King's College London

Sulforaphane (SFN) is an electrophilic isothiocyanate which can adduct cysteine thiols within proteins. Protein targets of SFN were immunoprecipitated from cardiac tissue of wildtype (WT) mice following in vivo treatment with the electrophile using a validated polyclonal antibody developed in-house to pan-specifically detect SFN adducted to cysteines. Combined with quantitative proteomics, this confirmed the non-receptor protein-tyrosine phosphatase, SHP2, as a target of SFN. SFN is intrinsically unstable at room temperature, therefore, a chemically stabilised variant developed by Evgen Pharmaceuticals (UK) known as Sulforadex (SFX-01), was used in subsequent experiments. Using a commercially available phosphatase activity assay, SFX-01 was shown to inhibit recombinant SHP2 in vitro, as well as that in cardiac tissue of mice administered SFX-01 in their drinking water for 4 days. We speculated that SFX-01 may be therapeutic in diseases where SHP2 is hyperactive, such as Noonan syndrome (NS). Indeed, using an NS mouse model, Ptpn11D61G/+, a mutation resulting in hyperactivity of the phosphatase, SFX-01 time-dependently inhibited cardiac SHP2 activity. 100 % of homozygous and ~50 % of heterozygous Ptpn11D61G/+ mice die mid-gestation due to severe skeletal or cardiac defects, with the remaining ~50 % surviving to adulthood where they show non-cardiac features of NS. To assess if SFN-induced inhibition of SHP2 in the homozygous or heterozygous foetus could improve embryonic development, breeding pairs consisting of WT only or NS only parents were administered SFX-01 before conception and continued during pregnancy. SFX-01 treatment induced SFN-protein labelling of foetal tissue but also reduced litter sizes born from NS breeding pairs and genotyping showed only WT mice were born. This adverse effect may be due to SFN increasing the phospho-activation of ERK, which is deleterious in embryonic development of NS foeti. However, SFX-01 had no adverse impact on the pregnancies of WT mice. Adult NS mice develop splenomegaly and myeloproliferative disease which can further develop into leukaemia. With this in mind, adult WT or NS mice were administered SFX-01 for 10 weeks to assess if prolonged treatment with the drug would inhibit SHP2 activity and reduce the incidence of myeloproliferative disease in the NS mouse model. Using whole blood cell staining, ultrasound and flow cytometry, lower total white blood cell count, spleen size and myeloid cell count in the blood, bone marrow and spleen of NS mice by SFX-01 was seen compared to water only controls. SHP2 activity was also attenuated in the spleen of both WT and NS mice, strongly suggesting this therapeutic action of SFX-01 was mediated by inhibition of SHP2 phosphatase activity. Unexpectedly, even though phosphatase activity was inhibited following 4-day or 10-week treatment with SFX-01, this occurred without evidence of an SHP2-SFN adduct in the tissue of WT or NS mice. Data from biochemical analyses involving biotinylated iodoacetamide (BIAM) labelling, the polyethylene glycol (PEG)-switch method or phenylarsinic acid (PAA)-binding, showed that SFN induced an inhibitory modification within SHP2 between two vicinal thiols within the active domain of the phosphatase, which to reiterate was not stable SFN adducts. Data from complementary studies using site-directed mutagenesis of cysteines supported the concept that SFN adducts to SHP2 and inhibits it, which is followed by a proximal cysteine thiol mediating its removal or truncation, with the resulting modification maintaining inhibition of the phosphatase. This SFN-induced inhibitory modification may be the formation of an intramolecular disulfide bond or perhaps the chemical modification of the SFN adduct to a dithiolethione. Additional data has also shown that an SFN adduct can transfer from one thiol to another, so-called ‘trans-thiolation’. Using bovine serum albumin with an SFN adduct, which had been purified by large-format gel filtration on a protein chromatograph, transfer of the adduct to other cysteine-containing molecules such as haemoglobin or glutathione was observed. SFX-01, a stabilised SFN variant in phase 2 clinical trials, inhibits WT SHP2 as well as a hyperactive Ptpn11D61G/+ mutant form expressed in many patients with NS. Consistent with this, SFX-01 significantly corrects the myeloproliferative disease found in Ptpn11D61G/+ NS mice. Thus, in conclusion, SFX-01 has potential as a new therapy for the treatment of NS.