COVID-19 infection alters kynurenine and fatty acid metabolism, correlating with IL-6 levels and renal status

Tryptophan is an amino acid most people know only as the thing in turkey that supposedly makes you sleepy. However, in COVID-19 patients, the body starts consuming tryptophan in a completely different way, and the enzyme responsible is one the immune system normally uses to starve pathogens. A team led by Angelo D'Alessandro in Denver found that this metabolic shift leaves a chemical signature in the blood that tracks almost perfectly with how sick a patient becomes. Here's the underlying logic. Viruses do not just infect cells; they commandeer the host's biochemical machinery to manufacture new copies of themselves. They need raw materials: free amino acids for proteins, lipids for membranes, and nucleotides for genetic material. Thomas and colleagues frame their entire study around this premise — that SARS-CoV-2, like other coronaviruses before it, would leave a metabolic fingerprint in the blood reflecting exactly this kind of hijacking. Their question was whether that fingerprint could tell us something clinically useful about severity, organ stress, or where to intervene. To find out, they ran a serum metabolomics study comparing thirty-three COVID-19-positive patients, confirmed by nucleic acid testing, to sixteen negative controls. They used both targeted and untargeted mass spectrometry-based metabolomics — methods that allow you to measure hundreds of small molecules simultaneously — and then correlated those metabolite levels against clinical markers of inflammation and kidney function. The most striking finding was what happened to tryptophan. In COVID-19 patients, tryptophan levels fell sharply while kynurenine — a downstream metabolite that tryptophan normally does not produce much of — rose dramatically. The enzyme doing this diversion is IDO1, indoleamine two, three-dioxygenase, which sits at a metabolic fork: it pulls tryptophan away from serotonin and protein synthesis and shunts it into the kynurenine pathway instead. The team used the kynurenine-to-tryptophan ratio as a proxy for IDO1 activity. More kynurenine relative to tryptophan means the enzyme is running harder. And it was running very hard. Tryptophan levels below 105 micromolar and kynurenine levels above 5.3 micromolar distinguished COVID-19-positive from negative sera with areas under the receiver operating characteristic curve greater than ninety-five percent. Those are remarkably clean numbers for a metabolic biomarker. These numbers correlated with how sick patients were. Thomas and colleagues divided the COVID-19 group by interleukin-six levels. Interleukin-six is the inflammatory cytokine most closely associated with severe COVID disease. They divided patients into low, which included five patients at or below ten picograms per milliliter, medium, which included ten patients from ten to sixty-five picograms per milliliter, and high, which included eighteen patients above ninety picograms per milliliter. As interleukin-six rose, tryptophan fell and kynurenine climbed in proportion. IDO1 activity, inferred from the metabolite ratio, inversely correlated with interleukin-six levels. The kynurenine pathway was not just activated; it was activated in lockstep with systemic inflammation. The immunological implications of this are genuinely complicated. IDO1 normally helps the immune system starve pathogens by depleting the tryptophan they need to replicate. Animal data cited by the authors also show that IDO activity limits lung inflammation and that genetic deletion of IDO worsens inflammatory lung pathology in mice. Therefore, IDO1 activation might be the immune system doing something useful. However, the downstream kynurenine metabolites, which the paper shows rising alongside kynurenine itself, including kynurenic acid, picolinic acid, and nicotinic acid, activate the aryl hydrocarbon receptor, and those downstream effects are broadly immunosuppressive. Proliferating immune cells trying to mount an antiviral response get starved of tryptophan. That's potentially a problem. In this context, IDO1 may be simultaneously limiting damaging lung inflammation and blunting the adaptive immune response needed to clear the virus. The authors are careful to state that causality cannot be established from observational data. But the correlation is real and it is tight. The kynurenine story is the centerpiece, but it sits inside a much larger metabolic disruption. Thomas and colleagues found widespread dysregulation of nitrogen metabolism — altered levels of most amino acids, not random noise but a coherent pattern of oxidative stress, protein breakdown, and impaired renal handling. Methionine sulfoxide and cystine were both elevated, and these are oxidized forms of sulfur-containing amino acids — signals that the blood's chemical environment had become significantly more oxidizing. Cysteine and taurine, by contrast, were decreased. Polyamines, specifically spermidine and acetyl-spermidine, were greatly increased, consistent with enhanced amino acid catabolism and accelerated cell turnover. Increases in acetyl-methionine and hydroxyproline pointed to active proteolysis, the large-scale breakdown of proteins into their component amino acids. The picture that emerges is one of systemic metabolic stress: the body breaking down its own proteins, generating oxidative byproducts, and struggling to clear the resulting nitrogen waste. Some of that nitrogen waste was landing in the kidney. Creatine and creatinine, which are standard markers of renal function, were elevated in COVID-19 patients. The mass spectrometry measurement of creatinine correlated with the clinical laboratory measurement with a Spearman correlation coefficient of 0.93. That near-perfect agreement between the metabolomics platform and the clinical test is not just a methods validation note; it tells you that the metabolomics data are capturing real biology happening in real patients. Then there is the carbon side of the equation. In addition to all the nitrogen disruption, Thomas and colleagues found altered carbon homeostasis: every patient in the study showed hyperglycemia, which is elevated circulating glucose, confirmed by clinical laboratory measurements. Free fatty acids, nearly all of the ones measured, were increased in all COVID-19 patients independent of interleukin-six level. At the same time, short- and medium-chain acylcarnitines, which are the molecules that ferry fatty acids into mitochondria for burning, were significantly decreased. That combination of more fatty acids in circulation and less being shuttled into energy production is consistent with a picture where lipids are being mobilized but not consumed for fuel. The authors note that other coronaviruses activate phospholipase A2 to mobilize free fatty acids for viral membrane formation. Whether that's happening here remains to be established causally, but the elevated free fatty acids fit that hypothesis. Now if we pull all of this together through the clinical correlations, the metabolomics picture becomes directly medically relevant. Several short- and medium-chain acylcarnitines, kynurenine, and methionine sulfoxide ranked among the top positive correlates of interleukin-six in Spearman correlation analyses. C-reactive protein, which is another standard inflammation marker, showed strong negative correlations with free fatty acids and tryptophan, and positive correlations with picolinic acid and inosine. Blood urea nitrogen and clinical creatinine both correlated positively with acylcarnitines and with oxidized purines like inosine. These are not isolated signals; they map metabolic shifts onto two clinically important axes simultaneously: systemic inflammation, tracked by interleukin-six and C-reactive protein, and renal impairment, tracked by blood urea nitrogen and creatinine. This matters because it suggests these metabolic signatures could serve as early biochemical indicators of worsening disease. They might flag patients at risk of cytokine storm through elevated kynurenine and oxidized amino acids, or identify patients headed toward renal complications through acylcarnitine and purine accumulation. The authors highlight IDO1 and its upstream regulators, particularly interferon signaling, as candidate therapeutic targets. Drugs that modulate IDO1 already exist in oncology contexts, where the goal is to relieve immunosuppression in tumors. Whether the same strategy would help, hurt, or do both in COVID-19, given IDO1's dual role in potentially limiting and impairing immunity, is a question this study raises but cannot answer. The honest limits here are worth naming. Thirty-three COVID-19-positive patients is a small cohort. Only six of those patients had renal parameters within normal reference ranges. This is an observational study, which means the correlations between metabolites and clinical markers indicate these things move together, not that one causes the other. What Thomas and colleagues have built is a signal-finding study — one that demonstrates the metabolomics approach works, identifies the pathways most worth pursuing, and generates a set of mechanistic hypotheses specific enough to test. The chemical signature of COVID-19 in the blood is coherent. The kynurenine pathway sits at its center, linking immune activation to amino acid metabolism to clinical markers of organ stress. Free fatty acids and glucose point to carbon hijacking. Oxidized amino acids and polyamines signal systemic protein breakdown. And all of it correlates with how inflamed and how renally compromised a patient becomes. The next step is larger cohorts, longitudinal measurements, and eventually intervention studies — testing whether the pathways this work identified can be targets, not just markers. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.