QnAs with Svetlana Mojsov, Joel Habener, and Lotte Bjerre Knudsen: Winners of the 2024 Lasker~DeBakey Clinical Medical Research Award

When the long-awaited results of a clinical trial of the wildly popular weight loss drug semaglutide were published in late 2023, the moment marked a milestone in cardiovascular medicine. A once-weekly injection of semaglutide slashed the incidence of death from cardiovascular causes and nonfatal heart attacks or strokes in nondiabetic patients with a history of heart disease as well as obesity or overweight (1). Only two years earlier, semaglutide, a type of drug called a glucagon-like peptide-1 (GLP-1) receptor agonist, had won approval from the US Food and Drug Administration for treating obesity, transforming the treatment of an epidemic poised to afflict more than half of the global population by 2035 (2, 3). The origin story of GLP-1 receptor agonists, which act on the brain to regulate appetite and food intake, can be traced back to the mists of early 20th-century science on human gut hormones. Decades of research underpin the development of this class of drugs, the discovery of which was inspired by the search for “incretins,” which are hormones that regulate blood glucose levels by triggering insulin release after a meal. From the initial discovery of the gene for GLP-1 hormones in the DNA of anglerfish to the successful development of GLP-1-based drugs for the treatment of type 2 diabetes and obesity, dozens of researchers played key roles in a sumptuous tale of scientific sleuthing and invention. Three of them have won the 2024 Lasker~DeBakey Clinical Medical Research Award for their singular contributions and sharply observed insights. The winners played a central role in discovering GLP-1 hormones, unraveling their intricate mechanisms of action, and developing effective drugs for diabetes, obesity, heart disease, and a growing list of human ailments: Joel Habener, of Harvard Medical School, and Svetlana Mojsov, of Rockefeller University, for identifying and characterizing GLP-1 and its role in glucose metabolism, and Lotte Bjerre Knudsen, of the Danish drug company Novo Nordisk, for developing a stable and long-acting version of GLP-1-based drugs. PNAS spoke to the trio about their plodding path to success as well as the protean power of the drugs they helped bring to patients.

GLP-1-based drugs have transformed the treatment of obesity. Image credit: iStock/Carolina Rudah.

[The tangled history of research on GLP-1 receptor agonists features several players that no single award can satisfactorily capture. For a comprehensive history of the underlying science, see here (4), here (5), here (6), and here (7)].

PNAS: A wealth of foundational work prefaces the saga of GLP-1-based drugs, but one point of entry into the sprawling story is the 1982 PNAS article reporting the cloning of the gene for proglucagon. How pivotal was this paper (8)?

Mojsov: The 1982 PNAS paper got things started, but it was the 1983 Nature paper by Bell et al. that was more influential for my own work (9). The Nature paper reported the cloning of the mammalian gene. The sequence of the gene showed the presence of two additional peptides that Graeme Bell named GLP-1 and GLP-2. Notably, Bell’s paper showed the predicted sequence of GLP-1 flanked by enzymatic cleavage sites known to release biologically active peptides from prohormones. It predicted the sequence of a 37-amino acid-long version of GLP-1 and compared it with the sequence of glucagon and other peptides in the same family. Looking at the figure, it struck me immediately that the predicted GLP-1 sequence had an expendable six amino acid extension at the amino terminus. Examining the sequence closely, I saw that the extension had an arginine residue at position 6. I also noted that cleavage after the arginine at position 6 would produce a 7-37 amino acid GLP-1 peptide with perfect alignment with amino acids in glucagon that were essential for its biological activity. That observation suggested that GLP-1 7-37 was the active sequence.

PNAS: Let’s rewind to give our readers context for that crucial finding. Joel, can you set the stage by telling us how you became interested in the GLP-1 hormone?

Habener: When I started an independent laboratory program at Mass General, we decided to work on glucose metabolism and the hormones involved in it. These hormones are made in the pancreas, which is really two organs: an endocrine part, which secretes hormones, and an exocrine part, which makes the enzymes that help digest food. Everybody knows about the major pancreatic hormone insulin, which is a lifesaving treatment for type 1 diabetes. But the pancreas produces two other major hormones, one of which is glucagon. At the time, nothing was known about how glucagon was made. We made a plan to isolate the endocrine cells from the rat pancreas and to use recombinant DNA technology to clone the gene that encodes the precursor of glucagon [or proglucagon].

Svetlana Mojsov. Image credit: Rockefeller University/Chris Taggart (photographer).

Fortunately, we had some knowledge in the lab. My postdoc had obtained marine animal tissues for research because he knew of a person whose brother was a commercial fisherman, and the animal we chose was the anglerfish. It was a trash fish in the United States but a delicacy called lotte in France. The endocrine cells of the fish are in a separate organ called the Brockmann body. Using the fish, we could bypass the tedium of isolating the islets from mammalian pancreas. This allowed us to have an enriched source of the messenger RNAs of the endocrine hormones.

Our goal was to search for the gene for proglucagon in pancreas extracts; previous work had suggested that there may be a proglucagon. We were eventually successful in cloning the anglerfish proglucagon (8) [reported in the 1982 PNAS article]. Significantly, the sequence of the anglerfish proglucagon also contained the sequence for another glucagon-related peptide. It was not glucagon but in the glucagon protein superfamily. It closely resembled a hormone called gastric inhibitory peptide (GIP), which was thought to be an incretin, a factor released in the bloodstream upon eating that augmented glucose-stimulated insulin secretion. Previous work had also suggested that there had to be a second, unidentified incretin, which was produced in the gut. We speculated that this new peptide in anglerfish might be the missing incretin.

PNAS: Svetlana, you joined the quest around this time.

Mojsov: I spent almost eight years studying and synthesizing glucagon, and I was fascinated by the physiology of glucose metabolism. I had developed strategies to synthesize glucagon using the solid phase method (10). [After] the glucagon gene sequence was published by Bell… I made highly pure GLP-1 peptides and used them to prove my idea about the biologically active form of GLP-1.

I had a hypothesis that the glucagon gene encoded the missing incretin, which Creutzfeldt had described vividly in a 1979 review article (11). It made sense that the glucagon gene encoded two peptides for regulating glucose metabolism: glucagon acts in the liver to regulate glucose metabolism through glycolysis and gluconeogenesis, and GLP-1 7-37 would be produced in the intestines to stimulate insulin and regulate glucose metabolism. It was long known that oral glucose administration stimulates insulin release to a much greater degree than intravenous administration. To pull all these strands together, I needed to show that GLP-1 7-37 is in fact made in the intestines and is active (12, 13). To do so, I developed antibodies and radioimmunoassays to isolate the 7-37 peptide. Eventually, I synthesized close to 900 milligram of pure GLP-1 7-37.

The next step was to separate the 1-37 version from the 7-37 version. I took advantage of ion-exchange high-pressure liquid chromatography. Working with scientists from Millipore, I was able to separate the two forms. I then needed to show that 7-37 is biologically active at physiological concentrations. For this, the perfused rat pancreas system, which Gordon Weir [at the Joslin Diabetes Center] developed, was critical. Working with Gordon’s technician, we showed that the 7-37 peptide was active in eliciting insulin secretion by the pancreas at picomolar concentrations. What was remarkable, if you look at the figure in the paper, is how parallel the lines were: as the GLP-1 7-37 levels went up, insulin secretion went up. In contrast, the 1-37 peptide was completely inactive (14). These experiments proved my hypothesis. At this point, it became clear that we needed to start clinical studies. [Another key player, Daniel Drucker, worked with Habener and Mojsov, to show that GLP-1 7-37 stimulated insulin secretion in an islet cell culture system in a glucose-dependent fashion (15–17)].

Habener: Meanwhile, studies with two versions of the GLP-1 peptide were going on, in parallel, in Copenhagen, in the group of Holst. Essentially, they arrived at the same conclusion (18). Together, these studies answered the question about what peptide to use in further studies in humans.

Mojsov: At that time, you didn’t need to get FDA approval for clinical studies if the material being tested was made at the same institution where the studies were carried out. The fact that I made the 7-37 peptide at Mass General, where the clinical studies were done, turned out to be critical. We only needed approval from the institutional review board. By 1989, the clinical studies in type 2 diabetic patients had shown a therapeutic effect for 7-37, but it took us a while to publish the work because the reviewers gave us such a hard time, and the paper bounced from journal to journal before finally landing in Diabetes Care (19). But the therapeutic effect was evident. Standing in front of the gamma counter, it was clear who was a placebo and who got the peptide.

PNAS: Lotte, tell us how you became part of the story.

Knudsen: My basic training was in biotechnology, and I never wanted to be an academic scientist. I wanted to do product-oriented, applied science. Denmark is a small country, and I’d had my eye on Novo Nordisk from early on. I started working in a part of the company that is now called Novonesis that was focused on industrial enzymes. At that time, I was mostly doing chemistry in the lab. Back then, my boss, who was an organic chemist and medical doctor, was asked to form a new group in the pharmaceutical part of the company. We were given the task to come up with new ideas for the treatment of type-2 diabetes. GLP-1 was one of those ideas. So, for a while, I was doing a lot of pharmacology work on GLP-1, and, after a few years, I became a project manager.

Lotte Bjerre Knudsen. Image credit: Søren Svendsen (photographer).

Many of the people who were working with me left to go back to the industrial enzymes part of the company, and I found myself surrounded by people who were looking for new medicines to treat type-2 diabetes. I also had colleagues who were looking for new medicines for obesity. Most pharma companies had retreated from obesity after disappointing attempts using different approaches. As the story around GLP-1, diabetes, and obesity started to unfold, I wondered why we couldn’t do both—diabetes and obesity—at the same time. That is probably one of the most important thoughts that I ever had—at a time when most people had shied away from obesity as a druggable problem (20).

The interesting thing is, back then, everyone was looking for small-molecule drugs. A lot of people tried to find small-molecule agonists for the GLP-1 receptor, and, as an interesting side note, my first and only PNAS paper was the report of a small-molecule agonist for the GLP-1 receptor (21). But multiple screens for small molecules to activate the GLP-1 receptor came up empty, and it soon became clear that we needed a peptide approach.

PNAS: One of your major contributions is the invention in 1997 of the peptide-based drug liraglutide, a modified version of GLP-1 that, unlike the rapidly degraded GLP-1, is stable and stays active for 24 hours. Tell us about the chemistry that made this possible.

Knudsen: My major contribution is to be the first one to suggest that we should apply the fatty acid acylation technology to GLP-1 and to successfully lead the invention of the first long-acting GLP-1 receptor agonist based on human GLP-1. The methodology was new. The abundant blood protein albumin has several roles, one of which is a transporter, and albumin has many fatty acid-binding sites. So the idea was to attach a fatty acid to the GLP-1 peptide to make it stable. [Rounds of tinkering with the fatty acid–peptide complex, to which a spacer was added, rendered a version that bound to albumin and was stable and long-acting]. The first paper reporting our work on the fatty acid derivatives of GLP-1 was in the Journal of Medicinal Chemistry in 2000 (22) (and the company filed a patent in 1996).

PNAS: Were there early clues that a GLP-1-based strategy might hold therapeutic promise for obesity?

Knudsen: There are two important papers on the role of GLP-1 in obesity that, in my mind, don’t get cited often enough. The first is a 1995 paper showing that GLP-1 can have a strong anorexic effect in animals (23). It described an animal model with a tumor that secreted glucagon, GLP-1, and GLP-2, and those animals simply starved themselves to death. A year later, another paper showed that when GLP-1 was directly administered into the rat brain, the rats completely stopped eating. This paper identified GLP-1 as a potential neurotransmitter (24). These two papers influenced my work in this area and set the stage for studying the role of GLP-1 in obesity.

Habener: In the late 1980s, we got together with David Nathan at Mass General and others for a clinical study of the 7-37 peptide. The aim was to test the effect of the peptide infusions on insulin and blood glucose levels in nondiabetic volunteers and people with type-2 diabetes after they had had a meal. From the beginning, we saw a problem with the diabetic patients who were given the GLP-1 7-37 infusion: they couldn’t finish the meal. They would get halfway through the meal and say they were full and couldn’t eat anymore. They would try but get nauseated. Previous studies in animals had shown that GLP-1 slowed down acid secretion and stomach emptying and caused early satiety. Looking back, those were early clues that GLP-1 receptor agonists cause food aversion. And if you don’t eat, you lose weight; it’s a no-brainer.

Mojsov: Back in the 1980s, there wasn’t any actual proof that hormones could regulate weight loss in people. The proof that GLP-1 could do so in people came in the mid-1990s through clinical studies in Germany using the GLP-1 7-36 amide version of the peptide. The clinical trial showed that participants who received the peptide lost weight. There were other relevant, preclinical studies. Steve Bloom -and colleagues had done experiments in rats showing that direct injection of GLP-1 in the hypothalamic regions led to a complete loss of appetite (24). Also, I had done experiments showing that the GLP-1 receptor had identical sequences in the brain and pancreas, suggesting that the appetite-suppressing effects in Bloom’s study were mediated by the same GLP-1 receptor that was present in the pancreas (25). Identical receptors in the pancreas and brain suggested that GLP-1(7-37) might also act on the brain for appetite control.

PNAS: Lotte, one of your own contributions was early work showing that GLP-1 is involved in altering food choice in animals—findings that bolster a role for GLP-1 in the brain’s reward mechanisms. Can you describe the work (26)?

Joel Habener. Image credit: MGH Photography Department.

Knudsen: In this paper, we reported that GLP-1 led to a shift in food choice in animals. It later turned out to be important also in people. In the study, we gave rats access to five different kinds of candy, and we saw that the rats on GLP-1 started eating less candy but more chow. There was a shift in food preference. We later went on to thoroughly characterize how liraglutide works on brain circuits.

PNAS: That work is described in an article published in 2014 in the Journal of Clinical Investigation (27). You showed how liraglutide acts on the brain to mediate weight loss.

Knudsen: This paper described the mechanism of action of liraglutide in treating obesity. We developed a new methodology to show how the drugs exert their effect on the brain by working on the circumventricular organs and communicating further into the hypothalamus, among other effects.

Everyone was skeptical about whether there should even be new medicines for obesity. So when we were preparing for our FDA advisory committee in 2014 for the approval of liraglutide for obesity, we had all this data on how GLP-1 works on well-defined neurons in the hypothalamus, in the hindbrain, and in the reward centers. It was very important to show that these medicines work on defined brain circuits and not ones that may be related to psychiatric or cardiovascular side effects, for example. So that has partly formed the basis of the work on the role of GLP-1 in treating addiction. But it’s important to say that there are no strong clinical studies that have demonstrated a role for GLP-1 in treating addiction. It’s a lot of animal studies and anecdotal reports in people.

PNAS: Despite the runaway commercial success of these drugs, they don’t work for everyone, when it comes to weight loss. What are your thoughts?

Knudsen: When you look at the data, especially the large randomized controlled trials, where we can ascertain that people actually take the medicines consistently, they do work for most people. That said, people have different underlying issues when it comes to food; there could be genetic, hormonal, emotional, and psychological influences on people’s relationship with food. Those influences could cause the drugs to work less well in some people.

Mojsov: Obesity is a complex disease. There is an active and ongoing pursuit now to combine pancreatic and intestinal peptides—either in one molecule or many—to increase potency and minimize the side effects of the GLP-1-based drugs. Take, for example, tirzepatide, which was developed by Richard DiMarchi and Matthias Tschöp. This drug combines the sequence of GLP-1 7-37 with GIP [the drug has been approved to treat type 2 diabetes and obesity and can result in up to 20% weight loss]. As more of these drugs are developed, I think some people will respond better to some than others. These efforts could result in targeted therapeutics tailored to specific individuals or populations.

Habener: The more we understand about the mechanism of action of GLP-1 agonists in multiple processes, better drugs can be designed. There are efforts to use ligands that can activate multiple receptors. There is even an effort to develop four-factor agonists, which also activate the fibroblast growth factor receptor, in addition to GLP-1 7-37, GIP, and glucagon. These efforts might help improve the drugs’ efficacy while minimizing side effects.

PNAS: Speaking of side effects, recent work in mice has revealed that it might be possible to target specific neurons in the hindbrain to boost weight loss using GLP-1 receptor agonists without triggering nausea and vomiting, which tend to deter people from taking the drugs (28).

Habener: Yes, that’s the goal of the pharmaceutical drug development process. To come up with a drug that works through the desired pathway. The idea would be to develop a drug that triggers only the appetite and satiety pathways without affecting the vomit-trigger pathway.

PNAS: Do you think the drugs could be targeted to true responders or patient subgroups? Is there any merit to the idea of subtyping obesity?

Knudsen: Of course, we are generally interested in subtyping obesity, but I don’t think there is very strong science right now to suggest that there are particular subtypes of obesity that are more or less responsive to GLP-1 drugs. When we started out in this area of research, we didn’t know much about the genetics of GLP-1. Now, there are Mendelian randomization studies that show that the GLP-1 receptor gene is associated with change in body weight, glucose metabolism, and even heart failure. However, that doesn’t mean there are subsets of people who will get a much bigger response to the treatment.

PNAS: One of the critiques against the use of drugs to treat obesity stems from the concern that overemphasis on drug treatment can downplay structural and environmental factors affecting obesity and flatten a multidimensional problem facing society. Do you have any thoughts on this concern?

Knudsen: It's important to note that the medicines should only be taken as prescribed by doctors and adhering to the label. According to the label, you must do diet and exercise at the same time. They are a part of the treatment. That said, I understand the concern. Still, you could think of parallels with hypertension and high cholesterol. The availability of medicines has created more awareness around those conditions and has been important to further study them. All the time I have been a scientist in pharma, I have heard about the need to treat obesity. And we are only seeing the epidemic of obesity worsening; it’s not going that well without medicines. I also really liked the way Oprah Winfrey phrased this in her show, when she said [and I paraphrase] that people who think there should be medicines for obesity and those who don’t should still be able to respect each other.

Habener: These are complex-trait diseases, which means there are both genetic and environmental influences. The prevalence of diabetes and obesity has gone up in the last 30 to 40 years. For the prevalence to go up so dramatically over that time-period, there has to be an environmental effect. It takes hundreds or thousands of years for genetic mutations to accumulate and genes to change. One possible, or even likely, environmental effect is diet. People’s diets have changed over time: with more ultraprocessed foods.

We should continue to encourage diet and exercise and energy expenditure. For now, diet and exercise alone are not as effective as the drugs. And I don’t expect that the drugs will harm the current exercise-based weight loss programs.

PNAS: The success of the GLP-1 drugs in treating diabetes, weight loss, and cardiovascular disease has shined a spotlight on the importance of prevention, as opposed to treatment. Some companies (notably Novo, Lilly) have launched efforts to develop preventive interventions. How would such preventive drugs work?

Knudsen: We don’t have a solution to the question of prevention, but we have an approach. We have a special unit within the company that works specifically on prevention, but they don’t have any programs that are launched yet. Right now, we are in the phase where we have effective treatments with a well-described risk-benefit ratio. The effective preventative tools we have are diet and exercise—and could we also mention the role of the food industry in the obesity epidemic?

PNAS: Your discoveries have ushered in what some have dubbed a revolution in treating obesity (29–31). How do you feel about the phenomenal role you played in bringing these combined efforts to fruition?

Knudsen: I would like to help inspire future generations of scientists. It’s easy to say now that we should make improved medicines for obesity, but we started this work more than 30 years ago. It’s a long and winding road when you tackle novel problems (32). Also, the evolution of science has been so profound in the last 20 years. I am excited for the future to see how we can tackle the progression of comorbidities in obesity. For example, when we say there’s a 20% risk reduction in cardiovascular disease, there is still 80% risk that we can work on. It’s never been a better time to be a scientist.

Mojsov: It's exciting that we are coming up with new treatments for obesity, which is increasing in prevalence. Once we solve obesity, I think a lot of other disorders will be taken care of. Physiologically, the role of GLP-1 is to integrate different factors controlling metabolism, and the cardiovascular benefits may stem from that role. As for all the other pathophysiological states in which GLP-1-based drugs appear to be beneficial, I think it may have something to do with the drugs’ central role in regulating glucose metabolism. I am especially excited that there will be a lot of basic science to understand the mechanisms by which GLP-1 influences human physiology.

Habener: I am surprised by how all of this has turned out; it seems like a fairy tale. It’s gratifying to see that our discoveries are helping improve the health of people throughout the world. Of course, we don’t know enough about the drugs’ serious side effects, which may emerge in the future, but it seems unlikely, given the number of people who have been taking the drugs. Hundreds of thousands of people have been taking exenatide [a GLP-1 receptor agonist approved for the treatment of type 2 diabetes] for almost 20 years without significant serious side effects, and the benefits far outweigh the risks. Nevertheless, we will have to see how all of this plays out.

The other thing I want to say is about awards. The days of the single investigator and the trusted lab technician are long gone. All this work on GLP-1 has involved hundreds of people. So it’s arbitrary to pick two or three people out of that arena of investigators for an award. Still, I’m humbled and thankful. I just wish I could share the award with more people.