Special Report 19

Retinoic Acid and (hind) brain development

During neurodevelopment, the spinal cord contains the highest RA levels, while forebrain, midbrain and hindbrain contain very little RA (Horton and Maden, 1995). As RA cannot be synthesised de novo by embryonic or adult organisms, developmental RA supply is produced in the target tissue from maternal dietary retinol uptake. Retinol dehydrogenases (RDH) produce retinaldehyde from retinol, which is further metabolised by retinaldehyde dehydrogenases (RADH) to RA. However, due to lack of RADH2 expression, embryonic brain tissue does not produce RA from retinaldehyde itself, but mesodermal somites flanking the neural tube produce RA. This diffuses into areas of neuroectodermal tissue, which will form segmental units for future hind-, mid- and forebrain development. In the cranial part of the neural tube, RA-metabolising CYP26A1 is substantially expressed converting RA to 4-hydroxy-RA and 4-oxo-RA, the substrates for glucuronidation and excretion. Due to RADH2-dependent RA formation in more caudal areas and CYP26A1-reliant RA metabolism a RA gradient spans across the future hindbrain. This gradient is thought to determine hindbrain formation as Vitamin A-deficient embryos display a complete lack of the caudal hindbrain (Maden et al., 1996; White et al., 2000; McCaffery et al., 2003). During hindbrain development seven to eight rhombomeres form that relate to later defined hindbrain areas. Individual rhombomeres contain specific expression patterns for transcription factors including Wnt family members (reviewed by Marshall et al., 1996; Rijli et al., 1998), which facilitate the identification of the missing rhombomeres numbers four to seven in experimentally-induced RA-deficiency (McCaffery et al., 2003). Thus, caudal hindbrain development is dependent on RA homeostasis. Both RA deficiency and RA excess can produce developmental abnormalities, as shown in Figure 7 below.

Figure 7: Retinoids and brain development

screen-shot-2016-11-23-at-14-52-39

Recently, a clinical hypothesis has been proposed that RA deficiency also causes underdevelopment of the hindbrain in humans (Emmett and West, 2014). This hypothesis was based on the observations that hearing loss is a global public health problem, mainly in low- and middle-income countries, paralleled by Vitamin A deficiency in such developing areas (WHO 2009, 2013). While it is now well established that other reasons, like ear infections triggered by lack of RA, contribute to hearing loss, Emmett and West (2014) proposed the scientifically plausible, but virtually unexplored causal relationship between hearing loss due to RA-deficiency-dependent underdevelopment of the inner ear in humans. There seems to be a critical threshold for proper inner ear development in mice with low retinoid intake causing immature and/or ectopic otic vesicles (Niederreither et al., 1999). These abnormalities are likely due to the loss of RA-dependent regulation of hindbrain development and the otic morphogenic process (White et al., 2000; Maden et al., 1996). To test the hypothesis that this mechanism contributes to hearing loss in the human population, a Vitamin A-supplemented population study is to be planned by Emmett and West (2014).

Besides hindbrain development, additional processes of neurodevelopment are affected by RA-deficiency:

  • Decreased neurite outgrowth (reviewed in McCaffery et al., 2003).
  • Neural crest cell apoptosis (reviewed in McCaffery et al., 2003).
  • Abnormal dorsoventral patterning of the anterior spinal cord (reviewed in McCaffery et al., 2003).
  • Anterior-posterior patterning of the forebrain (reviewed in Rhinn & Dollé, 2012).
  • Cell survival in the telencephalon (reviewed in Rhinn & Dollé, 2012).

That different processes of neurodevelopment can be targets for disturbed RA homeostasis is important for the human relevance of RA-dependent signalling pathways for brain development. Targeted genetic manipulation or experimental pharmacological interference causes defined neurodevelopmental, histopathological phenotypes in quail and rodent embryos. In humans, such brain developmental phenotypes cannot easily be studied. However, human mutations in the Stra6 gene, a RA-inducible gene that regulates cellular retinol uptake, causes – amongst other severe developmental defects – mental retardation across all living cases (Chassaing et al., 2009) implying necessity of RA also for human neurodevelopment beyond inner ear development. Precise pathophysiological processes underlying human mental retardation due to intracellular RA-deficiency are not known.