Ten years ago, a first draft of the sequence of the human genome was announced, amid promises of better drugs to come through improved understanding of disease, as highlighted by Kramer and Cohen (Nature Rev. Drug Discov. 3, 965–972; 2004)1. However, the tone of some recent commentaries reflects disappointment with the perceived outcome2,3. We believe that an objective discussion of this important topic has been hampered by both a lack of clarity about what constitutes a genomics drug and a lack of visibility of important genomics contributions to drug discovery. For example, recent articles have highlighted belimumab, a monoclonal antibody (mAb) that targets B lymphocyte stimulator (BLYS; also known as TNFSF13B/BAFF), as having life-altering potential for patients with systemic lupus erythematosus (SLE)4, but have failed to mention its clear-cut genomics origins.

We therefore searched the clinical pipelines of several genomics companies using the Thomson Reuters Integrity database and literature resources for drugs in which genomics data — for example, from sequencing, database mining, expression profiling or RNA-interference screening — seems to have had an essential role; that is, in the absence of the genomics data, a project would probably not have been initiated or progressed. As the drug development timeline is typically 10–15 years, we limited the search to drugs that are in Phase II or III trials. Here, we highlight 12 drugs that illustrate the wide-ranging impact of essential genomics data to drug discovery (Table 1). Subjective judgment was often needed during the selection, especially in cases in which pre-existing literature linked a target to the disease prior to genomics experiments.

Table 1 Selected genomics drugs in clinical development
Full size table

High-throughput sequencing of cDNA libraries was the first genomics technology. It produced partial sequences of expressed genes (expressed sequence tags; ESTs) that could be assembled and mined by bioinformaticians for genes that had specific expression in disease tissues. Several private and public organizations began populating databases with ESTs as a prelude to the full sequencing of the human genome, and these served as a rich source of novel drug targets. Thus, sequencing of a neutrophil monocyte-derived cDNA library led to the discovery of BLYS, a member of the tumour necrosis factor (TNF) family that induces B cell proliferation5,6. High levels of circulating BLYS are found in patients with SLE, and transgenic mice expressing BLYS develop a lupus-like disease. As mentioned above, the BLYS-specific mAb belimumab antagonizes the function of BLYS. Phase I and II trials of belimumab showed promising results7,8, and in two Phase III trials the primary end point was reportedly met in a large proportion of patients with SLE receiving high-dose belimumab4. These data provided the basis for its recent regulatory submission. The identification of BLYS by genomics prompted the search for its receptors9 and led indirectly to the development of atacicept, a soluble BLYS receptor that has also been tested in patients with SLE10.

During a race to identify new members of the TNF receptor family by database mining of ESTs, TNF-related apoptosis-inducing ligand (TRAIL) receptor 1 (TRAIL-R1; also known as DR4/TNFRSF10A)11 and TRAIL-R2 (also known as DR5/TNFRSF10B)12 were discovered. Binding of the ligand TRAIL activates these receptors and initiates the apoptosis cascade in tumour cell lines. Therefore, the disruption of the ligand–receptor interaction was of potential utility in the treatment of some cancers. TRAIL itself was identified by mining the EST database from the National Center for Biotechnology Information (NCBI)13 and led to the development of a recombinant protein therapeutic, dulanermin, which is currently in clinical trials for various cancers.

Mapatumumab is an agonistic human mAb that binds to TRAIL-R1 and induces apoptosis in cancer cells. It was among the first of several TRAIL receptor-specific mAbs to enter the clinic and has been tested in Phase I and II trials as a single agent and also in combination studies14. Apomab is a mAb agonist of TRAIL-R2 that is also in cancer clinical trials15. Thus, database mining of genome-wide cDNA sequence libraries led to the discovery of TRAIL, TRAIL-R1 and TRAIL-R2, and established these as targets for the development of biologics as genomics drugs for the treatment of cancer.

Cathepsin K is a cysteine protease that was identified in cDNA libraries derived from bone cells16. Its high, selective expression in osteoclasts implicated the protease in bone remodelling processes and genetics data (mutations in cathepsin K cause pycnodystosis) suggested that its inhibition might decrease pathological bone resorption in osteoporosis. Several cathepsin K inhibitors have entered clinical trials, but development has been difficult. However, a dose-finding study in postmenopausal women with low bone density showed that treatment with the cathepsin K inhibitor odanacatib increased bone mineral density in lumbar spine and in the hip17, and a Phase III trial is ongoing.

Access to the complete sequence of the human genome has revolutionized genetic studies by providing new markers and rapid gene-mapping methods. One of the most significant recent discoveries in bone disorders emerged from genetic studies in patients with sclerosteosis. A genome-wide linkage analysis in 22 affected families of the Afrikaneer population unveiled loss-of-function mutations in the SOST gene, which encodes sclerostin, a protein that is secreted by osteoblasts and negatively regulates bone formation18. The mutations cause severe bone overgrowth, suggesting that pharmacological inhibition of sclerostin may increase bone density. AMG-785, a mAb that targets sclerostin, has been tested in Phase I trials19 and is currently in Phase II trials for the treatment of osteoporosis.

At deCODE Genetics, genome-wide linkage scans have been conducted in Icelandic patients suffering from atherosclerosis. In one scan, a search for genes linked to peripheral arterial obstructive disease implicated the PTGER3 gene, which encodes the EP3 prostaglandin receptor20. A medicinal chemistry effort yielded a potent EP3 receptor antagonist, DG-041, which has progressed to Phase II clinical trials.

In a strategy intended to ensure the druggability of targets discovered using genetics, scientists from Oxagen restricted genotyping of patients to the family of G protein-coupled receptors (GPCRs)21, all members of which have now been identified in the human genome. During a genome-wide association study, the CRTH2 gene was found in a linkage region for asthma. CRTH2 encodes a GPCR for the prostaglandin PGD2, which had previously been linked to inflammatory diseases. A medicinal chemistry programme produced OC-000459, a CRTH2 antagonist, which has progressed to Phase II trials for the treatment of respiratory and gastrointestinal inflammatory disorders.

As part of an early genome-wide search for cancer genes focusing on mutations in the RAS–MAP kinase pathway, the sequencing of cancer patient cell lines through the coding exons of BRAF kinase revealed several gain-of-function mutations22. High-throughput sequencing in hundreds of primary human cancer cell lines then confirmed BRAF as an important oncogene in malignant melanoma. A potent selective inhibitor of a BRAF mutant (PLX-4032) was subsequently developed at Plexxikon23. Positive Phase I trial data with PLX-4032 prompted an extension trial in patients expressing the BRAF mutation, a large proportion of whom experienced tumour regression24. Phase II and III studies of PLX-4032 are ongoing.

Access to the complete sequences of human and mouse genomes has greatly advanced the use of mouse knockout technology in drug discovery. During a programme to study phenotypes of 5,000 gene knockouts, a new brain-specific isoform of tryptophan hydroxylase (TPH) was discovered25. Dysregulation of TPH in the gastrointestinal tract is associated with carcinoid syndrome, as well as with irritable bowel syndrome (IBS), and so the discovery of tissue-specific TPH isoforms prompted the search for serotonin inhibitors devoid of central nervous system-derived side effects. Two TPH inhibitors, LX-1031 and LX-1032, subsequently entered Phase II trials for the treatment of IBS (LX-1031) and carcinoid syndrome (LX-1032).

In summary, we have selected 12 drugs that, in our view, would probably not exist had it not been for genomics data. We argue that these drugs should be seen as the tip of the iceberg because this limited analysis only captured drugs originating from genomics companies and whose histories are documented in the literature. We conclude that the first decade of the genomics era has provided new, possibly valuable drug targets for some important diseases. Although genomics technologies are not substitutes for hypothesis-driven disease biology, medicinal chemistry and clinical testing, they have become firmly embedded in the drug discovery process and it seems likely that many more drugs in the foreseeable future will be genomics drugs.