Ubiquitin and Drug Discovery: Challenges and Opportunities

In 1977 ATP-dependent protein degradation in reticulocytes was demonstrated by Alfred Goldberg. Beginning in 1978, the ATP-dependent ubiquitin system for proteolysis was described by Avram Hershko, Irwin Rose and Aaron Ciechanover and colleagues, with the proteasome being characterized soon thereafter. Before this, the prevailing concept was that whatever protein degradation was taking place in cells was through lysosomes.

Naturally occurring cellular substrates for the ubiquitin-proteasome system (UPS) began to be discovered in 1987. In the mid 1990’s, in part due to development of proteasome inhibitors, there began an explosion in identified UPS targets. It is now universally appreciated that regulated protein degradation through the UPS is a rapid and exquisite means of controlling the levels of myriad cellular proteins and disposing of abnormal proteins. In fact, ubiquitin plays even broader roles in proteostasis. In addition to targeting for proteasomal degradation, it is integral to endocytosis and lysosomal targeting and intimately tied to autophagy (see review this issue by Cuervo and colleagues).

The role of ubiquitin goes beyond protein degradation. The ubiquitin signal in it various forms is critical to non-proteolytic processes such as DNA repair, NF-κB activation and regulation of transcription. Understanding how a single molecule added to proteins in different positions and in different linkages accomplishes so much remains a major challenge. We also now appreciate that ubiquitin is not the sole covalent polypeptide modifier of proteins. Other related polypeptides modify proteins through similar multi-enzyme cascades and many of them functionally intersect with ubiquitin.

Given the broad spectrum of UPS enzymes and substrates, it is not surprising that it is implicated in diseases from cancer to neurodegenerative disorders to HIV and other infectious diseases. Its importance in cancer is borne out through the FDA approval of the proteasome inhibitor bortezomib for treatment of multiple myeloma and other lymphoid malignancies and the recent FDA approval of carfilzomib.

There is, of course, intense interest in inhibiting enzymes of the ubiquitin system or enzymes that affect the ubiquitin system. Early work, including our studies with academic and biotech collaborators, helped established proof of principle for targeting the activity of the RING finger Mdm2/Hdm2 oncogene and the ubiquitin E1 in vitro and in cells. Moving beyond this to reagents suitable as therapeutics remains a major challenge. One area of promise is in development of agents that target the Nedd8 E1 that would thereby inactivate the hundreds of substrate-specific cullin ring ligase (CRL) E3s (see article this issue by Liao and colleagues). Such an agent, MLN4924, is now in Phase I clinical trials for both solid tumors and hematologic malignancies. By employing the principles and lessons learned in targeting the Nedd8 E1 there is the potential to develop potent inhibitors specific for the ubiquitin E1 and those of other ubiquitin-like modifiers. While there is great potential here, as was the case with proteasome inhibitors, the large number of proteins affected by such agents makes it difficult a priori to predict therapeutic outcomes. Another area of great interest and potential is in ubiquitin receptors and chaperones, such as p97/VCP (Cdc48 in yeast).

Substrate specificity in the ubiquitin system lies largely with the over 600 E3s, which act in combination with subsets of the ~35 E2s. Their activity is opposed by the ~100 deubiquitinating enzymes (DUBs). A challenge in drug development therefore lies not only in determining how to target these proteins, but how to do so with adequate specificity. The families of RING-type E3 domains, HECT domains and E2s each have marked structural similarity. Thus, as one moves beyond strategies aimed at disrupting E3-substrate interactions, approaches to targeting become problematic. As combinatorial pairs of E2s and E3s are matched with specific substrates and structural nuances important for function become apparent, we are hopefully approaching a time where drugs with a high level of specificity for individual E2s, E3s or E2–E3 pairs will be developed. Issues of specificity, as well as targetability, also pertain to the multiple classes of DUBs. However, there is immense interest in sorting out differences, both structural and functional, among DUBs and in targeting these enzymes. This is evident both from these Proceedings and from presentations at the 4^th Ubiquitin Drug Discovery and Diagnostic Conference in 2012. Therapeutics developed from the target-rich ubiquitin system has the potential to greatly enhance personalized medicine and treat previously intractable human diseases.

The reviews and primary research featured in this issue are representative of the exciting presentations at the 4^th Ubiquitin Drug Discovery and Diagnostic Conference^*. Aside from their intrinsic scientific interest, they collectively underscore the great interest in, and energy being devoted to, both advancing our basic knowledge and moving towards therapeutic development.