Early Stages of Drug Discovery in an Academic Institution and Involvement of Pharma for Advancing Promising Leads

Abstract

In this Viewpoint we provide a brief description of two efforts to develop drugs to treat diseases caused by tropical parasites (Malaria, human African trypanosomiasis, and Chagas disease). These efforts are largely based in a University setting but draw heavily on Pharma for a complete progression from drug hit discovery to advancement toward clinical trials. The first case is the development of protein farnesyltransferase inhibitors, and the second case is a series of benzthiazoles, the target of which is being investigated.

Drug leads from Pharma To Non-Pharma Institutions

The first example we give in this chapter is the story of protein farnesyltransferase (PFT) inhibitors as drug leads for treating tropical parasitic diseases. The Gelb and Glomset laboratories at the University of Washington (UW) discovered that specific proteins in eukaryotic cells are post-translationally modified by a 15-carbon farnesyl or a 20-carbon geranylgeranyl group (prenyl groups) [1]. When it became known that Ras proteins are farnesylated and that loss of farnesylation results in a reversal of oncogenic Ras-driven tumor growth, Pharma became very pro-active in the development of PFT inhibitors [2]. Given Gelb’s involvement in protein farnesylation and his ongoing work on anti-parasite drug discovery, the Gelb laboratory teamed up with molecular parasitologists Fred Buckner and Wesley Van Voorhis to explore protein prenylation in the malarial and trypanosomatid parasites [3]. In early work it was shown that a number of PFT inhibitors being developed by Pharma as cancer therapeutics were cidal to Plasmodium falciparum, Trypanosoma brucei, and Trypanosoma cruzi (the parasites that cause Falciparum malaria, human African trypanosomiasis (HAT), and Chagas disease). This was an exciting development because PFT inhibitors were advanced drug candidates, some even entered anti-cancer clinical trials, and thus these compounds were far along in the drug discovery process (some refer to this as “piggy-back” drug discovery).

Further work by the UW team provided a rational for why PFT inhibitors are much more toxic to parasites than to human cells. Humans contain a protein geranylgeranyltransferase-I that attaches the 20-carbon geranylgeranyl group to specific proteins. This enzyme can sometimes geranylgeranylate proteins that are normally farnesylated by PFT when cells are challenged with a PFT inhibitor. On the other hand, parasites lack this additional enzyme, and this may be the reason for parasite-selective toxicity of PFT inhibitors.

After we cloned the parasite PFT genes (alpha and beta subunits) it was apparent that the active sites of the parasite and human enzymes are highly conserved. It was thus very surprising to find that most of the Pharma-developed PFT inhibitors were orders of magnitude less potent against the parasite versus the human PFT. This shows that close structural similarities between parasite and human orthologues is not necessarily a “deal-breaker” in anti-parasite drug discovery (but of course toxicity due to inhibition of the human target needs to be explored early on in a drug discovery program).

Although our PFT inhibitor anti-parasite drug discovery program looked very promising in the early stages (we were well funded by the NIH and the Medicines for Malaria Venture and winner of an MMV Project of the Year), we could not progress PFT inhibitors into an anti-parasite clinical candidate. The downfall was that those PFT inhibitors developed by Pharma that had the best pharmacokinetic properties (for example high oral bioavailability) were not potent against parasite PFTs. We focused our work on the PFT inhibitors being developed by Bristol-Myers Squibb that were cidal to parasites in the low nanomolar range, but only available as injectable compounds. We were never able to obtain an orally-active analog by additional medicinal chemistry, and thus we could not achieve the target product profile of drugs for these parasites. At the same time, we could not generate structural analogs of the other series of PFT inhibitors (those with good pharmacokinetic properties) that gained potency on parasite PFTs.

PFT inhibitors have never emerged as approved drugs for cancer as it seems that the ability of oncogenic Ras to drive tumor growth can usually be restored by geranylgeranylation of this protein in the presence of a PFT inhibitor. Interestingly, PFT inhibitor have very recently been approved as drugs to improve the quality of life of Huntington-Gilford progeria patients (https://www.fda.gov/news-events/press-announcements/fda-approves-first-treatment-hutchinson-gilford-progeria-syndrome-and-some-progeroid-laminopathies). These patients have disease-causing mutations in nuclear lamin proteins, which are normally farnesylated.

Phenotypic Drug Discovery. From Pharma to Universities and Back Again.

Despite the attraction of target-based drug discovery, “old school” approaches that depend on phenotypic screening continue to be productive. For example, a review of 75 new first-in-class drugs developed between 1999 and 2008 revealed 37% were discovered by phenotypic screening versus 23% by target-based screening (the rest were natural substances or biologic drugs) [4]. In the early 2000s, the Genomics Institute of the Novartis Research Foundation (GNF) performed a high-throughput phenotypic screen of their compound library of 700,000 small molecules against cultures of Trypanosoma brucei [5]. The screen done at a compound concentration of 3.6 μM yielded 3889 primary hits causing >50% growth inhibition; the list was pared down to 1035 compounds (grouping into 115 distinct scaffolds) that passed the counter-screen for mammalian cell cytotoxicity. GNF generously invited our drug discovery group at UW (led by chemist, Michael Gelb) to review the hit list and select candidates for subsequent development. It was felt that for HAT, the only likely pathway for early drug development was via academic laboratories with public research funding. Professor Gelb and scientists at GNF applied the following additional filters to select candidates for hit-to-lead chemical optimization:

• Compliance with Lipinski’s rule of 5

• Avoidance of structural alerts for potential toxicity

• Avoidance of singletons

• Avoidance of molecules with >1 chiral center

• Chemical tractability

Seventeen starting scaffolds were subsequently reduced to 11 following experiments to select for a permeability through the blood brain barrier (necessary for treatment of late-stage HAT). Then over the course of 6 years, the chemistry group at UW generated >1500 analogs that were funneled through the biological screening pipeline led by the Buckner group [6]. Since the biochemical targets were unknown, compound design was guided by traditional medicinal chemistry principles; the compound properties and activities were optimized in iterative rounds of testing and synthesis. The products of this effort were three scaffolds demonstrated to cure mice with acute and/or chronic T. brucei infection [7–9]. However, at the time the lead candidates were being positioned for late-preclinical studies, it became evident that the demand for novel drug candidates for HAT was diminishing. This was due to the approval of fexinidazole as a new oral drug for HAT as well as the existence of another candidate in human trials, acoziborole (SCYX-7158), with potential for single-dose oral cures of HAT [10,11]. With this landscape, the UW group looked to repurpose the lead candidates from the HAT research towards Chagas disease.

Two of the three advanced compound series mentioned above were found to have potent in vitro activity against T. cruzi [7,12]. Despite the overlapping antiparasitic activity, some of the desired pharmacological properties differed for the two diseases. In particular, the need for CNS penetration is considered less important for Chagas disease while, at the same time, the intracellular niche of the replicating parasites demands that compounds distribute to and penetrate into a wide variety of host cells. In order to expand our expertise and capacity, the UW group joined an existing collaboration between University of Dundee, GlaxoSmithKline (Spain), and Drugs for Neglected diseases initiative that was working on Chagas disease drug discovery. Our desire was that the collaboration of academia, pharma, and a product development partnership would help bridge the chasm between our early drug discovery and late clinical development for this neglected tropical disease (NTD). The work is now primarily focused on the derivatives of the benzthiazole scaffold [8] with robust efforts in medicinal chemistry, pharmacology, toxicology, and animal models. One analog has been shown to eradicate parasites in the chronic model of T. cruzi infection, the only compound aside from the clinical drug, benznidazole, shown to have this activity as monotherapy. With the added resources and expertise brought to bear by the multidisciplinary collaboration, we are more optimistic than ever about advancing a drug candidate from the realm of academia into human trials for an NTD.