Gut ACE2 expression, tryptophan deficiency and inflammatory responses: the potential connection that should not be ignored during SARS-CoV-2 infection.

An “atypical” physiological function of angiotensin-converting enzyme 2 (ACE2) during COVID-19 is rarely mentioned. In the intestine, where ACE2 is highly expressed, it acts as a regulator of the homeostasis of dietary amino acids, particularly tryptophan (Trp). Trp catabolism through the kynurenine (Kyn) pathway plays an important role in regulating the balance between effector and regulatory immune responses; Trp catabolism is increased in inflammatory settings to attenuate excessive systemic immune activation and thus exerts protective effects against many diseases.,

Given that interaction with SARS-CoV-2 decreases the cell surface expression of ACE2 by endocytosis and ADAM17 cleavage, we performed a preliminary study and proposed a hypothesis that SARS-CoV-2 infection downregulates ACE2 expression in intestinal epithelial cells, reduces Trp absorption, and thus disrupts the local and systemic inflammatory responses. This hypothesis provides a possible reason for the digestive symptoms and hyperinflammatory phenotypes of patients with COVID-19.

We treated mice with recombinant SARS-CoV-2 spike protein (receptor binding domain [RBD]) to mimic SARS-CoV-2 infection and further challenged the mice with lipopolysaccharide (LPS) to induce a hyperinflammatory status. LPS stimulation reduced the serum Trp levels of the mice in both the control and RBD groups (Figure 1A), and this effect was accompanied by elevated Kyn levels and Kyn/Trp ratios (Figure 1B, Supplementary Figure 1A); these results suggested the increased catabolism of Trp to Kyn and higher levels of inflammation. Although the Kyn/Trp ratios in the RBD group were similar to those in the control group (Supplementary Figure 1A), the RBD-treated mice showed lower serum Trp and Kyn levels than the untreated mice regardless of LPS stimulation (Figure 1A and B).

Figure 1

Serum levels of Trp ( Relative serum levels of Trp (C) and Kyn (D) in healthy control subjects, non-COVID-19 pneumonia patients, nonsevere COVID-19 patients, and severe COVID-19 patients. (E) Representative immunofluorescence images of ACE2 (red) and panCK (green) counterstained with DAPI (blue) in the intestinal tissue of mice. Expression of ACE2 in MOD-K (F) and HNM460 cells (G). (H) Serum levels of Trp in wild-type (WT) and ACE2 knockout (ACE2KO) mice (n = 5). (I) Serum levels of cytokines in LPS-challenged mice treated with or without RBD and L-Trp (n = 5). (J) Serum levels of cytokines in WT and ACE2KO mice treated with or without LPS, RBD, or L-Trp (n = 5). (K) Survival of mice challenged with lethal levels of LPS and treated with or without RBD and L-Trp (n = 9).∗<0.05, ∗∗<0.01, ∗∗∗<0.001.

Supplementary Figure 1

Serum levels of Trp and its metabolites in mice and human. (A) The ratio of serum Kyn/Trp in mice treated with or without RBD and LPS. (n = 10). The ratio of serum Kyn/Trp (B), serum levels of serotonin (C), and quinolinate (D) in healthy, non-COVID-19 pneumonia patients, nonsevere COVID-19 patients, and severe COVID-19 patients. The data are representative of 3 (A) independent experiments, and analyzed by 1-way analysis of variant analysis (A), Mann-Whiney test (B-D). The data indicate the mean ± 95% confidence intervals. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. NS, not significant.

To further verify our results, we analyzed the metabonomic data from a published article. The serum levels of Trp and Kyn in COVID-19 patients were significantly lower than those in healthy persons and non-COVID-19 pneumonia patients; however, the Kyn/Trp ratios were comparable (Figure 1C and D, Supplementary Figure 1B). The levels of other important metabolites of Trp, such as quinolinate and serotonin, were also decreased in COVID-19 patients (Supplementary Figure 1C and D).

Immunofluorescence staining showed that ACE2 expression on the surface of intestinal epithelial cells was significantly decreased in RBD-treated mice (Figure 1E). Similarly, ACE2 expression on the surface of MOD-K and HNM460 cells, 2 kinds of intestinal epithelial cells from mice and humans, was also notably decreased after in vitro treatment with RBD and S1 protein (another SARS-CoV-2 spike protein) (Figure 1F and G). The serum Trp levels of ACE2-deficient mice (ACE2KO) were lower than those of wild-type mice, and RBD administration did not reduce the serum Trp levels in the ACE2KO mice treated with or without LPS (Figure 1H), suggesting a central role of ACE2 in the Trp deficiency induced by SARS-CoV-2 infection.

Furthermore, no significant difference was observed in the serum levels of inflammatory cytokines between the control animals and RBD-treated mice (Supplementary Figure 2A), indicating that RBD alone could not induce a systemic inflammatory response. However, RBD administration significantly enhanced LPS-induced hyperinflammation in mice, as evidenced by increased levels of proinflammatory cytokines (IFNγ, IFNα, IFNβ, TNFα, IL1β, and IL6), decreased levels of anti-inflammatory cytokines (IL10) (Figure 1I, Supplementary Figure 2B-E), and aggravated local infiltration of inflammatory cells and injury to the lung and liver (Supplementary Figure 3). The effect of RBD was negated or even reversed by L-Trp supplementation or ACE2 depletion (Figure 1I and J, Supplementary Figure 2B-E). When challenged with a sublethal dose of LPS (25 mg/kg), the RBD-treated group showed poorer survival than the control group (Figure 1K). Notably, resupplying mice with L-Trp greatly improved survival in the control and RBD-treated groups and abolished the difference between these groups (Figure 1K).

Supplementary Figure 2

Serum levels of cytokines in mice. (A) Serum levels of cytokines in mice treated with or without RBD. (n = 5). Serum levels of IFNγ (B), IL6 (C), TNFα (D), and IL10 (E) in LPS-challenged mice treated with or without RBD and L-Trp (n = 5). The data are representative of 2 independent experiments, and analyzed by unpaired Students t test (A) and 1-way analysis of variant analysis (B-E). The data indicate the mean ± 95% confidence intervals. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. NS, not significant.

Supplementary Figure 3

Histopathological analyses of lung and liver injury. Representative images of the hematoxylin-eosin staining of lung and liver paraffin sections; necrotic hepatocytes indicated by black arrow (A). Lung injury score (B) and necrotic hepatocyte counts (C) were used to assess target organ damage caused by hyperinflammation (n = 10). The data are representative of 2 (B and C) independent experiments, and analyzed by 1-way analysis of variant analysis (B and C). The data indicate the mean ± 95% confidence intervals. ∗P < .05, ∗∗P < .01, and ∗∗∗P < .001. NS, not significant.

Here, we found that Trp deficiency is a prominent characteristic and an important driving factor of the pathobiology of COVID-19. Nearly 50.5% of COVID-19 patients report digestive symptoms; Trp deficiency has been proven to enhance susceptibility to Dextran sulphate sodium, DSS-induced colitis. These findings allow us to infer that digestive symptoms of COVID-19 patients may be partly attributed to disrupted Trp homeostasis.

More importantly, by activating the aryl hydrocarbon receptor pathway and allowing the generation of regulatory T cells, Kyn plays a critical role in regulating disease tolerance and immune homeostasis. Our results suggested that alterations in the Kyn pathway impair the negative self-regulatory capacity of the immune system and thus promote the occurrence of hyperinflammation and cytokine storm syndrome, which makes COVID-19 more lethal.

Moreover, the Kyn pathway provides raw materials for the synthesis of nicotinamide adenine dinucleotide (NAD), which is important for regulating oxidative stress and DNA damage repair. Recent studies revealed that SARS-CoV-2 infection might decrease repletion of the NAD metabolome from Trp, and supplementation with NAD could effectively reverse cytokine storms and organ damage. Therefore, NAD supplementation may be a potential explanation for the therapeutic effect of L-Trp on SARS-CoV-2-related hyperinflammation.

Overall, as shown in Figure 2, attention should be given to the effect of ACE2-mediated Trp deficiency on the course of COVID-19. Further research is needed to pave the way for the development of new strategies to modulate COVID-19 by targeting Trp metabolism.

Figure 2

Schematic diagram of the role of SARS-CoV-2-induced Trp deficiency in the pathobiology of COVID-19.