Severe SARS-CoV-2 infection leads to changes in host metabolism, promoting viral replication, altering antiviral immune responses and is reported to cause long-term metabolic complications and sequel in infected individuals [64]. Currently, untargeted and targeted metabolomics have been used to analyze the plasma lipidome and metabolome of COVID-19 patients [65–69], samples across two waves of infection [70] and in healthcare workers exposed to the same risk of COVID-19 [71]. Other studies have investigated different approaches including assessment of biofluids, fecal samples, integrated proteomics and transcriptomic, as well as analyzing gender-specific and post-recovery metabolic changes for a holistic understanding of COVID-19 pathophysiology [72–79]. Emerging research provides compelling evidence that individuals experiencing severe SARS-CoV-2 infections often exhibit multiple metabolic disruptions. These disturbances encompass oxidative stress, altered concentrations of metabolites such as lactic acid, irregularities in energy production and amino acid metabolism, modifications in the metabolism of carnitines, ketone bodies, glucose, fatty acids, and metabolites associated with the tryptophan kynurenine pathway. Additionally, individuals with severe infections may also experience alterations in purine and leukotriene D4 metabolism, as well as imbalances in nutrient and diet-related components. In this study, we employed Kaplan-Meier survival analysis to evaluate the correlation between metabolic markers that were previously reported and patient outcomes. Our findings clearly revealed statistically significant differences in survival outcomes between individuals presenting with elevated and reduced levels of several metabolites. We found aberrant expression of amino acids, tryptophan and kynurenine, carnitine and arginine in severe COVID-19 patients. Here, we identified four metabolites based on multiple established machine learning models, which can distinguish between COVID-19 clinical phenotypes and predict mortality risk.
In this study, 609 targeted metabolites were analyzed in COVID-19 patients that included amino acids and their metabolites, tryptophan and kynurenine and their associated metabolites, SDMA, ADMA, 1-methylhistidine (1-MH), as well as carnitine palmitoyltransferase 1 and 2 enzymes indicators. A key finding reported by several COVID-19 studies is that many amino acids and their related metabolites are dysregulated following severe COVID-19 infection, the majority of which are significantly downregulated [80, 81]. Compared to these studies, in our COVID-19 patients, significant differences in serum amino acid levels between different severity groups and survivors and non-survivors of COVID-19 infection were observed. Amino acids play a key role in immune cell function, tissue regeneration and repair, while an abnormal amino acid metabolism causes neurological symptoms and multiorgan failure [82]. It is well reported that recovered COVID-19 patients have a certain degree of neurological sequelae and severe patients with COVID-19 may develop severe multiple organ failure during hospitalization. We identified that several amino acids and their related metabolites were significantly suppressed i.e. alanine, tryptophan, serine, glutamine, and histidine while upregulated levels of phenylalanine and its product tyrosine were observed in severe and critical COVID-19 cases. These amino acids are key players in energy metabolism, neurotransmitter production and metabolic homeostasis regulation [83, 84]. Multiple studies, which were focused on the role of amino acid metabolism in the progression of COVID-19 infection have discovered that the differentially expressed metabolites between the COVID-91 patients and uninfected individual were enriched with taurine and hypotaurine metabolic pathways [22, 68, 85], suggesting that an overactive taurine pathway can drive the excessive immune response in COVID-19 patients. Therefore, amino acid pathways are a promising targets for drug development that are required for viral replication and virulence.
Other amino acids, including tryptophan derivatives serotonin and tryptophan betaine, 3-indolepropionic acid, and kynurenine remain dysregulated in the severe/critical group compared to mild and asymptomatic groups. Multiple studies have revealed that the metabolome of COVID-19 patients, including products of the tryptophan/kynurenine pathway, reflects the severity of the disease and can be used to predict disease evolution [8, 86, 87]. It has been shown that Interleukin-6 (IL-6) levels were linked to tryptophan metabolism [21]. Kynurenine and arginine are known to be essential for the immunosuppressive activity of dendritic cells, which are critical immunomodulators [88]. Consistent with other studies, severe and critical COVID-19 patients showed dysregulation of tryptophan metabolism, increased markers of oxidative stress and renal dysfunction that correlated with the decreased lymphocyte count that was identified in this cohort [89, 90]. Indeed, several metabolite levels in tryptophan pathway correlated with clinical laboratory markers of inflammation and renal function [21]. Thus, their persistent dysregulation is most likely linked to the underlying molecular mechanism of long-COVID and requires further investigation and targeted interventions.
It is also worth noting significantly elevated levels of short chain acylcarnitines and CPT1 in accord with disease progression. Carnitine is a vitamin-like compound that plays an important role in fatty acid metabolism [91], mainly synthesized in the brain, liver and kidney and primarily stored in the skeletal muscle and heart [92]. Elevated acylcarnitines in COVID-19 patients have been proposed as activators of pro-inflammatory pathways [93], and their imbalance has been related to ATP depletion [94]. Our results support the fact that COVID-19 patients present an over utilization of lipid beta-oxidation pathway to supply the high energetic demand [68]. Thus, this could also suggest an important dysregulation of these metabolites especially the short chain acylcarnitines, which are fundamental for maintaining the optimal energetic status. Furthermore, random forest analysis showed that carnitine, acylcarnitines and CPT1 have excellent performance in survival outcome probability for COVID-19 patients. These results suggest that understanding the metabolic changes of carnitine, acylcarnitines and CPT1 during COVID-19 may advance monitoring disease progression and have a potential prognostic value.
An increase in ferritin level was identified according to disease progression and non-survivors had higher serum ferritin level compared to survivors, confirming enrichment of ferroptosis and energy metabolism pathways in patients with COVID-19. Indeed, the serum of patients with COVID-19 showed an iron imbalance [95] and significantly elevated ferritin levels were related to disease severity, development of acute respiratory distress syndrome (ARDS) and death in COVID-19 patients [96–99]. Furthermore, a study has demonstrated the ability of the COVID-19 infection to target hemoglobin, thus constituting a pivotal step in the pathogenesis of the disease [100]. Consequently, this process ultimately leads to the detachment of porphyrins from iron, subsequent release of iron into the circulation resulting in iron overload and subsequent elevation of ferritin levels [96]. As well documented, most patients with severe COVID-19 disease present with several extra-pulmonary manifestations, leading to multiorgan damage and failure, and the most affected organs are lungs, liver, and kidneys [101]. The multiple organ failure observed in severe COVID-19 patients typically presents around two weeks post-infection [102], and can coincide with the onset of ferroptosis [103].
Recent research has established the usefulness of machine learning (ML) models using routine laboratory test results as clinical decision support tools for COVID-19 diagnoses. ML has been demonstrated to play a significant role in understanding and combating the pandemic, particularly in predicting mortality risk and severity based on laboratory test results. In the present study, we employed multiple well-established ML models to predict the mortality risk model for COVID-19 patients, based on altered metabolites identified through our data and recent literature. This highlights the potential for ML models at present and artificial intelligence models (AI) in near future to provide a clinically valuable tool for predicting death risk in COVID-19 patients based on their metabolomic profile and suggested that research efforts should not overlook metabolic signatures of the disease.
The major limitation of this study is that the samples used were collected during the early days of the pandemic, which may not reflect the status of post vaccinated population cohort that is prevalent. This could potentially impact the generalization of the findings to more recent cases, as the vaccination may impact the metabolic changes observed. Further research is required to better understand the underlying mechanisms of the relationship between these metabolic markers and COVID-19 outcomes.