Participants were 9 adult patients with aHUS in remission, included in the International Registry of Recurrent and Familial Haemolytic Uremic Syndrome/Thrombotic Thrombocytopenic Purpura (HUS/TTP). The Registry was established in 1996 at the Aldo e Cele Daccò Clinical Research Center for Rare Diseases (Ranica, Bergamo) (villacamozzi.marionegri.it/seu). Samples collected were stored at the Mario Negri Institute Biological Resources Center, in Biobank for Rare Diseases and Kidney Diseases (Ranica, Bergamo).

Three patients carry pathogenic RVs in CFH gene (#1, p.R1210C; #2, p.R78G; #3, p.S1191L), three patients in C3 gene (#4, p.D61N; #5, p.S1063R; #6, p.T162R), and three patients in CFI gene (#7, p.P50R; #8, p.V412M; #9, p.C67W). No patients have anti-CFH autoantibodies. Table 1 summarizes clinical and genetic characteristics of the patients. Rare functional variants (missense, nonsense, indel, or splicing variants) with minor allele frequency (MAF) <0.001 in the Genome Aggregation Database (gnomAD, https://gnomad.broadinstitute.org/) were selected. Stop-gain, frameshift and splicing variants, and missense variants with published functional studies, were categorized as pathogenic variants (PV). The other variants were categorized as likely pathogenic variants (LPV) or variants of uncertain significance (VUS), using guidelines from the American College of Medical Genetics and Genomics (ACMG) and from the KDIGO conference on aHUS and C3G (11–13).

aHUS was diagnosed in patients who have had one or more episodes of microangiopathic hemolytic anemia and thrombocytopenia, with hematocrit (Ht) <30%, hemoglobin (Hb) <10 g/dL, serum lactate dehydrogenase (LDH) >500 IU/L, undetectable haptoglobin, fragmented erythrocytes in the peripheral blood smear, and platelet count <150,000/μL, associated with acute renal failure (serum creatinine >1.3 mg/dL, and/or urinary protein/creatinine ratio >200 mg/g, or an increase in serum creatinine or a urinary protein/creatinine ratio >15% compared to baseline levels) (14). TTP was ruled out on the basis of ADAMTS13 activity >10% and no anti-ADAMTS13 antibodies. Stx-E.Coli infection was ruled out based on negative assays for stx and eae genes (by PCR) or Shiga-toxin (Vero cell assay) in the stool and/or anti-Shiga-toxin antibodies (ELISA) and/or LPS O157, O26, O111, or O145 (ELISA) in serum.

Full remission was defined as the normalization of both hematological parameters (Ht >30%; Hb >10 g/dL; LDH <500 IU/L; platelets >150,000/μL) and renal function (serum creatinine <1.3 mg/dL for adults, urinary protein/creatinine ratio <200 mg/g).

Patients had not taken C5 inhibitory drugs for at least 8 weeks prior to blood draws. The protocol was approved by the Ethical Committee of the Azienda Sanitaria Locale Bergamo, Italy.

The human microvascular endothelial cell line of dermal origin (HMEC-1, a kind gift of Dr. Edwin Ades and Francisco J. Candal of CDC and Dr. Thomas Lawley of Emory University, Atlanta, GA) was cultured as described (14, 15). For the test, HMEC-1 were plated on glass coverslips and used when confluent. Cells were washed three times with test medium (HBSS: 137 mmol/l NaCl, 5.4 mmol/l KCl, 0.7 mmol/l Na₂HPO₄, 0.73 mmol/l KH₂PO₄, 1.9 mmol/l CaCl₂, 0.8 mmol/l MgSO₄, 28 mmol/l Trizma base pH 7.3, 0.1% dextrose; with 0.5% BSA) and then activated with 10 µM ADP (Sigma-Aldrich) for 10 minutes. At the end, cells were washed and then incubated for 4 hours with aHUS serum or with a pool of 10 control sera (NHS) run in each experiment diluted 1:2 with test medium (50% of serum final concentration) in the presence or absence of: the pan-complement inhibitor sCR1 (150 µg/ml, i.e. 600 nM), anti-C3d mAb conjugated with CR1_1-10 (1, 10, 100 nM, i.e. 0.290, 2.901, 29.008 µg/ml) or non-targeted CR1_1-10 (1, 10, 100 nM, i.e. 0.073, 0.726, 7.258 µg/ml), anti-C3d mAb conjugated with CR1_1-17 (1, 10, 100 nM, i.e. 0.389, 3.887, 38.874 µg/ml) or non-targeted CR1_1-17 (1, 10, 100 nM, i.e. 0.122, 1.219, 12.191 µg/ml), anti-C3d mAb conjugated with FH_1-5 (10, 100, 1000 nM, i.e. 2.171, 21.707, 217.068 µg/ml) or non-targeted FH_1-5 (10, 100, 1000 nM, i.e. 0.361, 3.607, 36.072 µg/ml). Table 2 summarizes structural characteristics of the complement inhibitors (anti-C3d mAb conjugated and unconjugated) used. The anti-C3d mAb, 3d8b, recognizes a shared epitope with iC3b, C3dg and C3d. The term “anti-C3d” is used when describing the activity of the C3d mAb fusion proteins.

At the end of the incubation step, HMEC-1 were washed, fixed in 3% paraformaldehyde, washed again and then blocked with PBS with 2% BSA for one hour. The cells were stained with FITC-conjugated rabbit anti-human C3c (Dako, that recognizes C3c, part of C3 and C3b, 1:300 final dilution in Dapi 1 µg/mL). In selected experiments, to confirm target presence on treated HMEC-1, cells have been stained with mouse anti-human C3d (3d8b, the same clone used for the anti-C3d conjugated complement inhibitors, 1.75 µg/ml) or the isotype control (mouse IgG1, k [MOPC-21] from murine myeloma, Sigma) followed by PE-conjugated rat anti-mouse light chain kappa secondary antibody (H139-52.1, Abcam, 1:100 final dilution in 1 µg/mL Dapi).

The fluorescent staining was acquired through the Apotome Axio Imager Z2 microscope (Zeiss) and the stained area was evaluated using Image J (NIH, Bethesda, MD). For each sample, 15 fields were analyzed, the highest and the lowest values were discarded and the mean was calculated on the remaining 13 fields.

Results were expressed in pixel² or as percentage of stained area relative to area measured after exposure to aHUS serum alone, which was set as 100%. Half maximal inhibitory concentration (IC₅₀) curves for C3d-targeted complement inhibitors were generated using non-linear regression in GraphPad PRISM v10 with normalized response (aHUS serum alone, 100%), values were expressed as mean ± SEM.

To assess whether the anti-C3d mAb (3d8b) alone, which targets complement inhibitors (CR1_1-10, CR1_1-17, and FH_1-5) to C3d, affects complement activation and deposition on HMEC-1 cells incubated with aHUS serum, preliminary experiments were conducted using aHUS serum added with 3d8b mAb (10, 100, and 1000 nM). C3 deposits on ADP-activated HMEC-1 exposed to aHUS serum added with 3d8b mAb, were comparable to those observed with serum alone, indicating that the 3d8b mAb alone did not affect complement activation and C3 deposition on HMEC-1 exposed to aHUS serum ( Supplementary Figure 1 ), and did not mask C3 detection on cells as well.

Experiments were performed as described above for C3 deposition evaluation. HMEC-1 cells were incubated with serum from aHUS patients in the presence or absence of anti-C3d mAb conjugated with CR1_1–10 or non-targeted CR1_1-10 (10 nM), anti-C3d mAb conjugated with CR1_1–17 or non-targeted CR1_1-17 (10 nM), anti-C3d mAb conjugated with FH_1–5 or non-targeted FH_1-5 (100, 1000 nM). At the end of the incubation step, fixed cells were stained with rabbit anti–human complement C5b-9 complex (Calbiochem, 1:200 final dilution) followed by FITC-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, 1:50 final dilution in 1 µg/mL Dapi). The fluorescent staining was acquired through the Apotome Axio Imager Z2 microscope (Zeiss) and the stained area was evaluated using Image J (NIH, Bethesda, MD). For each sample, 15 fields were analyzed, the highest and the lowest values were discarded and the mean was calculated on the remaining 13 fields.

Results were expressed as stained area in pixel².

Platelet adhesion and aggregation on HMEC-1 under flow condition was performed as described previously (16). HMEC-1 were plated on a glass slide and used when they formed a monolayer. Cells were activated with ADP (10 μM) for 10 minutes and then exposed for 4 hours to serum (aHUS or control serum pool) diluted 1:2 with test medium, in the presence or absence of anti-C3d conjugated with CR1_1-17 (10 nM) or with FH_1-5 (100 and 1000 nM), or sCR1 (150 µg/ml), or eculizumab (100 µg/ml, based on recommended C_through target level of 50–100 µg/ml). After a quick flush with HBSS, HMEC-1 were perfused with heparinized (10 IU/ml heparin) whole blood from healthy subjects (pre-labeled with the fluorescent dye mepacrine 10 μM) in a thermostatic flow chamber (37°C). A constant flow rate of 1500 sec^-1 (60 dynes/cm²) was applied to mimic the shear stress conditions found in the microcirculation (16). Three different healthy subjects were used as blood donors for all the experiments. The same blood donor was used for all the slides/conditions evaluated in each experiment. After 3 minutes, perfusion was stopped and the slide with the endothelial cell monolayer was dehydrated and fixed in acetone for 20 minutes. The integrity of the cell monolayer after the exposure to serum was verified in parallel slides in which cells were stained with May-Grunwald Giemsa.

The fluorescent staining was acquired through the Apotome Axio Imager Z2 microscope (Zeiss) and the stained area was evaluated using Image J (NIH, Bethesda, MD). For each sample, 15 fields were analyzed, the highest and the lowest values were discarded and the mean was calculated on the remaining 13 fields. Results were expressed as area covered by platelet aggregates, in pixel².

Data were reported as mean ± SD. The normality of value distribution was assessed using the D’Agostino-Pearson’s test. Comparisons of groups were performed with one-way analysis of variance followed by post hoc Student-Newman-Keuls test or Kruskal-Wallis test followed by post hoc Conover test, as appropriate. Statistically significant differences were assumed at a 5% level of probability. All calculations were made using MedCalc 10.0.1 statistical software (MedCalc Software, Mariakerke, Belgium).

We included in this study 9 aHUS patients in remission out of anti-C5 therapy who carried rare variants (RVs) in the CFH, or CFI, or C3 genes. Table 1 summarizes clinical and genetic characteristics of the patients.

Consistently with previously published reports (14–18), C3 deposition induced on ADP-activated HMEC-1 by serum from aHUS patients was significantly higher than that obtained with control serum pool, regardless of the genetic defect ( Figure 1 ). C3 deposits were evaluated by anti-C3c antibody that recognizes an epitope shared by intact C3 and C3b. The addition of the pan-complement inhibitor sCR1 at the concentration of 150µg/ml (600 nM), which has been shown to completely block complement activation (16), fully prevented the effect of aHUS serum on C3 deposits ( Figure 1 ). These data indicated that aHUS serum-induced C3 deposition entirely reflected complement activation.

To further confirm the above findings, we evaluated C3d deposited on ADP-activated HMEC-1 exposed to aHUS serum using the anti-C3d mAb 3d8b. This antibody recognizes a neoepitope shared by the complement fragments iC3b, C3dg, and C3d without binding to C3/C3b/C3c, and it has been documented to detect C3d deposition on biopsy specimens from various complement-related autoimmune kidney and skin conditions (9, 10). We observed a significant increase in C3d staining on ADP-activated cells exposed to aHUS serum compared to those exposed to control serum ( Supplementary Figure 2 ).

We then evaluated whether fusion proteins combining an anti-C3d monoclonal antibody (mAb) with complement inhibitors could efficiently block aHUS serum-induced complement activation on endothelial cell surface.

Both anti-C3d mAb conjugated with CR1_1–10 and soluble CR1_1–10 added to serum from aHUS patients, dose dependently reduced C3 deposition on ADP-activated HMEC-1 ( Figures 2A , 3 ). Notably, at the dose of 10nM, C3d-targeted CR1_1–10 was more effective than equimolar non-targeted CR1_1-10 ( Figure 2A ). The anti-C3d mAb alone had no effect on aHUS serum-induced C3 deposition on HMEC-1, indicating that it did not independently influence complement activation on HMEC-1 exposed to aHUS serum and did not interact with anti-C3c Ab staining ( Supplementary Figure 1 ).

Similar results were obtained with C3d-targeted CR1_1–17 added to aHUS serum, although it was more effective in reducing C3 deposition than non-targeted CR1_1-17 both at 1nM and 10 nM doses ( Figures 2B , 3 ).

Half maximal inhibitory concentrations (IC₅₀) on C3 deposition were 2.61 ± 0.48 nM and 1.05 ± 0.60 nM for C3d-targeted CR1_1–10 and C3d-targeted CR1_1-17, respectively. Of note the IC₅₀s of the non-targeted inhibitors were about 4-fold higher than the corresponding C3d-targeted ones (CR1_1-10: 8.16 ± 1.96 nM; CR1_1-17: 6.61 ± 1.67 nM).

At 10 nM concentration, neither the C3d-targeted FH_1-5, which combines the anti-C3d mAb with the complement-regulatory domain of CFH (SCRs 1-5), nor the non-targeted FH_1–5 reduced serum-induced C3 deposition on ADP-activated HMEC-1 ( Figures 2C , 3 ). Increasing the concentrations to 100 and 1000 nM, led to a partial but significant reduction of C3 deposits on HMEC-1 exposed to aHUS serum added with C3d-targeted FH_1-5, while equimolar non-targeted FH_1–5 had no effect ( Figures 2C , 3 ).

To compare the inhibitory potential of C3d-targeted complement inhibitors among patients with RVs in different genes, we expressed the results as percentage of C3 deposition induced by aHUS serum in the presence of complement inhibitors, relative to C3 deposition in their absence, which was set to 100%. The analysis of individual data revealed that, at the 1 nM dose, C3 deposition was minimally or not prevented on HMEC-1 exposed to sera from patients #1 and #3 – carrying the p.R1210C and p.S1191L RVs in CFH gene, respectively – with either C3d-targeted CR1_1–10 or C3d-targeted CR1_1-17 ( Figures 4A, B ). At the 10nM and 100nM doses, C3 deposits were nearly and completely in the normal range, respectively ( Figures 4A, B ). On the other hand, patients with either C3 or CFI RVs were more sensitive to the inhibitory effects of both C3d-targeted CR1_1–10 and C3d-targeted CR1_1-17 ( Figures 4A, B ). As for fusion proteins combining anti-C3d mAb with FH_1-5, the analysis of results in individual patients evidenced that a subgroup of patients was more sensitive to the inhibitory effect of C3d-targeted FH_1-5. Indeed, the compound was able to substantially reduce ex-vivo serum-induced C3 deposition even at the dose of 10 nM in patient #2 carrying the p.R78G RV in CFH gene, in patient #5 carrying the p.S1063R RV in C3 gene, and in patients #7 and #9, who carry the p.P50A and p.C67W RVs in CFI gene ( Figure 4C ).

As in aHUS the terminal product of the complement cascade is strongly activated on endothelium (14, 15), we then tested the efficacy of C3d-targeted complement inhibitors in preventing C5b-9 formation induced by aHUS serum on HMEC-1. For 3 patients, namely patient #1 (p.R1210C RV in CFH), #5 (p.S1063R RV in C3) and #7 (p.P50A RV in CFI), there were sufficient serum samples available.

Data showed a significant reduction of C5b-9 formation on cells exposed to sera added with either C3d-targeted CR1_1–10 or CR1_1-17 ( Figures 5A , 6 ), or C3d-targeted FH_1-5 ( Figures 5B , 6 ), as well as with non-targeted CR1_1-10, CR1_1-17, or FH_1-5 ( Figures 5A, B ). Notably, values of C5b-9 staining were comparable to those observed with control serum pool with C3d-targeted CR1_1–10 and CR1_1–17 at 10nM and with C3d-targeted FH_1–5 at 1000 nM concentration ( Figures 5A, B ).

In aHUS complement activation on endothelial cell surface is associated with loss of anti-thrombotic properties (16). Thus, we then moved to evaluate the effect of C3d-targeted CR1_1–17 and C3d-targeted FH_1–5 on platelet adhesion and aggregation on HMEC-1 exposed to aHUS serum and then perfused with control whole blood. C3d-targeted CR1_1–17 and FH_1–5 were tested with sera from the same 3 patients selected for the above-described experiments of C5b-9 formation.

As shown in Figure 7 , the area covered by platelet aggregates was about 10-fold wider on HMEC-1 pre-exposed to aHUS serum as compared with that observed on endothelium pre-exposed to control serum pool, confirming our previously published results (16). Addition of C3d-targeted CR1_1–17 at 10 nM to aHUS serum fully prevented platelet adhesion and aggregation induced on HMEC-1 by the exposure to aHUS serum ( Figures 7A, B ). Similar effects were achieved by C3d-targeted FH_1-5, at both 100 and 1000 nM ( Figures 7A, B ). The inhibitory effect was comparable to that achieved by the pan-complement inhibitor sCR1 as well as by the potent inhibitor of the complement terminal pathway, eculizumab.