Potentiation of complement regulator factor H protects human endothelial cells from complement attack in aHUS sera

Abstract

Mutations in the gene encoding for complement regulator factor H (FH) severely disrupt its normal function to protect human cells from unwanted complement activation, resulting in diseases such as atypical hemolytic uremic syndrome (aHUS). aHUS presents with severe hemolytic anemia, thrombocytopenia, and renal disease, leading to end-stage renal failure. Treatment of severe complement-mediated disease, such as aHUS, by inhibiting the terminal complement pathway, has proven to be successful but at the same time fails to preserve the protective role of complement against pathogens. To improve complement regulation on human cells without interfering with antimicrobial activity, we identified an anti-FH monoclonal antibody (mAb) that induced increased FH-mediated protection of primary human endothelial cells from complement, while preserving the complement-mediated killing of bacteria. Moreover, this FH-activating mAb restored complement regulation in sera from aHUS patients carrying various heterozygous mutations in FH known to impair FH function and dysregulate complement activation. Our data suggest that FH normally circulates in a less active conformation and can become more active, allowing enhanced complement regulation on human cells. Antibody-mediated potentiation of FH may serve as a highly effective approach to inhibit unwanted complement activation on human cells in a wide range of hematological diseases while preserving the protective role of complement against pathogens.

Graphical abstract

Figure 1.

Effect of anti-FH.07 on in vitro AP complement activation. (A) AP-mediated C3b deposition on immobilized zymosan incubated with 10% (vol/vol) NHS in the presence of anti-FH.07 (binding CCP18) or anti-FH.09 (binding CCP6), as determined by ELISA. C3b deposition without addition of mAbs was set to 100%. (B) As in panel A, but with immobilized LPS instead of zymosan. (C) IC₅₀ of intact or Fab′ fragments of anti-FH.07 in the AP-mediated C3b deposition assay on immobilized LPS. (D) Fluid phase regulation by FH was assessed with a fluid phase cofactor activity assay. Purified C3b was incubated with FH and FI, with addition of the indicated mAbs (anti-FH.11 binds CCP1-4). Cleavage of the α′-chain of C3b was visualized with SDS-PAGE run under reducing conditions, stained with PageBlue. Lanes were run on the same gel but were noncontiguous. (E) C3b/c concentration determined by ELISA after incubation of NHS with the indicated mAbs to assess fluid phase activation of C3. (F-G) Representative SPR sensograms (n = 2) of the binding of full-length FH (F) or a fragment of FH comprised of domains 18 to 20 (G) to immobilized anti-FH.07. For full-length FH, a twofold diluting concentration range starting at 200 nM was probed. For FH_18-20, a twofold diluting concentration range starting at 150 nM was probed. Injections were done in random order. The first 200 seconds of the 600-second dissociation phase are shown. Red lines depict the fits of a 1:1 association model. Data in panels A-C,E are presented as mean of n = 3 with standard deviation. ****P < .0001, 1-way ANOVA. ns, not significant; RT, room temperature.

Figure 2.

Effect of anti-FH.07 Fab′ fragment on binding of FH to C3 fragments. Binding of FH in the absence or presence of anti-FH.07 to C3b (A), iC3b (B), and C3d (C). C3b (2,045 RU), iC3b (2,039 RU) and C3d (2,055 RU) were amine-coupled to a CM5 chip to assess binding of FH (black lines) or FH in the presence of 4 µM anti-FH.07 Fab′ fragments (FH:07, orange lines). Double injections of a twofold diluting concentration range starting at 2 µM FH were probed in random order throughout. For ease of interpretation, only the sensograms of 2, 1, and 0.5 µM FH and FH:07 are shown. Sensograms were normalized for the molecular weight of FH (155 kDa) or the FH:anti-FH.07 Fab′ fragment complex (205 kDa), respectively. All graphs shown are representative of multiple experiments (n ≥ 2).

Figure 3.

Anti-FH.07 restores complement regulation in aHUS patient serum on E_S. (A) Schematic representation of the 20 CCP domains of FH with the location of the heterozygous mutations of each of the aHUS patients (numbered 1 to 11), as well as the domain targeted with anti-FH.07. (B) Incubation of E_S with serum from an aHUS patient carrying a heterozygous mutation in CFH (n = 2) or a serum pool from healthy individuals (NHS pool, n = 3). Hemolysis was measured at 412 nm and corrected for background at 690 nm. The hemolysis observed by incubation of E_S with 0.6% (wt/vol) saponin was set to 100%. (C) Addition of anti-FH.07 or an isotype control to 10% (vol/vol) serum of aHUS #1 or a NHS pool (n = 4). (D) Hemolysis observed in 11 aHUS sera (10%, vol/vol) with different CFH mutations before and after addition of anti-FH.07 or an isotype control; numbers correspond to the aHUS patient number in panel A. Addition of EDTA serves as the negative control of E_S lysis. Each point represents the mean of duplicates of a single experiment, and red lines indicate overall mean with standard deviation. (E) Comparison of the effect intact IgG and Fab′ fragments from a control IgG mAb and anti-FH.07 in the E_S hemolytic assay performed with 10% (vol/vol) aHUS#1 serum. Data in panels C-E are presented as mean with standard deviation. *P < .05, ****P < .0001, 2-way ANOVA.

Figure 4.

Anti-FH.07 protects E_Sfrom induced complement-mediated hemolysis. (A) E_S hemolysis was induced in 10% (vol/vol) NHS by inhibiting FH with anti-FH.09 (n = 3). Hemolysis was measured at 412 nm and corrected for background at 690 nm. The hemolysis observed by incubation of E_S with 0.6% (wt/vol) saponin was set to 100%. (B) Titration of anti-FH.07 in the induced E_S hemolysis assay (n = 3). For this E_S hemolytic assay, 0.067 µM (10 µg/mL) anti-FH.09 was used, resulting in 87% E_S lysis as indicated with the dashed lines in panel A. (C) Comparison of the inhibitory effect of anti-FH.07 and eculizumab in the E_S hemolytic assay performed in panel B (after normalization). Dashed line indicates 50% of the normalized E_S hemolysis. (D) IC₅₀ values of anti-FH.07 and eculizumab in the induced E_S hemolytic assay as in panel C (n = 3) using either a NHS pool or a pool of sera from 4 healthy donors lacking the CFHR1 gene (CFHR1^−/− pool). (E) Incubation of E_R with 10% (vol/vol) NHS serum in the presence of eculizumab or anti-FH.07 at the indicated concentrations (n = 2). All data are presented as mean with standard deviation. *P < .05, ****P < .0001, 2-way ANOVA.

Figure 5.

Anti-FH.07 increases complement regulation on human cell surfaces. (A) Immunofluorescent imaging of C3b deposition on HUVECs incubated with 10% (vol/vol) NHS, HI-NHS, or aHUS-like NHS in the absence or presence of anti-FH.07. For each condition, a representative image of independent experiments (n = 3) is shown. (B) Relative C3b deposition on HUVECs (from panel A), calculated by quantifying the mean fluorescence intensity (MFI) using a maximum intensity projection of the whole coverslip (10 × 10 × 15 images per coverslip). Anti-C3.19 MFI of NHS was set to 1. Bars represent mean with standard deviation (n = 3). (C) Immunofluorescent imaging of C3b deposition on HUVECs following incubation with 10% (vol/vol) serum of 5 aHUS patients in the absence of presence of an isotype control mAb or anti-FH.07. In panels A,C, HUVECs were stained with Hoechst (turquoise, nuclei), WGA (magenta, plasma membrane), and anti-C3.19 (green, C3b deposition). Scale bars in panels A,C represent 100 µm. *P < .05, 1-way ANOVA.

Figure 6.

Bactericidal activity of NHS remains intact in the presence of anti-FH.07. (A) E coli CFUs/mL were determined after incubation in 75% (vol/vol) NHS, NHS preincubated with anti-FH.07, NHS preincubated with eculizumab, or HI-NHS. (B) CFUs/mL of 2 clinical isolates of N meningitidis serogroup B (left) and serogroup W (right), determined after incubation in 75% (vol/vol) NHS, NHS preincubated with anti-FH.07, NHS preincubated with eculizumab, or HI-NHS. All data are presented as mean with standard deviation (n = 3). **P < .01, ****P < .0001, 2-way ANOVA.