Despite clinicians having a variety of therapeutic options for rheumatoid arthritis (RA), very few patients achieve long-term clinical remission, so the need for novel effective and safe therapeutic options remains.

Groundbreaking publications from GSK have identified that a drug’s clinical success can be increased 2.6-fold if there is a genetic basis for its mechanism in disease. The chance of success increases further when a causal gene is known. Genome-wide association studies (GWAS) are the main source of genetic information for chronic polygenic diseases, like RA.

However, GWAS have not yet delivered on their promise to revolutionise drug discovery. This may be because 90% of disease-associated variants are in the non-coding Dark Genome, making it difficult to prove their functional effects. Thus, GWAS remain an untapped resource, and potential clinically successful targets are being missed. Nucleome’s cutting edge and proprietary platform combines machine learning and lab-based 3D genome analysis tools to identify the causal variant and its linked gene, even from complex loci with multiple genes and variants.

The Nucleome platform identified a variant from an RA-associated locus and nominated NTP-464 as the affected gene. It predicted the disease-associated allele would close chromatin at the locus and that an agonist therapeutic modality is required to restore the affected NTP-464’s function. We validated the platform’s prediction in the wet lab using genomic, in vitro and in vivo tools.

For genetic target validation (GTV), we found the NTP-464 variant sits in an enhancer region specific to B cells. Next, we recruited donors heterozygous for NTP-464’s variant and measured chromatin openness and RNA expression across numerous primary human immune cell types. This analysis revealed that the presence of the disease-associated allele resulted in closed chromatin and decreased RNA expression of NTP-464 in B cells and mature monocyte-derived dendritic cells, but not T cells or NK cells. We hypothesised that this disease-associated variant destroys the link between an enhancer and the NTP-464 promoter by closing chromatin, and that this genetic mechanism happens in a cell-type specific manner.

We collaborated with the Oxford BioBank to recruit donors of known genotype and measure NTP-464 protein expression on the surface of immune cells. Like the chromatin and RNA data, we observed a decrease in NTP-464 protein expression in the presence of the disease-associated variant in B cells but not T cells. Although this genetic mechanism of locus control appears mostly restricted to B cells, we observe target expression in T cells, including Tregs. These data validate both the GTV data and the platform’s prediction of loss of functional activity in B cells.

To validate the platform’s prediction of an agonist modality, we applied a tool agonist antibody to a suite of in vitro assays in primary cells. Treatment with the tool agonist impaired B cell activation (as measured by CD69 protein expression), B cell IL-6 secretion and B cell proliferation. Additionally, agonism of NTP-464 expressed on Tregs enhances their suppression of other CD4+ T cells. Early analysis of how agonism of NTP-464 may mediate its anti-inflammatory effects point to the generation of the enzyme, IDO1. IDO1 converts tryptophan to kynurenine, with subsequent anti-inflammatory effects. In vivo treatment in a murine collagen-induced arthritis model with the tool agonist antibody reduced disease severity, inflammation and bone damage compared to controls.  

While NTP-464 represents the story of a single genetic locus, Nucleome’s platform can also map key pathological mechanisms through an integrative network analysis of multiple loci. Indeed, using the NuML and 3D genome mapping capabilities, we have created gene clusters to identify 17 molecular mechanisms genetically associated with RA. One such molecular mechanism is the tryptophan metabolism pathway, offering convergent lines of evidence for targeting this pathway in RA.

Nucleome’s platform correctly predicted the functional effects of an RA-associated variant and its linked gene. Comprehensive in vitro and in vivo analysis with a tool molecule further recommend this target for the management of RA.