An HFSP fellowship that sowed the seeds for PROTAC drug development

In 2009, Alessio Ciulli was awarded an HFSP Short-Term Fellowship^* to support his 3-month research visit to Yale University to start a collaboration between his new lab and the lab of Professor Craig Crews at Yale. Crews, Ciulli and co-workers designed potent small molecules targeting the von Hippel-Lindau (VHL) E3 ligase, aiding development of much improved proteolysis targeting chimeric molecules (PROTACs). Today, only a few years after the beginning of this crucial collaboration, targeted protein degradation has emerged as a transformative new modality of therapeutic intervention that is making an impact across the pharmaceutical industry. The following article by Alessio Ciulli briefly recounts how HFSP support helped to seed these important advances, and highlights its impact both academically and commercially.

Targeted protein degradation using small molecules has emerged as a transformative new modality of pharmacological intervention into biological systems, with increasingly recognized therapeutic potential.  To induce intracellular depletion of a target protein, the protein must be recruited in proximity to an E3 ubiquitin ligase, complex enzymatic machineries that naturally catalyse the transfer of ubiquitin to substrate proteins. Ubiquitin ‘tagging’ leads to the tagged protein being degraded by the proteasome, the major ‘trash bin’ responsible for the recycling of most cellular proteins.

One approach to induce targeted protein degradation is to design bifunctional small molecules called Proteolysis-Targeting Chimeras (PROTACs). PROTACs comprise a ligand that binds the target protein of interest, and a ligand for an E3 ubiquitin ligase, joined by a linker. Formation of a ternary complex between the PROTAC, the ligase and the target protein results in the latter being poly-ubiquitylated and subsequently degraded by the proteasome. First described by Raymond Deshaies, Craig Crews and Kathleen Sakamoto in a PNAS article published in 2001¹¹, this approach remained on the sideline for a long time, largely because first-generation molecules that had been designed all incorporated long peptidic recognition moieties for the E3 ligase. This led to low cell permeability, limited stability and poor ‘drug-like’ properties for these molecules, which limited their biological applications. Many in the field recognized the need to replace these peptide-based moieties with smaller, more drug-like molecules. However, this meant disrupting a tight protein-protein interaction (PPI), a task esteemed too challenging as PPIs were considered ‘undruggable’ targets to small molecules. Hence, the dream remained elusive for a long time. Despite the risk, and against the odds, Crews and I decided to team up to tackle this challenge head-on, with the support of HFSP.

It all started back in September 2007, when Craig and I met for the first time in Manchester on the occasion of a symposium entitled “SCIBS: Selective small-molecule Chemical Intervention in Biological Systems”, sponsored by the UK Biotechnology and Biological Sciences Research Council (BBSRC). At the meeting, Craig presented his lab’s work broadly at the interface of chemistry and biology, including his pioneering work on PROTACs. I then spoke about my research on fragment-based and structure-based drug design, working as a College Research Fellow at Cambridge in collaboration with Chris Abell and Tom Blundell, co-founders of Astex Pharmaceuticals and two of the recognized pioneers in the field of fragments. Throughout the rest of the conference, Craig and I had stimulating conversations about our own research, and how we could work together on something new. From this beginning, we began to explore funding opportunities and came across the Human Frontier Science Program. The HFSP Short-Term Fellowship^* scheme seemed a perfect fit to our purposes: 1) to obtain pilot results in order to establish the new research collaboration; 2) to bring together complementary expertise available within our laboratories, so that I obtain training in new biological techniques while sharing my know-how in small-molecule design. It felt like a perfect 'win-win' situation from the start, and it rapidly crystallized into a proposal entitled “Discovery of small molecules disrupting the VHL:HIF-1 α interaction: towards next-generation PROTACs”, which I drafted and submitted for consideration by HFSP, with Craig’s support.

Figure 1. Crystal structure of first-generation inhibitor designed to disrupt the VHL-HIF interaction. Inhibitor is shown as sticks, with cyan carbon atoms bound to VHL (green surface, key residues labeled as sticks and with yellow carbons), as determined by X-ray crystallography (PDB 3ZRC²). Hydrogen bonds between the inhibitor and the protein are highlighted as purple dashed lines. HIF-1α  peptide (PDB 1LM8, shown as pink carbons) is superposed to reveal both the mimicry and differences in binding modes. Oxygen atoms are colored red, nitrogen, blue, and sulfur, dark yellow.

The proposal was successful, and I was able to visit Yale and Craig’s lab for three months in early 2009. I enjoyed a truly intense and enriching ‘sabbatical’ at Yale. In the Crews lab, I was exposed to several biochemical and cell biology techniques that were new to me, and learnt how to address biological questions. Equally, the exchange allowed me to share my experience in fragment-based drug design with other members of the group and within the departments, which helped to push their science in a new direction. Possibly more importantly, the fellowship exposed me to how the Crews lab integrated chemistry with biology in an academic setting. It was clear to me while I was there that I had made a good choice. I believe in life everything happens for a reason, and it was perhaps fitting that, while at Yale, I was awarded a major individual fellowship to start my own research group. I shortly returned to the UK to set up my own academic lab at the interface of chemistry and biology, studying and targeting PPIs with small molecules. The collaboration seeded with HFSP funding continued and expanded when I returned to Cambridge.

The rest is history. In just a few years, we achieved the goal that we had set ourselves from the outset and discovered first-in-class, micromolar affinity and non-peptidic small-molecule ligands for the VHL E3 ligase (Figure 1). These discoveries were published in three back-to-back papers in 2012^{2, 3 ,12}, and two patents were filed. The molecules were jointly designed between the two labs, with important early contributions from Julien Michel and Bill Jorgensen on the computational design³. The Crews lab helped primarily in the synthesis and assaying of the compounds², with key contributions from PhD student Dennis Buckley amongst others. My lab helped primarily in the biophysical screening and crystallographic characterization, driving structure-guided design and supporting fragment-based deconstruction work¹². Key contributions came from my group’s first postdoc and Marie-Curie Fellow, Inge van Molle.

The impact of our discovery should not be underestimated.  It promoted and bolstered interest in this area, and big pharma started to invest in it. GlaxoSmithKline first got involved in 2012, and subsequently founded a Discovery Platform Unit (DPU) in Protein Degradation (2013). Shortly afterwards, Yale University and Crews founded the spin-off company Arvinas to further develop the PROTAC technology.  Academically, my group published further fragment-based and structure-guided studies, leading to the optimization of VHL ligands with nanomolar affinities^5,8. In the late spring of 2015, a series of four back-to-back papers from my lab and others^1,10,13,14, all published within the timeframe of one month, reported dramatically improved activities and specificities of PROTACs, including compounds based on our VHL ligand, against different targets. Parallel discoveries that phthalimide-based immune modulatory drugs (IMiDs) such as thalidomide, lenalidomide and pomalidomide, bind to the E3 ligase cereblon (CRBN)^4,6,9 provided new ligands for a different E3 ligase, and these were also successfully incorporated into PROTACs.  Since then, Merck, Genentech, Novartis, Roche and more recently Boehringer Ingelheim, in a major collaboration with my lab now at Dundee University (http://www.fiercebiotech.com/biotech/boehringer-strikes-deal-to-develop-...), have all strongly committed to this promising and exciting new area.

With such a swell of excitement, many anticipate that more companies will be following suit. PROTACs are particularly attractive because of their modularity, their sub-stoichiometric mode of action, which relieves the need to fully occupy a target’s active site, and because potentially any cellular target could be degraded provided a binding ligand for that target is available or can be developed. The renaissance in interest on PROTACs is reminiscent of the boom in kinase drug discovery that followed the discovery of Gleevec in the early 2000s, in spite of many years of quiet and sceptic industrial approach to that field. I am convinced that targeted protein degradation is set to rapidly establish itself as one of the most disruptive new chemical modalities of therapeutic intervention in the future - and I am sure many share this belief.

Undoubtedly, the discovery of small molecules with high affinity, specificity and with crystallographically defined binding modes for E3 ubiquitin ligases, as is the case with our VHL ligands, greatly contributed to these major developments in the PROTAC area.  In a distinct approach, VHL ligands in their own right have potential to impact also as VHL inhibitors. My lab has recently developed a potent, selective and cell-active VHL inhibitor that we have named VH298⁷. We have characterized the compound as a selective chemical probe of the hypoxic signaling pathway, providing a new chemical tool to study biological systems, per se an attractive starting point for drug development. VH298 will be made available to the scientific community through the newly established Chemical Probes Portal (http://www.chemicalprobes.org/) and via commercial vendor catalogues, and it is anticipated that it will lead to many new unforeseen discoveries.

It is unlikely that the successes and advances highlighted here would have happened if it was not for the enlightened decision of HFSP to support my grant application to visit Yale to begin a collaboration with Prof. Crews. For this, I feel privileged and I am immensely grateful.

^{*The HFSP Short-Term Fellowship program was terminated in 2010.}