Reduced circulating FABP2 in patients with moderate to severe COVID-19 may indicate enterocyte functional change rather than cell death

Abstract

The gut is of importance in the pathology of COVID-19 both as a route of infection, and gut dysfunction influencing the severity of disease. Systemic changes caused by SARS-CoV-2 gut infection include alterations in circulating levels of metabolites, nutrients and microbial products which alter immune and inflammatory responses. Circulating plasma markers for gut inflammation and damage such as zonulin, lipopolysaccharide and β-glycan increase in plasma along with severity of disease. However, Intestinal Fatty Acid Binding Protein / Fatty Acid Binding Protein 2 (I-FABP/FABP2), a widely used biomarker for gut cell death, has paradoxically been shown to be reduced in moderate to severe COVID-19. We also found this pattern in a pilot cohort of mild (n = 18) and moderately severe (n = 19) COVID-19 patients in Milan from March to June 2020. These patients were part of the first phase of COVID-19 in Europe and were therefore all unvaccinated. After exclusion of outliers, patients with more severe vs milder disease showed reduced FABP2 levels (median [IQR]) (124 [368] vs. 274 [558] pg/mL, P < 0.01). A reduction in NMR measured plasma relative lipid-CH3 levels approached significance (median [IQR]) (0.081 [0.011] vs. 0.073 [0.024], P = 0.06). Changes in circulating lipid levels are another feature commonly observed in severe COVID-19 and a weak positive correlation was observed in the more severe group between reduced FABP2 and reduced relative lipid-CH3 and lipid-CH2 levels. FABP2 is a key regulator of enterocyte lipid import, a process which is inhibited by gut SARS-CoV-2 infection. We propose that the reduced circulating FABP2 in moderate to severe COVID-19 is a marker of infected enterocyte functional change rather than gut damage, which could also contribute to the development of hypolipidemia in patients with more severe disease.

Figure 1

Plasma FABP2 but not zonulin is decreased in patients with more severe COVID-19 compared to patients with milder COVID-19. (A) Plasma FABP2 levels in patients with milder (n = 16) or more severe COVID-19 (n = 15), respectively. (B) Plasma zonulin levels in patients with milder (n = 16) or more severe (n = 14) COVID-19, respectively. Student’s t-test used for group comparisons, mean ± SEM indicated with red lines.

Figure 2

(A,B) Illustrative expanded NMR spectra from patients in the moderately severe Covid group with differing FABP2 levels. (A) ID 112, FABP2 38 pg/mL, (B) ID 433, FABP2 195 pg/mL. The peaks marked with * were excluded from the analysis and include EDTA peaks (3.635–3.600 ppm; 3.235–3.220 ppm, 3.18–3.07 ppm; 2.72–2.68 ppm; 2.58–2.53 ppm) and ethanol peaks (quartet 3.685–3.635 ppm, triplet 1.20–1.06 ppm). Lac, lactate; L–CH₂, lipid-CH₂ peak; L–CH₃, lipid-CH₃ peak. (C) Principal Component Analysis showing no difference between the mild [1, pink] and moderately severe [2, green] groups. (D) Partial Least Squares Discriminant Analysis showing no difference between the mild [1, pink] and moderately severe [2, green] groups. (E) Variable in Importance Projection plot showing L–CH₂ and L–CH₃ to be the top discriminating metabolites between the mild [1] and moderately severe [2] groups.

Figure 3

(A) Correlation Pearson R matrix showing association between gut biomarkers and plasma NMR metabolites for the milder group (N = 15), after two outliers were removed, as determined by ROUT (ID 503, 291). (B) Associations between gut biomarkers and plasma NMR metabolites for the more severe group (N = 13), after two outliers were removed, as determined by ROUT analysis (ID 506, 562). Blue indicates a positive association, red indicates a negative association. Stronger associations indicated by deeper colour.