Sulfate is essential for healthy growth and development (Dawson, 2011). Supplied from the diet and catabolism of sulfur-containing amino acids, intracellular sulfate is transformed into 3′-phosphoadenosine 5′-phosphosulfate (PAPS) which is the universal sulfate donor for sulfate conjugation (sulfonation) to a wide range of endogenous and exogenous molecules via sulfotransferases (Mulder and Jakoby, 1990). Sulfonation biotransforms the physiological properties of molecules, including: (i) inactivation of steroids, thyroid hormone and neurotransmitters (Darras et al., 1999; Richard et al., 2001; Dawson, 2012); (ii) detoxification of xenobiotics and certain pharmaceutical drugs (McCarver and Hines, 2002); and (iii) maintenance of tissue structure and function via the sulfate content of glycosaminoglycans (Yamaguchi, 2001). Sulfatases remove sulfate from substrates to maintain the required balance of sulfonated to unconjugated substrates (Hanson et al., 2004). Disturbances in any of these sulfation pathways has the potential for adverse physiological consequences.

In recent years, interest in human sulfate biology has expanded following animal studies that show adverse phenotypes linked to sulfate biology genes (Langford et al., 2017). Neurological dysfunction is one of the most predominant phenotypes linked to disturbances of sulfate biology. This is not surprising when considering that sulfonation biotransforms numerous target molecules in the brain, including inactivation of thyroid hormone, metabolism of neurotransmitters (e.g., serotonin, noradrenaline and dopamine) and modulation of neurosteroid actions on GABA_A, N-methyl-D-aspartate (NMDA) glutaminergic and ơ-opioid receptors (Rivett et al., 1982; Kríz et al., 2008). Sulfonation of proteoglycans (e.g., heparan sulfate) and cerebroside sulfate also plays an important role in maintaining the structure and function of brain tissue (Schwartz and Domowicz, 2018). Animal studies have also shown that hyposulfataemia leads to reduced sulfonation capacity and abnormal behavioral phenotypes, including impaired memory, increased anxiety and seizures in mice (Dawson et al., 2003; Dawson et al., 2004; Dawson et al., 2005). More recent studies have shown that reduced transfer of sulfate across the blood-brain barrier of neonatal mice leads to: (i) decreased heparan sulfate levels in the subventricular zone (SVZ) and rostral migratory stream; (ii) impairment in perineuronal net formation (which contains sulfonated proteoglycans) in the hippocampus and somatosensory cortex; (iii) increased neural stem cell proliferation and decreased neuronal maturation in the SVZ; (iv) seizures; and (v) impaired behavioral phenotypes, including impaired long-term spatial memory and defects in social interaction and social memory (Zhang et al., 2020). Despite the diverse roles for sulfate during neurodevelopment, most human sulfate biology genes have not yet been considered in clinical genetics.

To date, 91 genes are known to contribute to maintaining sulfate homeostasis, including those encoding: sulfate transporters; PAPS synthetases and transporters; enzymes in the pathway of sulfate generation; cytosolic and membrane-bound sulfotransferases; and sulfatases (Langford et al., 2017). However, only 24 of these genes are currently considered in clinical genetics, with 16 of those genes linked to neurological dysfunction (Table 1). Animal studies have shown additional sulfate biology genes are linked to adverse neurological phenotypes (Langford et al., 2017), suggesting that the number of clinically reportable genes is currently underestimated. This warrants further investigation into the role of sulfate biology genes in human neurophysiology.

In this study, we curated the temporal and spatial mRNA expression patterns of all known sulfate biology genes in the developing human fetal brain between 4 and 17 weeks gestation, corresponding to a critical time for nervous system formation in utero. We show moderate to abundant mRNA expression levels of 62 genes, suggesting a critical role of sulfate in early neurodevelopment. Additionally, we collated 18 sulfate biology genes implicated in atypical neurological function from the animal and human research literature which are currently not captured in clinical resources. Overall, these findings provide data for future investigations of sulfate biology genes associated with neurological dysfunction.

We obtained data of gene expression in the human fetal brain from the online Human Developmental Biology Resource (HDBR) Expression (Lindsay et al., 2016) (https://www.hdbr.org/) from 6 November 2021 to 12 January 2022. The HDBR is one of the rare human fetal tissue databases providing gene expression under 8 weeks post-conception compared to other databases (Johnson et al., 2009; Sunkin et al., 2013), which is essential for getting a better understanding of brain development since neurulation starts after 3 weeks post conception (O'Rahilly and Müller, 1994). Taking in account that brain development is dynamic, HDBR reports anatomical regionalisation gene expression results, which is also essential for mapping the differential pattern of gene expression related to selected brain regions compared to using full brain tissues (Ono et al., 2017). Data were available for four brain regions (cerebrum, brainstem, diencephalon and cerebellum) from 4 to 17 weeks post-conception. The gene expression levels for each tissue sample were averaged for each set of technical replicates, and then quantile normalized within each set of biological replicates using limma. Finally, they are averaged for all biological replicates. Specifically, we used transcripts per million (TPM) mapped reads-normalized RNA-Seq data to plot results for each gene. Of the 91 sulfate biology genes, 16 were abundantly expressed (>100 TPM) in at least one brain region at one gestational time point, 46 genes had moderate (>10 to 100 TPM) mRNA expression, and 15 genes had lower (≥0.1 to 10 TPM) mRNA expression levels (Figures 1–4). Fourteen genes (SLC13A1, SLC26A3, SLC26A9, SULT1A4, SULT1C3, SULT2A1, SULT2B1, SULT6B1, CHST4, CHST13, GAL3ST2, HS3ST6, ARSH and ARSJ) had undetectable or negligible (<0.1 TPM) mRNA expression levels (data not shown). The limitations of these data are that mRNA levels may: (i) not be representative of encoded protein level; (ii) not be predictive of functional impact when abnormal, for example altered expression of highly-expressed genes does not necessarily result in a more severe neurodevelopmental phenotype than altered expression of lowly-expressed genes; (iii) be redundant as other family member genes may compensate for altered expression level of a particular gene; (iv) not identify any sex-specific effects as the data are derived from pooled samples, which warrants future investigations to compare sulfate biology gene expression between male and female samples. Furthermore, our focus on genes expressed in the brain doesn’t take into account any of the sulfate biology genes expressed in other tissues that may influence neurodevelopment, such as the thyroid or adrenal gland or testis that could theoretically influence neurodevelopment through endocrine effects. Nonetheless, the temporal and spatial mRNA expression profiles of sulfate biology genes reported in the present study provide information for the next phase of research into understanding the regulation of sulfate homeostasis in the developing brain.

Despite the mRNA expression of 62 sulfate biology genes in the fetal brain (Figures 1–4), only 12 of these genes are considered for neurological dysfunction in routine clinical genetics (Table 1). As an approach to inform clinical geneticists and potentially expand the list of clinically reportable genes, we undertook a systematic online search (10–23 September 2021). We used PubMed and Medline to identify sulfate biology genes reported to be associated with atypical neurological phenotypes in humans using the search terms: (“brain” or “cerebellum” or “cerebral” or “neurodevelop”) and (“gene” or “dna” or “gene expression regulation” or “gene expression” or “gene genetic regulation”) and (“sulphate” or “sulfate” or “SO4”). We included articles that associated any of the sulfate biology genes with neurological conditions. All references used in this study were peer-reviewed research articles published in English between January 2001 and September 2021. In addition, we used OMIM (https://www.ncbi.nlm.nih.gov/omim/?term=) to expand our search prior to 2001 and to include any research articles not captured in our literature search. Overall, our research literature and OMIM searches identified 42 research articles that report the involvement of sulfate biology genes in neurological dysfunction.

From the curated list of research articles, we identified 18 candidate neurological dysfunction genes that are not yet clinically reportable (Table 2). Ten of these genes were reported from animal studies (SLC13A1, SLC13A4, SLC35B2, CDO1, CHST12, CHST15, NDST2, NDST3, SULF2 and UST), 2 from human and animal studies (ARSG and SULF1) and 6 from human research studies (SQOR, TST, SULT4A1, CHST10, HS6ST2 and ARSK). Thirteen of these genes are abundantly expressed in the human brain between 4 and 17 weeks gestation, whereas SQOR, CHST12, NDST2 and ARSG are expressed at low level (Figures 1–4) and SLC13A1 mRNA level is negligible (data not shown). The SLC13A1 gene is primarily expressed in the proximal tubule of the kidney where it mediates sulfate reabsorption (Ullrich and Murer, 1982; Lee et al., 2005). Disruption of human SLC13A1 and mouse Slc13a1 causes renal sulfate wasting and hyposulfataemia, which is proposed to reduce sulfate supply to the brain (Dawson et al., 2003; Bowling et al., 2012). Sulfide:quinone oxidoreductase (SQOR) is an enzyme in the minor pathway of sulfate generation (Dawson, 2013) and its loss leads to excess hydrogen sulfide levels that cause encephalopathy and Leigh disease (Friederich et al., 2020). Individuals with SQOR mutations first present with coma in early childhood (Friederich et al., 2020), suggesting a physiological role for SQOR in the brain beyond the 17-week gestational end point used to assess gene expression in the present study. Increasing CHST12, NDST2 and ARSG mRNA levels in brain at developmental ages beyond 17 weeks gestation (Humphries et al., 1998; Hiraoka et al., 2000; Khateb et al., 2018), also explains their low mRNA levels presented in this study. Nonetheless, the 18 genes listed in Table 2, warrant further research to determine whether these candidate neurological dysfunction genes are to be considered in clinical genetics.

In this study, we show the moderate to abundant mRNA expression of 62 sulfate biology genes in the human fetal brain between 4 and 17 weeks post-conception, a critical period in nervous system formation. This relatively large number of genes highlights the physiological relevance of sulfate in neurodevelopment and is a testament to the genetic complexity of maintaining sulfate homeostasis. However, sulfate biology is underappreciated in clinical settings, with only 12 genes routinely considered for certain neurological conditions. Our finding of an additional 18 genes reported in the research literature with neurological dysfunction, warrants future studies to validate the pathogenetics of these genes in human brain. Expanding the list of clinically reportable sulfate biology genes has the potential for improved genetic counselling, and potentially for developing therapeutic approaches towards clinical treatments. Overall, this study curated a list of sulfate biology genes involved in early brain development that provide information for future research of the physiological roles and regulation of sulfate homeostasis in human brain.