Complement lectin pathway activation is associated with COVID-19 disease severity, independent of MBL2 genotype subgroups

Abstract

Introduction: While complement is a contributor to disease severity in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections, all three complement pathways might be activated by the virus. Lectin pathway activation occurs through different pattern recognition molecules, including mannan binding lectin (MBL), a protein shown to interact with SARS-CoV-2 proteins. However, the exact role of lectin pathway activation and its key pattern recognition molecule MBL in COVID-19 is still not fully understood.

Methods: We therefore investigated activation of the lectin pathway in two independent cohorts of SARS-CoV-2 infected patients, while also analysing MBL protein levels and potential effects of the six major single nucleotide polymorphisms (SNPs) found in the MBL2 gene on COVID-19 severity and outcome.

Results: We show that the lectin pathway is activated in acute COVID-19, indicated by the correlation between complement activation product levels of the MASP-1/C1-INH complex (p=0.0011) and C4d (p<0.0001) and COVID-19 severity. Despite this, genetic variations in MBL2 are not associated with susceptibility to SARS-CoV-2 infection or disease outcomes such as mortality and the development of Long COVID.

Conclusion: In conclusion, activation of the MBL-LP only plays a minor role in COVID-19 pathogenesis, since no clinically meaningful, consistent associations with disease outcomes were noted.

Keywords: COVID-19; MBL2 genotypes; lectin pathway activation; lectin pathway of complement; mannose binding lectin (MBL); severe acute respiratory coronavirus 2 (SARS-CoV-2).

Figure 1

Graphical summary of the study. Figure created with BioRender.com.

Figure 2

Distribution of cases and controls of the BUD and CAM cohorts for biomarker and genetic analysis. A total of 467 individuals were enrolled in Budapest (BUD), while 262 individuals were enrolled in Cambridge, UK (CAM). Individuals were stratified according to five severity groups for biomarker analysis (HC, MILD, HOSP, HOSP+O₂, ICU), while genetic analysis were performed in only three severity groups (HC, MOD, SEV), merging MILD and HOSP into the moderate (MOD), and HOSP+O₂ and ICU into the severe (SEV) group. BUD, Budapest/Hungarian cohort; CAM, Cambridge cohort; N, number; HC, Healthy controls; MILD, patients not requiring hospitalization; HOSP, Hospitalized patients not requiring ventilation; HOSP+O₂, Hospitalized patients requiring ventilation; ICU, Intensive care unit patients; MOD, moderate patients group; SEV, severe patients group.

Figure 3

Levels of lectin pathway activation products in healthy controls and COVID-19 patients. (A) MASP-1/C1-INH complex and (B) C4d levels were measured in EDTA plasma samples and stratified according to severity in healthy controls and the merged COVID-19 cohort. Differences between severity groups were determined using Kruskal-Wallis test with Dunn’s multiple comparison post-hoc test. Non-significant differences are not marked, and asterisks indicate significant results (* p<0.05, ** p<0.01, *** p<0.001). HC, Healthy controls; MILD, patients not requiring hospitalization; CONV, patients not requiring hospitalization and sampled during convalescence; HOSP, Hospitalized patients not requiring ventilation; HOSP+O₂, Hospitalized patients requiring ventilation; ICU, Intensive care unit patients.

Figure 4

Frequencies of long MBL2 haplotype combinations in cases and controls. (A) Stratification of the merged COVID-19 cohorts (CAM and BUD) according to cases and controls (controls: green, cases: grey). (B) Cases stratified according to disease severity (MODERATE: blue, SEVERE: red). n, number.

Figure 5

MBL levels (CAM) and Lectin pathway activity (BUD) in healthy controls and COVID-19 patients, stratified according to disease severity. (A) MBL levels measured in the Cambridge COVID-19 cohort and healthy individuals and (B) Lectin pathway activity measured in the Hungarian COVID-19 cohort and healthy controls stratified according to disease severity. Differences between severity groups were determined using Kruskal-Wallis test with Dunn’s multiple comparison post-hoc test. Non-significant differences are not marked, and asterisks indicate significant results (* p<0.05, ** p<0.01). HC, Healthy controls; MILD, patients not requiring hospitalization; CONV, patients not requiring hospitalization and sampled during convalescence; HOSP, Hospitalized patients not requiring ventilation; HOSP+O₂, Hospitalized patients requiring ventilation; ICU, Intensive care unit patients; LP, lectin pathway.

Figure 6

MBL levels (CAM) and Lectin pathway activity (BUD) in healthy controls and COVID-19 patients, stratified according to MBL2 genotype groups and disease severity. Short MBL2 haplotype combinations were merged into four different groups: 1) MBL high (YA/YA), 2) MBL intermediate (YA/XA+XA/XA), 3) MBL low (YA/0 (including YA/YB, YA/YC and YA/YD)), and 4) MBL deficient (XA/0+0/0 (including 0/0 for allele A, XA/YB, XA/YC and XA/YD)). After genetic grouping, groups were furthermore stratified according to disease severity, and MBL (A) as well as Lectin pathway activity levels (B) were compared within groups using ANOVA with Dunn’s multiple comparison post-hoc test. Asterisks indicate significant results between severity groups (* p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001). HC, Healthy controls; HOSP, Hospitalized patients not requiring ventilation; HOSP+O₂, Hospitalized patients requiring ventilation; ICU, Intensive care unit patients; MOD, moderate patients group; SEV, severe patients group; LP, lectin pathway; ns, non-significant.

Figure 7

MASP-1/C1-INH complex and C4d levels in healthy controls and COVID-19 patients, stratified according to MBL2 genotype groups and disease severity. Short MBL2 haplotype combinations were merged into four different groups: 1) MBL high (YA/YA), 2) MBL intermediate (YA/XA+XA/XA), 3) MBL low (YA/0 (including YA/YB, YA/YC and YA/YD)), and 4) MBL deficient (XA/0+0/0 (including 0/0 for allele A, XA/YB, XA/YC and XA/YD)). After genetic grouping, groups were furthermore stratified according to disease severity, and MASP-1/C1-INH complex (A) as well as C4d levels (B) were compared within groups using ANOVA with Dunn’s multiple comparison post-hoc test. Asterisks indicate significant results between severity groups (* p<0.05, ** p<0.01). HC, Healthy controls; HOSP, Hospitalized patients not requiring ventilation; HOSP+O₂, Hospitalized patients requiring ventilation; ICU, Intensive care unit patients; MOD, moderate patients group; SEV, severe patients group; ns, non-significant.

Figure 8

Binding of MBL to SARS-CoV-2 spike protein. Recombinant SARS-CoV-2 spike protein was immobilised on a 96-well plate in different concentrations (0-5 pmol/well). Recombinant MBL (rMBL; 1000 ng/mL), BssSA (5% in TBST-Ca²⁺) or serum from individuals with known MBL2 genotypes (10% in TBST-Ca²⁺) were incubated on the plates with captured spike protein and bound MBL was detected. Genotypes YA/YA (n=5), YA/XA (n=3), XA/XA (n=3), YA/0 (n=4), XA/0 (n=2) and 0/0 (n=3) were tested, while average MBL levels of the individuals in each group are included in the legend (x̄). Results are presented in mean ± SD. The same results were observed in two independent experiments. OD, optical density; S-protein, SARS-CoV-2 spike protein; rMBL, recombinant MBL; x̄, average concentration; BSA, bovine serum albumin; TBST-Ca, Tris-HCl buffer containing calcium.

Figure 9

Relationship between complement markers, MBL genotype and mortality in COVID-19. Levels of complement markers MBL (A), MASP-1/C1-INH complex (B) and TCC/C5b-9 (C) in the Cambridge cohort and Lectin pathway activity (F), MASP-1/C1-INH complex (G) and TCC/C5b-9 (H) of the Budapest cohort were stratified according to survival (alive vs. deceased). Differences between the two groups were analysed using the Mann-Whitney test, while asterisks indicate significant differences (* p<0.05, ** p<0.01, *** p<0.001). Non-significant differences are indicated (ns). Forest plot displaying Odds ratios (ORs) with 95% confidence intervals (CIs) for COVID-19 related death in individuals, stratified according to the MBL2 A allele carrier state (D) Cambridge, (E) Budapest) as well as according to the MBL2 genotype groups (I) Cambridge, (J) Budapest). Odds ratios of 0 or infinite are not indicated on the forest plots. p-values are presented as non-corrected for multiple testing. Threshold for significance taking into account multiple testing are p=0.0050 for the MBL2 A allele, and p=0.0083 for the short MBL2 haplotype (Benjamini-Hochberg correction). MASP-1/C1-INH, MASP-1/C1-INH complex; LP, lectin pathway activity.