Mouse Models of Musculocontractural Ehlers-Danlos Syndrome

Musculocontractural Ehlers-Danlos syndrome (mcEDS) is a subtype of EDS caused by mutations in the gene for carbohydrate sulfotransferase 14 (CHST14) (mcEDS-CHST14) or dermatan sulfate epimerase (DSE) (mcEDS-DSE). These mutations induce loss of enzymatic activity in D4ST1 or DSE and disrupt dermatan sulfate (DS) biosynthesis. The depletion of DS causes the symptoms of mcEDS, such as multiple congenital malformations (e.g., adducted thumbs, clubfeet, and craniofacial characteristics) and progressive connective tissue fragility-related manifestations (e.g., recurrent dislocations, progressive talipes or spinal deformities, pneumothorax or pneumohemothorax, large subcutaneous hematomas, and/or diverticular perforation). Careful observations of patients and model animals are important to investigate pathophysiological mechanisms and therapies for the disorder. Some independent groups have investigated Chst14 gene-deleted (Chst14-/-) and Dse-/- mice as models of mcEDS-CHST14 and mcEDS-DSE, respectively. These mouse models exhibit similar phenotypes to patients with mcEDS, such as suppressed growth and skin fragility with deformation of the collagen fibrils. Mouse models of mcEDS-CHST14 also show thoracic kyphosis, hypotonia, and myopathy, which are typical complications of mcEDS. These findings suggest that the mouse models can be useful for research uncovering the pathophysiology of mcEDS and developing etiology-based therapy. In this review, we organize and compare the data of patients and model mice.

Dermatan 4-O-sulfotransferase-1 (D4ST1) and DSE are essential enzymes for the biosynthesis of dermatan sulfate (DS). D4ST1 and DSE are encoded by CHST14 and DSE genes, respectively [16,17]. Chondroitin sulfate (CS) and DS are classified as glycosaminoglycans (GAGs), which are linear polysaccharide chains consisting of repeating disaccharide units. The biosynthesis of DS and CS is initiated by the synthesis of a tetrasaccharide linker region, glucuronic acid-β1-3galactose-β1-3galactose-β1-4xyloseβ1-O-(GlcUA-Gal-Gal-Xyl-), on serine residues of specific core proteins of proteoglycans [18][19][20][21]. CS contains disaccharide units consisting of GlcUA and N-acetylgalactosamine (GalNAc), which are commonly sulfated at C-4 and/or C-6 of the hydroxy group on GalNAc residues [22][23][24]. DS differs from CS by containing iduronic acid (IdoUA) in place of GlcUA [22][23][24]. DSE catalyzes the conversion from GlcUA to IdoUA in CS/DS [17,22]. D4ST1 catalyzes the transfer of a sulfate group from the sulfate donor 3′-phosphoadenosine 5′-phosphosulfate to the C-4 position of the GalNAc residue in DS chains in vivo and in vitro [16,25,26] (Figure 1). DS and CS chains frequently exist as CS/DS hybrid chains in mammalian cells and tissues and covalently attach to a core protein to form a proteoglycan. Decorin and biglycan are representative core proteins of these GAG chains [27,28]. These proteoglycans connect collagen fibrils via GAG chains and influence the structure and functions of the extracellular matrix [27,29]. DS was not detected in the urine or skin fibroblast of patients with mcEDS-CHST14 [30,31]. DS was absent or significantly decreased in skin fibroblasts and urine derived from patients with mcEDS-DSE [7,13,32]. Thus, it is thought that the loss or a decrease in DS is a critical cause of mcEDS symptoms.
Animal models are essential for investigating human disease mechanisms and developing therapies. Chst14 and Dse gene-deleted mice are expected to be model animals for mcEDS-CHST14 and mcEDS-DSE, respectively. In this review, we organize the published data regarding mcEDS and model animals.

Dse Gene-Deleted Mouse
A Dse gene-deleted mouse was generated by Maccarana et al. in 2009 [34]. A structural analysis of the CS/DS chains revealed a reduction, but not dissipation, of IdoUA blocks, which are characteristic of DS in the whole body and the skin of Dse homozygous gene-deleted mice (Dse -/-) [34]. Although the Dse gene was deleted and DSE expression was not detected in Dse -/-, partial epimerase activity was observed in some tissues, such as skin, spleen, lung, kidney, and brain tissues, which suggested an influence of enzyme activity of DSEL [34]. The birth rate of Dse -/mixed C57BL/6-129/SvJ genetic background mice was slightly lower than that of Mendelism (offspring of heterozygous breeding pairs: 33% wildtype, 49% heterogeneous gene-deleted, and 18% Dse -/-) [34]. Conversely, perinatally lethality was observed in Dse -/mice with a pure C57BL/6 genetic background [35]. Pups of Dse -/mice were smaller (20 to 30% lighter body weight) than their wildtype littermates [34]. Adult Dse -/mice were also smaller (a 5 to 10% shorter crown-to-rump length and 10% lighter) than their wildtype littermates [34]. Kinked tails were observed in both the pups and adults of the mice [34,35]. The embryos of Dse -/mice showed an abdominal wall defect with herniated intestines [35]. The thickness of the skin epidermal layer was increased in newborns of Dse -/pups compared with heterozygote and wildtype littermates [35]. Newborns of Dse -/mice showed an increase in keratin 5 protein expression in the skin basal layer and thicker epidermal layers compared with heterozygote and wildtype littermates [35]. Contents of collagen and decorin, a representative core protein of DS-proteoglycans, were not significantly different from those of the wildtype mice [34]. However, the reduced tensile strength of the skin (41% reduction), with an altered ultrastructure of collagen fibrils, such as irregular outlines and a shift toward thicker fibrils, was observed in Dse -/mice [34]. The immune response against ovalbumin and the migration of dermal dendritic cells to skin-draining lymph nodes were decreased in Dse -/mice [36]. After the ovalbumin injection, there were fewer CD4+ T cells and CD8+ T cells in the spleens of Dse -/mice than in their wildtype littermates [36]. In Dse -/mice, there were also fewer ovalbumin-specific B cells and immunoglobulins than in wildtype littermates [36].

Dsel Gene-Deleted Mouse
Bartolini et al. generated a Dsel gene-deleted mouse (Dsel -/-) in 2012 [37]. The gene expression of Dsel is highest in the kidneys and brain, which are tissues with the lowest expression of Dse in mice [38]. Epimerase activity and IdoUA content decrease in the brain and kidneys of mutated mice compared with wildtype littermates [37]. However, Dsel -/mice show no anatomical, histological, or morphological abnormalities [37]. The lifespan of the Dsel -/mice was normal [37]. The body weights and lengths of mutated and wildtype littermates were indistinguishable [37]. The brain extracellular matrix architecture did not change in the Dsel -/mice [37].  [34,37]. They generated a double knockout (DKO) mouse with Dse and Dsel genes (Dse -/-; Dsel -/-) by crossbreeding these gene-deleted mice in 2015 [39]. This DKO mouse had complete loss of the IdoUA residue of CS/DS in 2-day-old pups, which suggested a correlation between DSE and DSEL in mice [39]. Dse heterogenous knockout (Dse +/-); Dsel -/mice were mated, and the embryos showed Mendelian ratios on E13.5-E19.5 [39]. Major organs, including the brain, spinal cord, liver, and lungs, in DKO embryos appeared normal [39]. However, the newborns of the DKO mice died in the perinatal period after birth [39]. Some Dse +/-; Dsel -/and DKO embryos showed umbilical hernia, exencephaly, and a kinked tail [39]. Although the development of secondary lymphoid organs, such as the lymph nodes and the spleen, were normal, the binding of the chemokine (C-X-C motif) ligand 13, an organogenetic modulator of lymphoid, to the DKO embryonic fibroblasts was impaired [39]. The phenotype of connective tissues in DKO mice remains unclear.
Yoshizawa et al. analyzed the prenatal period of the mice [43]. Embryos of Chst14 -/-(E18.5) mice had a shorter crown-rump length than their Chst14 +/+ and Chst14 +/littermates, whereas body weight was unchanged [43]. The gene expression of Chst14 and the content of DS were significantly decreased in the placentas of Chst14 -/embryos [43]. Because the placenta is known to be an organ that is composed of fetal (derived from the epiblast of the Chst14 -/embryo) and maternal components (derived from the endometrium of the Chst14 +/mother mouse), and thus Chst14 and DS derived from the maternal placenta were detected, the gene expression of Chst14 and the content of DS are not completely absent in the analysis of whole placenta. The placentas of Chst14 -/embryos exhibited reduced weight and ischemic and/or necrotic-like changes [43]. The percentage volume of the placental villus on the labyrinth zone was significantly decreased in the placentas of Chst14 -/embryos [43]. In the placental villus (fetal placenta), an alteration in the vascular structure, with an abnormal structure of the basement membrane of the capillaries, was observed [43]. The basement membranes of capillaries showed decreased thickness and/or rupture [43]. The gene expressions of decorin and biglycan were not different between the whole placentas of Chst14 -/and wildtype mice [43]. In immunostaining, biglycan was observed at the capillary basement membranes in both Chst14 -/and wildtype mice, whereas decorin was not detected [43]. The gene expression of collagens (Col1a, Col2a, Col3a, Col4a, and Col15) was unchanged [43]. These findings suggest that Chst14 is involved in placental development, capillary functions, and fetal survival.
Hirose et al. performed a detailed histological analysis of the skin from Chst14 -/mice [46]. The tensile strength of the skin was decreased to one-fourth in Chst14 -/mice compared with Chst14 +/+ mice [46]. The thicknesses of the epidermis and dermis were not different between Chst14 -/and Chst14 +/+ [46]. In the dermal reticular layer, the boundary between adjacent collagen bundles was unclear, showing loosened collagen bundles and scattered collagen fibers in Chst14 -/mice, whereas the boundary was clear in Chst14 +/+ mice [46]. In transmission electron microscopy, collagen fibrils were oriented in various directions to form disorganized collagen fibers in Chst14 -/mice, whereas these fibrils were oriented uniformly to form well-organized fibers in Chst14 +/+ mice [46]. In scanning electron microscopy, disorganized collagen fibrils, which were twisted and turned in various directions, were found in the skin of Chst14 -/mice [46]. Through transmission electron microscopy-based cupromeronic blue staining to visualize GAG chains, rod-shaped linear GAG chains were attached, at one end, to collagen fibrils and protruded outside of the fibrils, in contrast to the round and wrapping collagen fibrils in Chst14 +/+ mice [46]. These results suggest that DS deficiency reduces tissue strength with the deformation of GAG chains and collagen fibrils of the extracellular matrix in mice.
Nitahara-Kasahara et al. generated two strains of Chst14 gene-deleted mice with CRISPR/Cas9-mediated mutations (−1 bp or +6 bp/−10 bp). In these strains, phenotypes similar to other Chst14 -/strains, such as growth impairment and skin fragility, were observed [42,44,46]. These mutations in Chst14 induced the loss of D4ST1 activity and DS deletion [42]. The CS amount was increased sixfold in the skeletal muscles of the mutant mice [42]. The expression of decorin, a core protein of DS, was significantly decreased in the skeletal muscles of Chst14 -/mice [47]. Although decorin was present in the muscle perimysium of wildtype mice, decorin was distributed in both the muscle perimysium and endomysium of Chst14 -/mice [47]. Mutations in Chst14 induced myopathy-related phenotypes and thoracic kyphosis [42]. The grip strength of the mutant mice was significantly and consistently weaker than that of heterogeneous mutant and wildtype mice at 2 to 12 months of age [42]. The mutant mice showed significantly decreased motor functions, such as voluntary activity and running speed [42]. In pathological analysis, Chst14 deficiency induced myopathy phenotypes, including a predominance of small muscle fiber sizes and type I muscle (slow muscle) fibers, compared with Chst14 +/+ and Chst14 +/mice [42,47]. Thoracic kyphosis was detected in 1-year-old Chst14 mutant mice [42]. It was unclear whether the cause of the thoracic kyphosis was an influence of skeletal muscle or a direct effect of DS deficiency in the bone.
Akyuz et al. analyzed the function of DS in the mouse nervous system [44]. The ablation of Chst14 did not affect brain weight or gross anatomy [44]. In cultured cells of Chst14 -/mice, Schwann cells formed longer processes and showed increased proliferation compared with Chst14 +/+ cells [44]. Neurite lengths of cerebellar neurons and motoneurons were significantly increased in the cultured cells of Chst14 -/- [44]. After femoral nerve injury, functional recovery in Chst14 -/mice was significantly accelerated [44]. In Chst14 -/mice, the axonal growth rate of motoneurons was accelerated after the injury, suggesting an enhanced regeneration rate [44]. After the injury, the gene expression of CS/DS proteoglycans, such as decorin, biglycan, epiphycan, neuron-glial antigen 2, and tenascin-c, was decreased in the lumbar spinal cord of Chst14 -/and Chst14 +/+ compared with that of uninjured mice [44]. In uninjured mice, gene expression of biglycan, tenascin-c, and epiphycan was changed in Chst14 -/compared with Chst14 +/+ , suggesting an influence of DS deficiency on the composition ratio of core proteins [44]. Rost et al. analyzed the spinal cords of Chst14 -/mice [48]. No differences were found in the size of the spinal cords or the numbers of microglia and astrocytes [48]. After severe compression injury of the spinal cord, regeneration was reduced in Chst14 -/mice compared with Chst14 +/+ littermates [48]. Li et al. analyzed cognitive functions and hippocampal synaptic plasticity in Chst14 -/mice [49]. Chst14 -/mice showed deficits in spatial learning and memory, with the reduced expression of hippocampal proteins related to synaptic plasticity [49]. Protein expressions of synaptic proteins in the hippocampus, such as growth-associated protein 43, synaptophysin, N-ethylmaleimide-sensitive factor, N-methyl-D-aspartate receptor (NMDAR) 1, NMDAR2A, NMDAR2B, glutamate receptor 1, and postsynaptic density protein 95, was significantly decreased in Chst14 -/mice [49]. The phosphorylation of the intracellular signaling molecules Akt, mTOR, and S6 proteins in the hippocampi was significantly decreased in Chst14 -/mice [49].

Phenotypic Similarities and Differences of Patients with mcEDS and the Mouse Models
It is essential to understand the phenotypic similarities and differences between patients with mcEDS and the mouse models for a deep understanding of the pathophysiology. Most patients with mcEDS (>90%) have various symptoms, such as specific craniofacial (large fontanelle with delayed closure, downward slanting palpebral fissures, and hypertelorism), skeletal (characteristic finger morphologies, joint hypermobility, multiple congenital contractures, progressive talipes deformities, and recurrent joint dislocation), cutaneous (hyperextensibility, fine/acrogeria-like/wrinkling palmar creases, and easily bruised) and ocular (refractive errors) features [14]. Large subcutaneous hematomas, constipation, cryptorchidism, hypotonia, and motor developmental delay are also common (>80%) [14]. In particular, some skeletal, cutaneous, and vascular symptoms (e.g., progressive spinal and joint deformity and subcutaneous hematomas) seriously reduce the activities of daily living (ADL) and quality of life (QOL) of patients [4,14].
On the basis of the causative gene of mcEDS, Chst14 or Dse gene-deficient mice are expected to be typical animal models of the disease. However, mutations in the DSEL gene may be linked to bipolar disorder in humans [50]. Because an association between DSEL and mcEDS has not been reported, Dsel -/mice and Dse -/-; Dsel -/mice may not reflect the pathological mechanisms of mcEDS; however, they may be a useful model for investigating the function of IdoUA residue in CS/DS in vivo.
Epimerase activity was completely extinguished, not in Dse -/mice, but in Dse -/and Dsel -/-DKO mice, and thus, complementation between DSE and DSEL was suggested [34,39]. Conversely, Müller et al. reported that DSEL cannot compensate for the loss of DSE functions in mcEDS-DSE [7]. The amounts of DS disaccharides were decreased by 90.7% and 80.0% in the conditioned medium and the cell fraction from the cultured fibroblasts of a patient with mcEDS-DSE compared with a healthy control subject, respectively [7]. Similarly, although the complete loss of epimerase activity was not observed, Dse -/mice showed mcEDS-like phenotypes, such as skin fragility and dermal collagen fibril deformity [34]. These reports suggest that DSE deficiency induces a dramatic decrease in DS and the phenotypes of mcEDS-DSE and Dse -/because insufficient compensation with DSEL is inadequate to improve the symptoms.
Comparisons between phenotypes of patients with mcEDS-CHST14 or mcEDS-DSE and Chst14 and Dse gene-deficient mice are shown in Table 1. The embryonic lethality of Chst14 -/mice was higher than that of Dse -/mice with a mixed C57BL/6-129/SvJ genetic background [34,43,44]. The body weight of the embryo or pups was reduced by 20-30% in Dse -/mice compared with their wildtype littermates, but it was unchanged in Chst14 -/mice [34,43]. Body weights of adult Dse -/and Chst14 -/mice were significantly decreased compared with wildtype littermates [34,42,44]. In humans, the influence of mutations in DSE or CHST14 on embryonic survival has not been reported. Patients with mcEDS-CHST14 showed mild prenatal growth impairment, with a mean birth length of −0.5 SD and a median of −0.6 SD and a mean birthweight of −0.6 SD and a median of −0.67 SD [51]. Postnatal growth was also mildly impaired in terms of slenderness, with a mean height of −0.9 SD and a median of −0.6 SD and a mean weight of −1.5 SD and a median of −1.4 SD [51]. Patients with mcEDS-DSE had a normal birth weight [13]. In adults, a mildly decreased body weight (−1.2 SD) and normal height were observed [13]. Table 1. Comparisons between reported phenotypes of the model mice and symptoms of mcEDS.
The numbers in the parenthesis show the frequency of the symptom among the patients. The square brackets indicate the references. +, positive; −, negative; ?, unclear; *, dependent on genetic background; #, in embryos; $, in adults; †, capillary deformities were observed in the placenta; §, tail deformities were also observed.
Skin fragility was observed in 90% and 29% of patients with mcEDS-CHST14 and mcEDS-DSE, respectively [14]. Hyperextensibility of the skin was detected in 100% and 57% of patients with mcEDS-CHST14 and mcEDS-DSE, respectively [14]. The tensile strength of the skin was decreased by 41% and 75% in Dse -/and Chst14 -/mice compared with wildtypes, respectively [34,46]. In the pathological observation of the skin specimens, dispersed collagen bundles were similarly observed in patients with mcEDS-CHST14 or mcEDS-DSE as well as in Chst14 -/mice [11,13,46,52]. Pathological observation of the skin specimens of Dse -/mice did not mention the dispersion of skin collagen fibrils, but showed an increased diameter of collagen fibrils [34].
One of the most serious symptoms of mcEDS was a large subcutaneous hematoma, which 81% and 67% of patients with mcEDS-CHST14 and mcEDS-DSE had experienced, respectively [4,14]. A subcutaneous hematoma-like phenotype has not been reported in mcEDS model mice. The evaluation of hemorrhage requires attention because the hemostatic system of mice is generally stronger than that of humans. The structure and function of blood capillaries were also regarded as causative factors of subcutaneous hematoma. In Chst14 -/embryos, an abnormal structure of the capillary basement membrane in the placenta was observed [43]. Further analysis of blood capillaries in model animals and patients is expected to clarify the mechanisms of subcutaneous hematoma.
Neuromuscular symptoms were detected in some mcEDS patients. Hypotonia was observed in 86% of patients with mcEDS-CHST14 and 75% of those with mcEDS-DSE [14]. A tethered spinal cord was observed in 39% of patients with mcEDS-CHST14 [14]. Relationships between the cranial nervous system and mcEDS-DSE are unclear, with only a few case studies. Chst14 -/mice showed hypotonia-related phenotypes, such as decreased grip strength, voluntary activity, and running speed [42]. Furthermore, Chst14 -/mice showed deficits in spatial learning and memory, while brain weight was unchanged [38,49]. No differences were found in the size of the spinal cords or the numbers of microglia or astrocytes in Chst14 -/mice [48]. The neuromuscular phenotype of Dse -/mice has not been clarified.

Consideration of the Pathogenic Processes of mcEDS
Analysis of Chst14 -/mice has made steady progress, and models of severe mcEDS symptoms (e.g., skin fragility and skeletal deformity) have been investigated. Only one report has focused on the skin phenotype of Dse-deleted mice [34]. Gene mutations in CHST14 and DSE induced DS deficiency in common [7,11]. DS deficiency is a common pathophysiological cause of mcEDS-CHST14 and mcEDS-DSE. Because DSE and CHST14 are both related to DS synthesis, the phenotype of Chst14 -/mice may also be useful to understand the mechanisms of mcEDS-DSE symptoms.
Neuromuscular symptoms were detected in some mcEDS patients. Hypotonia w observed in 86% of patients with mcEDS-CHST14 and 75% of those with mcEDS-DSE [1 A tethered spinal cord was observed in 39% of patients with mcEDS-CHST14 [14]. Re tionships between the cranial nervous system and mcEDS-DSE are unclear, with only few case studies. Chst14 -/-mice showed hypotonia-related phenotypes, such as decreas grip strength, voluntary activity, and running speed [42]. Furthermore, Chst14 -/-mi showed deficits in spatial learning and memory, while brain weight was unchang [38,49]. No differences were found in the size of the spinal cords or the numbers of micr glia or astrocytes in Chst14 -/-mice [48]. The neuromuscular phenotype of Dse -/-mice h not been clarified.

Consideration of the Pathogenic Processes of mcEDS
Analysis of Chst14 -/-mice has made steady progress, and models of severe mcED symptoms (e.g., skin fragility and skeletal deformity) have been investigated. Only o report has focused on the skin phenotype of Dse-deleted mice [34]. Gene mutations CHST14 and DSE induced DS deficiency in common [7,11]. DS deficiency is a comm pathophysiological cause of mcEDS-CHST14 and mcEDS-DSE. Because DSE and CHST are both related to DS synthesis, the phenotype of Chst14 -/-mice may also be useful understand the mechanisms of mcEDS-DSE symptoms.
Patients with mcEDS-DSE, mcEDS-CHST14, and Chst14 -/-mice showed various com mon symptoms, such as growth impairment, skin fragility, spinal and talipes deformitie hypotonia, and motor developmental delay [11,13,14,[42][43][44]46,51,52]. Growth impairme and skin fragility were also reported in Dse -/-mice [34]. The pathogenic process of sk fragility has been well investigated in patients and mouse models. Patients with mcED DSE and mcEDS-CHST14, and Chst14 -/-and Dse -/-mice, showed skin hyperextensibili and fragility with collagen fibril deformity [11,13,34,46,52]. The extracellular matrix is i volved in tissue strength [53]. Collagen is known as an important component of the extr cellular matrix, and mutations in fibrous collagen genes (Col1a1, Col1a2, Col3a1, Col5a1, Col12a1) induce tissue fragility [1]. It was suggested that DS deficiency induced skin h perextensibility and fragility with structural changes of the GAG chains and collagen brils in the mice [46]. These reports suggest that DS is an important factor for tiss strength and morphology ( Figure 2).  Mechanical changes of the tissues caused by DS deficiency are insufficient to explain all pathogenic processes of the symptoms of the patients and the mice. GAGs act as coreceptors at the cell surface and interact with cell growth factors [54]. DS is an activating factor and/or co-receptor of fibroblast growth factors (FGFs) in vitro [55,56]. The suppression of FGF signaling via mutations in FGF receptor 3 causes achondroplasia, which involves growth impairment [57][58][59][60]. These reports suggest that DS influences the activity of cell growth factors and is related to growth impairment in mcEDS ( Figure 2).
Patients with mcEDS and Chst14 -/mice showed hypotonia and motor developmental delay [14,42]. Motor functions are provided by skeletal muscles and the nervous system. In the skeletal muscles of Chst14 -/mice, Nitahara-Kasahara et al. found muscle growth suppression with the upregulated expression of myostatin, a negative regulator of protein synthesis in the muscle [47]. The skeletal muscle of Chst14 -/mice showed fibrosis with increased expressions of transforming growth factor-β (TGF-β) and a decreased expression of decorin [47]. A DS deficiency induced cells to form longer processes and increased the proliferation of cultured Schwann cells from Chst14 -/mice, suggesting that DS was a regulatory factor of neural development [44]. Peripheral nervous system functions for motor functions in mice have not been elucidated.
These reports suggest that DS is essential for tissue strength, morphology, cell growth, and motor functions (Figure 2).

Conclusions
In this review, we discussed the reported phenotypes of Chst14 -/and Dse -/mice, which were expected to be model animals of mcEDS, and the relationships with symptoms of mcEDS and pathogenic processes. Careful observations of patients and model animals are important to investigate the pathophysiological mechanisms and therapeutic strategies of rare diseases. There are still some unexplained pathogenic processes and molecular mechanisms of mcEDS and the model mice. A more detailed analysis of the mechanisms of the phenotypes and therapies is expected in the future.