John Denis McGarry, PhD: A Remembrance of a Master Metabolic Physiologist

John Denis McGarry, PhD (“Denis” to all who knew him), was the most charming and engaging man that I have ever known. He was also one of the most thoughtful and impactful scientists of his generation.

Although remarkably talented, Denis was a humble man. This quality can be traced to his birth and upbringing in Widnes, a working-class town in northwest England near Liverpool. His father Cornelius McGarry was English, and his mother Sarah O’Dell was Irish. Cornelius worked in a chemical production plant making sulfuric acid, and his mother cared for Denis and his two sisters Ann and Pat. They were “quite poor but thrifty,” according to Denis’s wife Angela. Denis went to a local public school until the age of 11 and then won a scholarship to attend St. Francis Xavier’s College Catholic boys school from age 11 until graduation at age 18. During this period, Denis visited a research laboratory at the University of Liverpool and was intrigued by what he saw. Denis was an excellent student and was able to secure scholarships to attend The University of Manchester, where he earned both his Bachelor of Science and PhD degrees, the latter in 1966. His career-long focus on regulation of intermediary metabolism began with his PhD studies with Brian Hodgson at The University of Manchester, where he studied the metabolic fates of radiolabeled propionate in the bacterium Moraxella lwoffi (1,2). He also performed a postgraduate fellowship with Hugh K. King at the University of Liverpool and then moved with Professor King to The University College of Wales, Aberystwyth, where he performed studies on bacterial lipoamide dehydrogenase (3).

Denis met Angela Caldwell, also from Widnes, through friends, and they married in 1967. They moved to Dallas, Texas, in 1968, for Denis to perform a postdoctoral fellowship in the laboratory of Daniel W. Foster, MD, at the The University of Texas Southwestern Medical Center. Little did the young couple realize that the move would be permanent! Denis’s talent was immediately obvious, and he was promoted to the faculty in the Departments of Internal Medicine and Biochemistry in 1969, at which time he and Dan Foster decided to codirect a research laboratory. The McGarrys raised three beautiful children in Dallas—Margaret, John, and Denise, who have in turn produced five grandchildren—Shane, Alexander, Tristan, Rhys, and Tessa. By 1977, Denis had ascended rapidly to the rank of professor of internal medicine and biochemistry, and in 1996 he was named the Clifton and Betsy Robinson Chair in Biomedical Research. The McGarry/Foster partnership flourished for 33 years and led to a series of impactful scientific discoveries, as will be summarized. For his leading role in this work, Denis McGarry earned numerous awards and honors, including the Lilly Outstanding Scientific Achievement Award and Banting Medal for Scientific Achievement from the American Diabetes Association, the Joslin Medal, and the David Rumbough Award for Scientific Excellence and the Gerold & Kayla Grodsky Basic Research Scientist Award from JDRF.

Denis was a mesmerizing public speaker and conversationalist who was suddenly and ironically stricken with expressive aphasia (an inability to speak) in 2001. The cause was found to be a brain tumor, which led to his untimely death on 27 January 2002 at the age of 61. Reflecting the broad esteem and respect that he enjoyed, his many friends and colleagues came together upon his death to establish and fund the John Denis McGarry, Ph.D. Distinguished Chair in Diabetes and Metabolic Research in his honor at The University of Texas Southwestern Medical Center. Fittingly, Denis’s long-time collaborator Dan Foster held the McGarry Chair until his recent passing on 20 January 2018.

Denis came to Dallas in 1968 to work with Dan Foster on the ketogenesis pathway and its regulation in normal and diabetic states. Denis’s deep knowledge of metabolic biochemistry blended well with Dan Foster’s expertise in clinical diabetes, and they decided to focus on regulation of fatty acid oxidation and ketone production by the two key endocrine hormones, glucagon and insulin. Their interest in endocrine regulation of the pathway was spurred by their irrepressible friend and Dallas colleague, Roger Unger, who had established the world’s first radioimmunoassay for glucagon and defined many of the hormone’s physiologic functions (impressively, Unger, McGarry, and Foster all went on to earn the highest prize of the American Diabetes Association, the Banting Medal, during their careers).

In an elegant series of studies influenced by prior work from Mayes and Felts (4) and Krebs and Hems (5), Denis perfected the surgical methods required to perfuse rat liver, and he used this model system and liver extracts to compare metabolic fates of the 18-carbon fatty acid oleate and the 8-carbon fatty acid octanoate in livers from fed, fasted, and diabetic rats (6–8). He found that oleate oxidation to ketones occurred at a much higher rate in livers from diabetic or fasted compared with fed rats, whereas octanoate oxidation was similar across these conditions. He also noted that the low rate of oleate oxidation in liver of fed rats was balanced by an increase in fatty acid esterification and storage. It was known at the time that the carnitine acyltransferase (CAT) system was required to transport the long-chain fatty acid oleate through the mitochondrial membrane to engage with the fatty acid oxidation machinery, whereas octanoate could cross the mitochondrial membrane independent of the CAT system. These studies therefore suggested that the CAT system might be a regulatory site for control of long-chain fatty acid oxidation in response to changes in nutritional status and hormonal milieu. In support of this idea, Denis showed that inclusion of a CAT inhibitor, d-decanoylcarnitine, during perfusion of livers from ketotic rats abolished ketone production and caused oleate to be shunted into the esterification pathway and triglyceride formation (9).

The concept that carbohydrate ingestion serves to suppress fatty acid oxidation, referred to as “carbohydrate sparing of hepatic fatty acid oxidation,” had been prevalent for decades prior to Denis’s work (10). This idea, coupled with Denis’s early studies, framed a burning question—what is the biochemical factor that links changes in the glucagon-to-insulin ratio (high in fasting/diabetes, low in the fed state) and carbohydrate metabolism to control of long-chain fatty acid oxidation at the level of the CAT system? An important clue was that all the manipulations used by Denis and his collegues to enhance fatty acid oxidation, including alloxan diabetes, fasting, or infusion of an insulin-neutralizing antibody, resulted in a striking depletion of liver glycogen (11). Denis became convinced that a product of glycogen degradation and glucose catabolism might serve to modulate fatty acid oxidation. Beginning in 1975, he and his fellows performed what Denis later described as an “exhaustive search over a period of some 18 months” (12) for a regulatory factor from the classical glucose metabolic pathways capable of regulating fatty acid oxidation in liver homogenates, a survey that included intermediates of glycogen metabolism, glycolysis, the pentose monophosphate shunt, and the tricarboxylic acid cycle (TCA) cycle. None served to alter the oxidation of ¹⁴C-labeled oleate to ketones (12,13). Denis expanded his thinking about the problem, and as he liked to say in recounting the moment years later, “it was then that the penny dropped.” Using the reciprocal regulation of glycogen degradation and synthesis in the fasted and fed states as a model, Denis began to think of factors that might be utilized for reciprocal regulation of lipogenesis and fatty acid oxidation. He landed on malonyl CoA, which is linked to carbohydrate metabolism via the TCA cycle intermediate citrate, which is cleaved by ATP-citrate lyase to form acetyl CoA and oxaloacetate. Acetyl CoA is then converted to malonyl CoA by acetyl CoA carboxylase (ACC). Importantly, malonyl CoA is the immediate precursor of fatty acid synthesis (lipogenesis) by the fatty acid synthase (FAS) enzyme complex. Could it be that it was also the elusive glucose-derived factor responsible for reciprocal regulation of fatty acid oxidation?

In a classic article published in the The Journal of Clinical Investigation in 1977, Denis, his postdoctoral fellow Dr. Guy Mannaerts from Belgium, and Dan Foster demonstrated clearly that malonyl CoA inhibits the oxidation of oleate to produce ketones (13). The inhibition of long-chain fatty acid oxidation by malonyl CoA was shown to be specific, as the related metabolites acetyl, propionyl, and methylmalonyl CoAs were without effect. The inhibitory effect of malonyl CoA was absent in liver extracts when octanoate was used as the substrate, strongly suggesting that malonyl CoA acts at the level of the CAT system, which is required for entry of long-chain but not medium-chain fatty acids into the mitochondria. Malonyl CoA inhibition of fatty acid oxidation could also be demonstrated in isolated mitochondria, further supporting the idea that malonyl CoA acts upon a protein residing on the solvent exposed mitochondrial membrane. Some years earlier, Kopec and Fritz (14,15) had purified and characterized carnitine palmitoyltransferase 1 (CPT1) and carnitine palmitoyltransferase 2 (CPT2) as components of the CAT system, demonstrating that CPT1 is responsible for generating long-chain fatty acylcarnitines from long-chain acyl CoAs on the outer aspect of the mitochondrial membrane, whereas CPT2 catalyzes the reconversion of acylcarnitines to acyl CoAs after transport of acylcarnitines into mitochondria. With this framework in hand, Denis concluded that “since malonyl CoA would not be expected to gain access to CATII (CPT2), which is situated on the inner aspect of the inner mitochondrial membrane…., the more likely site of interaction would appear to be the CATI (CPT1) step” (13). Shortly thereafter, McGarry, Leatherman, and Foster (16) proved that CPT1 was the site of inhibition of hepatic fatty acid oxidation by malonyl CoA via several key experiments, including the following: 1) malonyl CoA was shown to inhibit ∼50% of total carnitine palmitoyltransferase activity (representing the combined CPT1 and CPT2 activity in mitochondria), and 2) malonyl CoA blocked the oxidation of palmitoyl CoA (the CPT1 substrate) but had no effect on oxidation of palmitoylcarnitine (the CPT2 substrate).

Put quite simply, the discovery of regulation of fatty acid oxidation by malonyl CoA is now a foundational pillar for understanding of metabolic regulation under different nutritional and physiologic conditions. For example, since 1977, we have learned a great deal about mechanisms that result in reciprocal activation of glycolysis and suppression of gluconeogenesis in the fasted-to-fed transition, including the role of fructose-2,6-bisphosphate for activation of phosphofructokinase and mechanisms of transcriptional upregulation of key glycolytic enzymes (17–19). Thanks to the brilliant studies of Denis and his colleagues, these events can be neatly tied to inhibition of fatty acid oxidation and activation of lipid esterification in the fed state via increases in the levels of malonyl CoA and inhibition of the CAT system.

With the emergence of molecular biology methods in metabolic research, the McGarry/Foster laboratory (20–26) and others (27–31) made further important contributions to our understanding of the CAT system and its regulation through-out the 1980s and 1990s. This included work showing that different isoforms of CPT1 are expressed in liver (CPT1a), skeletal muscle (CPT1b), and brain (CPT1c), with some tissues such as heart “switching” from one isoform to another as a function of developmental stage. These isoforms were also shown to have different sensitivities for malonyl CoA inhibition. The studies included mutagenesis experiments that defined protein domains of CPT1 required for its catalytic activity, regulation by malonyl CoA, and insertion in the outer aspect of the mitochondrial membrane. Fellows and students that contributed strongly to this productive phase of work and who enjoyed close relationships with Denis included Victoria Esser, Nicholas Brown, Scott Mills, Keith Woeltje, Shawn Adams, Charles Britton, and Brian Weis, all of whom enjoyed outstanding technical support from Murphy Daniels and Petra (“Pete”) Contreras.

Although the CAT system and its regulation was a central focus of Denis’s scientific career, his outstanding instincts as a metabolic physiologist allowed for occasional exploration of tangential topics. One of these excursions involved your faithful author, then (in 1980) a bright-eyed but naive trainee just entering graduate school after predoctoral internships with Jens Holst in Copenhagen and Roger Unger in Dallas. Denis framed an irresistible question for me—if glucose is regarded as the major direct precursor for liver glycogen synthesis, why are glycogen synthesis rates in the perfused liver or isolated hepatocytes very low in the presence of high levels of glucose alone but much higher when gluconeogenic precursors such as fructose, pyruvate, or alanine are provided? Using a combination of methods including tracing of radiolabeled substrates into glycogen in living rats and enzymologic measurements, we demonstrated that a significant portion of liver glycogen synthesized in the postprandial state comes from an “indirect” pathway involving gluconeogenic precursors and not only from a “direct” pathway in which glucose is phosphorylated by glucokinase and converted to UDP-glucose for glycogen synthesis (32–35). Masamichi (“Mitchie”) Kuwajima, a fellow in the laboratory, also made important contributions to the project in its later stages. Similar findings were obtained in animal experiments from other groups (36,37) and in elegant studies in humans by Gerald Shulman and colleagues (38). Denis also became interested in the role of fatty acids in control of insulin secretion. He and his then fellows Daniel Stein and Robert Dobbins demonstrated that depletion of fatty acids caused strong impairment of insulin secretion in response to glucose and other secretagogues (39,40) and that the insulinotropic potency of fatty acids varied by chain length and degree of saturation (41). Finally, in collaboration with my laboratory (42–45), as well as those of other good friends Marc Prentki and Barbara Corkey (46,47), Denis explored the role of the malonyl CoA/CAT system in regulation of metabolism and insulin secretion in pancreatic islets, with many interesting observations.

The foregoing lasting contributions would easily warrant this article of remembrance. But Denis was not finished. Displaying his remarkable command of systems physiology, long before the advent of today’s machine learning tools, Denis began to think about the metabolic derangements of diabetes from a new perspective, leading ultimately to the publication of a seminal and broadly impactful editorial in Science in 1992 entitled “What if Minkowski had been ageusic? An alternative angle on diabetes” (48). The premise of the article was that the field’s long-standing focus on diabetes as a disease of insulin resistance and hyperglycemia may have obscured appreciation of a role for dyslipidemia in disease pathogenesis. The reference to Oskar Minkowski, a major figure in 19th century diabetes research at the University of Breslau (now Wrocław, Poland), comes from studies that he conducted in pancreatectomized dogs. Minkowski noticed that urine samples from these dogs attracted many flies, purportedly leading him (gulp!) to taste the urine and comment upon its sweetness. The thrust of the logic developed in Denis’s editorial is that if Minkowski had been unable to taste (aguesia) and had to rely instead on his sense of smell, he would have noticed the distinct aroma of ketones, the byproducts of fatty acid oxidation, possibly altering perspectives on metabolic derangements in diabetes over the ensuing century. The article then moves to a systematic and logical framing of a model in which hyperinsulinemia is an early step in emerging diabetes, which via its effects to suppress fatty acid oxidation and promote lipid storage, contributes to ectopic accumulation of fat in peripheral tissues such as liver, skeletal muscle, heart, and the pancreatic islets. Unlike adipose tissue, these organs are not designed for fat storage, such that accumulation of lipids in nonadipose tissues leads to “lipotoxicity,” manifest as impairments in metabolism, increased oxidative stress, and deterioration of key tissue functions. These concepts, while worked on by others at the time, had never been framed in such a lucid and complete way, and the article has influenced nearly every laboratory working in the field of diabetes since its publication. Denis himself devoted significant effort to following up on his ideas in the later stages of his career, including use of nuclear magnetic resonance imaging to distinguish the inter- and intramyocellular fat depots in obese, insulin-resistant subjects from those in insulin-sensitive subjects (49,50), an approach also championed by Shulman and colleagues (51). The Minkowski article has remained a guiding light for the field and secures the legacy of J. Denis McGarry as a generational thought leader and contributor of discoveries and ideas, whose impact is still felt strongly 16 years after his untimely death.

I close with some personal memories. First, what was it like to have Denis McGarry as a mentor? Denis was meticulous and believed in training at a very deep level of detail, for example, actually supervising pipetting technique to ensure clear and precise results! He also insisted on maintenance of detailed laboratory notebooks, which he would review on a regular, sometimes daily basis. He was almost maniacally interested in new data, and the only disagreement that we ever had during my graduate training came when he once grabbed my data off the scintillation counter in the early morning, after I had worked very late the night before to generate the samples. While these characteristics could be mildly maddening, the PhD I earned with Denis truly equipped me to think critically and independently, and I know the same to be true for his many other graduate students and postdoctoral trainees. Moreover, we all had the spectacular privilege of working with a walking encyclopedia of metabolic knowledge, who loved to discuss and develop new ideas. He also wrote beautifully and with great flair, always on a legal pad with a freshly sharpened pencil that he kept perched on his ear when not in use. When I would bring my labored prose to him for review from my electric typewriter, he would switch to a red pen for corrections, and I eventually learned scientific writing from those fields of red. My 4 years in his laboratory remain among the best of my life.

And what was it like to have Denis as a friend? I was particularly fortunate in this regard, as we interacted closely both in my graduate student years (1980–1984) and then again as faculty colleagues in the Departments of Biochemistry and Internal Medicine and the Touchstone Diabetes Center at The University of Texas Southwestern Medical Center, Dallas (1987–2002). As I mentioned earlier, he was blessed with spectacular charm, elegance, and conviviality. Imagine a cross between Cary Grant and Peter Lawford, with deep scientific knowledge and a twinkle in his eye! He loved nothing more than to sit down over a meal (Italian food was a favorite) and a bottle of wine to engage in wide-ranging conversations of science, philosophy, religion, or any other interesting topics of the day. He had a way of making everyone in his company feel special, with an inspiring natural grace and sense of compassion. Whether at one of his favorite restaurants in Dallas, at his warm and welcoming home with Angela, or at a bistro or tavern in a faraway place, Denis made us feel that we were lucky to be scientists, lucky to be alive, and lucky to be in his company.