Complement Activation in Kidneys of Patients With COVID-19

Abstract

Most patients who became critically ill following infection with COVID-19 develop severe acute respiratory syndrome (SARS) attributed to a maladaptive or inadequate immune response. The complement system is an important component of the innate immune system that is involved in the opsonization of viruses but also in triggering further immune cell responses. Complement activation was seen in plasma adsorber material that clogged during the treatment of critically ill patients with COVID-19. Apart from the lung, the kidney is the second most common organ affected by COVID-19. Using immunohistochemistry for complement factors C1q, MASP-2, C3c, C3d, C4d, and C5b-9 we investigated the involvement of the complement system in six kidney biopsies with acute kidney failure in different clinical settings and three kidneys from autopsy material of patients with COVID-19. Renal tissue was analyzed for signs of renal injury by detection of thrombus formation using CD61, endothelial cell rarefaction using the marker E-26 transformation specific-related gene (ERG-) and proliferation using proliferating cell nuclear antigen (PCNA)-staining. SARS-CoV-2 was detected by in situ hybridization and immunohistochemistry. Biopsies from patients with hemolytic uremic syndrome (HUS, n = 5), severe acute tubular injury (ATI, n = 7), zero biopsies with disseminated intravascular coagulation (DIC, n = 7) and 1 year protocol biopsies from renal transplants (Ctrl, n = 7) served as controls. In the material clogging plasma adsorbers used for extracorporeal therapy of patients with COVID-19 C3 was the dominant protein but collectin 11 and MASP-2 were also identified. SARS-CoV-2 was sporadically present in varying numbers in some biopsies from patients with COVID-19. The highest frequency of CD61-positive platelets was found in peritubular capillaries and arteries of COVID-19 infected renal specimens as compared to all controls. Apart from COVID-19 specimens, MASP-2 was detected in glomeruli with DIC and ATI. In contrast, the classical pathway (i.e. C1q) was hardly seen in COVID-19 biopsies. Both C3 cleavage products C3c and C3d were strongly detected in renal arteries but also occurs in glomerular capillaries of COVID-19 biopsies, while tubular C3d was stronger than C3c in biopsies from COVID-19 patients. The membrane attack complex C5b-9, demonstrating terminal pathway activation, was predominantly deposited in COVID-19 biopsies in peritubular capillaries, renal arterioles, and tubular basement membrane with similar or even higher frequency compared to controls. In conclusion, various complement pathways were activated in COVID-19 kidneys, the lectin pathway mainly in peritubular capillaries and in part the classical pathway in renal arteries whereas the alternative pathway seem to be crucial for tubular complement activation. Therefore, activation of the complement system might be involved in the worsening of renal injury. Complement inhibition might thus be a promising treatment option to prevent deregulated activation and subsequent collateral tissue injury.

Keywords: COVID-19; complement—immunological term; endothelial injury; kidney; lectin pathway activation.

Figure 1

Involvement of complement components in COVID-19. Workflow for the analysis of clotted material isolated from plasma adsorber material collected from treated COVID-19 patient (A). Coverage rates of complement factors C3, collectin 11, and mannose-binding protein-associated serine protease 2 (MASP-2) analyzed by MALDI-TOF (B).

Figure 2

Renal injury in kidney biopsies from patients infected with SARS-CoV-2. Representative pictures of a PAS-stained COVID-19 renal biopsy showing acute tubular injury with swollen and vacuolized cells (A) and glomerular microthrombi in HE-stained section (B, arrows). In situ hybridization for SARS-CoV-2 mRNA showed few positive signals in tubular and endothelial localization (C, pink signals marked by arrows). Immunohistochemistry for SARS-CoV-3 spike protein detected sporadic positive signals in renal biopsies from patients with COVID-19 (D, brown signals marked by arrows). As positive control, we employed HEK293T cells transfected with a plasmid expressing SARS-CoV-2 spike protein, fixed with formalin, embedded in paraffin followed by staining of sections by immunohistochemistry using anti-SARS-CoV-2 spike antibody (E, brown staining).

Figure 3

Endothelial rarefaction and proliferative activity in kidney biopsies from patients infected with SARS-CoV-2. Renal proliferation, as assessed by PCNA-staining (A) and ERG-positive endothelial cells (B) were evaluated by immunofluorescence staining analyzed by confocal microscopy and quantification by ZEN-software in COVID-19 biopsies (COV, n = 7) compared to renal control biopsies taken 1 year after transplantation (Ctrl, n=7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5) and disseminated intravascular coagulation (DIC, n = 7). Representative pictures of immunofluorescence double staining showing PCNA (green staining) and ERG (red staining) double staining were shown in biopsies of Ctrl (C), ATI (D), HUS (E), DIC (F), and COV (G). *p < 0.05; ***p < 0.001.

Figure 4

Frequency and localization of CD61-positive platelets in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS, and DIC. Frequency and amount of CD61-positive platelets was analyzed in glomerular capillaries (A), peritubular capillaries (C) and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7), and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of CD61-positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatching in the bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for CD61 were shown for glomerular (B, brown staining), peritubular (D, brown staining), and arterial localization (F, brown staining).

Figure 5

Frequency and vascular localization of MASP-2 deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS, and DIC. Frequency and amount of mannose-binding protein-associated serine protease 2 (MASP-2) deposition was analyzed in glomerular capillaries (A), peritubular capillaries (C) and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7), and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of MASP-2–positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for MASP-2 were shown for glomerular (B, brown staining), peritubular (D, brown staining) and arterial localization (F, brown staining).

Figure 6

Frequency and vascular localization of C1q deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS, and DIC. Frequency and amount of C1q deposition were analyzed in glomerular capillaries (A), peritubular capillaries (C), and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7) and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of C1q-positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for C1q were shown for glomerular (B, brown staining), peritubular (D, brown staining) and arterial localization (F, brown staining).

Figure 7

Frequency and vascular localization of C3c deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS and DIC. Frequency and amount of C3c deposition were analyzed in glomerular capillaries (A), peritubular capillaries (C), and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7) and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of C3d-positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for C3c were shown for glomerular (B, brown staining), peritubular (D, brown staining) and arterial localization (F, brown staining).

Figure 8

Frequency and vascular localization of C3d deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS, and DIC. Frequency and amount of C3d deposition were analyzed in glomerular capillaries (A), peritubular capillaries (C) and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7) and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of C3d-positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for C3d were shown for glomerular (B, brown staining), peritubular (D, brown staining) and arterial localization (F, brown staining).

Figure 9

Frequency and vascular localization of C5b-9 deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS, and DIC. Frequency and amount of C5b-9 deposition were analyzed in glomerular capillaries (A), peritubular capillaries (C) and renal arteries (E) in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7) and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of C5b-9–positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for C5b-9 were shown for glomerular (B, brown staining), peritubular (D, brown staining) and arterial localization (F, brown staining).

Figure 10

Frequency and tubular localization of C3c, C3d, and C5b-9 deposition in renal biopsies with COVID-19 compared to Ctrl, ATI, HUS and DIC. Frequency and amount of tubular C3c (A), C3d (C), and C5b-9 (E) deposition were analyzed in renal control biopsies taken 1 year after transplantation (Ctrl, n = 7), biopsies with acute tubular injury (ATI, n = 7), hemolytic uremic syndrome (HUS, n = 5), disseminated intravascular coagulation (DIC, n = 7), and COVID-19 (COVID-19, n = 9) using immunohistochemistry and semi-quantitative scoring. The proportion of complement-positive stained cases that belongs to cases with comorbidities with known involvement of complement activation is marked by hatched bars representing the COVID-19 cohort. Representative pictures of COVID-19 biopsies positive for tubular C3c (B, brown staining), C3d (D, brown staining), and C5b-9 (F, brown staining) were shown.