Congenital spinal deformities are due to anomalous development of the vertebrae including failure of formation and segmentation during embryogenesis, and frequently associated with malformations of other organs (e.g. spinal cord, heart, lung, kidney, etc.) 
. Patients with congenital spinal deformities normally require observation and may need to undergo surgery in case of curve progression. The etiology and mechanism of congenital vertebral deformities have not yet been fully elucidated. In this regard, various animal models have hinted at the involvement of low oxygen, valproic acid, boric acid, cigarette smoking, fetal alcohol, hyperthermia, and vitamin B6 deficiency in the pathogenesis of congenital spinal deformities, although the exact cellular and molecular mechanisms are not well understood 
. The phenomenon that gestational vitamin A deficiency disrupts embryonic development has been known for over 80 years. Teratogenic effects of vitamin A deficiency and excess both involve skeletal morphogenesis, and these abnormalities include hypoplastic cranial bones, defects of the thyroid, cricoid and tracheal cartilages as well as agenesis of the neural arch of cervical vertebrae 1 and ectopic bone in the dorsal regions of C1, malformation of the sternal and pelvic regions 
. These abnormalities have been widely observed in human and animal models. However, the relationship between the congenital spinal deformities and vitamin A deficiency has been poorly understood. In the present study, we found that VAD is associated with congenital spinal deformities in rats as confirmed by anteroposterior radiography. Our findings support the hypothesis that maternal VAD could induce vertebral anomalies in the offspring. Our study has also observed other skeletal abnormalities include loss of the neural arch of cervical vertebrae 1, malformation of sternum, loss of the ribs, rib fusion, abnormal of forelimb and hindlimb, malformation of the pelvic regions, which have also been observed by others 
. Another important finding of our study is that VAD repressed the expression RALDHs and RARs, which are important components of RA signaling, in the liver and vertebral body of VAD rats. Although the number of rats in each group is small, the findings are novel and may be relevant to clinical practice, and the difference between the study and control group is very prominent and statistically significant.
Vitamin A is a dietary requirement because of the body's inability to synthesize sufficient quanitities. VAD is still a major public health problem worldwide, which has contributed substantially to the health threat among young children and women of reproductive age in developing coutries. VAD in pregnant women is associated with night blindness, severe anaemia, wasting, malnutrition, and reproductive and infectious morbidity, and increased risk of mortality 1–2 years following delivery 
. Several previous studies have demonstrated that VAD induces congenital heart, ocular tissues, respiratory, urogenital and circulatory systems, and skeletal deformities in experimental animals including monkeys, rabbits, rats, mice and hamsters 
. Excess dietary vitamin A, on the other hand, has been shown to also cause teratogenesis. Teratogenic targets of vitamin A excess were the heart, the skull, skeleton, limbs, brain, eyes, CNS, as well as craniofacial structures 
. However, toxicity of vitamin A excess from food sources is rare. Vitamin A supplies must be regulated to avoid teratogenic consequences from deficiency or over-intake and adequate maternal vitamin A level is important for newborn. To verify the reduction of vitamin A in mice receiving VAD diet, we measured the vitamin A level of mothers before mating and their offspring at postnatal day 1, and confirmed the VAD status during pregnancy in the VAD group as previously described 
. Previous studies have found that no reduction in food intake and body weight was observed in adult VAD rats 
. So in our study, there was no difference in body weight between the two groups during the pre-mating and parturition period. Our results show that VAD has little effect on the body weight of the neonate. Although much of skeletal abnormalities were reported as part of the VAD syndrome, interestingly, congenital spinal deformities were not reported in VAD model. The present study now clearly demonstrates that VAD during pregnancy also can induce congenital spinal deformities. As most individuals with congenital spinal deformities have additional congenital abnormalities (spinal cord, heart, lung, kidney, etc.), the underlying defect could be caused by VAD 
. Here, we demonstrated that maternal VAD induced vertebral anomalies in the offspring are associated with other skeletal deformities and also may be associated with abnormalities in other organs. The skeletal immaturity of the neonatal rats and the breed of Wistar rats and only by X-ray may account for this low incidence rate on skeletal and vertebral anomalies.
How might VAD be acting to interrupt signalling? RA is the active form of vitamin A and plays a crucial role in stimulating nuclear receptor signaling during development. RA synthesis is a two-step process. The first step, oxidation of retinol to retinaldehyde, is catalyzed by several members of the alcohol dehydrogenase family (Adh1, Ad3,and Adh4); the second step, oxidation of retinaldehyde to RA, is catalyzed by three members of the aldehyde dehydrogenase family (RALDH1, RALDH2, RALDH3) and is irreversible 
. RA synthesis is controlled both spatially and temporally and three RALDHs identified as catalysts for the second step are expressed in dynamic, nonoverlapping spatiotemporal patterns, indicating that this step is tissue- and time-restricted 
. In vitro studies have shown that RA serves as a ligand for two families of nuclear receptors, namely, RARs and retinoid X receptor (RXRs), both of which bind DNA as heterodimers and directly regulate gene expression 
. Previous studies have demonstrated that RA is required for embryogenesis and functions through RARs, whereas RXRs are undetectable in mouse embryos 
. In these studies, RA was shown to be essential for the development of several organs, including the hindbrain, spinal cord, heart, eye, skeleton, forelimb buds, lung, pancreas and genitourinary tract. Moreover, recent studies showed that loss of RA signaling results in left-right asymmetry of somites in mouse, chick, or zebrafish embryos, in which one side has fewer somites than the other 
. Administration of RA maternally to RA-deficient mouse embryos also restores normal axial turning and normal spinal column development 
. These findings suggest that RA signaling is implicated in somitogenesis as well as development of other organs and VAD induced reduction RA signaling may be a common cause of congenital spinal deformities and other organ defects.
Another important finding of our study was that VAD caused a decrease in the mRNA levels of RALDH1, RALDH2, RALDH3, RAR-α, RAR-β and RAR-γ in the liver and RALDH1, RALDH2, RALDH3, RAR-αand RAR-β in the vertebral body. Precious studies have demonstrated that RA was synthesized during vertebrate evolution by these three RALDHs (RALDH1, RALDH2 and RALDH3). RA also mediates its effects on embryogenesis exclusively through RARs but not RXRs 
. Robinson et al 
using a transcriptomic approach, they compared RA-exposed and nonexposed rat embryos to identify overlapping and nonoverlapping effects of RA on RNA expression, and their finding had indicated that 845 genes were identified to be significantly time-dependent altered by RA in parallel with morphpological and RA induced upregulation of expression of three enzymes, CYP26A1, CYP26B1, and DHRS3, which are known involved in the breakdown of RA. On the other hand, many reports have demonstracted that VAD suppressed RALDHs, RARs and RXRs mRNA expression in many organs of VAD rats 
. Moreover, the expressions of RAR-α and RAR-β are also severely downregulated in the VAD embryo in the avian 
. Our results also have observed that VAD suppressed RALDHs and RARs mRNA expression in the liver and vertebral body of VAD neonatal rats at the day of birth. Therefore, our findings suggest that maternal VAD during pregnancy induced congenital spinal deformities in offspring might also be associated with decrease in RA signalling caused by VAD.
A previous study has shown that scoliosis occurs in 74% of birds with vitamin B6 deficiency, while mild VAD did not influence expression of scoliosis. The authors found that the serum retinol levels of birds with mild VAD were 18±6 µg/dl. In their study, the birds were not severely deficient in vitamin A because the investigators used corn oil as the source of dietary lipid 
. However, in our study, we used the modified AIN-93G diet without any source of vitamin A and the mean serum retinol concentration in the adult females in the VAD group was 0.51±0.57 µmol/L, which was significantly lower than that of the control group (p<0.01).
The findings of the present study may have implications for understanding the etiology and mechanism of human congenital spinal deformities. Given that VAD is still a serious and prevalent public health problem in the world especially among pregnant women and children and dietary vitamin A is only available from limited sources, including vegetables in the form of β-carotene or meat as retinyl esters, VAD may be an important cause of human congenital spinal deformities 
. Three signaling pathways have been proposed to regulate the segmentation clock, namely the Notch, Wnt, and Fgf pathways. Thus far, mutations in the Notch ligand DLL3, LFNG,MESP2, Notch ligand JAGGED1 and polymorphisms in the Tbx6 gene have been identified to be associated with congenital scoliosis in humans 
. Strikingly, most of these genes are implicated in the modulation of segmentation clock. Our study shows that RA signaling is implicated in the pathogenesis of human congenital spinal deformities. The role of RA signaling in the regulation of segmentation clock, however, awaits further investigation. Moreover, not all rats in the VAD group develop congenital spinal deformities, suggesting that defects in RA signaling may only confer partial susceptibility.
In conclusion, we here present evidence that congenital spinal deformities may occur in neonatal rats whose mothers were exposed to VAD diet throughout the entire pregnancy. In mothers and neonatal rats, decreases in serum retinoid levels and reduced mRNA expression of RALDHs and RARs in liver and vertebral were noted. These results suggest that vertebral birth defects may be caused by VAD and a defect in RA signaling pathway during somitogenesis. This raises the possibility that an environmentally VAD induced reduction RA signaling paly an important role in the etiology of a wide range of human congenital spinal deformities. To our knowledge, this is the first report presenting the possible association between prenatal VAD and congenital scoliosis. However, large-scale epidemiological studies are still needed to confirm such association in humans. Moreover, effects of partial instead of absolute deprivation of vitamin A on spine malformation should also be examined. Future investigation into the molecular linkage between VAD and somitogenesis will also provide more insight into this phenomenon.