ADAMTS7-Mediated Complement Factor H Degradation Potentiates Complement Activation to Contributing to Renal Injuries

Abstract

Significance Statement

Complement factor H (CFH) dysfunction by an incomplete underlying mechanism causes various complement-mediated renal injuries. We identified metalloprotease ADAMTS7 as a novel binding protein of CFH that further degrades CFH and potentiates complement activation. ADAMTS7 deficiency alleviated CFH degradation and renal pathologies in lupus nephritis and renal ischemia-reperfusion injury in mice, but without affecting complement-dependent bactericidal activity. The investigation revealed a novel mechanism to explain CFH dysfunction in complement-mediated renal injuries. ADAMTS7 would be a promising target for anticomplement therapies that would potentially avoid increased risk of infection, which is the drawback of current strategies.

Background

The dysfunction of complement factor H (CFH), the main soluble complement negative regulator, potentiates various complement-induced renal injuries. However, insights into the underlying mechanism of CFH dysfunction remain limited. In this study, we investigated whether extracellular protease-mediated degradation accounts for CFH dysfunction in complement-mediated renal injuries.

Methods

An unbiased interactome of lupus mice kidneys identified CFH-binding protease. In vitro cleavage assay clarified CFH degradation. Pristane-induced SLE or renal ischemia-reperfusion (I/R) injury models were used in wild-type and ADAMTS7^−/− mice.

Results

We identified the metalloprotease ADAMTS7 as a CFH-binding protein in lupus kidneys. Moreover, the upregulation of ADAMTS7 correlated with CFH reduction in both lupus mice and patients. Mechanistically, ADAMTS7 is directly bound to CFH complement control protein (CCP) 1–4 domain and degraded CCP 1–7 domain through multiple cleavages. In mice with lupus nephritis or renal I/R injury, ADAMTS7 deficiency alleviated complement activation and related renal pathologies, but without affecting complement-mediated bactericidal activity. Adeno-associated virus–mediated CFH silencing compromised these protective effects of ADAMTS7 knockout against complement-mediated renal injuries in vivo.

Conclusion

ADAMTS7-mediated CFH degradation potentiates complement activation and related renal injuries. ADAMTS7 would be a promising anticomplement therapeutic target that does not increase bacterial infection risk.

Introduction

Complement factor H (CFH) is the main plasma regulatory protein in the complement system.¹ CFH dysfunction potentiates complement activation, which is associated with various complement-mediated renal injuries.^2,3 To date, a series of studies have revealed that genetic mutations,^4–6 autoantibodies,^7,8 conformational alterations,⁹ and CFH-interacting proteins^10,11 may cause CFH dysfunction. Of interest, the downregulation of CFH protein expression has also been observed in the circulation of patients with complement-mediated renal injuries (e.g., lupus nephritis [LN]^12,13 and renal ischemia-reperfusion [I/R] injury¹⁴), even in the absence of genetic mutations. Deficiency of CFH aggravated complement activation and renal I/R injury or LN progression in animal models.^14,15 However, insights into the underlying mechanism of CFH reduction during complement-mediated diseases remain limited. Of note, CFH, and other complement factors and regulatory proteins, are mainly localized proximal to the extracellular matrix (ECM) microenvironment.^1,16 Metalloprotease-mediated ECM remodeling (e.g., MMP-9 and collagen degradation) modulates the development of renal pathologies.^17,18 In this study, we aimed to investigate whether extracellular protease-mediated degradation accounts for CFH reduction and dysfunction in complement-mediated renal injuries.

Methods

Materials

An antibody against human CFH (PAA635Hu01) used for immunoprecipitation assays and measuring ADAMTS7-mediated cleavage was purchased from Cloud-Clone Corp. (Wuhan, China). An antibody against CFH from human or mice (DF6889) used for western blot analysis, immunohistochemistry, and solid-phase binding assays was purchased from Affinity Biosciences (Jiangsu, China). An antibody against ADAMTS7 (WG-04133) was purchased from ABclonal (Wuhan, China). Antibodies against His-Tag (66005-1-Ig), Flag-Tag (66008-2-Ig), β-tubulin (10094-1-AP), and GAPDH (10494-1-AP) were purchased from Proteintech Group, Inc. (Wuhan, China). Anti-FLAG-M2 antibody (F1804) was purchased from Sigma-Aldrich (St. Louis, MO). PE-conjugated anti-His-tag antibody was purchased from BioLegend, Inc. (San Diego, CA). Normal mouse IgG (sc-2025) and Protein A/G plus-agarose (sc-2003) were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Normal rabbit IgG (2729S) was obtained from Cell Signaling Technology (Boston, MA). DAPI (4′,6-diamidino-2-phenylindole), Alexa Fluor 633-conjugated goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody, and Alexa Fluor 555-conjugated goat anti-mouse IgG (H+L) cross-adsorbed secondary antibody were purchased from Thermo Fisher Scientific (Rochester, NY). HRP affinipure goat anti-mouse lgG (H+L; E030110-01) and HRP affinipure goat anti-rabbit lgG (H+L; E030120-01) were purchased from EarthOx Life Sciences (San Francisco, CA). Human C3b protein (A114), human CFH protein (A136), and human complement factor I (CFI) protein (A138) were purchased from CompTech (Tyler, TX). Human ADAMTS7 ELISA kits (abx251352) and mouse ADAMTS7 ELISA kits (abx518999) were purchased from Abbexac (Cambridge, United Kingdom). Human CFH ELISA kits (CSB-E08931h), mouse CFH ELISA kits (CSB-E08933m), mouse C3 ELISA kits (CSB-E08667m), and mouse C5a ELISA kits (CSB-E08514m) were purchased from Cusabio Technology LLC. (Wuhan, China). Pristane (MB0318) was purchased from Meilun Biotechnology Co. (Suzhou, China). Zymosan (Z867766) was purchased from Macklin (Shanghai, China). Rabbit erythrocytes (S28338) were purchased from Yuanye Bio-Technology Co. (Shanghai, China). Periodic acid-Schiff stain (G1280) was purchased from Solarbio Inc. (Beijing, China).

Animals

Mice were bred/kept at Peking University Health Science Center Animal Facilities under specific pathogen-free conditions. The animals were housed in an air-conditioned environment, with a 12-hour light-dark cycle and free access to food and water. All animal studies adhered to the guidelines of the Animal Care and Use Committee of Peking University Health Science Center. ADAMTS7^−/− mice were generated on the C57BL/6 strain, and homozygous knockout and littermate wild-type mice were obtained by interbreeding the heterozygotes.¹⁹ MRL/lpr mice homozygous for spontaneous mutation in the Fas gene on the C57BL/6 background were purchased from Shanghai Laboratory Animal Center, Chinese Academy of Sciences (Shanghai, China). Isoflurane (1.5%–2%) was used to anesthetize the mice before surgery or blood collection, and anesthesia was carefully monitored to avoid pain or discomfort.

Evaluation of Renal Function

Mouse renal function was evaluated by the indexes including serum creatinine, BUN, and proteinuria. Serum samples were collected for measurement of creatinine and BUN using commercial kits (C011-2 for creatinine; C013-2-1 for BUN, Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Urine samples were collected from mice housed in metabolic cages. Urinary albumin concentrations were measured with a mouse albumin ELISA kit (E-90AL, NeoBioscience, Shenzhen, China) and normalized to creatinine concentrations in the same urine (C011-2-1, Nanjing Jiancheng Bioengineering Institute, Nanjing, China).

The Interactomic Analysis of CFH-Binding Proteins

The kidney tissue of MRL/lpr mice (24 weeks old) was extracted by NETN buffer (20 mM Tris-HCl [pH 8.0], 150 mM NaCl, 2 mM EDTA, and 0.5% NP-40) with protease inhibitor, followed by incubation with recombinant CFH protein (30 µg/ml) purchased from CompTech for 2 hours at 4°C. The tissue lysate was equally divided for incubation with anti-CFH antibody and anti-rabbit IgG antibody overnight at 4°C. Then, protein A/G agarose was added to pull down proteins for 2 hours at 4°C. Next, protein A/G agarose was washed three times and vortexed with IP elution buffer (5 mM glycine/HCl, pH 2.0) for 10 min. Finally, 1 M ammonium hydrogen carbonate was added to the supernatant until the pH was 8.0. The eluted proteins were further identified by HPLC-mass spectrometry (Core Facilities of Peking University Health Science Center). For further validation of the CFH-binding proteins, the eluted samples were subjected to western blotting.

Human Studies

The study was conducted in accordance with the ethical guidelines of the 1975 Declaration of Helsinki, and the study protocol was approved by the Scientific and Ethics Committee of the Peking University First Hospital. All experiments were performed in accordance with relevant guidelines and regulations. The Chinese participants including healthy volunteers (n=29) and patients with LN (n=29) were recruited at Peking University First Hospital. Informed consent was obtained from all study participants. The demographic characteristics of the subjects are listed in Supplemental Table S1. Peripheral blood samples were centrifuged at 1000g for 5 minutes after standing at RT for at least 30 minutes to obtain the serum. Serum ADAMTS7 and CFH were measured by ELISA kits.

Coimmunoprecipitation Analysis

Full-length human CFH and various fragments (complement control protein [CCP] 1–4 [amino acids 19–264], CCP 5–9 [amino acids 265–507], CCP 10–18 [amino acids 508–1104], and CCP 19–20 [amino acids 1105–1230]) fused with a His-tag were subcloned into pcDNA3.1 plasmids using a ClonExpress II One-Step Cloning Kit (C112-02, Vazyme, Nanjing, China). Full-length rat ADAMTS7 and various fragments (prodomain [amino acids 26–246], metalloproteinase plus disintegrin-like and cysteine-rich domains [amino acids 238–711], spacer-1 plus 3 TSP repeats [amino acids 703–1007] and spacer-2 plus 4 C-terminal TSP repeats of ADAMTS-7 [amino acids 999–1595]) fused with a C-terminal Flag-tag were subcloned into pcDNA3.1 plasmids using the One-Step Cloning Kit. HEK293T cells were cotransfected with JetPEI DNA transfection reagent (Polyplus-transfection, Strasbourg, France), with plasmids encoding the various human CFH fragments or the distinct rat ADAMTS7 domains for 48 hours. Cell lysates were incubated with anti-His or anti-Flag antibodies at 4°C overnight, followed by precipitation with protein A/G agarose beads. After being washed and eluted with 2× SDS loading buffer, the immunoprecipitates were resolved by SDS-PAGE and immunoblotted with anti-Flag or anti-His antibodies. Species-matched IgGs served as a negative control.

Solid-Phase Binding Assay

The solid-phase binding assays were conducted as previously described.^20,21 Briefly, 96-well microtiter plates were coated with 100 µl of recombinant CFH (4 µg/ml), CCP 1–4 fragment (4 µg/ml), or BSA (4 µg/ml, as a negative control) in 0.1 M carbonate buffer (pH 9.6) overnight at 4°C. After three washes with 0.1% Tween 20/PBS (PBST), the wells were blocked with 2% BSA/PBST for 30 minutes at RT. They were washed twice and then incubated with 100 µl of various concentrations of recombinant Flag-tagged ADAMTS7 full-length or F4 domain protein (0–20 µg/ml) in 1% BSA/PBS at 37°C for 1 hour. Equal amounts of BSA were applied as negative control. After five washes, the plates were incubated for 1 hour with anti-Flag M2 monoclonal antibody (Sigma-Aldrich, 1:1000). After an additional five washes, the cells were then incubated for 1 hour with HRP-conjugated goat-anti-mouse IgG antibody (1:5000). The plates were further incubated with TMB/E solution (EMD Millipore, Billerica, MA), and 0.25 M HCl was subsequently added to stop the reaction. The HRP activity was determined by measuring the absorbance at 450 nm using a Variskan microplate reader (Thermo). For determination of the effect of ADAMTS7 on the interaction between CFH and C3b, 96-well microtiter plates were coated with 5 µg/ml C3b overnight at 4°C. Recombinant CFH (2 µg/ml) was preincubated with increasing amounts of ADMTS7 or BSA (0–20 µg/ml) for 1 hour at 37°C and then the mixture was added to C3b-coated plates for another 1-hour incubation at 37°C. After three washes, the CFH binding to C3b was measured by ELISAs as previously described using anti-CFH antibody and HRP-conjugated anti-rabbit IgG secondary antibody sequentially.

Mammalian Two-Hybrid Assay

The full-length sequence and fragments flanking the functional domains of human CFH (CCP 1–4 [amino acids 19–264], CCP 5–9 [amino acids 265–507], CCP 10–18 [amino acids 508–1104], and CCP 19–20 [amino acids 1105–1230]) were amplified by PCR and subcloned into the pBIND vector using the pEASY-Uni Seamless Cloning and Assembly Kit (CU101-01, TransGen Biotech., Beijing, China). The full-length sequence and fragments flanking the functional domains of rat ADAMTS7 (prodomain [amino acids 26–246], the metalloproteinase plus disintegrin-like and cysteine-rich domain [amino acids 238–711], the spacer-1 plus 3 TSP repeats [amino acids 703–1007], and the spacer-2 plus 4 C-terminal TSP repeats of ADAMTS7 [amino acids 999–1595]) were amplified by PCR and subcloned into the pACT vector.¹⁹ HEK293T cells were cotransfected with the JetPEI DNA transfection reagent with the target and bait constructs together with the luciferase reporter plasmid pG5-luc at a ratio of 1:1:1. Forty-eight hours after transfection, the cells were harvested, and cell lysates were used for luciferase activity assays with a Dual-Luciferase Reporter Assay System (Promega, Madison, WI).

ADAMTS7-Mediated Cleavage Assay and Identification of Cleavage Site

ADAMTS7-mediated cleavage substrate assay was performed as previously described.²² CFH (2 µg) was incubated with increasing amounts of ADAMTS7 (0–0.4 µg) at 37°C for 2 hours in 20 µl of digestion buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM CaCl₂, 2 mM ZnCl₂, and 0.05% Brij-35, pH 7.5). To further validate the potential cleavage sites on CFH by ADAMTS7, two strategies were used, respectively. First, the proteins in cleavage assays were denatured with 5× SDS sample buffer at 95°C for 5 minutes and were subsequently resolved on 8% gradient SDS-PAGE gels, followed by direct Coomassie brilliant blue staining or western blotting using an anti-CFH antibody (PAA635Hu01, Cloud-Clone Corp., Wuhan, China). The potential cleavage bands observed with Coomassie brilliant blue staining were excised from the SDS-PAGE gels. The extracted cleavage fragments were labeled with a dimethyl moiety at the N-terminus before trypsin digestion. Second, the samples of ADAMTS7-cleaved CFH were directly performed with the dimethyl moiety labeling followed by trypsin digestion and mass spectrometry. Through mass spectrometric analysis (Peking University Health Science Center Core Facilities), the initial amino acids of the peptides with N-terminal dimethyl moieties were considered potential cleavage sites. In addition, ³⁹⁷GYNQNYGR peptides as N-terminus of ADAMTS7-cleaved CFH fragments were directly evaluated in human serum samples by parallel reaction monitoring-based targeted mass spectrometry (Peking University Health Science Center Core Facilities).

C3b Deposition on Zymosan and Lipopolysaccharide

C3b deposition on zymosan and LPS was evaluated as previously described.²³ Zymosan (100 µg/ml) or LPS (40 µg/ml) was coated on 96-well microtiter plates in PBS overnight at RT. After washes with PBS containing 0.1% (wt/vol) Tween-20 (PBST), the sera from C57BL/6 mice (10 µl per sample), first diluted in 100 µl of Veronal buffer (3 mM barbital, 1.8 mM sodium barbital, and 145 mM NaCl, pH 7.4) containing 0.05% (wt/vol) gelatin, 5 mM MgCl₂ and 10 mM EGTA, was incubated with different amounts of BSA or ADAMTS7 (0–20 µg/ml) in coated 96-well microtiter plates. C3b deposition was detected by ELISAs as previously described using anti-C3 antibody and HRP-conjugated goat-anti-rabbit IgG antibody.

Measurement of CFH Cofactor Activity

A fluid phase assay to measure CFH activity as a cofactor of CFI was to explore the cleavage of C3b after the previous study.²⁴ C3b (1 µg), CFI (500 ng/ml), and CFH (10 µg/ml) were incubated in 20 µl of reaction buffer containing 25 mM HEPES (pH 7.4) and 150 mM NaCl at 37°C for 1 hour. To evaluate the effect of ADAMTS7, CFH was incubated with ADAMTS7 (10 µg/ml) at 37°C for 2 hours before the addition of C3b and CFI. Degradation of C3b was visualized by SDS-PAGE under reducing conditions and staining with Coomassie brilliant blue. C3b was cleaved by CFI with assistance of CFH into iC3b fragments with molecular weights at 68-kDa and 43-kDa.

E_R Hemolytic Assay

The hemolytic assays were modified after the previous study.²³ Sera from mice or healthy donors were diluted at a 1:10 ratio with AP-CFTD buffer (2.5 mM barbital, 1.5 mM sodium barbital, 144 mM NaCl, 7 mM MgCl₂, 10 mM EGTA, pH 7.4). The diluted sera were preincubated with BSA (20 µg/ml) or ADAMTS7 (20 µg/ml), followed by incubation with an equal volume of rabbit erythrocytes (E_R, 1×10⁸ cells/ml in AP-CFTD buffer) at 37°C for 60 minutes while shaking. Lysis was stopped with 100 µl of ice-cold VBS-EDTA buffer (2.5 mM barbital, 1.5 mM sodium barbital, 144 mM NaCl, 2 mM EDTA, pH 7.4), followed by centrifugation. The absorbance of the supernatants at OD542 was determined and is expressed as percentage of the 100% lysis control, while rabbit erythrocytes were lyzed by ddH₂O as a 100% lysis control. The sera were inactivated at 56°C for 30 minutes as a negative control.

Mouse Model of Pristane-Induced Lupus and Evaluation of Renal Injuries

Eight- to ten-week-old ADAMTS7^−/− female mice and their female WT littermates were treated with a single intraperitoneal (i.p.) injection of 0.5 ml pristane. The mice were killed at 1, 3, or 6 months after pristane injection, to further evaluate ADAMTS7 expression levels in serum and kidneys through a commercially available ELISA kit (abx518999, Abbexac, Cambridge, United Kingdom), real-time PCR or western blotting, respectively. Moreover, after 6 months of pristane induction, urine albumin to creatinine and serum creatinine and BUN were measured to evaluate renal function. Moreover, kidneys were fixed in 4% paraformaldehyde for at least 24 hours, embedded in paraffin, and cut into 3-mm sections. Sections were stained with Periodic acid–Schiff (PAS). The glomerular and tubular-interstitial pathologies were evaluated by two independent investigators in a blinded manner as previously described.^25,26 Briefly, glomerular pathology was assessed by examining 20 glomerular cross-sections (gcs) per kidney and scoring each glomerulus on a semiquantitative scale: 0=normal (35–40 cells/gcs); 1=mild [glomeruli with few lesions showing slight proliferative changes, mild hypercellularity (41–50 cells/gcs), and/or minor exudation]; 2=moderate [glomeruli with moderate hypercellularity (51–60 cells/gcs), including segmental and/or diffuse proliferative changes, hyalinosis, and moderate exudates]; and 3=severe (glomeruli with segmental or global sclerosis and/or severe hypercellularity [>60 cells/gcs], necrosis, crescent formation, and heavy exudation). Tubular-interstitial pathology was determined in 10 randomly selected high-power fields (>400) of infiltrates and damaged tubules on a scale of 0–3: 0=normal, 1=mild, 2=moderate, and 3=maximum.

Adeno-Associated Virus–Mediated Gene Knockdown In Vivo

For generation of in vivo knockdown of CFH, CFH-specific small short hairpin RNA (shRNA) was cloned and packaged into an adeno-associated virus 9 (AAV9) vector (Hanbio, Shanghai, China), respectively. AAV9-shRNA-NC/CFH was packaged with pAAV-RC and pHelper using the triple-plasmid transient transfection method (HB infusion Kit; Hanbio Bio.). Viral particles were applied in SLE and I/R injury mouse models after purification. The shRNA-NC sequence was as follows: 5′-GATCCCCTAAGGTTAAGTCGCCCTCGCTCGAGCGAGGGCGACTTAACCTTAGGTTT TTA-3’. The shRNA-CFH sequence was as follows: 5′-GATCCGCAGAAACAGACCAGGAATACCTCGAGGTATTCCTGGTCTGTTTCTGCTTTTTA-3′. Eight- to ten-week-old mice were randomly divided and administered 150 µl of AAV9-shRNA-NC/CFH (2.5×10¹¹ virus particles) through tail vein injection. At 2 weeks after injection, the transduced mice suffered from 6 months of pristane induction or renal I/R injury. Then, the efficiency of CFH or ADAMTS7 knockdown was evaluated by western blot analysis.

Renal I/R Injury Model and Evaluation of Renal Injuries

I/R injury surgery was performed as previously described.²⁷ After administration of isoflurane anesthesia, male ADAMTS7^−/− mice and WT littermates (8–12 weeks old) were subjected to laparotomy through a small flank incision to exposure both kidneys; thereafter, both renal pedicles were dissected and a vascular clamp was applied for 30 min (bilateral renal I/R injury). After the ischemic period, clamps were released to induce blood reperfusion in kidneys, and 1 ml of 37°C saline was administered by intraperitoneal injection after surgery for volume repletion. In sham operations, both kidneys were exposed without induction of ischemia. These animals were euthanized on day 1 after renal I/R injury, and kidneys were harvested for further examination. Serum creatinine and BUN were measured to evaluate renal function. Kidneys were fixed in 4% paraformaldehyde for at least 24 hours, embedded in paraffin, and cut into 3-mm sections. The severity of morphological renal damage was assessed by using an arbitrary score on the basis of PAS-stained kidney sections, after a modification of a previous protocol.²⁸ In brief, the extent of four typical I/R injury–associated damage markers (e.g., dilatation, denudation, intraluminal casts, loss of brush border membrane, and cell flattening) was expressed in arbitrary units in a range of 0–5, according to the percentage of damaged tubules: 0, none; 1, 0% to 10%; 2, 11% to 25%; 3, 26% to 45%; 4, 46% to 75%; and 5, 76% to 100%. Each section was evaluated by two independent investigators in a blinded manner, and the data are presented as the mean of five randomly chosen nonoverlapping fields.

Statistical Analyses

All results are presented as the means±SEM. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA). For statistical comparisons, we first evaluated whether the data were normally distributed using the Shapiro-Wilk normality test. Then, we applied Student t test, with Welch's correction if equal standard deviations were not assumed through an F-test, for two-group comparisons of normally distributed data. In addition, the Brown-Forsythe test was used to assess equal variances among data from more than two groups; we applied ordinary ANOVA or Welch ANOVA when equal variances were assumed or not assumed, respectively.

Nonparametric tests were used when the data were not normally distributed. In all cases, a statistically significant difference was present when the two-tailed probability was < 0.05. The details of the statistical analysis applied to each experiment are presented in the corresponding figure legends.

Results

ADAMTS7 Is Associated with a Decline in CFH in Lupus Nephritis

Because CFH dysfunction-potentiated complement activation is involved in SLE or LN progression,²⁹ we first used mice with lupus to explore the potential proteases acting on CFH. As expected, 24-week-old SLE-prone MRL/lpr mice exhibited the significant renal dysfunction and injuries (Supplemental Table S2 and Supplemental Figure S1A-D). Of note, CFH protein expression was markedly downregulated in renal tissues from MRL/lpr mice, compared with C57BL/6 control mice (Supplemental Figure S1E), whereas the mRNA level did not alter significantly (Supplemental Figure S1F), suggesting the downregulation of CFH mainly at protein level. Considering the direct interaction of proteases and substrates, we accordingly performed an interactomic analysis by using purified CFH pulldown assay and mass spectrometry of renal tissues from MRL/lpr mice to screen the potential degrades possibly resulting in CFH reduction (Figure 1A). Consequently, 66 proteins in whole kidney lysate were exclusively precipitated by anti-CFH antibody rather than the control IgG (Data set S1). Considering that CFH is endogenously located extracellularly or on the cell surface, GO analysis further enriched 17 extracellular or membrane CFH-binding proteins: five proteases, four matrix proteins, five cytokines or growth factors, and three cell surface receptors. Among the five extracellular proteases, the expression of ADAMTS7, a member of the disintegrin and metalloproteinase with thrombospondin motif (ADAMTS) family, was induced the most at the mRNA level in lupus kidneys compared with the controls (Supplemental Figure S1G).

Figure 1.

ADAMTS7 is associated with the decline of CFH in LN. (A) The workflow shows the experimental procedure for the interactomic analysis of CFH in kidneys from the lupus group of 24-week-old female MRL/lpr mice. (B) The Venn diagram (left) displays the numbers of proteins immunoprecipitated by anti-CFH antibody or rabbit IgG. The table (right) includes CFH-interacting extracellular or membrane proteins. (C) Representative western blot analysis and quantitation of ADAMTS7 protein in kidneys from 24-week-old female C57BL/6 (C57) and MRL/lpr mice. n=4, *P<0.05 by unpaired two-tailed Student t test. (D) Immunohistochemical staining of ADAMTS7 and CFH in kidneys from 24-week-old female C57BL/6 (C57) and MRL/lpr mice. Nuclei were counterstained with hematoxylin, and rabbit IgG was applied as a negative control. Scale bar=50 μm. (E) The relative quantification of ADAMTS7 and CFH immunohistochemical staining. n=3, *P<0.05 by unpaired two-tailed Student t test. (F and G) ELISA measurements of serum ADAMTS7 and CFH from healthy controls (n=29) and patients with LN (n=29). *P<0.05 by unpaired two-tailed Student t test. (H) Correlation analysis of serum ADAMTS7 and CFH. n=58, P=0.0006 by Pearson correlation analysis.

In addition, western blotting showed that ADMATS7 protein expression was markedly increased in kidneys from lupus mice (Figure 1C), whereas renal immunohistochemical staining showed increased ADAMTS7 accompanied by decreased CFH (Figure 1, D and E). Meanwhile, the serum ADAMTS7 was significantly elevated together with the decline of serum CFH in MRL/lpr mice compared with C57BL/6 control mice (Supplemental Figure S1H-I). These results indicated the potential relevance of ADAMTS7 to the reduction of CFH protein in mouse kidneys and circulation from the lupus group.

To further investigate the correlation of ADMATS7 and CFH in humans, we measured serum ADAMTS7 and CFH in patients with LN by using ELISAs. Serum ADAMTS7 was significantly elevated, whereas CFH was markedly decreased in serum from patients with LN compared with the healthy controls (Figure 1, F and G). In addition, serum ADAMTS7 levels were negatively associated with CFH, suggesting a possible role of ADAMTS7 in CFH reduction in patients with LN (Figure 1H).

ADAMTS7 Directly Interacts with CFH

Next, we accordingly validated the interaction of recombinant CFH protein with ADAMTS7 in kidney lysates from lupus mice (Figure 2A). Meanwhile, an endogenous interaction between ADAMTS7 and CFH in lupus kidneys was confirmed by the coimmunoprecipitation (Co-IP) assay (Supplemental Figure S2A-B). Moreover, an additional Co-IP assay identified a similar specific interaction between ADAMTS7 and CFH in HEK293 cells overexpressing Flag-ADAMTS7 and His-CFH (Supplemental Figure S2C-D). The solid-phase binding assay further indicated that purified ADAMTS7 directly interacted with CFH in a dose-dependent manner (Figure 2B).

Figure 2.

ADAMTS7 directly interacts with CFH. (A) The pull-down assay of renal lysates (100 µg) from 24-week-old female MRL/lpr mice using purified CFH (30 µg/ml). (B) The solid-phase binding assay of ADAMTS7 and CFH. Increasing amounts of ADAMTS7 or BSA were incubated with CFH-coated ELISA plates. n=3, *P<0.05 versus respective 0 µg/ml point by one-way ANOVA followed by the Bonferroni test. (C) Schematic illustration of the CFH constructs used to map the corresponding domains that bind to the ADAMTS7. (D) Mammalian two-hybrid analysis of the CFH and ADAMTS7 interaction were performed using HEK293T cells were cotransfected with various CFH domains and full-length ADAMTS7. n=4, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (E) Schematic illustration of ADAMTS-7 structures used to map the corresponding domains, including the ADAMTS7 prodomain (TS7-F1, aa 26–246), the metalloproteinase plus disintegrin-like and cysteine-rich domain (TS7-F2, aa 238–711), the spacer-1 plus 3 TSP repeats (TS7-F3, aa 703–1007) and the spacer-2 plus 4 C-terminal TSP repeats of ADAMTS-7 (TS7-F4, aa 999–1595). (F) Mammalian two-hybrid analysis of the CFH and ADAMTS7 interaction were performed using HEK293T cells cotransfected with various ADAMTS7 domains and the CCP1-4 domain. n=4, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (G) Co-IP was performed to explore the interaction of the ADAMTS7-F4 fragment with CFH domains in the transfected HEK293T cells. (H) The solid-phase binding assay of ADAMTS7-F4 and CFH CCP 1–4. Increasing amounts of ADAMTS7-F4 or BSA were incubated with CCP 1-4-coated ELISA plates. n=3, *P<0.05 versus respective 0 µg/ml point by one-way ANOVA followed by the Bonferroni test.

To further explore the binding domains of CFH and ADAMTS7, we first subcloned the CCP module 1–4, CCP 5–9, CCP 10–18 and CCP 19–20 domains of CFH (Figure 2C) and characterized, which part of CFH is responsible for the ADAMTS7 interaction by performing a mammalian two-hybrid dual-luciferase reporter assay. As a consequence, the CCP 1–4 domain, the complement regulation domain of CFH, but not other domains, was shown to bind to ADAMTS7 (Figure 2D). To conversely characterize the binding motif of ADAMTS7 responsible for this interaction, we assessed the binding of the CFH CCP 1–4 domain with different structural motifs of ADAMTS7 in a similar assay. The spacer-2 plus 4 C-terminal TSP repeats of ADAMTS-7 (F4, aa 999–1595), rather than other domains, interacted with the CCP 1–4 domain of CFH (Figure 2, E and F). Of interest, we previously reported that other ADAMTS7 substrates, thrombospondin-1 (TSP-1) and cartilage oligomeric matrix protein (COMP), also bound to the same domain of ADAMTS7.^19,22 In addition, the direct interaction of the CFH CCP 1–4 domain with the general substrate-bound domain of ADAMTS7 was further confirmed by Co-IP experiments using HEK293 cells overexpressing various His-tagged CFH domains and Flag-tagged ADAMTS7 motifs (Figure 2G and Supplemental Figure S2E) and solid-phase binding assays using recombinant CCP 1–4 domain and ADAMTS7-F4 (Figure 2H). Collectively, CFH is directly bound to the substrate-bound domain of ADAMTS7 through its complement regulation domain CCP 1–4.

CFH Is a Novel Substrate of ADAMTS7

Subsequently, we verified whether CFH was a novel substrate of ADAMTS7. The in vitro cleavage assay was performed by using purified CFH and ADAMTS7 proteins. As shown by Coomassie brilliant blue staining of SDS-PAGE gels, reduced full-length CFH (approximately 155 kDa) accompanied by an increasing cleavage fragment (approximately 100 kDa) was observed after ADAMTS7 addition in a dose-dependent manner (Figure 3A). In accordance, the approximately 100 kDa bands were collected and extracted from SDS-PAGE gels, followed by the labeling of a dimethyl moiety at the N-terminus before trypsin digestion. Through mass spectrometry, a dimethyl moiety-labeled peptide, exclusively and repetitively identified in three independent experiments, revealed a scissile bond at N396-G397 in the CCP 7 domain, indicating the cleavage site by ADAMTS7 and suggesting the observed cleavage band (approximately 100 kDa) was C-terminal fragments produced by ADAMTS7 (Figure 3B). Of interest, no obvious bands were observed at approximately 50 kDa, which was predicted as the molecular weight of the N-terminal fragment produced from the cleavage at the N396-G397 bond by ADAMTS7, suggesting the extra degradation of this cleavage fragment by ADAMTS7. In accordance, we further performed the dimethyl moiety labeling followed by mass spectrometry directly on the samples of ADAMTS7-cleaved CFH without resolving on SDS-PAGE gels. As a result, besides N396-G397 in the CCP 7 domain, we additionally identified three scissile bonds at N136-D137 in the CCP2 domain, N179-S180 in the CCP3 domain, and N305-T306 in the CCP5 domain, respectively (Supplemental Figure S3). These mass spectrometry data collectively suggested that ADAMTS7 directly degraded the CCP 1–7 domain of CFH through multiple cleavages (Figure 3C). Furthermore, western blotting using an antibody against the C-terminus of CFH validated ADAMTS7 cleavage of CFH in the in vitro cleavage assay (Figure 3D). In addition, we validated the potential cleavage in humans, as shown by the elevation of approximately 100 kDa fragments in lupus sera (Figure 3E). Meanwhile, mass spectrometry further identified more abundant N-terminal peptides of ADAMTS7-cleaved CFH fragment starting at G397 in lupus patients compared with the healthy controls, suggesting the enhanced ADAMTS7-cleaved CFH in lupus patients (Figure 3F).

Figure 3.

ADAMTS7 cleaves CFH. (A) CFH (2 µg) was incubated with increasing amounts of ADAMTS7 (0–20 µg/ml) for 2 hours. The reaction mixture was resolved by 8% SDS-PAGE gel, followed by Coomassie blue staining. (B) The MS/MS spectrum of CFH peptide specifically generated in the presence of ADAMTS7. n=3 independent experiments. (C) Schematic indication of the potential cleavage sites of CFH by ADAMTS7. (D) Representative western blotting and quantification of CFH (2 µg) cleavage by increasing amounts of ADAMTS7 (0–20 µg/ml) for 2 hours. n=3, *P<0.05 by one-way ANOVA followed by Bonferroni test. (E) Western blot and quantification of cleaved CFH fragment and full-length CFH proteins in sera (2.5 µl/sample) from healthy controls and LN patients. n=3, *P<0.05 by unpaired two-tailed Student t test. (F) Left: The representative MS/MS spectrum of N terminal peptides of cleaved CFH in LN patients sera. Right: Relative quantification of N terminal peptides of cleaved CFH by parallel reaction monitoring-based targeted mass spectrometry in sera from controls and lupus patients. n=5, *P<0.05 by unpaired two-tailed Student t test.

ADAMTS7 Inhibits CFH Functions and Potentiates Alternative Pathway Activation

CFH mainly affects two functions to inhibit the alternative pathway of complement activation. On the one hand, CFH directly interacts with C3b through its complement regulation domain CCP 1–4 to cause the dissociation of C3b and Bb in alternative pathway C3 convertases in a process termed decay-accelerating activity. Because ADAMTS7 directly degraded CCP 1–7 domain, a solid-phase binding assay showed that ADAMTS7 disrupted the interaction of CFH with C3b in a dose-dependent manner (Figure 4A). On the other hand, CFH also serves as a cofactor of CFI, which degrades C3b to yield inactive C3b (iC3b). In accordance, cofactor activity assays showed that ADAMTS7 significantly suppressed the cofactor activity of CFH for CFI, while excluding the direct effect of ADAMTS7 on C3b degradation (Figure 4B). Furthermore, zymosan and LPS were applied to activate the complement alternative pathway in normal mouse sera. The addition of ADAMTS7 dose-dependently increased C3b deposition during incubation of normal mouse sera on immobilized zymosan and LPS (Figure 4, C and D), whereas C5a production in zymosan- or LPS-treated sera was also enhanced by ADAMTS7 (Figure 4, E and F). Alternatively, we performed hemolytic assays using rabbit erythrocytes (E_R) with diluted mice or human sera. Consequently, exogenous ADAMTS7 significantly enhanced complement-mediated lysis of E_R in both mouse and human sera (Figure 4, G and H). Conclusively, ADAMTS7 inhibited the complement regulatory activity of CFH, thereby potentiating alternative pathway activation.

Figure 4.

ADAMTS7 impairs CFH function and potentiates alternative pathway in complement activation. (A) Solid-phase binding assay of the competitive inhibition of ADAMTS7 on the interaction between C3b and CFH. C3b (0.5 µg per well) was coated on ELISA plates overnight. Increasing amounts of ADAMTS7 or BSA were incubated with CFH (2 µg/ml) at 37°C for 1 hour. Then, the mixture was added to C3b-coated wells, followed by the evaluation of the CFH-C3b interaction. n=3, *P<0.05 versus respective 0 µg/ml point by one-way ANOVA followed by the Bonferroni test. (B) The cofactor activity of CFH with CFI was embodied in the cleavage of C3b to iC3b. Under reducing conditions, iC3b could be visualized as 68-kDa and 43-kDa bands by Coomassie blue staining. n=4, *P<0.05, ns, no significance by one-way ANOVA followed by the Bonferroni test. (C and D) Alternative pathway-mediated C3b deposition on immobilized zymosan (C) or LPS (D) incubated with 10% (vol/vol) C57BL/6 mouse serum with increasing amounts ofrecombinant ADAMTS7 or BSA (as a negative control), as determined by ELISAs. C3b deposition without the addition of ADAMTS7 or BSA was set to 1. n=3, *P<0.05 versus respective 0 µg/ml point by one-way ANOVA followed by the Bonferroni test. (E and F) Alternative pathway-mediated C5a production induced by zymosan (10 mg/ml) or LPS (2 mg/ml) in C57BL/6 mouse serum with recombinant ADAMTS7 or BSA (20 µg/ml), as determined by ELISAs. n=3, *P<0.05 by one-way ANOVA followed by the Bonferroni test. (G and H) ER hemolysis was induced in 10% (vol/vol) mouse (G) or human (H) serum with or without recombinant ADAMTS7 or BSA (20 µg/ml). The hemolysis observed by incubation of ER with ddH2O was set to 100%, while sera inactivated by heat were applied as negative controls. n=4, *P<0.05 by one-way ANOVA followed by the Bonferroni test.

ADAMTS7 Deficiency Mitigates Complement Activation and Development of Lupus Pathologies

To verify the modulatory effect of ADAMTS7 on complement activation in vivo, we used mouse pristane-induced SLE model. Similar to those in MRL/lpr mice, both circulating and renal ADAMTS7 levels were markedly upregulated in the mice during the process of pristane induction (Supplemental Figure S4A-C), whereas the obvious reduction in CFH was also observed in sera and kidneys from the lupus group (Supplemental Figure S4C and Supplemental Figure S5A). To further explore the cell source of upregulated ADAMTS7, we sorted endothelial cells (CD144⁺), epithelial cells (CD326⁺), fibroblasts (PDGFRβ⁺), and leukocytes (CD45⁺) from normal and lupus kidneys by flow cytometry for real-time PCR. As a result, ADAMTS7 was greatly upregulated exclusively in endothelial cells from lupus kidneys (Supplemental Figure S4D).

We first compared complement modulation between pristane-induced wild-type (WT) and ADAMTS7^−/− mice (Supplemental Table S3). Of note, ADAMTS7 deficiency reversed the lupus-related decrease in both circulating and renal CFH (Supplemental Figure S5A and Figure 5A). In accordance with the alteration of sera CFH, circulating C3 was significantly reduced in pristane induction, but was reversed by ADAMTS7 deficiency (Supplemental Figure S5B). By contrast, pristane induction led to obvious C3 deposition in kidneys from the lupus group, but this effect was suppressed by ADAMTS7 deficiency (Figure 5, B and C). Because CFH with the analogous function of complement receptor 1 also inhibits immune complex deposition in glomeruli,^15,30 we further investigated whether ADAMTS7-degraded CFH affected renal IgG deposition. Consequently, pristane-induced glomerular IgG deposition was significantly downregulated by ADAMTS7 deficiency-rescued CFH (Supplemental Figure S5C-D). Meanwhile, ADAMTS7 deficiency did not affect serum anti-dsDNA antibody levels (Supplemental Figure S5E), in line with that ADAMTS7-cleaved CFH would be downstream of autoantibody production. Taken together, these data indicated that ADAMTS7 deficiency rescued CFH functions and reversed the pathologically increased complement activation in lupus mice.

Figure 5.

ADAMTS7 deficiency alleviates the development of renal injuries in SLE. (A) Representative western blot and quantification of CFH protein in kidneys from the WT and ADAMTS7^−/− mice with or without 6 months of pristane induction. n=4, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (B) Immunofluorescence staining of C3 deposition in renal tissues from the WT and ADAMTS7^−/− mice with or without 6 months of pristane induction. Scale bar=25 μm. (C) The relative quantification of C3 deposition on renal cross-sections. n=6, *P<0.05 by Mann–Whitney test. (D) Serum creatinine (Cre) and BUN levelsin the WT and ADAMTS7^−/− mice with or without 6 months of pristane induction. n=6, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (E) Urinary albumin-creatinine ratios (UACR) in the WT and ADAMTS7^−/− mice with or without 6 months of pristane induction. n=6, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (F) Representative images of renal sections stained with PAS. Scale bar=50 μm. (G) Quantification of glomerular and tubular-interstitial injuries. n=6, *P<0.05 by Mann–Whitney test.

Then, the characteristic pathologies of LN were accordingly evaluated. Impaired renal function was improved in the pristane-induced ADAMTS7^−/− mice compared with the WT littermates (Figure 5, D and E). In accordance, pristane-induced renal injuries, including both glomerular and tubular-interstitial regions, were ameliorated in the absence of ADAMTS7 (Figure 5, F and G). In addition, knockout of ADAMTS7 suppressed renal inflammation, as evidenced by the downregulation of inflammatory cytokines (Il1b, Il6, and Tnf) and type I IFN signaling molecules (Isg15, Mx1, and Irf7) in pristane-induced kidneys (Supplemental Figure S5F-G). Collectively, ADAMTS7 deficiency alleviated renal pathologies in mice with pristane-induced lupus. The increased risk of bacterial infection is a potential side-effect of anticomplement therapies. In accordance, we further explored whether ADAMTS7 deficiency affected complement-mediated serum bactericidal activity. Consequently, there was no difference in bacterial lysis induced by complements between WT and ADAMTS7^−/− sera, even under the conditions of SLE (Supplemental Figure S6A-B). Furthermore, we examined whether the intervention on ADAMTS7 affected the resistance to Salmonella typhimurium (5×10³ CFU/mouse) invasions in WT and ADAMTS7^−/− mice. The bacterial burdens in spleens and livers were evaluated 48 hours after bacterial infection. Both bacterial loads exhibited comparable between WT and ADAMTS7^−/− mice (Supplemental Figure S6C-D). These results suggested that the intervention on ADAMTS7 may avoid the increased risk of infection.

ADAMTS7 Participates in Lupus-Related Renal Injuries by Targeting CFH

To confirm whether CFH mediated the effects of ADAMTS7 in lupus pathologies, we intravenously injected AAV9 encoding CFH shRNA into the WT and ADAMTS7^−/− mice, followed by pristane induction (Supplemental Table S4). The liver is the major source of CFH protein. In accordance, we verified the successful knockdown of CFH in livers, kidneys, and circulation after AAV transduction (Supplemental Figure S7A-B). Consistent with the results shown in Figure 5, ADAMTS7 deficiency rescued the levels of serum CFH and C3 (Supplemental Figure S7B-C), decreased renal C3 and IgG deposition (Figure 6, A and B and Supplemental Figure S7D-E), improved renal function (Figure 6C), and mitigated renal injuries and inflammation (Figure 6, D and E and Supplemental Figure S7F-G) in lupus mice. Of note, AAV-mediated CFH silencing abolished these protective effects of ADAMTS7 deficiency (Figure 6 and Supplemental Figure S7D-G), implying that CFH mediated the effects of ADAMTS7 in complement activation and lupus-related renal pathologies.

Figure 6.

CFH mediates the effect of ADAMTS7 on the development of renal injuries in SLE. Eight- to ten-week-old female WT and ADAMTS7^−/− mice were transduced with AAV9-shRNA-NC (negative control) or AAV9-shRNA-CFH. Two weeks later, pristane was applied to induce the infected mice. (A) Immunofluorescence staining of C3 deposition in renal tissues from the AAV-transduced WT and ADAMTS7^−/− mice after 6 months of pristane induction. Scale bar=25 μm. (B) The relative quantitation of C3 deposition on renal cross-sections. n=6, *P<0.05, ns, no significance by Mann–Whitney test. (C) Serum creatinine (Cre) and BUN levels in the AAV-transduced WT and ADAMTS7^−/− mice with 6 months of pristane induction. n=6, *P<0.05, ns, no significance by two-way ANOVA followed by the Bonferroni test. (D) Representative images of renal sections stained with PAS. Scale bar=50 μm. (E) Quantification of glomerular and tubular-interstitial injuries. n=6, *P<0.05, ns, no significance by Mann–Whitney test.

ADAMTS7-Potentiated Complement Activation Is Also Involved in Renal I/R Injury Dependent on CFH

To further validate the role of ADAMTS7 in local kidneys, in addition, we used the renal I/R injury model. Distinct from SLE, the complement-mediated pathologies of I/R injury are mainly limited to the kidneys, especially tubular and interstitial injuries.³¹ Renal ADAMTS7 expression—especially in endothelial cells, epithelial cells, and leukocytes—were markedly upregulated in mice with I/R injury, whereas serum ADAMTS7 levels displayed no obvious alteration, supporting the local pathologies of renal I/R injury (Supplemental Figure S8A-E). In addition, CFH was significantly decreased in injured kidneys (Supplemental Figure S8D-E), suggesting the correlation of ADAMTS7 with CFH reduction in kidneys after I/R injury. Furthermore, we compared complement modulation between I/R injury-treated WT and ADAMTS7^−/− mice (Supplemental Table S5). Consequently, ADAMTS7 deficiency compromised I/R injury-induced downregulation of CFH expression and tubular/interstitial C3 deposition in kidneys (Figure 7, A–C). In accordance, the impaired renal function, tubular injuries, and cell apoptosis in I/R injury-treated mice were significantly improved by ADAMTS7 deficiency (Figure 7, D–H).

Figure 7.

ADAMTS7 deficiency mitigates renal complement activation and I/R injury. (A) Representative western blot and quantification of CFH protein in kidneys from the WT and ADAMTS7^−/− mice with or without I/R injury. n=4, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (B) Immunofluorescence staining of C3 deposition in renal tissues from the WT and ADAMTS7^−/− mice with or without renal I/R injury. Scale bar=50 μm. (C) The relative quantification of C3 deposition on renal cross-sections. n=6, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (D) Serum Cre and BUN levels in the WT and ADAMTS7^−/− mice with or without renal I/R injury. n=6, *P<0.05 by two-way ANOVA followed by the Bonferroni test. (E) Representative images of renal sections stained with PAS. Scale bar=50 μm. (F) Quantification of tubular injuries. n=6, *P<0.05 by Mann–Whitney test. (G) Representative images of renal sections with TUNEL staining. Scale bar=50 μm. (H) Quantification of TUNEL-positive cells. n=6, *P<0.05 by two-way ANOVA followed by the Bonferroni test.

Furthermore, CFH silencing by AAV transduction circumvented the suppressive effect of ADAMTS7 deficiency on tubular/interstitial C3 deposition and related injuries (Supplemental Table S6 and Supplemental Supplemental Figure S9). Thus, ADAMTS7 facilitated complement activation and consequently caused I/R injury-related renal injuries through CFH.

Discussion

In the current study, we found that metalloproteinase ADAMTS7 directly interacted with and degraded CFH, subsequently potentiating complement activation. Using LN or renal I/R injury mouse models, we observed the upregulation of ADAMTS7 mainly in renal endothelium.

As we previously reported that ADAMTS7 was upregulated after H₂O₂ and proinflammatory cytokine stimuli,³² ADAMTS7 upregulation in both models was possibly due to oxidative stress and inflammatory response, which mediated the pathogenesis of SLE and renal I/R injury.^33–37 Moreover, we demonstrated that ADAMTS7 deficiency rescued CFH degradation and alleviated complement-mediated renal pathologies, but without affecting complement-dependent serum bactericidal activity. Therefore, ADAMTS7, which degrades CFH, would be a promising anticomplement therapeutic target for related renal injuries, potentially reducing the risk of infection.

One major finding of the current study is that ADAMTS7-mediated CFH degradation potentiated complement activation. CFH is a critical inhibitor of alternative pathways in the complement system, whereas CFH dysfunction resulting in complement activation leads to various complement-mediated diseases. Loss-of-function mutations and autoantibodies are the major genetic and acquired drivers of CFH dysfunction, respectively.³⁸ In addition, direct protein-protein interactions, such as binding with annexin A2 or pentraxin 3, modulate CFH functions and complement activation.^10,11,39 In this study, we identified protease-mediated cleavage or degradation as a novel modulatory pathway of CFH functions. CFH exerts its complement regulatory activities in the fluid phase and on the host surface depending on its abundance in circulation and local deposition in various organs, respectively.⁴⁰ The N-terminal CCP 1–4 domain of CFH has C3bBb decay-accelerating and CFI cofactor regulatory functions, whereas the C-terminal CCP 19–20 domain mainly mediates CFH localization in the kidneys to prevent complement attack.^1,41 ADAMTS7 is a genetically identified novel locus associated with human coronary atherosclerosis.⁴² We previously demonstrated mechanistically that ADAMTS7 disturbs vascular homeostasis by cleavage of its substrates COMP and TSP-1.^19,22,32 In addition, ADAMTS7 mediates the pathogenesis of osteoarthritis by the formation of a positive feedback loop with TNF-α.⁴³ In this study, we extended the spectrum of ADAMTS7 functions and cleavage substrates, as CFH is a novel substrate of ADAMTS7. ADAMTS7 directly degraded CFH CCP 1–7 domain through multiple cleavages. In accordance, the degradation of CCP 1–7 by ADAMTS7 would cause the residual cell surface fragment lacking CCP 1-4 domain, and subsequent CFH dysfunction on the renal cell surface, as indicated by the increase in C3 deposition in renal tissues from mice with lupus and I/R injury. Moreover, ADAMTS7-degraded CCP 1–7 directly impaired the functions of CFH in the fluid phase, as evidenced by ADAMTS7-suppressed CFH cofactor activity and ADAMTS7-potentiated complement activation in mouse and human sera. Through these potential mechanisms, ADAMTS7 causes CFH dysfunction and complement activation in the fluid phase and on the local cell surface (Figure 8). Of note, a recent study found that activated coagulation factor XI (FXIa) also neutralized CFH by cleavage at a distinct site (R341/R342) in the CCP 6 domain.⁴⁴ However, whether FXIa-cleaved CFH is involved in the pathogenesis of complement-mediated diseases in vivo was not further clarified. Another finding of our study is that ADAMTS7 is a potential target for anticomplement therapy. As the first anticomplement drug approved by the US Food and Drug Administration in 2007, the humanized antihuman C5 monoclonal antibody eculizumab currently has been applied in therapies of complement-mediated disorders, including LN,⁴⁵ atypical hemolytic uremic syndrome,⁴⁶ C3 glomerulopathy (C3G)⁴⁷, and paroxysmal nocturnal hemoglobinuria (PNH).⁴⁸ To date, more than 30 anticomplement drugs or agents mainly targeting activation factors in distinct stages of complement cascades have been designed and are undergoing different stages of clinical trials.^49,50 Alternatively, complement regulatory proteins could also serve as therapeutic targets, given their inhibitory activities on complement activation. By targeting CFH, recombinant human CFH fragments (compsorbin and Mini-FH, AMY-201; Amyndas) have exhibited high efficacy in preclinical models of PNH or transplantation-induced inflammation,^51,52 whereas the phase 2a ReGAtta study (NCT04643886) is being conducted to investigate whether intravitreal delivery of GEM103 (Gemini Therapeutics), a full-length recombinant CFH protein, is safe and can slow the progression of nonexudative age-related macular degeneration (AMD) in patients with loss-of-function CFH gene variants. Our current study identified intervention with ADAMTS7 as a novel potential strategy for anticomplement therapy targeting CFH, with preserving the intact complement-mediated serum bactericidal activity to potentially avoid the increase in infection. The indications of intervention on ADAMTS7 in other complement-mediated diseases, such as PNH, C3G, aHUS, and AMD, are dependent on the existence of ADAMTS7-related CFH dysfunction, which requires further validation.

Figure 8.

Schematic illustration of ADAMTS7-potentiated complement activation. Under physiological conditions, CFH serves as a negative regulator in complement activation through decay-accelerating activity and the CFI cofactor function. In complement-mediated diseases, ADAMTS7 directly binds to and degrades CFH, further neutralizing CFH functions and potentiating complement activation in fluid phase and on cell surface.

Supplementary Material

jasn-34-291-s001.pdf ^{(9.1MB, pdf)}

jasn-34-291-s002.xlsx ^{(13.1KB, xlsx)}

Disclosures

P. Xu reports Patents or Royalties: Beijing Proteome Research Center. All remaining authors have nothing to disclose.

Funding

This research was supported by funding from the National Natural Science Foundation of China (NSFC; grants 81922009 and 81921001) and the Scientific Research Starting Foundation of Peking University (BMU2022RCZX011).

We were grateful to the grant support from the National Natural Science Foundation of China (NSFC, 82170499, 31930056, and 91839302) and the Key R&D Program of Sichuan Province (2021YFSY0038).

Author Contributions

Y. Fu, Z. Ma, Y. Tan, and X.-J. Zhou conceptualized the study; Z. Ma and C. Mao were responsible for data curation; Y. Fu, Y. Jia, Z. Ma, and C. Mao were responsible for formal analysis; Z. Ma and C. Mao were responsible for investigation; Y. Fu, Z. Ma, C. Mao, Y. Tan, P. Xu, and Q.-H. Zou were responsible for methodology; Y. Fu and W. Kong were responsible for project administration; W. Kong, Y. Tan, Q.-H. Zou, X.-J. Zhou, and P. Xu were responsible for collecting resources; Y. Fu and W. Kong were responsible for funding acquisition; Y. Fu and W. Kong were responsible for supervision; Y. Fu, Y. Jia, W. Kong, and Z. Ma were responsible for validation; Y. Fu and W. Kong were responsible for visualization; Y. Fu, W. Kong, Z. Ma, and C. Mao wrote the original draft; and Y. Fu and W. Kong reviewed and edited the manuscript.

Data Sharing Statement

All data are available in the article or the supplementary materials.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/A479 and http://links.lww.com/JSN/A480.

Supplemental Methods

Supplemental Table 1. Characteristics of the healthy controls and patients with LN.

Supplemental Table 2. Characteristics of C57 and MRL/lpr mice.

Supplemental Table 3. Characteristics of saline- or pristane-induced WT and ADAMTS7^-/- mice.

Supplemental Table 4. Characteristics of pristane-induced WT and ADAMTS7^-/- mice transduced with AAV-shRNA-NC or AAV-shRNA-CFH.

Supplemental Table 5. Characteristics of WT and ADAMTS7^-/- Mice suffering sham or renal I/R injury.

Supplemental Table 6. Characteristics of renal I/R injury-treated WT and ADAMTS7^-/- mice transduced with AAV-shRNA-NC or AAV-shRNA-CFH.

Supplemental Table 7. Primers used for RT-qPCR analysis on mouse gene expression.

Supplemental Figure 1. The expression of ADAMTS7 is induced the most in lupus nephritis.

Supplemental Figure 2. ADAMTS7 directly interacts with CFH.

Supplemental Figure 3. The MS/MS spectrum of CFH peptides specifically generated in the presence of ADAMTS7 using samples without revolving on SDS-PAGE gels.

Supplemental Figure 4. The elevation of ADAMTS7 correlates with the decline in CFH in kidneys from mice with lupus.

Supplemental Figure 5. ADAMTS7 deficiency alleviates the development of renal injuries in SLE.

Supplemental Figure 6. Targeting ADAMTS7 does not influence on complement-mediated bacterial killing.

Supplemental Figure 7. CFH mediates the effect of ADAMTS7 on the development of renal injuries in SLE.

Supplemental Figure 8. The elevation of ADAMTS7 correlates to the decline in CFH in kidneys with ischemia-reperfusion injury.

Supplemental Figure 9. CFH mediates the effect of ADAMTS7 on the development of ischemia-reperfusion injury.

Supplemental References

Supplemental Excel File I. The proteins immunoprecipitated by anti-CFH antibody.