Skip to main content
European Commission logo print header

Sulfation pathways in Health and Disease: SUPA-HD

Final Report Summary - SUPA-HD (Sulfation pathways in Health and Disease: SUPA-HD)

Sulfation pathways are of main importance for healthy human physiology. They consist of the triad of enzymatic activation, transfer and removal of a sulfate moiety to and from a multitude of small and large biomolecules. One class of highly bio-active compounds that undergo sulfation and de-sulfation processes are steroid hormones. A prerequisite for sulfation to occur is sulfate activation, i.e. the transformation of the inert oxy-anion sulfate to the universal sulfate donor 3’-phospho-adenosine-5’-phosphosulfate (PAPS). This complex reaction is catalysed by so-called PAPS synthases and the human genome contains just two genes for PAPS synthases, PAPSS1 and PAPSS2. Genetic defects in the gene for PAPSS2 have been implicated in bone and cartilage malformation of varying severity as well as a steroid sulfation defect resulting in clinically apparent androgen excess. The first research objective (RO1) of this project was to illuminate this genotype-phenotype relation in novel ways. RO2 was to find and characterise novel protein interaction partners for human PAPS synthases and RO3 was to try to develop novel analytical methods for the canonical nucleotides APS, PAPS and PAP.

Prof Arlt’s international reputation was instrumental for the first part of RO1 as we were referred the case of two boys with delayed growth and bone mal-formation who were identified as compound-heterozygous for two mutations in the gene for PAPSS2. We characterised the severity of these mutations in a cellular model and measured residual enzymatic activity for the novel miss-sense mutation G270D, but no activity for the novel frameshift mutation W462Cfs*3 that resulted in a severely truncated protein. We observed only low expression of PAPSS2-G270D due to degradation by the proteasome, a finding that may also apply to other clinically observed mutations in PAPSS2. We also performed for the first time an in vivo DHEA challenge test in these patients and their parents thereby testing for hepatic DHEA sulfation capacity (which clearly was low in the patients and partially in their mother). This study was published in the Journal of Clinical Endocrinology and Metabolism (JCEM 2015 Apr;100(4):E672-80 PubMed ID: 25594860) with Jon Mueller in a prominent co-author position.

More point mutations were characterised for their stability in cells by fluorescence microscopy, proteasome inhibitors and western blotting as well as fluorescence-monitored in vivo unfolding. This method uses a FRET probe (AcGFP1-PAPSS2-mCherry) over-expressed in cells that are then step-wise heated up by an infrared laser; while the FRET probe is monitored in real time. We could classify our mutants into those that were strongly compromised in protein stability and those affected only to a minor extend. These biochemical classes seem to align also with the phenotypic appearance. Surprisingly, it is already the PAPSS2 wild-type protein that showed a strong tendency towards aggregation when expressed in cells. Moreover, we observed this effect only at temperatures about 10 degree Celsius higher than we previously had measured for the recombinant PAPSS2 protein. Now we need to identify the factors that account for this obvious difference.

RO2 of this project was to find novel interaction partners for PAPS synthases. Here we used a HEK293 Flp-In cell line that stably expressed a Strep-HA-tagged version of PAPSS2 for pull down experiments with concomitant protein identification by LC-MS/MS. A total of 433 proteins were identified. Reassuringly, it was PAPSS1 that we found highly enriched in this dataset as we had predicted in our proposal, confirming the formation of PAPSS1-PAPSS2 heterodimers that Dr Mueller had reported already in 2010 (BBRC 2010 May 7;395(3):420-5 PubMed ID: 20382111). However, the remaining proteins were more difficult to interpret as their enrichment in PAPSS2-overexpressing cells was only marginally higher than in the negative controls, but we have seen 37 proteasomal subunits and ubiquitin ligases to be enriched in the PAPSS2-pull down. Subsequently we established that PAPSS2 is subject to proteasomal turnover, being sensitive to MG-132 treatment, an established proteasome inhibitor. It remains to be determined what biological significance this highly interesting finding has by searching for the responsible ubiquitin ligase and the biological stimuli that modulate PAPSS2 protein turnover.

PAPSS2 mutations as described above actually should give more broad, more systemic defects if many sulfation pathways are compromised. Instead, some specific PAPSS-SULT “pairing” was suggested by the apparent SULT2A1 deficiency described above. Hence we tested whether PAPS synthases and the SULT2A1 sulfotransferase physically interact with each other. We could not see an interaction in GFP pulldown assays, suggesting any interaction to be transient. Using cysteine-reactive cross-linking chemistry (BM(PEG)3) we could detect a protein band at 100 kDa that was immunoreactive towards a SULT2A1 and a PAPSS2 antibody. As these molecules have masses of 33 kDa and 69 kDa the 100 kDa band is strongly suggestive of a cross-linked protein interaction. In an orthogonal approach we performed proximity-ligation assays (PLA) where the endogenous SULT2A1 and PAPSS2 proteins in HepG2 cells were labelled with antibodies. A multistep staining procedure involving secondary antibodies with DNA oligonucleotides attached to them would only give a strong fluorescent signal if the two proteins were less than 40 nm apart from each other. Clear PLA signals were obtained for SULT2A1 and PAPSS2, but also for SULT2A1 and PAPSS1. This study also involves PLA studies with more SULT antibodies and a thorough bioinformatics analysis of this interaction including macromolecular docking, virtual alanine scanning as well as molecular dynamics. This study is near to completion with an imminent publication (Mueller et al, manuscript in preparation).

The third part of this project dealt with nucleotide analytics. The canonical nucleotides APS, PAPS and PAP are usually not reported when people measure nucleotide or cofactor levels. We spent quite some effort in developing a LC-MS/MS method for these molecules, but struggled with impurities in commercial standards. Furthermore, we could not achieve chromatographic separation between APS and PAP and their multiple reaction monitoring (one could also say, their fragmentation pattern) in the mass spectrometry was indistinguishable. Hence, we returned to an alternative method that includes derivatisation of the adenosine moiety and separation on a μHPLC. Using this method we measured PAPS and PAP levels to be at about 60 μM in the adrenal NCI-295 cell line and APS levels to be lower than 1 μM. Most certainly, it will be of interest to determine these concentrations in more cell lines and under varying physiological conditions.

The current project has allowed Dr Mueller to re-frame his research into the mechanism and function of human PAPS synthases in the exciting context of steroid metabolism. Together with the scientist-in-charge, Prof Arlt, and Dr Paul Foster we could publish a scholarly article on the regulation of steroid hormone action by sulfation and de-sulfation processes (Endocrine Reviews). The feedback we got from colleagues clearly demonstrated how this article raised awareness of this important process. Furthermore, in September 2015 an international conference on this up-and-coming topic, SUPA (Steroid Sulfation Pathways), was held, organised and led by Dr Mueller. Taken together, this fellowship has very effectively strengthened European expertise in this specific area, Sulfation Pathways in Health and Disease.