There was no complete crystal structure of the FHR1 protein in the protein data bank (PDB) database. We performed homology modeling of SCR3-5 of FHR1*A and FHR1*B by using the ModBase database (22) to predict the effects of coding variants of FHR1 on the structure of its encoded protein. The C-terminal region (SCR18-20) of FH (PDB accession code: 3SW0), which showed a high sequence identity with the C-terminal region of FHR1, was defined as the template structure, and the models of FHR1*A and FHR1*B were analyzed via PyMOL software version 1.7.0.0. The homology modeling results were further evaluated with the given quality criteria (23).

The codon-optimized (Homo sapiens) DNA sequences for CFHR1*A and CFHR1*B (fused to a hexa-histidine tag at the C-terminus), which had been synthesized and ligated into the pTT5 vector, were customized and provided by Genescript. The expression vectors were then transiently transfected into 293F cells using polyethyleneimine according to the manufacturer’s instructions (PEI, Polysciences Inc.). Then, the 293F cells were cultured in serum-free HEK293 cell complete medium (Sino Biological Inc.) in an incubator with 5% CO₂ at 37°C for four days. After centrifugation and filtering, the harvested cell supernatant was then applied to a Ni Sepharose column (GE Healthcare) to obtain the His-tagged FHR1 proteins. The purified FHR1*A and FHR1*B proteins were finally concentrated in PBS by using an Amicon Ultra Centrifugal Filter (Merck). After purification and concentration, FHR-1 proteins were stored in PBS in -80°C.

The recombinant FHR1*A and FHR1*B proteins were analyzed by Coomassie blue staining and Western blotting. Briefly, the FHR1 proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and then, the gels were stained with Coomassie blue staining solution. For Western blotting, after SDS-PAGE, the FHR1 proteins were transferred to polyvinylidene fluoride membranes (PVDF membranes, Millipore). After blocking, the membranes were immunoblotted with mouse anti-His tag (Origene) or mouse anti-human FHR1 antibody (R&D, cross-reacts with FH), followed by anti-mouse IgG-HRP antibodies (Santa Cruz). Finally, the membranes were developed with a chemiluminescent HRP substrate (Millipore) according to the manufacturer’s instructions.

To assess the capacities of FHR1*A and FHR1*B to bind C3b, we conducted solid-phase C3b binding assays. Serial dilutions of the recombinant FHR1*A and FHR1*B proteins (2.5 μg/ml-0.078 μg/ml in PBS) were coated onto plates (Thermo Fisher Scientific) and incubated at 4°C for more than 16 hours. After washing with 0.1% PBST and blocking with 1%BSA/PBST at 37°C for 1 hour, C3b (2 μg/ml) was added and incubated at 37°C for 1 hour. Then, FHR1-bound C3b was detected using a C3c polyclonal antibody (Dako), followed by the addition of an anti-rabbit IgG-alkaline phosphatase antibody (Santa Cruz). In the reverse setting, 100nM C3b (Complement Tech) dissolved in TBS (140mM NaCl, 2mM CaCl₂, 1mM MgCl₂ and 10mM Tris, pH 7.4) were immobilized in the microtiter plates at 4°C overnight. After blocking with 3% milk in 0.02%TBST for 2 hours at 25°C, serially diluted FHR1-A and FHR1-B proteins (10μg/ml-1.25μg/ml) were added and incubated for 30min at 37°C. And mouse anti-human FHR1 antibody (R&D, cross-reacts with FH) was added as primary antibody, followed by adding AP-conjugated goat anti-mouse IgG antibody (Sigma-Aldrich). Finally, the plates were developed with alkaline phosphatase chromogenic substrate (Sigma-Aldrich), and the optical density was read at 405 nm.

Binding of FHR1 with surface-bound C3b was detected as previously reported (24). In brief, Saccharomyces cerevisiae strain (Mingzhoubio B47202) cells were resuspended in TBS and added in the microtiter plates 200μl per well for about 45min RT for deposition. After washing away the extra cells with 0.02%TBST and blocking with 3% milk in 0.02%TBST for 2 hours at 25°C, 100nM C3b or TBS were added and incubated for 1 hour at 37°C. And after incubation with serial dilution of FHR1*A and FHR1*B (5μg/ml-1.25μg/ml), mouse anti-human FHR1 antibody (R&D, cross-reacts with FH), and AP-conjugated goat anti-mouse IgG antibody (Sigma-Aldrich) were added for incubation in succession. In the last, alkaline phosphatase chromogenic substrate were added and the optical density was read at 405 nm.

The capacities of FHR1*A and FHR1*B to bind to necrotic cells were examined by flow cytometry using human umbilical vein endothelial cells (HUVECs) as previously reported (18, 25). HUVECs were purchased from ScienCell Corporation (ScienCell, Carlsbad, CA) and cultured according to the manufacturer’s protocols. Briefly, 10⁶ HUVECs/ml were incubated at 90°C for 10 min in 1%BSA/PBS and after blocking with 1%BSA/PBS, cells were then incubated with serial dilutions of the recombinant FHR1 proteins (FHR1*A or FHR1*B; 0.313 μg/ml-0.078 μg/ml in 1%BSA/PBS) at 37°C for 30 min. After washing with PBS, the HUVECs were stained with a mouse anti-human FHR1 antibody (JHD10, Hycult Biotech, cross-reacts with FHR2 and FHR5) or mouse IgG1 [CT6] isotype control (Abcam) followed by incubation with an Alexa Fluor^® 647-conjugated anti-mouse IgG antibody (Cell Signaling Technology) as the secondary antibody. The necrosis induction was measured by propidium iodide and Annexin V staining (BD556547). Double positive for both Annexin V and PI cells were determined as necrotic cells. Cell acquisition was performed by a FACScan flow cytometer (BD), 10,000 cells were measured for each sample and data were analyzed by FlowJo (v10).

To compare the capacity of FHR1*A and FHR1*B to compete with FH to bind C3b, FH (9 μg/ml) was immobilized on the surface of a microtiter plate (Thermo Fisher Scientific) at 4°C overnight. After blocking with 1% BSA/PBST at 37°C for 1 hour, serial dilutions of the recombinant FHR1*A or FHR1*B proteins (2.5 μg/ml-0.039 μg/ml) as well as C3b (0.25 μg/ml) were added to the plates. After shaking for 30 sec, the plates were incubated at 37°C for 1 hour. After washing with 0.1%PBST, FH-bound C3b was detected with a C3c polyclonal antibody (Dako), followed by incubation with an alkaline phosphatase-conjugated anti-rabbit IgG antibody as the secondary antibody (Sigma-Aldrich). The plates were then developed with alkaline phosphatase chromogenic substrate (Sigma-Aldrich), and the optical density was read at 405 nm.

To investigate the effect of FHR1 on the cofactor activity of FH, the generation of C3b cleavage products by complement factor I (CFI) was assessed by SDS-PAGE and Western blotting. Two serially diluted isoforms of FHR1 (125 μg/ml-62.5 μg/ml), alone or in combination with FH (15.64 ng), were mixed with C3b (1.1 μg) and CFI (100 ng) and then incubated at 37°C for 30 min. After SDS-PAGE and Western blotting, the C3b degradation product (α’ 43 kDa) was detected with a C3c polyclonal antibody (Dako), followed by an HRP-conjugated anti-rabbit IgG antibody (Jackson ImmunoResearch Labs). The membranes were developed with a chemiluminescent HRP substrate (Millipore), and the signals of the C3b degradation product (α’ 43 kDa) were analyzed by ImageJ.

To evaluate the potential effect of FHR1 on the FH-mediated regulation of the C3bBb, C3 convertase assay was performed as previously described (24). In brief, at first 5μg/ml C3b (Complement Tech) were immobilized in the microtiter plates. After blocking with 3% milk in 0.02%TBST for 2 hours at 25°C, 4μg/ml factor B (Millipore), 8μg/ml properdin (Millipore), 0.2μg/ml factor D (Millipore), 20μg/ml BSA (Sigma-Aldrich), altogether with 100nM factor H (Complement Tech) and serially diluted FHR1*A or FHR1*B (300nM-500nM) were added. After incubation, anti-factor B goat polyclonal antibody (Sigma-Aldrich) was added for incubation, followed by adding AP-conjugated rabbit anti-goat IgG antibody (Sigma-Aldrich). Finally, after incubated with alkaline phosphatase chromogenic substrate, the optical density was read at 405 nm.

As previously reported (18), we evaluated the effect of FHR1 as a driver of monocytic inflammation. Briefly, peripheral blood mononuclear cells (PBMCs) were freshly isolated from the whole blood of three independent healthy donors by the density gradient centrifugation method using Ficoll Pague Plus (GE Healthcare Life Science). Then, the monocytes were indirectly magnetically labeled and isolated from the PBMCs using the Classic Monocyte Isolation Kit (Miltenyi Biotec). Then, 96-well microtiter plates (Thermo Fisher Scientific) were coated with recombinant-FHR1*A, FHR1*B, and BSA for 1 hour at 37°C. Then, 1x10⁵ monocytes were incubated in the plates with immobilized FHR1 in complete medium with 10% normal human serum (NHS) for 20 hours at 37°C in an incubator with 5% CO₂. Finally, the levels of the inflammatory factors IL-1β and IL-6 in the supernatants after 20 hours of incubation were measured by enzyme-linked immunosorbent assay (ELISA) using IL-1β (R&D) and IL-6 (R&D) ELISA kits according to the manufacturers’ protocols.

Continuous variables with normal distribution are expressed as mean ± standard deviation and independent t-test or One way-ANOVA is used for comparison between groups. For non-normally distributed variables, the median and interquartile range (IQR) are used for description, while the Kruskal-Wallis test or Mann-Whitney U test is used for comparison. Statistics were performed using SPSS25.0 software and graphing were performed by GraphPad Prism 7.0 software. A two-tailed P value less than 0.05 was considered statistically significant.

The structural models of FHR1*A and FHR1*B were successfully established via Modbase, and we selected the most reliable models for further analysis according to the quality criteria provided by the database ( Figures 1A, B ). After aligning the models of FHR1*A and FHR1*B ( Figure 1C ), we discovered that the overall spatial structures were similar. Although the secondary structure of the β-fold in SCR3 was not altered by the amino acid changes at p.139 and p.157, the conformations of FHR1*A and FHR1*B were still slightly different due to the three different amino acids, and SCR3 of FHR1*B was more prone to SCR4-5 than that of FHR1*A. To better analyze the conformation, we set p.189Lys as the vertex and p.189Lys-p.194Pro as the edge to measure the angles formed by different amino acids in SCR3 ( Table 1 ). The angles in FHR1*B are generally smaller than those in FHR1*A, ranging from 6.3° to 8.9° ( Figures 1D, E ).

We expressed recombinant FHR1*A and FHR1*B proteins with His-tag in eukaryotic expression system and purified by Ni Sepharose affinity chromatography. On SDS-PAGE gels, the recombinant FHR1*A and FHR1*B proteins showed two clear bands at approximately 40 kDa, which were detected with the His-tag antibody and FHR1 monoclonal antibody; these results were consistent with two previously reported glycosylated forms of FHR1 (42 kDa and 37 kDa, Supplemental Figure 1 ) (26). Thereafter, the recombinant FHR1*A and FHR1*B proteins were used in the following experiments.

Since the C-terminus of FHR1 showed 98% sequence identity with that of FH and the C-terminus of FH contained a C3b binding site, we explored the capacity of the two isoforms of FHR1 to bind to C3b. In the binding assay, both isoforms, namely, FHR1*A and FHR1*B, bound C3b in a dose-dependent manner. Moreover, at the same concentrations (either 1.25 μg/ml or 2.5 μg/ml), FHR1*B exhibited significantly higher C3b binding capacity than FHR1*A (FHR1*B vs. FHR1*A: at 1.25 μg/ml concentration, P=0.021; at 2.5 μg/ml concentration, P=0.003, Figure 2A ). In the reverse setting, dose-dependent binding of the FHR-1 isoforms to immobilized C3b was measured, and FHR1*B also showed higher capacity of binding to C3b than FHR1*A (FHR1*B vs. FHR1*A: at 1.25μg/ml concentration, P=0.002; at 2.5μg/ml concentration, P=0.001; at 5μg/ml concentration, P=0.002; at 10μg/ml concentration, P<0.001, Figure 2B ).

In surface-bound C3b binding assay, in which C3b was bound to Saccharomyces cerevisiae strain, we found that the FHR1*B isotype also has a stronger binding ability than FHR1*A to surface-bound C3b. (FHR1*B vs. FHR1*A: at 1.25μg/ml concentration, P=0.006; at 2.5μg/ml concentration, P<0.001; at 5μg/ml concentration, P=0.001, Figures 2C, D ).

FHR1 was previously reported to bind to necrotic cells (13); therefore, we compared the capacity of the FHR1 isoforms to bind to necrotic HUVECs. We observed that both FHR1 isoforms bound to necrotic HUVECs ( Figure 3 ) in a dose-dependent manner. In addition, at the same concentrations (0.313 μg/ml-0.078 μg/ml), FHR1*B showed significantly increased capacity to bind to necrotic HUVECs compared with FHR1*A (FHR1*B vs. FHR1*A: at 0.313 μg/ml concentration, P=0.001; at 0.156 μg/ml concentration, P<0.001; at 0.078 μg/ml concentration, P=0.001).

Since FHR1 was found to function as a competitive antagonist of FH to modulate activation of the complement cascade (27), we next explored whether FHR1*A and FHR1*B showed different levels of competition with FH. First, we evaluated the competition of FHR1*A and FHR1*B with FH for binding C3b. We found that FHR1*A and FHR1*B competed with FH for binding C3b in a dose-dependent manner, which was similar to previously reported observations of FHR1 (27). Moreover, within the concentration range of 0.078 μg/ml to 2.5 μg/ml, FHR1*B exhibited stronger competition with FH for binding C3b than FHR1*A, as FHR1*B decreased the binding by 10.49% (0.078 μg/ml FHR-1) and 30.52% (2.5 μg/ml FHR-1), while FHR1*A decreased the binding by only 1.11% and 10.73% ( Figure 4A ).

Next, we performed a cofactor assay to assess the deregulation of FHR1 by FH in terms of its CFI cofactor activity. As previously reported, unlike FH, FHR1 showed no intrinsic cofactor activity for CFI ( Figures 5A, B ). Here, we observed that FHR1 competed with FH in a dose-dependent manner, which led to a reduction in the CFI cofactor activity of FH. At the same concentrations (either 125 μg/ml or 62.5 μg/ml), FHR1*B showed a higher capacity for de-regulating FH-mediated CFI cofactor activity than FHR1*A (FHR1*B vs. FHR1*A: at 62.5 μg/ml concentration, P<0.001; at 125 μg/ml concentration, P<0.001; Figures 5C, D ).

Given the negative regulation of FH on C3 convertase formation into consideration, we investigated the potential different effect of FHR1*B and FHR1*A on FH mediated regulation of the solid-phase C3 convertase. We found that FH inhibited the C3bBb assembling, while both isoforms of FHR1 competed with FH and reversed the inhibition. Moreover, FHR1*B showed more powerful deregulation effect on FH than FHR1*A, regarding the FH mediated regulation of the solid-phase C3 convertase. (FH vs. TBS, P<0.001. FHR1*B vs. FHR1*A: at 300nM concentration, P=0.009; at 400nM concentration, P=0.005; at 500nM concentration, P=0.007, Figure 4B ).

Recently, FHR1 was reported to play a proinflammatory role. We therefore compared the effects of FHR1*A and FHR1*B on monocytes. We found that monocytes incubated with FHR1*B secreted higher levels of IL-1β and IL-6 than monocytes incubated with FHR1*A. (FHR1*B vs. FHR1*A: IL-1β, P<0.001; IL-6, P<0.001; Figure 6 ).