Issue #68: The beginnings of rational drug design

Written by Nobel Laureate Professor Peter Doherty

Much of biology is about the complementarity of shapes, and the chemical interactions that allow small regions on protein 1 (P1) and protein 2 (P2) to bind together. Our P1 might be the peptide (a protein bit) hormone insulin that binds to P2 proteins (receptors) on, say, our liver cells. The ‘chemical’ signal that goes across the cell membrane as consequence of insulin binding ‘instructs’ the liver to store glucose as glycogen and decrease blood glucose levels. Most relevant to COVID-19 and vaccine efficacy (#51), P1 would be an antibody – or immunoglobulin (Ig) – molecule, while P2 is the spike protein (SP) of SARS-COV-2. I’m going to summarise a little of what I wrote earlier in 2020 and use some scientific terms here. If you’re unfamiliar with immunology, maybe go back and take a look at #18- #22 of these essays.

Virus neutralisation happens when a unique variable region (V) at the tip of the combined Ig heavy and light chains Ig (IgVSP) recognises the ‘micro-anatomy’ of the SARS-CoV-2 SP receptor-binding domain (RBD) and attaches with sufficient ‘affinity’ to hold these two proteins together. The consequence of that ‘fatal attraction’ (for the virus) is that the spike protein-specific IgVSP ‘covers’ the SP RBD and prevents the virus attaching to our cell-surface ACE-2 molecules. That ‘steric inhibition’ blocks virus access to the cytoplasm of our respiratory epithelial cells and prevents them from being taken over to become SARS-CoV-2 producing ‘factories’.

When I first started seriously in immunology five decades back, we used to talk about ‘antigen/antibody’ binding as a ‘lock and key’ interaction. That’s how we (with Rolf Zinkernagel) in fact illustrated the comparable T cell receptor (TCR) ‘altered self’ interaction when we discovered the basic mechanism underlying ‘immune surveillance of self’ by killer T cells, the ‘hit men’ of immunity that ‘zap’ the virus-producing ‘factory’ cells (#33, #34). Drawing ‘ping-pong balls’ and ‘tennis balls’ with a few squiggles on the surface reflected the limitations of the technology of the time. Scientifically, that was clearly, and massively, unsatisfactory. Reform came with a spectrum of advances in molecular biology and technology that were, by the 1980s, allowing the chemistry of such P1/P2 interactions to be defined at the level of microanatomy and chemistry. This was the beginning of rational drug design.

The ‘heroes’ who drove this advance from the 1980s were the structural biologists (#67), scientists who came from a physics background. Of course, they had a lot of help from biologists and chemists. Of particular importance was the development of monoclonal antibodies, the marvellous mAbs, a story I’ll start to tell in detail next week. Here, it’s sufficient to say that a ‘mAb’ is an Ig with a single specificity that serves, in fact, as a reproducible chemical reagent.

There’s a great Australian story that illustrates how this worked to enable the design of Relenza, the very first ‘small molecule’ drug for stopping influenza virus infection. As background, the influenza viruses have two major proteins on their surface, the hemagglutinin (HA) and neuraminidase (NA). Less restrictive than the SP RBD/ACE-2 attachment that allows a SARS-CoV-2 molecule to invade, the influenza HA binds to rather ubiquitous sugars, sialic acids (SAs), that are prominently expressed in the respiratory tract. But there it has a problem. Once new flu virus particles (virions) are made, how do they avoid that SA attachment and get away from the cell? That’s the job of the viral NA, an enzyme that cleaves the HA/SA interaction and allows the virions to escape, then infect other cells.

What happened with Relenza – the ancestor of the more familiar oral drug Tamiflu (Relenza is inhaled using a ‘puffer’) – was that ANU virologist and chemist Graeme Laver made lots of purified influenza virus NA. Graeme then mixed this painstakingly acquired NA with an NA-specific mAb, and went about making co-crystals of these two proteins bound together. In Shakespeare’s Romeo and Juliet, Mab is the Midwife Queen of the Fairies and, when it comes to making crystals that will reveal protein structure when bombarded with X-rays from a high energy source, there’s still a bit of magical chemical midwifery involved!

The crystals then went to structural biologist Peter Coleman, who is now at the WEHI (1G, Royal Parade) on the Parkville Campus, but was then further north at number 343 Royal Parade in the CSIRO Division of Protein Chemistry. Working in a tradition dating back to former University of Adelaide Physics Professor William Bragg and his Australian born son Laurence, who (having moved to the UK) shared the 1915 Nobel Prize for Chemistry, Peter and his CSIRO colleagues defined the ‘micro-anatomical’ chemistry of the interaction that enables the mAb and the NA to form a strong attachment.

That information was then passed to young carbohydrate chemist Mark Von Itzstein at the Victorian Pharmacy College (384 Royal Parade), who designed the ‘small molecule’ drug, Relenza, that prevents the NA from breaking the HA/SA bond and releasing newly made flu virus from the surface of an infected cell. Looking down an electron microscope at an influenza virus infected cell that has been treated with Tamiflu or Relenza, you can see the virions clustered like bunches of grapes on the cell surface. Mark now heads the Glycomics Institute at Griffith University and is, of course working to develop SARS-CoV-2 antivirals.