers in TgFD9; IkbkapΔ20/flox mice compared with controls (Fig. 4C). Sensory neurons in the DRG of adult mice are classified into two major classes based on their morphology and cytochemical characteristics: large light type A cells, representing 25–30% of total neurons, and small dark B cells, which give rise to unmyelinated fibers that conduct pain and thermal perception, and represent 70–75% of total neurons (53–55). Notably, consistent with the reduction of unmyelinated fibers observed in FD patients, the relative proportion of small dark B cells in lumbar DRG was significantly reduced in TgFD9; IkbkapΔ20/flox mice at 12 months of age compared with controls (Fig. 4D–F). This sensory neuronal loss is consistent with both the decrease in temperature and pain perception and the reduced number of unmyelinated fibers observed in FD patients (51). To assess the sensory innervation of skin in TgFD9; IkbkapΔ20/flox mice, we analyzed the epidermal nerve fibers (ENFs) in agonal biopsies of hind paw skin from TgFD9; IkbkapΔ20/flox and control 12-month-old mice. Cutaneous innervation in humans and mice is assessed using vertical skin sections immunolabeled against protein gene product (PGP) 9.5, a pan-axonal marker (56–58). PGP9.5 ENF density is markedly reduced in human FD (9). As shown in Figure 4G AND H, in TgFD9; IkbkapΔ20/flox plantar hind paws, the axons were fragmented and few crossed the dermal-epidermal junction to innervate the epidermis. Remarkably, TgFD9; IkbkapΔ20/flox mice had much fewer PGP9.5-immunoreactive axons compared with control mice (Fig. 4I). These results demonstrate that TgFD9; IkbkapΔ20/flox mice have proximal and distal deficiencies of both sympathetic and somatic small fibers, as reported in FD patients (5,51,52).

Figure 4.

Sensory and sympathetic neuropathology in adult TgFD9; IkbkapΔ20/flox mice. (AC) Sympathetic deficits in adult TgFD9; IkbkapΔ20/flox mice. Representative H&E-stained sections of SCG in control (A) and TgFD9; IkbkapΔ20/flox (B) mice at 18 months of age. Note that large neurons (arrowheads) are abundant in control (A) and that small neurons (arrows) are prevalent in TgFD9; IkbkapΔ20/flox SCG (B). (C) Volumes and total neuronal counts in SCGs from 18-month-old controls (black, n = 4) and TgFD9; IkbkapΔ20/flox (grey, n = 3) mice were calculated from serial paraffin sections spanning the whole ganglia and are expressed as percentage of controls. (DF) Sensory deficits in adult TgFD9; IkbkapΔ20/flox mice. H&E staining of DRG in control (D) and TgFD9; IkbkapΔ20/flox (E) at 12 months of age. An arrowhead indicates a large light A cell, and an arrow points to a small B cell. (F) Percentage of large light (A cells) and small dark (B cells) neuronal cells over the total number of cells were scored and counted in defined areas encompassing 40–50 neurons in 12-month-old controls (black, n = 4) and TgFD9; IkbkapΔ20/flox (grey, n = 3) mice. (G and H) Evaluation of PGP9.5-positive ENF in 12-month-old TgFD9; IkbkapΔ20/flox and control plantar hindpaw skin. (I) Quantification of ENF density in 12-month-old control (black, n = 10) and TgFD9; IkbkapΔ20/flox (grey, n = 8) plantar hindpaws. Data are expressed as mean ± SD for volumes and neuronal counts. Data are expressed as mean ± SEM for ENF density. **P < 0.01, ***P < 0.001.

Figure 4.

Sensory and sympathetic neuropathology in adult TgFD9; IkbkapΔ20/flox mice. (AC) Sympathetic deficits in adult TgFD9; IkbkapΔ20/flox mice. Representative H&E-stained sections of SCG in control (A) and TgFD9; IkbkapΔ20/flox (B) mice at 18 months of age. Note that large neurons (arrowheads) are abundant in control (A) and that small neurons (arrows) are prevalent in TgFD9; IkbkapΔ20/flox SCG (B). (C) Volumes and total neuronal counts in SCGs from 18-month-old controls (black, n = 4) and TgFD9; IkbkapΔ20/flox (grey, n = 3) mice were calculated from serial paraffin sections spanning the whole ganglia and are expressed as percentage of controls. (DF) Sensory deficits in adult TgFD9; IkbkapΔ20/flox mice. H&E staining of DRG in control (D) and TgFD9; IkbkapΔ20/flox (E) at 12 months of age. An arrowhead indicates a large light A cell, and an arrow points to a small B cell. (F) Percentage of large light (A cells) and small dark (B cells) neuronal cells over the total number of cells were scored and counted in defined areas encompassing 40–50 neurons in 12-month-old controls (black, n = 4) and TgFD9; IkbkapΔ20/flox (grey, n = 3) mice. (G and H) Evaluation of PGP9.5-positive ENF in 12-month-old TgFD9; IkbkapΔ20/flox and control plantar hindpaw skin. (I) Quantification of ENF density in 12-month-old control (black, n = 10) and TgFD9; IkbkapΔ20/flox (grey, n = 8) plantar hindpaws. Data are expressed as mean ± SD for volumes and neuronal counts. Data are expressed as mean ± SEM for ENF density. **P < 0.01, ***P < 0.001.

FD phenotypic features in TgFD9; IkbkapΔ20/flox embryos

Since many FD features are present at birth, we analyzed TgFD9; IkbkapΔ20/flox embryos at late gestation and compared them with control embryos of the same embryonic stage. At E18.5, TgFD9; IkbkapΔ20/flox embryos (Fig. 5B) appeared smaller and had body lengths intermediate between controls (Fig. 5A) and IkbkapΔ20/flox embryos (Fig. 5C). Comparative skeletal analyses between E18.5 TgFD9; IkbkapΔ20/flox, IkbkapΔ20/flox and control embryos (Fig. 5A–C) confirmed that as in FD, there are no obvious skeletal malformations at this stage in either TgFD9; IkbkapΔ20/flox or IkbkapΔ20/flox embryos. However, a slight delay in ossification is observed in IkbkapΔ20/flox and TgFD9; IkbkapΔ20/flox embryos. In particular, there is a delay in ossification of parietal and intraparietal cranial bones, as well as in the phalanges of front and hindlimbs. The body weight of E18.5 TgFD9; IkbkapΔ20/flox embryos is significantly reduced compared with control embryos and the placenta size of TgFD9; IkbkapΔ20/flox embryos is not rescued by the presence of the TgFD9 transgene (Fig. 5D). To assess whether the sensory and sympathetic loss observed in TgFD9; IkbkapΔ20/flox adult mice is due to abnormal embryonic development of sensory and sympathetic ganglia and/or to progressive postnatal neurodegeneration, serial transverse and coronal sections of E18.5 TgFD9; IkbkapΔ20/flox embryos spanning thoracic and lumbar DRG, stellate ganglia (SG) and SCG were analyzed (Fig. 5E–J). The volumes of SG (Fig. 5E and F), SCG (Fig. 5I) and DRG (Fig. 5G, H and J) were already significantly reduced in TgFD9; IkbkapΔ20/flox embryos compared with their WT littermates. Notably, the number of trkA/calcitonin gene-related peptide (CGRP) positive nociceptive neurons was already significantly reduced in the DRG of TgFD9; IkbkapΔ20/flox E18.5 embryos compared with control embryos (Fig. 5K–M). However, the neuronal density in SCG (Fig. 5I) and in SG (data not shown) did not differ significantly between TgFD9; IkbkapΔ20/flox and WT embryos, suggesting that the neuronal loss observed in adults occurs primarily postnatally.

Figure 5.

FD phenotypic features in TgFD9; IkbkapΔ20/flox embryos. (AC) Control littermate (A), TgFD9; IkbkapΔ20/flox (B) and IkbkapΔ20/flox (C) E18.5 skeletal preparations photographed at the same magnification. (D) Weights of E18.5 controls (black, n = 5) and TgFD9; IkbkapΔ20/flox (grey, n = 3) embryos and placentas. (E and F) H&E-stained sections of SG of control (E) and TgFD9; IkbkapΔ20/flox (F) E18.5 embryos. (G and H) H&E-stained sections of DRG of control (G) and TgFD9; IkbkapΔ20/flox (H) E18.5 embryos. (I) Volumes and total neuronal counts in E18.5 SCG. (J) Volume of L1 and T2 DRG in controls (black, n = 5) and TgFD9; IkbkapΔ20/flox (grey, n = 3) embryos was normalized to body weight and are displayed as percentage of controls. (KM) CGRP-positive neurons in E18.5 L1 DRG from control (K) and TgFD9; IkbkapΔ20/flox (L) embryos. Quantification of total CGRP neurons in control (black, n = 3), and (grey, n = 3) embryos (M). Data are presented as percentage of controls. For all graphs, data are expressed as mean ± SD. **P < 0.01, ***P < 0.001.

Figure 5.

FD phenotypic features in TgFD9; IkbkapΔ20/flox embryos. (AC) Control littermate (A), TgFD9; IkbkapΔ20/flox (B) and IkbkapΔ20/flox (C) E18.5 skeletal preparations photographed at the same magnification. (D) Weights of E18.5 controls (black, n = 5) and TgFD9; IkbkapΔ20/flox (grey, n = 3) embryos and placentas. (E and F) H&E-stained sections of SG of control (E) and TgFD9; IkbkapΔ20/flox (F) E18.5 embryos. (G and H) H&E-stained sections of DRG of control (G) and TgFD9; IkbkapΔ20/flox (H) E18.5 embryos. (I) Volumes and total neuronal counts in E18.5 SCG. (J) Volume of L1 and T2 DRG in controls (black, n = 5) and TgFD9; IkbkapΔ20/flox (grey, n = 3) embryos was normalized to body weight and are displayed as percentage of controls. (KM) CGRP-positive neurons in E18.5 L1 DRG from control (K) and TgFD9; IkbkapΔ20/flox (L) embryos. Quantification of total CGRP neurons in control (black, n = 3), and (grey, n = 3) embryos (M). Data are presented as percentage of controls. For all graphs, data are expressed as mean ± SD. **P < 0.01, ***P < 0.001.

These results corroborate the hypothesis that in FD the development of sensory and sympathetic ganglia is already compromised during embryogenesis and that significant neurodegeneration, in particular in sympathetic ganglia, occurs mainly postnatally.

Modulation of IKAP protein levels in vivo by splicing modification of the human FD transgene

We previously demonstrated that the plant cytokinin kinetin (6-furfurylaminopurine) is a potent splicing enhancer for IKBKAP both in vitro and in vivo. Kinetin treatment dramatically improved IKBKAP splicing in FD lymphoblast and fibroblast cell lines (39,40) with IKAP protein levels returning to normal after 1 week in culture with kinetin (39,40). Moreover, we proved that oral administration of 400 mg/kg/day of kinetin for 30 days improved IKBKAP splicing and also human IKAP protein expression in our transgenic mouse model TgFD (41); however, we did not examine dose response. Therefore, prior to initiating preclinical trials in our new TgFD9; IkbkapΔ20/flox model, we sought to determine if we can precisely modulate IKAP protein levels in a dose-dependent manner in vivo in TgFD9. On the basis of the calculated average daily dietary consumption of our mice, we formulated three different mouse chows to dose each experimental group of mice with 100, 200 or 400 mg/kg/day of kinetin. Kinetin formulated chow was administered to dams beginning at birth. At weaning, mice were genotyped and weaned onto the same dose of kinetin chow previously administered to the dams. To assess IKBKAP mRNA splicing, RT-PCR assays were performed on total RNA extracted from the liver of the mice carrying the TgFD9 transgene. As predicted, the percent of exon 20 inclusion increased in a dose-dependent manner in mice treated with increasing dose of kinetin (Supplementary Material, Fig. S3). To determine whether the increase in the level of WT IKBKAP transcript correlated with higher production of IKAP protein, we performed western blot analysis on liver tissues from the treated TgFD9 mice, and we saw a corresponding dose-dependent increase in IKAP protein (Supplementary Material, Fig. S3). These results show, for the first time, that IKBKAP mRNA splicing, and therefore the level of human IKAP protein, can be tightly modified in vivo in a dose-dependent manner by oral administration of kinetin, thus setting the stage for preclinical efficacy trials in our new TgFD9; IkbkapΔ20/flox mouse model.

Discussion

FD is an incurable disease with only 50% of patients reaching 40 years of age (12). The leading causes of death for FD patients are unexplained sudden death, aspiration pneumonias and respiratory insufficiency. Current treatments are supportive and far from optimal. The danger of aspiration during vomiting has been reduced with fundoplication (59), a surgical procedure that prevents gastroesophageal reflux, but vomiting is then replaced by severe nausea and violent retching. Treatment of ‘dysautonomic crises’ involves benzodiazepines and clonidine (1), which are not always effective and leave the patient severely hypotensive and extremely sedated. Mineralocorticoids and alpha-1-agonists are commonly used to treat orthostatic hypotension (1), but they worsen hypertension and have increased the prevalence of renal failure in patients with FD (60). Spinal fusion is used to halt progressive deformity in children with FD that have scoliosis but the procedure may worsen thoracic compliance and pulmonary function (61). Clearly, improved therapies are needed.

All FD patients possess at least one copy of the IVS20+6T>C mutation that leads to mis-splicing of IKBKAP mRNA and reduced levels of IKAP/ELP1 protein. 99.5% of patients are homozygous for the major mutation, and five patients are compound heterozygotes for two missense mutations (15,16). Therefore, developing a treatment that directly targets the molecular defect will benefit all FD patients. Moreover, unlike the existing supportive treatments, the development of drugs that directly target the molecular mechanism represents a new, precisely targeted treatment approach in which a single drug might prove efficacious for several disease symptoms. The advantage of targeting splicing defects for pharmacologic treatment is that, unlike with loss-of-function mutations, the cells retain the capacity to make functional protein. In FD, exon 20 inclusion is favorable in some tissues, which leads to sufficient IKAP/ELP1 levels. However, in neurons, exon 20 is skipped in the vast majority of transcripts, leading to significantly reduced IKAP/ELP1 expression. It is quite possible that a very slight shift in the balance of splicing may be sufficient to raise cellular levels of IKAP/ELP1 above threshold levels, and achieve therapeutic benefit.

To date, several promising therapeutic candidates for FD have been identified. All of these therapeutic approaches target the underlying molecular defect, either by improving IKBKAP mRNA splicing, such as kinetin (39,40) and RECTAS (45), or by increasing IKBKAP expression, such as phosphatidylserine (43,44). Although our previous FD mouse models faithfully recapitulate key features of FD (37), they present some significant limitations. First, the severity of the phenotype in these models results in a high rate of perinatal lethality, therefore reducing the number of viable offspring. Secondly, and most importantly, since these models do not carry the FD point mutation, they cannot be used to test compounds that increase IKAP/ELP1 expression by modulating splicing efficiency. In this manuscript, we describe the generation and characterization of the first FD mouse model that can be used to evaluate in vivo the efficacy of compounds that increase IKAP/ELP1 expression by either modulating splicing efficiency or boosting IKAP expression at the transcriptional level.

In addition, the phenotype of the new TgFD9; IkbkapΔ20/flox mouse further supports our previous finding that increasing the amount of functional IKAP/ELP1 protein, in this case by expressing human IKAP/ELP1 from a transgene, dramatically improves the severe FD phenotype that we observed in the IkbkapΔ20/flox mouse (37). This strongly suggests that our hypothesis, namely that increasing IKAP/ELP1 levels by splicing modification will improve disease, is correct. The rigorous characterization of this mouse revealed that TgFD9; IkbkapΔ20/flox is an excellent model for FD. In fact, it recapitulates many of the primary FD phenotypic hallmarks, including reduced growth (47), reduced number of fungiform papillae on the tongue (10,49), spinal abnormalities (48) and sensory and autonomic impairment. The examination of E18.5 TgFD9; IkbkapΔ20/flox embryos showed that the pathogenesis of the disease starts during the embryogenesis. E18.5 TgFD9; IkbkapΔ20/flox embryos are smaller than the control littermates and they display reduction in the volume of sensory DRG and sympathetic ganglia. These results are consistent with the fact that FD patients show early signs of the disease already at birth.

The creation of a mouse model that recapitulates the pathological features of the disease while preserves the tissue-specific mis-splicing pattern seen in FD patients will allow us for the first time to determine the clinical effectiveness of the available treatments for FD. In fact, the quantitative assessment of the FD features here described will allow us to determine which treatment and dosages are necessary to have in vivo efficacy on neurodegeneration.

Our work represents a major breakthrough in our efforts to develop a model system to test the efficacy of existing therapeutics as we work to bring new therapies to the clinic. Even though FD phenotypes are present at birth, there is continued neurodegeneration throughout life that eventually leads to severe ataxia and sudden death. The creation of this new model will allow the field to determine, for the first time, if improvement of IKBKAP splicing leading to increased IKAP/ELP1 protein might slow or even halt the progressive neurodegeneration that characterizes this devastating disease.

Materials and Methods

Generation of the TgFD9; IkbkapΔ20/flox mouse model and genotyping

The detailed descriptions of the original targeting vector to generate Ikbkapflox allele as well as the strategy to generate the IkbkapΔ20 allele have been previously published (37,62). Furthermore, the detailed description of the generation of the transgene with the complete genomic IKBKAP sequence and carrying the FD splice IVS20+6T>C mutation can be found in our previous manuscript by Hims et al. (50). To create the TgFD9; IkbkapΔ20/flox mouse, we crossed the previously generated TgFD9 transgenic mouse line carrying the human IKBKAP gene with the FD major splice mutation with the mouse line heterozygous for the Ikbkapflox allele (Ikbkapflox/+). The resulting progeny was genotyped to detect the presence of the TgFD9 transgene and of the Ikbkapflox allele. As expected, the Ikbkapflox allele and the TgFD9 transgene segregated independently; therefore around one-fourth of the F1 mice carried both the TgFD9 transgenic and Ikbkapflox alleles (TgFD9; Ikbkapflox/+). Subsequently, we crossed the TgFD9; Ikbkapflox/+ mice with the mouse line heterozygous for the IkbkapΔ20 allele (IkbkapΔ20/+). The resulting progeny was genotyped to detect the presence of the TgFD9 transgene as well as both Ikbkapflox and IkbkapΔ20 alleles. The expected mendelian ratio of TgFD9; IkbkapΔ20/flox mice in the F2 progeny was 1 in 8 (12.5%). However, the actual ratio obtained at weaning age was around 1 in 12 (49/589 = 8.3%), indicating that about 60% of them survive postnatally. The mice used for this study were housed in the animal facility of Massachusetts General Hospital (Boston, MA, USA) and in the animal facility of The University of Tennessee (Memphis, TN, USA), provided with constant access to a standard diet of food and water, and maintained on a 12-h light/dark cycle, and all experimental protocols were approved by the Institutional Animal Care and Use Committee of the Massachusetts General Hospital and of the University of Tennessee, and were in accordance with NIH guidelines.

For routine genotyping of progeny, genomic DNA was prepared from tail biopsies and PCR amplification reactions were carried out using the following primers: Ikap1F (5′-TGATTGACACAGACTCTGGCCA-3′) and Ikap4R (5′-CTTTCACTCTGAAATTACAGGAAG-3′) to discriminate the Ikbkap alleles and TgProbe1F (5′-GCCATTGTACTGTTTGCGACT-3′) and TgProbe1R (5′-TGAGTGTCACGATTCTTTCTGC-3′) to discriminate the TgFD9 transgene.

RNA isolation from mouse tissues and reverse transcription

Mice were euthanized and brain, DRG, liver, lung, kidney and heart tissues were removed. Tissues were homogenized in ice-cold TRI reagent (Molecular Research Center, Inc., Cincinnati, OH, USA), using a TissueLyser (Qiagen). Total RNA was extracted using the TRI reagent procedure provided by the manufacturer. The yield, purity and quality of the total RNA for each sample were determined using a Nanodrop ND-1000 spectrophotometer. Reverse transcription was performed using 1 µg of total RNA, Random Primers (Promega), and Superscript III reverse transcriptase (Invitrogen) according to the manufacturer's protocol.

Quantitative real-time PCR

WT mouse Ikbkap and WT human IKBKAP mRNA expression levels were estimated in mouse brain and liver by quantitative real-time PCR (qRTPCR) analysis using LightCycler® 480 Instrument (Roche). Mouse Ikbkap was analyzed using iQ SYBR-Green supermix (BioRad) and the mouse-specific Ikbkap primers, mm_Ikbkap1F (5′-GGACACAAAGCTTATACTACAGATGCC-3′) and mm_Ikbkap1R (5′-CTTGGGGTTATGGTCATGAATCAG-3′), which are specific to the WT Ikbkap spliced isoform. To assess the human IKBKAP WT transcript, we used the human IKBKAP-specific TaqMan probe hsIKBKAP-Taq_Probe_E20 (5′-TTCATCATCGAGCCCTGGTTTTAGCTCA-3′) and the primers hsIKBKAP Taq_E19f (5′-CCCCAGGACACAAAGCTTGT-3′) and hsIKBKAP Taq_E20-21bis_r (5′-TAAGTTTGTCCAACCACTTCCG-3′). Mouse Rpl113a gene was used as an internal normalization, mRpl13a F (5′-CCACCCTATGACAAGAAA-3′), mRpl13a R (5′-CTGCTTCTTCTTCCGATA-3′) and TaqMan mRPL13a probe (5′-CTGCTGCTCTCAAGGTTGTTCG-3′). qPCR was performed using the following thermal cycling protocol: 1 cycle at 95°C for 3 min; and 40 cycles at 95°C for 10 s, and at 60°C for 50 s. Data were analyzed using the LightCycler® 480 SW 1.5 software.

Splicing analysis of IKBKAP isoforms

PCR was performed using the cDNA equivalent of 100 ng of starting RNA in a 30-µl reaction, using GoTaq® green master mix (Promega) and 30 amplification cycles (94°C for 30 s, 58°C for 30 s, 72°C for 30 s). Human-specific IKBKAP primers Ex19F (5′-CCTGAGCAGCAATCATGTG-3′) and Ex23R (5′-TACATGGTCTTCGTGACATC-3′) were used to amplify human IKBKAP isoforms. PCR products were separated on 1.5% agarose gels and stained with ethidium bromide. The relative amounts of WT and MU (Δ20) IKBKAP spliced isoforms in a single PCR were determined using ImageJ, using the integrated density value for each band as previously described (41,50), and the relative proportion of the WT isoform detected in a sample was calculated as a percentage.

Protein isolation and western blot analysis

Protein extracts were obtained by homogenizing the liver in RIPA buffer (Tris–HCl 50 mm, pH 7.4; NaCl 150 mm; NP-40 1%; sodium deoxycholate 0.5%; SDS 0.1%) containing protease inhibitor cocktail (Sigma), DTT (100 µM) and PMSF (100 µM). Insoluble debris were discarded after centrifugation and protein concentration was determined using Pierce® BCA Protein Assay Kit (Thermo Scientific). 50 µg of protein was separated on NuPage 4–12% Bis–Tris Gel (Invitrogen) and transferred into nitrocellulose membrane (Thermo Scientific). Membrane was blocked in 5% non-fat milk for 1 h at room temperature and incubated overnight at 4°C with rabbit polyclonal antibody against the C-terminus region of the human IKAP protein (Anaspec, 1:2000) and with the rabbit polyclonal antibody against actin (Sigma, 1:2000). Membranes were washed and incubated with secondary antibodies for 1 h at room temperature. Protein bands were visualized by chemiluminescence (Pierce® ECL Western Blotting Substrate, Thermo Scientific) followed by exposure to autoradiographic film.

Counts of fungiform papillae

Fungiform papillae on the tongues were visualized by methylene blue staining, as described (37). In brief, tongues were dissected out in phosphate-buffered saline (PBS) and fixed in 4% paraformaldehyde (PFA) for 2 weeks. Tongues were rinsed with PBS and stained by incubation in 0.5% methylene blue solution for 20 min at room temperature, followed by three brief washes in PBS. Fungiform papillae were counted on both sides of the median fissure in the anterior part of the tongue, as well as in middle and posterior regions of the tongue under a Zeiss stereomicroscope.

Measurement of Cobb angle

Computed tomographic (CT) images were taken on the Siemen's Inveon system with 360 projections over 360° using a 80 kVp 500 µA X-ray tube on a 125 mm detector while the mice were under 1.5–2.5% isoflurane gas anesthesia. The images were reconstructed into 110 µm isotropic voxels (512 × 512 × 768 matrix) by a modified Feldkamp reconstruction algorithm (COBRA, Exxim Computer Corporation). The severity of the spinal deformities was evaluated with the use of the Osirix image processing software program by a blinded observer. The magnitude of the curvature was measured on posteroanterior radiographs by determining the Cobb angle (θ), which is the angle derived from the positions of the most-tilted vertebrae above and below the apex of the curvature (63). This angle is formed by the intersection of two lines plotted at the end-vertebrae of the curve deformity. One line is parallel to the endplate of the superior end-vertebra and the other is parallel to the endplate of the inferior end-vertebra.

Histological analysis

For histology, embryos and tissues were collected and fixed in 4% PFA in PBS. DRG and SCG were fixed for 12 h and embryos were fixed for 1 week; incubated for 24 h at 4°C in PBS containing 0.25 M sucrose and 0.2 M glycine; dehydrated; cleared with toluene and embedded in paraffin. Paraffin blocks were sectioned at 7 µm, mounted in superfrost slides (Fisher), and stained with hematoxylin and eosin (H&E).

Volumetric determination and neuronal counts in sympathetic and sensory ganglia

Volumes of sympathetic (SCG) and sensory (DRG) ganglia were determined essentially as described (37). In brief, H&E-stained serial paraffin sections (7 µm) spanning the whole ganglia were analyzed under a Zeiss stereomicroscope, and width and length were measured every fifth section. Volumetric measurements were performed by calculating and adding the volumes between every section analyzed. For neuronal counts, neurons with clearly visible nucleoli were counted from photomicrographs of H&E-stained paraffin sections and total neuronal numbers were estimated based on the total volume of the ganglia. For counts of DRG neurons, large light and small dark cells were scored and counted in defined areas encompassing 40–50 neurons each (64).

Immunohistochemistry

For immunohistochemistry on paraffin sections, slides were deparaffinized, rehydrated and incubated with 0.3% H202 in methanol for 20 min to quench endogenous peroxidase. Sections were then washed with PBS, blocked for 1 h with 4% BSA, 0.2% Triton X-100 in PBS and incubated at 4°C for 48 h with primary rabbit polyclonal antibody anti-CGRP (PC205L Calbiochem 1:200) in 0.4% BSA; 0.2% Triton X-100 in PBS. After several washes in PBS, primary antibody detection was carried out using the Vector ABC kit according to the manufacturer's instructions, followed by incubation with Fast DAB with metal enhancer (Sigma) or DAB brown substrate (BD Biosciences).

ENF measurement

At sacrifice, skin was removed from both plantar hind paws using a razor blade. Skin was fixed for 12 h in 2% periodate–lysine–paraformaldehyde (PLP) solution, rinsed with 0.l M Sorensons Phosphate Buffer and then preserved in 20% Glycerol/0.l M Sorensons Phosphate Buffer until being blocked and processed into 50-µm sections. Free-floating skin sections were then immunohistochemically labeled against PGP 9.5 using standard methods used for clinical diagnosis (65). Sections were incubated at room temperature overnight with the primary antibody against PGP9.5 (1:1200 dilution, Chemicon). They were then rinsed three times in PBS and incubated with biotinylated goat anti-rabbit IgG for 1 h (1:100 dilution, Vector Labs) and placed in avidin–biotin complex for 1 h (Vector Labs). The reaction product was visualized using the chromagen/peroxidase substrate (Vector SG substrate kit, Vector Lab). In order to confirm that there were no nonspecific immunoreactions, sections were incubated with primary or secondary antisera alone. Images were collected on a microscope Zeiss Axio Scope.A1 microscope using 100× oil magnification. Six sections were measured for each footpad. ENF data were presented as the mean number of fibers per mm2 of epidermis.

Skeleton staining of E18.5 embryos

E18.5 embryos were fixed in 95% ethanol for 1 h. Skin and viscera were removed and the carcass was stained overnight in 95% ethanol, containing alcian blue. Carcasses were then transferred back into 95% ethanol for 5 h, placed in 2% KOH for 24 h and stained with alizarin red S overnight. Skeletons were then cleared in 1% KOH; 20% glycerol for 4 weeks, with several solution changes. Once skeletons were cleared, they were stored in glycerol and photographed for documentation.

Kinetin administration via specially formulated diet

TgFD9 transgenic mice were fed special AIN-76A rodent diet containing kinetin at a concentration of 1.07 2.14 or 4.28 g/kg of mouse chow (Research Diets, Inc., NJ, USA), corresponding to a daily consumption of 100, 200 and 400 mg kinetin/kg body weight, from birth until 9 weeks of age. In the control group, TgFD9 transgenic mice were fed AIN-76A diet without kinetin. The mice were given water ad libitum, and changes in body weights were monitored on a weekly basis.

Statistical analysis

Statistical analysis was performed by two-tailed unpaired Student's t-test. P > 0.05 was scored as not significant.

Supplementary Material

Supplementary Material is available at HMG online.

Funding

This work was supported by National Institute of Health grants (R01NS036326 to S.A.S., R01NS061842 to I.D.), and by the Dysautonomia Foundation, Inc. E.M. was recipient of a postdoctoral grant from the Rotary Foundation.

Acknowledgments

We thank Dr Felicia Axelrod and Dr Horacio Kaufmann of the Dysautonomia Treatment and Evaluation Center at New York University Medical School for their longstanding collaboration and helpful discussions. We are also grateful to David Brenner of the Dysautonomia Foundation, Inc. for his support; Dr James Gusella for his helpful discussions and Dr David Schoenfeld for assistance with statistical analysis.

Conflict of Interest statement: The authors have no conflict of interest to declare.

References

1
Axelrod
F.B.
(
2005
)
Familial dysautonomia: a review of the current pharmacological treatments
.
Expert. Opin. Pharmacother.
 ,
6
,
561
567
.
2
Axelrod
F.B.
,
Nachtigal
R.
,
Dancis
J.
(
1974
)
Familial dysautonomia: diagnosis, pathogenesis and management
.
Adv. Pediatr.
 ,
21
,
75
96
.
3
Pearson
J.
(
1979
)
Familial dysautonomia (a brief review)
.
J. Auton. Nerv. Syst.
 ,
1
,
119
126
.
4
Pearson
J.
,
Pytel
B.
(
1978
)
Quantitative studies of ciliary and sphenopalatine ganglia in familial dysautonomia
.
J. Neurol. Sci.
 ,
39
,
123
130
.
5
Pearson
J.
,
Pytel
B.A.
,
Grover-Johnson
N.
,
Axelrod
F.
,
Dancis
J.
(
1978
)
Quantitative studies of dorsal root ganglia and neuropathologic observations on spinal cords in familial dysautonomia
.
J. Neurol. Sci.
 ,
35
,
77
92
.
6
Geltzer
A.I.
,
Gluck
L.
,
Talner
N.S.
,
Polesky
H.F.
(
1964
)
Familial dysautonomia; studies in a newborn infant
.
N. Engl. J. Med.
 ,
271
,
436
440
.
7
Mahloudji
M.
,
Brunt
P.W.
,
McKusick
V.A.
(
1970
)
Clinical neurological aspects of familial dysautonomia
.
J. Neurol. Sci.
 ,
11
,
383
395
.
8
Axelrod
F.B.
,
Pearson
J.
(
1984
)
Congenital sensory neuropathies. Diagnostic distinction from familial dysautonomia
.
Am. J. Dis. Child.
 ,
138
,
947
954
.
9
Hilz
M.J.
,
Axelrod
F.B.
,
Bickel
A.
,
Stemper
B.
,
Brys
M.
,
Wendelschafer-Crabb
G.
,
Kennedy
W.R.
(
2004
)
Assessing function and pathology in familial dysautonomia: assessment of temperature perception, sweating and cutaneous innervation
.
Brain
 ,
127
,
2090
2098
.
10
Brunt
P.W.
,
McKusick
V.A.
(
1970
)
Familial dysautonomia. A report of genetic and clinical studies, with a review of the literature
.
Medicine (Baltimore)
 ,
49
,
343
374
.
11
Gadoth
N.
,
Mass
E.
,
Gordon
C.R.
,
Steiner
J.E.
(
1997
)
Taste and smell in familial dysautonomia
.
Dev. Med. Child Neurol.
 ,
39
,
393
397
.
12
Axelrod
F.B.
(
2004
)
Familial dysautonomia
.
Muscle Nerve
 ,
29
,
352
363
.
13
Lehavi
O.
,
Aizenstein
O.
,
Bercovich
D.
,
Pavzner
D.
,
Shomrat
R.
,
Orr-Urtreger
A.
,
Yaron
Y.
(
2003
)
Screening for familial dysautonomia in Israel: evidence for higher carrier rate among Polish Ashkenazi Jews
.
Genet. Test.
 ,
7
,
139
142
.
14
Maayan
C.
,
Kaplan
E.
,
Shachar
S.
,
Peleg
O.
,
Godfrey
S.
(
1987
)
Incidence of familial dysautonomia in Israel 1977–1981
.
Clin. Genet.
 ,
32
,
106
108
.
15
Leyne
M.
,
Mull
J.
,
Gill
S.P.
,
Cuajungco
M.P.
,
Oddoux
C.
,
Blumenfeld
A.
,
Maayan
C.
,
Gusella
J.F.
,
Axelrod
F.B.
,
Slaugenhaupt
S.A.
(
2003
)
Identification of the first non-Jewish mutation in familial dysautonomia
.
Am. J. Med. Genet. A
 ,
118A
,
305
308
.
16
Slaugenhaupt
S.A.
,
Blumenfeld
A.
,
Gill
S.P.
,
Leyne
M.
,
Mull
J.
,
Cuajungco
M.P.
,
Liebert
C.B.
,
Chadwick
B.
,
Idelson
M.
,
Reznik
L.
et al
. (
2001
)
Tissue-specific expression of a splicing mutation in the IKBKAP gene causes familial dysautonomia
.
Am. J. Hum. Genet.
 ,
68
,
598
605
.
17
Cuajungco
M.P.
,
Leyne
M.
,
Mull
J.
,
Gill
S.P.
,
Lu
W.
,
Zagzag
D.
,
Axelrod
F.B.
,
Maayan
C.
,
Gusella
J.F.
,
Slaugenhaupt
S.A.
(
2003
)
Tissue-specific reduction in splicing efficiency of IKBKAP due to the major mutation associated with familial dysautonomia
.
Am. J. Hum. Genet.
 ,
72
,
749
758
.
18
Hawkes
N.A.
,
Otero
G.
,
Winkler
G.S.
,
Marshall
N.
,
Dahmus
M.E.
,
Krappmann
D.
,
Scheidereit
C.
,
Thomas
C.L.
,
Schiavo
G.
,
Erdjument-Bromage
H.
et al
. (
2002
)
Purification and characterization of the human elongator complex
.
J. Biol. Chem.
 ,
277
,
3047
3052
.
19
Otero
G.
,
Fellows
J.
,
Li
Y.
,
de Bizemont
T.
,
Dirac
A.M.
,
Gustafsson
C.M.
,
Erdjument-Bromage
H.
,
Tempst
P.
,
Svejstrup
J.Q.
(
1999
)
Elongator, a multisubunit component of a novel RNA polymerase II holoenzyme for transcriptional elongation
.
Mol. Cell
 ,
3
,
109
118
.
20
Li
Y.
,
Takagi
Y.
,
Jiang
Y.
,
Tokunaga
M.
,
Erdjument-Bromage
H.
,
Tempst
P.
,
Kornberg
R.D.
(
2001
)
A multiprotein complex that interacts with RNA polymerase II elongator
.
J. Biol. Chem.
 ,
276
,
29628
29631
.
21
Kim
J.H.
,
Lane
W.S.
,
Reinberg
D.
(
2002
)
Human Elongator facilitates RNA polymerase II transcription through chromatin
.
Proc. Natl Acad. Sci. USA
 ,
99
,
1241
1246
.
22
Pokholok
D.K.
,
Hannett
N.M.
,
Young
R.A.
(
2002
)
Exchange of RNA polymerase II initiation and elongation factors during gene expression in vivo
.
Mol. Cell
 ,
9
,
799
809
.
23
Creppe
C.
,
Malinouskaya
L.
,
Volvert
M.L.
,
Gillard
M.
,
Close
P.
,
Malaise
O.
,
Laguesse
S.
,
Cornez
I.
,
Rahmouni
S.
,
Ormenese
S.
et al
. (
2009
)
Elongator controls the migration and differentiation of cortical neurons through acetylation of alpha-tubulin
.
Cell
 ,
136
,
551
564
.
24
Svejstrup
J.Q.
(
2007
)
Elongator complex: how many roles does it play?
Curr. Opin. Cell Biol.
 ,
19
,
331
336
.
25
Huang
B.
,
Johansson
M.J.
,
Bystrom
A.S.
(
2005
)
An early step in wobble uridine tRNA modification requires the Elongator complex
.
RNA
 ,
11
,
424
436
.
26
Esberg
A.
,
Huang
B.
,
Johansson
M.J.
,
Bystrom
A.S.
(
2006
)
Elevated levels of two tRNA species bypass the requirement for Elongator complex in transcription and exocytosis
.
Mol. Cell
 ,
24
,
139
148
.
27
Bauer
F.
,
Hermand
D.
(
2012
)
A coordinated codon-dependent regulation of translation by Elongator
.
Cell Cycle
 ,
11
,
4524
4529
.
28
Bauer
F.
,
Matsuyama
A.
,
Candiracci
J.
,
Dieu
M.
,
Scheliga
J.
,
Wolf
D.A.
,
Yoshida
M.
,
Hermand
D.
(
2012
)
Translational control of cell division by Elongator
.
Cell Rep.
 ,
1
,
424
433
.
29
Cheishvili
D.
,
Maayan
C.
,
Cohen-Kupiec
R.
,
Lefler
S.
,
Weil
M.
,
Ast
G.
,
Razin
A.
(
2011
)
IKAP/Elp1 involvement in cytoskeleton regulation and implication for familial dysautonomia
.
Hum. Mol. Genet.
 ,
20
,
1585
1594
.
30
Close
P.
,
Hawkes
N.
,
Cornez
I.
,
Creppe
C.
,
Lambert
C.A.
,
Rogister
B.
,
Siebenlist
U.
,
Merville
M.P.
,
Slaugenhaupt
S.A.
,
Bours
V.
et al
. (
2006
)
Transcription impairment and cell migration defects in elongator-depleted cells: implication for familial dysautonomia
.
Mol. Cell
 ,
22
,
521
531
.
31
Johansen
L.D.
,
Naumanen
T.
,
Knudsen
A.
,
Westerlund
N.
,
Gromova
I.
,
Junttila
M.
,
Nielsen
C.
,
Bottzauw
T.
,
Tolkovsky
A.
,
Westermarck
J.
et al
. (
2008
)
IKAP localizes to membrane ruffles with filamin A and regulates actin cytoskeleton organization and cell migration
.
J. Cell Sci.
 ,
121
,
854
864
.
32
Jackson
M.Z.
,
Gruner
K.A.
,
Qin
C.
,
Tourtellotte
W.G.
(
2014
)
A neuron autonomous role for the familial dysautonomia gene ELP1 in sympathetic and sensory target tissue innervation
.
Development
 ,
141
,
2452
2461
.
33
George
L.
,
Chaverra
M.
,
Wolfe
L.
,
Thorne
J.
,
Close-Davis
M.
,
Eibs
A.
,
Riojas
V.
,
Grindeland
A.
,
Orr
M.
,
Carlson
G.A.
et al
. (
2013
)
Familial dysautonomia model reveals Ikbkap deletion causes apoptosis of Pax3+ progenitors and peripheral neurons
.
Proc. Natl Acad. Sci. USA
 ,
110
,
18698
18703
.
34
Abashidze
A.
,
Gold
V.
,
Anavi
Y.
,
Greenspan
H.
,
Weil
M.
(
2014
)
Involvement of IKAP in peripheral target innervation and in specific JNK and NGF signaling in developing PNS neurons
.
PLoS ONE
 ,
9
,
e113428
.
35
Hunnicutt
B.J.
,
Chaverra
M.
,
George
L.
,
Lefcort
F.
(
2012
)
IKAP/Elp1 is required in vivo for neurogenesis and neuronal survival, but not for neural crest migration
.
PLoS ONE
 ,
7
,
e32050
.
36
Cuajungco
M.P.
,
Leyne
M.
,
Mull
J.
,
Gill
S.P.
,
Gusella
J.F.
,
Slaugenhaupt
S.A.
(
2001
)
Cloning, characterization, and genomic structure of the mouse Ikbkap gene
.
DNA Cell Biol.
 ,
20
,
579
586
.
37
Dietrich
P.
,
Alli
S.
,
Shanmugasundaram
R.
,
Dragatsis
I.
(
2012
)
IKAP expression levels modulate disease severity in a mouse model of familial dysautonomia
.
Hum. Mol. Genet.
 ,
21
,
5078
5090
.
38
Heemskerk
J.
,
Tobin
A.J.
,
Bain
L.J.
(
2002
)
Teaching old drugs new tricks. Meeting of the Neurodegeneration Drug Screening Consortium, 7–8 April 2002, Washington, DC, USA
.
Trends Neurosci.
 ,
25
,
494
496
.
39
Slaugenhaupt
S.A.
,
Mull
J.
,
Leyne
M.
,
Cuajungco
M.P.
,
Gill
S.P.
,
Hims
M.M.
,
Quintero
F.
,
Axelrod
F.B.
,
Gusella
J.F.
(
2004
)
Rescue of a human mRNA splicing defect by the plant cytokinin kinetin
.
Hum. Mol. Genet.
 ,
13
,
429
436
.
40
Hims
M.M.
,
Ibrahim
E.C.
,
Leyne
M.
,
Mull
J.
,
Liu
L.
,
Lazaro
C.
,
Shetty
R.S.
,
Gill
S.
,
Gusella
J.F.
,
Reed
R.
et al
. (
2007
)
Therapeutic potential and mechanism of kinetin as a treatment for the human splicing disease familial dysautonomia
.
J. Mol. Med. (Berl.)
 ,
85
,
149
161
.
41
Shetty
R.S.
,
Gallagher
C.S.
,
Chen
Y.T.
,
Hims
M.M.
,
Mull
J.
,
Leyne
M.
,
Pickel
J.
,
Kwok
D.
,
Slaugenhaupt
S.A.
(
2011
)
Specific correction of a splice defect in brain by nutritional supplementation
.
Hum. Mol. Genet.
 ,
20
,
4093
4101
.
42
Axelrod
F.B.
,
Liebes
L.
,
Gold-Von Simson
G.
,
Mendoza
S.
,
Mull
J.
,
Leyne
M.
,
Norcliffe-Kaufmann
L.
,
Kaufmann
H.
,
Slaugenhaupt
S.A.
(
2011
)
Kinetin improves IKBKAP mRNA splicing in patients with familial dysautonomia
.
Pediatr. Res.
 ,
70
,
480
483
.
43
Bochner
R.
,
Ziv
Y.
,
Zeevi
D.
,
Donyo
M.
,
Abraham
L.
,
Ashery-Padan
R.
,
Ast
G.
(
2013
)
Phosphatidylserine increases IKBKAP levels in a humanized knock-in IKBKAP mouse model
.
Hum. Mol. Genet.
 ,
22
,
2785
2794
.
44
Keren
H.
,
Donyo
M.
,
Zeevi
D.
,
Maayan
C.
,
Pupko
T.
,
Ast
G.
(
2010
)
Phosphatidylserine increases IKBKAP levels in familial dysautonomia cells
.
PLoS ONE
 ,
5
,
e15884
.
45
Yoshida
M.
,
Kataoka
N.
,
Miyauchi
K.
,
Ohe
K.
,
Iida
K.
,
Yoshida
S.
,
Nojima
T.
,
Okuno
Y.
,
Onogi
H.
,
Usui
T.
et al
. (
2015
)
Rectifier of aberrant mRNA splicing recovers tRNA modification in familial dysautonomia
.
Proc. Natl Acad. Sci. USA
 ,
112
,
2764
2769
.
46
Chen
Y.T.
,
Hims
M.M.
,
Shetty
R.S.
,
Mull
J.
,
Liu
L.
,
Leyne
M.
,
Slaugenhaupt
S.A.
(
2009
)
Loss of mouse Ikbkap, a subunit of elongator, leads to transcriptional deficits and embryonic lethality that can be rescued by human IKBKAP
.
Mol. Cell. Biol.
 ,
29
,
736
744
.
47
Kamboj
M.K.
,
Axelrod
F.B.
,
David
R.
,
Geffner
M.E.
,
Novogroder
M.
,
Oberfield
S.E.
,
Turco
J.H.
,
Maayan
C.
,
Kohn
B.
(
2004
)
Growth hormone treatment in children with familial dysautonomia
.
J. Pediatr.
 ,
144
,
63
67
.
48
Kaplan
L.
,
Margulies
J.Y.
,
Kadari
A.
,
Floman
Y.
,
Robin
G.C.
(
1997
)
Aspects of spinal deformity in familial dysautonomia (Riley–Day syndrome)
.
Eur. Spine J.
 ,
6
,
33
38
.
49
Smith
A.
,
Farbman
A.
,
Dancis
J.
(
1965
)
Absence of taste-bud papillae in familial dysautonomia
.
Science
 ,
147
,
1040
1041
.
50
Hims
M.M.
,
Shetty
R.S.
,
Pickel
J.
,
Mull
J.
,
Leyne
M.
,
Liu
L.
,
Gusella
J.F.
,
Slaugenhaupt
S.A.
(
2007
)
A humanized IKBKAP transgenic mouse models a tissue-specific human splicing defect
.
Genomics
 ,
90
,
389
396
.
51
Axelrod
F.B.
,
Iyer
K.
,
Fish
I.
,
Pearson
J.
,
Sein
M.E.
,
Spielholz
N.
(
1981
)
Progressive sensory loss in familial dysautonomia
.
Pediatrics
 ,
67
,
517
522
.
52
Pearson
J.
,
Pytel
B.A.
(
1978
)
Quantitative studies of sympathetic ganglia and spinal cord intermediolateral gray columns in familial dysautonomia
.
J. Neurol. Sci.
 ,
39
,
47
59
.
53
Lawson
S.N.
,
Waddell
P.J.
(
1991
)
Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons
.
J. Physiol.
 ,
435
,
41
63
.
54
Tandrup
T.
(
1993
)
A method for unbiased and efficient estimation of number and mean volume of specified neuron subtypes in rat dorsal root ganglion
.
J. Comp. Neurol.
 ,
329
,
269
276
.
55
Gjerstad
M.D.
,
Tandrup
T.
,
Koltzenburg
M.
,
Jakobsen
J.
(
2002
)
Predominant neuronal B-cell loss in L5 DRG of p75 receptor-deficient mice
.
J. Anat.
 ,
200
,
81
87
.
56
Johnson
M.S.
,
Ryals
J.M.
,
Wright
D.E.
(
2008
)
Early loss of peptidergic intraepidermal nerve fibers in an STZ-induced mouse model of insensate diabetic neuropathy
.
Pain
 ,
140
,
35
47
.
57
Myers
M.I.
,
Peltier
A.C.
(
2013
)
Uses of skin biopsy for sensory and autonomic nerve assessment
.
Curr. Neurol. Neurosci. Rep.
 ,
13
,
323
.
58
Oaklander
A.L.
(
2006
)
Chapter 31 Neuropathological examination of peripheral nerves in painful neuropathies (neuralgias)
.
Handb. Clin. Neurol.
 ,
81
,
463-XII
.
59
Vinograd
I.
,
Udassin
R.
,
Beilin
B.
,
Neuman
A.
,
Maayan
C.
,
Nissan
S.
(
1985
)
The surgical management of children with familial dysautonomia
.
J. Pediatr. Surg.
 ,
20
,
632
636
.
60
Pearson
J.
,
Gallo
G.
,
Gluck
M.
,
Axelrod
F.
(
1980
)
Renal disease in familial dysautonomia
.
Kidney Int.
 ,
17
,
102
112
.
61
Karol
L.A.
,
Johnston
C.
,
Mladenov
K.
,
Schochet
P.
,
Walters
P.
,
Browne
R.H.
(
2008
)
Pulmonary function following early thoracic fusion in non-neuromuscular scoliosis
.
J. Bone Joint Surg. Am.
 ,
90
,
1272
1281
.
62
Dietrich
P.
,
Yue
J.
,
E
S.
,
Dragatsis
I.
(
2011
)
Deletion of exon 20 of the Familial Dysautonomia gene Ikbkap in mice causes developmental delay, cardiovascular defects, and early embryonic lethality
.
PLoS ONE
 ,
6
,
e27015
.
63
Kubota
K.
,
Doi
T.
,
Murata
M.
,
Kobayakawa
K.
,
Matsumoto
Y.
,
Harimaya
K.
,
Shiba
K.
,
Hashizume
M.
,
Iwamoto
Y.
,
Okada
S.
(
2013
)
Disturbance of rib cage development causes progressive thoracic scoliosis: the creation of a nonsurgical structural scoliosis model in mice
.
J. Bone Joint Surg. Am.
 ,
95
,
e130
.
64
Tandrup
T.
(
1995
)
Are the neurons in the dorsal root ganglion pseudounipolar? A comparison of the number of neurons and number of myelinated and unmyelinated fibres in the dorsal root
.
J. Comp. Neurol.
 ,
357
,
341
347
.
65
Lauria
G.
,
Lombardi
R.
,
Camozzi
F.
,
Devigili
G.
(
2009
)
Skin biopsy for the diagnosis of peripheral neuropathy
.
Histopathology
 ,
54
,
273
285
.

Author notes

These authors contributed equally to this work.
Present address: RxBio, Inc., Memphis Tennessee, USA.
Co-senior authors.