GPCR Assay Strategy, Bias, and Translational Drug Discovery

Scientific Abstract

A recurring problem in GPCR drug discovery is that the assay used to find molecules is not always the assay that predicts what those molecules will do in a living system. This conversation follows Martin Marro, head of cell pharmacology in obesity research at Eli Lilly, through more than two decades of confronting that gap across small molecules, peptides, aptamers, and antibodies.

Marro describes a Gq coupled receptor program in which four lead series with sub nanomolar calcium potency showed little activity ex vivo and in vivo. Calcium, IP1, and competition binding all failed to predict efficacy. Persisting with an unexplored cAMP hypothesis revealed super agonists with up to threefold greater Emax than the natural ligand, and that readout tracked in vivo efficacy. The discussion extends to biased GLP1 receptor agonists that favor membrane retention and faster recycling, antibody discovery against hard to drug GPCRs, and how the field learned to define biased agonism rigorously.

The thread began long before GPCRs: Marro's first research was transgenic wheat in Argentina, aimed at better bread, a quiet nod to a grandfather who was a baker.

About Martin

Martin Marro is Head of Cell Pharmacology and Executive Director of the Obesity Research Group at Eli Lilly in Boston, where he is building an in vitro pharmacology lab focused on cell based signaling assays for target and drug discovery. He trained at the International Centre for Genetic Engineering and Biotechnology in Trieste, then joined GSK in Stevenage as a Marie Curie Industrial Fellow working on aptamers against chemokine receptors.

Over 15 years at Novartis in Horsham and Cambridge, Massachusetts, he led GPCR programs spanning gastrointestinal and cardiovascular metabolic disease, including the discovery of small molecule GLP1 receptor agonist leads and biased fatty acid conjugated GLP1 analogs. At Tectonic Therapeutic, his lab contributed to the discovery of the RXFP1 agonist TX45, and he led the company's first GPCR antagonist antibody program from hit to an optimized lead, later declared a development candidate.

Scientific Themes of the Conversation

- Why fast, far from equilibrium readouts can mislead lead optimization

- Pathway selection, super agonism, and Emax as a driver of in vivo efficacy

- Setting a high bar for what counts as true biased agonism

- Receptor trafficking, internalization, and recycling as design parameters for GLP1 agonists

- Antibodies and aptamers as modalities for difficult GPCR targets

- Orphan and class B receptors as enduring challenges in the field

Key Insights from the Conversation

Potency in the primary assay is not the same as efficacy in the animal

Four lead series reached sub nanomolar potency in a calcium assay, yet showed little activity ex vivo and in vivo despite good pharmacokinetic properties. IP1 and competition binding did not close the gap either. The lesson was that a clean screening cascade can still be measuring the wrong thing.

Persisting with an unpopular hypothesis rescued a nearly dead program

As a junior lab head, Marro proposed pausing the program to examine alternative pathways, biased agonism, and heterodimerization. Management declined because of the program's urgency, so he explored cAMP signaling in the background. The data eventually showed that super agonism in the cAMP assay predicted in vivo efficacy, and the molecule went on to reach the clinic.

Maximal response can matter more than potency

The molecules that worked were less potent in cAMP than in calcium, which initially made them look like weaker candidates. What set them apart was an Emax up to threefold greater than the natural ligand. That shift in perspective meant accepting a program built on rightward shifted potencies.

Calcium is a less specific readout than it looks

Calcium can arise from multiple sources, including voltage gated channels, so a calcium signal cannot always be attributed to Gq alone. Longer incubation assays such as IP1 allow readouts closer to equilibrium. Choosing the readout is a scientific decision, not only a throughput decision.

Biased agonism deserves a strict definition

Around 2007, many Novartis scientists doubted that published examples reflected true ligand bias rather than system bias. An internal GPCR convention set a demanding standard: a reversal in rank order of potency or efficacy between two assays, or a compound inactive in one of them. As tools and analytical methods have matured, the field can now separate these effects more confidently.

Keeping GLP1 receptors at the membrane can shape the therapeutic profile

Building on work from Stephen Bloom's lab on exendin 4 analogs, Marro's team studied fatty acid conjugated GLP1 agonists that were more biased than semaglutide, with weaker internalization and faster recycling. In diet induced obese mice, the lead matched semaglutide weight loss with less frequent dosing, and no emesis was observed.

Moving to a new modality means relearning the flowchart

Shifting from small molecules to antibodies required new ways of defining selectivity, assessing biophysical properties early, and treating affinity maturation as the core of hit to lead work. Marro describes it as a genuine learning curve, given how few antibodies against GPCRs exist.

Episode Timeline

Timestamps were generated using AI for readability.

- 00:00 Introduction and building a new obesity research lab in Boston

- 02:35 A fellowship discovered the day before the deadline

- 07:26 Aptamers, SELEX, and chemokine receptors at GSK

- 15:47 Doubting a career choice after a Nobel Prize announcement

- 26:16 Sub nanomolar hits that failed in vivo

- 32:26 The cAMP hypothesis that resurrected the program

- 37:42 Biased GLP1 agonists, trafficking, and weight loss

- 44:12 Antibodies against hard to drug GPCRs at Tectonic

- 51:03 From green to red: transgenic wheat to GPCR drug discovery

- 54:26 Advice for young scientists

Selected Quotes

"I was intrigued by the fact that a pill that would fit in the palm of your hand could treat symptoms and in some cases cure diseases."

Martin Marro

"It was inevitable to ask myself whether I had made the right decision. I reflected on the fact that my time in industry was going to come back at some point, and I had to be patient."

Martin Marro

"My contributions managed to resurrect a project that was almost dead, just trying to think outside the box and persist with my idea."

Martin Marro

"Most experiments won't work. Most drugs won't make it. The most important thing is not what you face, but how you react to that."

Martin Marro

Guest Link

- LinkedIn: https://www.linkedin.com/in/martinmarro/