Kei Semba and Ken-ichi Yamamura*
Division of Developmental Genetics, Institute of Resource Development and Analysis, Kumamoto University, Kumamoto 860-0811, Japan
Received Date: May 08, 2013; Accepted Date: May 22, 2013; Published Date: May 24, 2013
Citation: Semba K, Yamamura Ki (2013) Etiology of Caudal Regression Syndrome. Human Genet Embryol 3:107. doi: 10.4172/2161-0436.1000107
Copyright: © 2013 Semba K, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Human Genetics & Embryology
Caudal regression syndrome (CRS) is a rare congenital disorder in which lumbosacral anomalies are combined with anorectal and urogenital malformations. However, the molecular mechanisms of human CRS are not yet known. Trauma, nutritional problems, toxic agents, and genetics are suggested in the etiology of CRS. To the best of our knowledge, linkage studies of families affected exclusively by CRS or total sacral agenesis have not been conducted. In spite of the small number of familial cases reported, some specific genes have been shown to cause defined phenotypes. Environmental factors also may act as an enhancer in the etiology for CRS. There are several mutant mice that are considered as models for CRS, showing characteristic vertebral, anorectal, and urogenital abnormalities. Understanding the mechanisms for CRS development gives us valuable information to understand better what mutations may cause or contribute to CRS in humans. This review highlights the current evidence that pinpoints the link to the etiology of CRS.
Caudal regression syndrome; Etiology; Mutation; Diabetes; Animal model
Caudal regression syndrome (CRS) is characterized by sacrococcygeal or lumbosacrococcygeal agenesis, of variable extent, most often accompanied by multiple musculoskeletal abnormalities of the pelvis and legs. In addition, various other malformations, vertebral and non-vertebral, have occasionally been reported to be associated with the complex [1-5]. CRS occurs at a rate of approximately one per 25,000 live births [6,7]. The condition is caused by some factor or set of factors present during weeks 3-7 of fetal development. Formation of the sacrum/lower back and corresponding nervous system is usually nearing completion by week 4 of development. Owing to abnormal gastrulation, the mesoderm migration is disturbed. This disturbance results in symptoms varying from minor lesions of the lower vertebrae to more severe symptoms such as complete fusion of the lower limbs, also known as sirenomelia or mermaid syndrome. CRS is a heterogeneous disorder with respect to its etiology and developmental pathogenesis. In fact, caudal regression is caused by various gene mutations in mice (Table 1) [8-33]. In addition, it is clear that environmental factors and an underlying genetic predisposition are involved in CRS.
Table 1: List of CRS model mutant mice.
CRS is a rare and usually sporadic disorder. It comprises developmental anomalies of the caudal vertebrae, neural tube, urogenital and digestive organs, and hindlimbs; the precursors of which are derived from the caudal eminence. This may result in various types of anorectal malformations (ARMs), agenesis of spinal segments (usually sacral or lumbosacral), and multiple visceral anomalies. In the most severe cases, the lower limbs are fused [1-5]. Diagnosis can be made in the first trimester by noting the short crownrump length. Distal vertebral anomalies and fetal spinal anatomy may be seen by obstetric ultrasonography, and in the intrauterine period, amnioinfusion may be important, especially in cases associated with oligohydramnios. It may help in detecting associated genitourinary and gastrointestinal anomalies. The superiority of lumbosacral magnetic resonance imaging for diagnosis, both antenatal and postnatal, and for classification, is generally accepted today. It also readily detects various associated anomalies [34-46].
The simple form of CRS is sacral agenesis. Renshaw classified the spectrum of sacral agenesis into five types based on type of defect and articulation between bones . Type I has total or partial unilateral sacral agenesis; type II has variable lumbar and total sacral agenesis and the ilia articulate with the sides of the lowest vertebra; type III has variable lumbar and total sacral agenesis, and the caudal end plate of the lowest vertebra rests above fused ilia or an iliac amphiarthrosis; type IV has fusion of soft tissues in both lower limbs; and type V, also known as sirenomelia or mermaid syndrome, has a single femur and tibia.
Several associated anomalies of other systems are frequently present and complicate the picture of CRS. For example, CRS can be found in association with OEIS complex (omphalocele, exstrophy of the bladder, imperforate anus, and spinal defects) [48,49].
VATER association was originally named in the early 1970s with the description of seven patients who had at least three of the following features: vertebral defects, anal atresia, tracheoesophageal fistula, and radial and renal dysplasia . Shortly thereafter, additional features, such as cardiac malformations and additional limb abnormalities, were added, and the condition was called VACTERL syndrome, which remains the more prevalent term according to our survey [50-56]. However, the precise definition has remained uncertain .
In 1981 Guido Currarino described the triad of hemisacrum, ARM (rectoperineal fistula, rectourethral fistula, rectovestibular fistula, or rectocloacal fistula) and presacral mass (teratoma, anterior meningocele, rectal duplication, or a combination of these). Currarino syndrome (CS) has great phenotype variability. Many cases present with asymptomatic hemisacrum without other anomalies, whereas others present with a complete spectrum of malformations and other associated anomalies [58-60].
Ocular findings can also be described together with CRS, including hypertelorism, epibulbar dermoid, anophthalmia infantile glaucoma, and punctal atresia .
Three genetic and association studies on CRS suggested the involvement of genes such as HOXD13 , CYP26A1 , and HLXB9 . Although mutations in SHHand GLI3 are associated with human syndromes that may include ARM as part of the phenotypic spectrum, such as holoprosencephaly type 3 (OMIM #142945) and Pallister-Hall syndrome (OMIM #146510), SHH and GLI3 mutations have not been identified in CRS. However, mutations in the human HOXD13gene, which is an SHH target during gut morphogenesis, are associated with limb malformations, including synpolydactyly 1 (SPD1; OMIM #186000), brachydactyly types D (BDD; OMIM #113200) and E (BDE; OMIM #113300), syndactyly type V (OMIM #186300), and a novel brachydactyly-syndactyly syndrome . Thus, HOXD13 may not only be implicated in limb malformations, but also in the development of gut and genitourinary structures.
Retinoic acid (RA) is a key determinant of vertebrate embryo patterning and organogenesis [66,67]. The spatiotemporal distribution of embryonic RA results from regulated expression of RA-synthesizing enzymes (RALDH1, RALDH2, and RALDH3)  and RA-catabolizing enzymes (CYP26A1, CYP26B1, and CYP26C1) [69-72]. Differences in the genotype and allele frequency of each single nucleotide polymorphism (SNP) in 49 CRS patients and 132 controls were analyzed using 8 SNPs . However, both single-locus and haplotype analyses revealed no association with increased risk for CRS, although two intronic polymorphisms affected the intron splicing efficiency. The relationship between CRS risk and CYP26A1genotype requires further study with a larger number of genotyped subjects.
Three previous reports have shown HLXB9 to be a major causative gene for CS [73-75]. HLXB9 has three exons and encodes a protein of 403 amino acids. A polymorphic GCC repeat coding for a polyalanine stretch occurs in exon 1 of HLXB9. To date, no gene other than HLXB9 has been shown to be involved in sacral malformations. Several studies reaffirm the lack of associations between HLXB9 mutations and several caudal regression variants. Thus, many other genes will be involved in CRS. In contrast to humans, a null mouse mutation of the HLXB9 gene showed agenesis of the dorsal pancreas but no skeletal truncation . It is possible that gain of function, but not loss of function, may induce CRS-like disease with sacral malformations.
In addition, three studies have shown an association between chromosomal abnormalities 18p11.2 deletion and micro duplication at 22q11.21 and CRS, and between duplication-deletion of distal 7q and CS (Table 2) [77-79]. In the 18p11.2 region, several genes such as CDH2, TTR, SMAD2, SMAD7, SMAD4, DCC, TCF4, FECH, NEDD4L, MC4R, and TNFRSF are found. In 22q11.21, genes such as SMARCB1, MIF, GSTT1, ADORA2A, CHEK2, NF2, TIMP3, HMOX1, MYH9, LGALS1, APOBEC3G, PDGFB, EP300, XRCC6, CYP2D6, and PPARA are found. It may be that a gene dosage effect involving the 22q11.2 region is involved in predisposition to association with VACTERL, but it is not clear whether these genes are directly involved in CRS. In the third case, there is a de novo 10.3-Mb duplication of 7q34-q35 (chromosome 7: 138,293,371-148,443,994) as well as an 8.8-Mb deletion on 7q36 (chromosome 7: 148,472,027-157,265,994) . The deleted region contains 70 RefSeq genes including HLXB9, SHH and EN2, while the duplication contains82 genes (Table 2). Loss of HLXB9 may account for the CS phenotype. Pathogenesis of human CRS is still unknown. Further investigations are warranted to determine the genetic causes of this complex malformation.
|Disease||Gene or genetic position||Association or evidence||Reference|
Table 2: Genetic research list for diseases related and similar to caudal regression syndrome.
Pathogenesis of human CRS is still unknown. Further investigations are warranted to determine the genetic causes of this complex malformation.
Overt type 1 diabetes around conception carries a marked risk of embryopathy (neural tube defects, cardiac defects, and CRS), whereas later in gestation, severe and unstable type 1 maternal diabetes carries a higher risk of intrauterine growth restriction, asphyxia, and fetal death. CRS is seen in 0.1-0.25:10,000 of normal pregnancies [6,7]. In contrast, about 1% of newborn infants born to diabetic mothers exhibit CRS, indicating a risk at least 250 times higher than that in the offspring of nondiabetic mothers . As a result, 15-25% of mothers who have children with CRS have insulin-dependent diabetes mellitus [81-86].
The possible role of the prediabetic state in the etiology of CRS is obscure. Landauer et al.  demonstrated that defective development of the tail skeleton orits complete absence (“rumplessness”) was induced inchickens by injection of insulin.
Apart from hyperglycemia, other environmental factors are presumably involved, as demonstrated by the rare incidence of CRS compared with that of diabetes. Certainly not all children born with CRS have diabetic mothers. However, it is not yet clear which other factors are involved. Further clinical investigations may clarify the relationship between maternal diabetes and CRS.
Various mouse mutants show that caudal agenesis occurs as a result of hypodevelopment of the anterior–posterior axis (Table 1) [8-33]. These mutant mice can be classified into four categories. Mutants with CRS phenotypes can be caused by mutation in: (1) Shh-related related signaling molecules such as Hoxa13, Hoxd13 [88-90] and Bmp4 [10,11,17,19]; (2) Cdx2 and its downstream signaling molecules such as Cyp26a1 and T; (3) Wnt and its related genes such as Wnt3a, Acd and Lrp6; and (4) Danforth’s short tail (Sd).
Shh protein is a regulator of mesodermal patterning during many aspects of development [91,92]. It acts locally on adjacent visceral mesoderm to induce expression of bone morphogenetic protein and the Hox genes . Shh appears to signal an epithelial–mesenchymal interaction in the developing hindgut. Hoxd13 is an SHH target during gut morphogenesis. Hoxd13 natural mutant mice and mice with genetically engineered inactivation of the most 5’ genes of the Hoxd cluster show disorganization of the anorectal region [94-96].
Cdx2 encodes homeodomain transcription factors related to the Drosophila gene caudal. Expression of Cdx2 in the embryo proper initiates at E8.5 in all germ layers of the posterior embryo, extending caudally into the base of the allantois and rostrally into the posterior neural plate, hindgut endoderm and unsegmented paraxial mesoderm. Expression of Cdx2 continues into the tail bud, the posterior neural plate and the endoderm, and is eventually confined to the hindgut endoderm located posterior to the foregut/midgut junction from E12.5 onwards, and perdures in the adult intestinal epithelium . Deficiency in Cdx genes causes body truncations and malformations of the caudal spine and urorectal system, a phenotype reminiscent of CRS in humans [12,25].
Savory et al.  demonstrated thatT, Wnt3a and Cyp26a1are direct Cdx2 targets. In accordance with this, mutant mice for these genes showed similar CRS phenotypes. T encodes a transcription factor, which was the first T-box gene to be molecularly characterized. Heterozygous T mutant mice exhibit a variably shortened tail, whereas homozygous T mutant mice die between E9.5 and E10.5. T-deficient embryos display an overall loss of mesoderm caused by failure of epiblast cells to ingress through the primitive streak. Consequently, embryos have impaired axial development and allantoic defects; the latter of which are thought to be the cause of death [98-100]. Homozygous T mutant mice die during early development; therefore, this T mutant cannot be used to analyze the effect of T on other tissues. Interestingly, the hypomorphic mutant was created using an inducible miRNA-based in vivo knockdown of Brachyury. These mice survive until birth and show skeletal defects as well as urorectal malformations resembling CRS .
Cyp26a1−/−mice showed caudal truncation, vertebra transformation, and hindbrain mis-patterning. Externally, Cyp26a1−/− mice exhibited a mermaid-like deformity (sirenomelia), with the caudal part of the body severely truncated and the anterior portion remaining relatively normal. The hindlimbs were typically fused at the midline and the tail was missing. Spina bifida was also apparent. Postmortem examination of abdominal organs of Cyp26a1−/− mice revealed aplasia or hypoplasia of the urogenital system, including the kidneys. The posterior end of the hindgut was often terminated as an appendix. Histological analysis of embryos confirmed the aplasia or hypoplasia of the urogenital system. The kidneys were often fused, resulting in the development of horseshoe kidney . These phenotypes in Cyp26a1−/− mouse mimicked those generated by excess RA administration. As described before, further investigation is required to provide evidence that CYP26A1 has implications for the pathogenesis of human CRS.
The Wntfamily, which encodes cysteine-rich secreted glycoproteins,consists of at least 16 members in mice [101,102]. Wnt-3aexpression is detected first at 7.5 days post-coitum (dpc), extending along much of the length of the streak. Wnt-3anull mutant embryos lack all but the first seven to nine somites, have a disrupted notochordand pronounced central nervous system dysmorphology, and most die before 12.5 dpc . A mutation in the Wnt3a gene results in a lack of caudal somites, a disrupted notochord, and failure to form a tail bud [11,104]. Interestingly, Wnt3a mutants fail to express T  and T mutants fail to express Shh . However, the downstream mechanisms via T are still not entirely clear.
Adrenocortical dysplasia (acd) is a spontaneous autosomal recessive mouse mutant with developmental defects in organs derived from the urogenital ridge. Keegan et al.  identified a splice donor mutationin a gene, which is the mouse ortholog of a newly discovered telomeric regulator. This gene, termed Acd, is a component of the TRF1 protein complex that controls telomereelongation by telomerase and is widely expressed in adult tissues and throughout embryonic development. On the DW/J background, a striking caudal regression and limb malformation phenotype is observed in mutant embryos, resembling CRS in humans. Interestingly, expression of Wnt3aand Dll1 at E11.5 is reduced in the tip of the developing tail. The extent of reduction in expression of Wnt3aand Dll1 correlates with severity of the abnormal morphology of the tail bud. This suggests that Acd functions upstream of Wnt3a and Dll1 in the developing tail bud directly or indirectly. The caudal truncation observed in some acd homozygotes is similar to that observed in Wnt3a–/Wnt3avt or Wnt3avt/Wnt3avtmice .
The LRP6 gene encodes a member of the low-density lipoprotein receptor (LDLR) gene family  and Lrp6 is required during Wnt/b-catenin signaling, acting as a co-receptor for Wnt molecules. LRP6 is internalized with caveolin in human cell lines, and this endocytic pathway is required for WNT3A-induced internalization of LRP6 and accumulation of b-catenin .Targeted disruption of Lrp6 results in widespread developmental defects that include caudal truncation, limb defects and urogenital anomalies . The specific developmental defects in homozygous embryos are remarkably similar to mice carrying mutations in the Wnt3a gene [11,103]. LRP6 mutant mice do not exhibit all of the Wnt phenotypes reported, for example, the early embryonic lethality associated with Wnt-3mutants . Therefore, LRP6 may be involved in signaling by only a subset of vertebrate Wnts. In any case, LRP6 is required for efficient signaling by several Wnt proteins.
Finally, Danforth’s short tail (Sd) mutant mice accurately model human caudal malformation and regression syndromes. In recent studies, three laboratories including our research group independently identified the Sd mutation as a ETn retrotransposon insertion upstream from the Ptf1a gene [30-33]. Ptf1a encodes a cell-type-restricted basic helix–loop–helix transcription factor required for development of the pancreas and cerebellum [110,111]. We demonstrated an insertion of one type of retrotransposon 12 kb upstream of the Ptf1a gene. This resulted in ectopic expression of Ptf1a gene in the caudal region of the embryo and downregulation of Cdx2 and its downstream targets, Cdx2, T, Cyp26a1 and Wnt3a, leading to characteristic phenotypes in Sd mouse. The shortened tail phenotype of the Sd mutant is similar to that of the Cdx2, T, Cyp26a1, Wnt3a, Acd and Lrp6 mutants. Human CRS is a heterogeneous disorder with respect to its etiology and developmental pathogenesis, thus, Sd mutant mice are considered the best model for human CRS. In addition, Sd mutants show phenotypic similarity in terms of abnormalities in the spine, hindgut, and urogenital system. Taken together, these results suggest that the phenotypes observed in Sd mice are caused by combination of partial deficiency of Cdx2 and its downstream target genes.
One puzzling question remains. As described before, a null mutation of the Hlxb9 gene in mice showed agenesis of the dorsal pancreas but no skeletal truncation . Various types of mutation, such as insertion, deletion or substitution, were found resulting in nonsense mutation, frame shift mutation, or abnormal splicing . Thompson et al. showed that both Ptf1a and Hlxb9 were expressed in the pancreas, and that Ptf1a was required for Hlxb9 expression . Although it is not clear whether PTF1A is involved in human CRS or CS, the evidence from this study implies that misexpression of PTF1A and possibly HLXB9 in the caudal abnormalities may cause CRS. Further studies using Sd mice will provide insight into the development of human CRS.
Animal experiments have shown that CRS-like syndrome could be induced by RA, adriamycin, and hyperglycemia [113-118]. However, the exact etiology is unknown. In this syndrome, chromosomal studies have been normal, with few exceptions. Similar to human CRS, the offspring of diabetic mice also exhibit poor development of the lower part of the body, including genitourinary dysmorphogenesis . Vestigial tail (Wnt3avt) mutants provide evidence that Wnt3a, a gene that controls the development of the caudal region, is directly involved in the pathogenic pathway of RA-induced caudal regression. The mutants further show that the increased susceptibility of embryos of diabetic mice to RA involves enhanced down regulation of Wnt3a. This positive interaction between RA and maternal diabetes may have implications for humans in suggesting increased susceptibility to environmental teratogens during diabetic pregnancy .
As discussed above, CRS is a heterogeneous disorder with respect to its etiology and developmental pathogenesis. In addition, both an underlying genetic predisposition and environmental factors are involved in CRS. The etiological mechanisms can be summarized as follows. As shown in Figure 1, at least 4 pathways seem to be involved in CRS. One is the Shh signaling pathway in which the Shh downstream target, Hoxd13, is involved in caudal formation. Second, the Cdx2 pathway including T, Cyp26a1, and Wnt3a, plays an important role in caudal formation. Third, the Acd pathway involving also Wnt3a is responsible for caudal formation. Fourth, the Lrp6 pathway is involved through regulation of Wnt/b-catenin signaling. At present, the pathway of Hlxb9 is totally unknown. In any case, it is plausible that CRS is caused by a combination of partial deficiency of the molecules described in Figure 1.
Recently, next-generation sequencing technology has advanced rapidly. Using the newest techniques, whole-genome sequencing from five persons can be done within 10 days. Thus, it is possible to sequence whole genomes from 100 CRS patients. However, it is difficult to assess whether variations found in patients are responsible for the development of CRS. Therefore, there will be a strong demand for production of genetically engineered mice with particular genetic alterations.