Introduction: C3 is central for all complement activation pathways, thus making it an attractive therapeutic target. Many C3-targeted agents are under extensive development with one already approved for clinical use. However, most, if not all, C3 inhibitors are human or nonhuman primate C3-specific, making evaluating their efficacies in vivo before a clinical trial extremely difficult and costly. Methods: We first studied the compatibility of human C3 in the rat complement system, then developed a C3 humanized rat using the CRISPR/Cas9 technology. We thoroughly characterized the resultant human C3 humanized rats and tested the treatment efficacy of an established primate-specific C3 inhibitor in a model of complement-mediated hemolysis in the C3 humanized rats. Results: We found that supplementing human C3 protein into the C3-deficient rat blood restored its complement activity, which was inhibited by rat factor H or compstatin, suggesting that human C3 is compatible to the rat complement system. The newly developed C3 humanized rats appeared healthy and expressed human but not rat C3 without detectable spontaneous C3 activation. More importantly, complement-mediated hemolysis in the C3 humanized rats was also inhibited by compstatin both in vitro and in vivo. Conclusion: The successfully developed C3 humanized rats provided a much-desired rodent model to evaluate novel C3 inhibitors in vivo as potential drugs.

The complement system is a key effector arm of innate immunity. Its activation not only directly clears invading pathogens but also bridges innate and adaptive immune responses [1]. However, excessive complement activation damages self-tissues, leading to many pathological conditions, such as paroxysmal nocturnal hemoglobinuria (PNH) [2] and age-related macular degeneration (AMD) [3]. Complement component 5 (C5) inhibitors have been developed and are successfully used to treat several complement-mediated diseases [4] as they potently inhibit the assembly of membrane attack complexes (MACs, C5b-9) that directly form “holes” on the cell surface to cause damage. In the case of PNH, MAC formation on patient red blood cells leads to intravascular hemolysis and sequential complications [5]. Besides these MAC-mediated effects, upstream complement activation products C3b/iC3b-mediated extravascular hemolysis is another major pathological mechanism in PNH patients, and C5 inhibitors have no effect on this pathogenic process [6]. Consequently, despite the potent inhibition of MAC formation and prevention of intravascular hemolysis by the C5 inhibitors [7], many PNH patients with the treatment still need frequent blood transfusions to manage their anemia [8].

Complement component 3 (C3) is upstream of C5 in the complement activation cascade. Targeting C3 inhibits the activations of both C3 and the downstream C5, which concurrently prevents both extra- and intravascular hemolysis [9]. Moreover, C3 is the central component of all complement activation pathways, making it an attractive therapeutic target. Indeed, a small, PEGylated peptide (pegcetacoplan) that directly binds to C3 to inhibit its activation by C3 convertases has been developed and approved for treating patients with PNH [10]. It effectively prevented both extra- and intravascular hemolysis and significantly reduced the requirement for blood transfusions in the treated patients [10]. This same C3 inhibitor was also recently approved for treating dry AMD [11], demonstrating its potential in treating complement-mediated diseases. However, this C3 inhibitor is not optimal, and new C3-targeting reagents with better pharmacokinetics and stronger potency are desired. Since most C3 inhibitors, including pegcetacoplan [12], only recognize human or nonhuman primate C3, testing and evaluating these inhibitors in vivo before a clinical trial are difficult and costly.

In attempting to address this issue, a C3 humanized mouse was developed by replacing the mouse C3 gene with the equivalent human C3 gene [13]. However, the resultant C3 humanized mice developed renal complications due to spontaneously augmented complement activation and died several months after birth [13]. This outcome was potentially due to a dysregulated interaction of the human C3 protein with multiple mouse complement proteins, and the failure of mouse factor H, the most potent complement alternative pathway activation inhibitor in the blood [14], to control the mouse-human hybrid complement system. Since the rat complement system is much closer to human than the mouse complement system [15], we hypothesized that a C3 humanized rat could be developed to test novel C3 inhibitors in vivo instead of using nonhuman primates before a clinical trial.

In this project, we first studied the compatibility of human C3 in the rat complement system using the C3 knockout (KO) rats that we previously developed [16, 17]. We then generated a human C3 knock-in (hC3 KI) rat using CRISPR/Cas9 technology that expressed human but not rat C3. The resultant hC3 KI rats were healthy without detectable spontaneous C3 activation. The male rats possessed complement activities that were inhibited by a primate-specific C3 inhibitor both in vitro and in vivo, suggesting that these new rats could serve as a much-needed animal model for the development of novel C3 inhibitors as new therapeutics.

Ethics Statement

This study protocol was reviewed and approved by the Institutional Animal Care and Use Committees of the University of Michigan and Cleveland Clinic (Approval #2398&2557).

Generation of Human C3 KI Rats

Cas9 was used to induce a double-strand chromosome break near the rat C3 initiation codon. Then the human C3 cDNA sequence was inserted into the exon coding for the start of rat C3 DNA via the DNA donor. In the fertilized rat egg, homology-directed repair between the DNA donor and the chromosome inserted the human C3 cDNA sequence into rat C3 exon 1 so that rats will express human C3 in place of rat C3. Briefly, a ribonucleoprotein complex composed of 60 ng/μL single guide RNA (sgRNA, Synthego, CA, USA) that targets 5′ TTA​CCA​TGG​GAC​CCA​CGT​CA (PAM:GGG) 3′, 50 ng/μL wild-type (WT) Cas9 protein (Sigma-Aldrich, MO, USA), and 10 ng/uL of circular plasmid DNA donor were microinjected [18] into fertilized eggs collected from inbred Lewis rats (LEW/Crl, Charles River Laboratory, MA, USA). The rat genomic DNA used for homology-directed repair that included 1,905 bp immediately upstream of the rat C3 initiator methionine (ATG) and 1,987 bp downstream of the ATG was synthesized by GenScript (NJ). AsiSI and AgeI restriction sites were placed between the upstream and downstream arms of homology. The human C3 cDNA (catalog no. RC215069, OriGene, MD, USA) was cloned into the restriction sites. Genotyping to identify correctly targeted founder rats used PCR primer pairs outside of the arms of homology and within the C3 cDNA sequence. Expression of rat C3 was prevented by the insertion of human cDNA and a potent human growth hormone polyadenylation sequence after the human cDNA. The C3 mosaic founder rats were bred to WT Lewis rats, and the pups were genotyped to identify human C3 KI rats by targeted amplicon sequencing.

Evaluation of Human C3 Transcripts in Human C3 KI Rats

To confirm C3 gene expression in hC3 KI, RNA was extracted from homogenized hC3 KI rat liver, spleen, and retina with RNeasy mini kit (Qiagen, Germany), and the genomic DNA was removed via DNase I (Thermo Fisher, MA, USA). Then the RNA was used for cDNA synthesis via SuperScript II (Thermo Fisher). One set of human C3 primers was designed and used to amplify human C3 transcripts in hC3 KI rats to examine human C3 expression via gel electrophoresis.

Forward (5′-->3′)Reverse (5′-->3′)
hC3 CCGAGCCGTTCTCTACAATTAC AGGTGGGATGTCCTCTTTCT 
Forward (5′-->3′)Reverse (5′-->3′)
hC3 CCGAGCCGTTCTCTACAATTAC AGGTGGGATGTCCTCTTTCT 

Evaluation of Human and Rat C3 Proteins in the Human C3 KI Rats

To demonstrate that human C3 was expressed at the protein level and that rat C3 was knocked out in the hC3 KI rats, 50 ng human C3 or rat C3 (Complement Tech, TX, USA) were first boiled in 4 × Laemmli sample buffer (Bio-Rad, CA, USA) for 5 min and loaded on to a 4–20% gradient page gel (GenScript, NJ, USA). The proteins were detected by anti-human complement C3 goat IgG fraction (MP Biomedicals, OH, USA), followed by HRP-conjugated donkey anti-goat IgG (H + L) polyclonal antibody (Southern Biotech, AL, USA).

Next, 10 μL of 100-fold diluted hC3 KI rat plasma, normal human plasma, and WT rat plasma were examined under the same condition as described above. 100-fold diluted human C3-depleted serum (Complement Tech) and C3 KO rat plasma [17] were used as negative controls.

Measurement of Human C3 Levels in the Plasma and Ocular Fluids of hC3 KI Rats

To assess human C3 levels in hC3 KI rats, a C3 ELISA kit that only detects human C3, but not rat C3, was used per the manufacturer’s manual (Hycult Biotech, PA, USA). Briefly, 1:1,000 diluted hC3 KI rat plasma or WT rat plasma or 1:50,000 diluted human plasma was added into the pre-coated plate for 1 h at room temperature. For human C3 measurements in the ocular fluids, vitreous and aqueous humor collected from the hC3 KI rats were diluted 10-fold before being added into the same ELISA plates. Followed by three times washing, a tracer antibody was added for 1 h at room temperature. The binding signal was developed by TMB. A standard curve was generated using purified human C3.

Complement Hemolytic Assays

To assess the compatibility of human C3 in rat complement system, 0–100 μg/mL human C3 (Complement Tech) was added into rabbit red blood cells (Erabb), followed by incubating with 20% C3 KO rat serum in the presence of 0.5 mm Mg2+ and 10 mm EGTA GVB0 buffer (Mg-EGTA GVB) at 37°C for 20 min. The background of complement-independent hemolysis was determined by incubating 10 mm EDTA with 20% serum and the same amount of Erabb. To determine the hemolysis, 200 μL of GVBE (GVB0 + 10 mm EDTA) were added after the reaction, the samples were centrifuged, and the OD414 of the supernatant was read by a spectrophotometer (Molecular Devices, CA, USA). The percentage of hemolysis (hemolysis%) was calculated using the following equation: Hemolysis rate (%) = [(A-B)/(C-B)] × 100%. A = OD414 reading of sample in Mg-EGTA GVB, B = OD414 reading of sample with 10 mm EDTA, C = OD414 reading of maximum hemolysis induced by H2O.

To examine the protective role of rat factor H in the hybrid human C3 rat complement system, 0–3.4 μm rat factor H (Complement Tech) or human factor H (Complement Tech) were added into Erabb in the presence of 5% C3 KO rat serum supplemented with 125 μg/mL human C3. To determine the effect of compstatin [12] in the hybrid human-rat complement system, 0–50 µm compstatin (Tocris Bioscience, Bristol, UK) was added into Erabb in the presence of 20% C3 KO rat serum supplemented with 50 μg/mL human C3.

To evaluate classical pathway activities of the hC3 KI rats, 80% hC3 KI rat serum or WT rat serum or normal human serum were incubated with antibody-sensitized sheep red blood cells (EShA) in GVB++ (GVB0 + 0.15 mm Ca2+ + 0.5 mm Mg2+) for 20 min at 37°C. To evaluate the alternative pathway activities of the hC3 KI rat, 80% hC3 KI rat serum or WT rat serum or normal human serum were incubated with Erabb in Mg-EGTA GVB buffer for 1 h at 37°C. The hemolysis was determined by OD414 readings, and the hemolysis% was calculated as described above.

To evaluate the inhibitory effect of compstatin in hC3 KI rat serum, 50% hC3 KI rat serum, 10% normal human serum, or 20% WT rat serum were incubated with Erabb in the presence of 0–5 µM compstatin in Mg-EGTA GVB buffer for 1 h at 37°C.

In vivo Complement Activity Studies of hC3 KI Rats

2–3 × 109 rabbit red blood cells (Erabb) were intravenously injected into C3 KO rats supplemented with 500 μg human C3 or into human C3 KI rats. The rats were euthanized after 30 min (for C3 KO rats supplemented with human C3) or 60 min (for hC3 KI rats), and blood was collected by cardiac puncture. Levels of released hemoglobin in the serum were evaluated via measuring OD414.

Evaluation of Spontaneous Complement Activation in the hC3 KI Rats

To assess spontaneous complement activation in the naïve hC3 KI rats, human C3a levels were evaluated via BD cytometric bead array (BD Biosciences, NJ, USA), per manufacturer’s instruction. Briefly, 1:1,000 diluted hC3 KI plasma was incubated with capture beads for 2 h at room temperature. After 1 mL wash buffer, human anaphylatoxin PE detection reagent was added for 1 h. After washing, the beads were evaluated via flow cytometry. Cobra venom factor (Complement Tech) was added in hC3 KI rat serum to activate the complement system for 30 min at 37°C in GVB++, and C3a levels were assessed as described above.

Renal Function Analysis Using Urine Samples Collected from the hC3 KI and WT Rats

WT and hC3 KI rat urine was collected at 10 and 11 weeks of age, and the urine levels of creatinine, albumin, and total nitrogen were evaluated using an automatic Roche Cobas 6000 analyzer following manufacturer-provided protocols.

Human C3 Restores C3 KO Rat Complement Activity in vitro

Human C3 protein shares approximately 78% similarity with rat C3 protein; therefore, it might be compatible with the rat complement system. Our previously developed C3 KO rats [16, 17] enabled us to test whether human C3 is functionally compatible with the rat complement system for the first time. In brief, we added 50 or 100 μg/mL of purified human C3 protein to C3 KO rat serum and measured the resulting complement alternative pathway activity in a conventional hemolytic assay. The results demonstrated that human C3 protein restored the complement activity of the C3-deficient rat serum (Fig. 1a), providing the first evidence indicating that human C3 is compatible with the rest of the rat complement system.

Fig. 1.

Human C3 restores C3 KO rat complement activity that is controlled by rat factor H and compstatin, a primate-specific C3 inhibitor. a 1 × 105 rabbit red blood cells (Erabb) were incubated with 20% wild-type (WT) or C3 KO serum, supplemented with 50 or 100 μg/mL human C3 in Mg-EGTA GVB buffer at 37°C for 20 min. The hemolysis% was evaluated and calculated as described in the Methods. b 0–3.4 μm of human or rat factor H were added into 5% C3 KO rat serum that was supplemented with 125 μg/mL human C3 together with Erabb in the Mg-EGTA GVB buffer. Hemolysis was evaluated and calculated by OD414 readings after 1 h of incubation at 37°C. c 0–50 µm of compstatin were added into 20% C3 KO rat serum supplemented with 50 μg/mL of human C3 together with Erabb in the Mg-EGTA GVB buffer at 37°C for 20 min. 20% WT rat serum was used as a positive control, and 20% C3 KO rat serum was used as a negative control. d Erabb were intravenously injected into C3 KO rats that were supplemented with 500 μg human C3 each. The rats were euthanized after 30 min, and blood was collected by cardiac puncture. Levels of free hemoglobin in plasma were evaluated by measuring OD414. **p < 0.01.

Fig. 1.

Human C3 restores C3 KO rat complement activity that is controlled by rat factor H and compstatin, a primate-specific C3 inhibitor. a 1 × 105 rabbit red blood cells (Erabb) were incubated with 20% wild-type (WT) or C3 KO serum, supplemented with 50 or 100 μg/mL human C3 in Mg-EGTA GVB buffer at 37°C for 20 min. The hemolysis% was evaluated and calculated as described in the Methods. b 0–3.4 μm of human or rat factor H were added into 5% C3 KO rat serum that was supplemented with 125 μg/mL human C3 together with Erabb in the Mg-EGTA GVB buffer. Hemolysis was evaluated and calculated by OD414 readings after 1 h of incubation at 37°C. c 0–50 µm of compstatin were added into 20% C3 KO rat serum supplemented with 50 μg/mL of human C3 together with Erabb in the Mg-EGTA GVB buffer at 37°C for 20 min. 20% WT rat serum was used as a positive control, and 20% C3 KO rat serum was used as a negative control. d Erabb were intravenously injected into C3 KO rats that were supplemented with 500 μg human C3 each. The rats were euthanized after 30 min, and blood was collected by cardiac puncture. Levels of free hemoglobin in plasma were evaluated by measuring OD414. **p < 0.01.

Close modal

Activation of the Human-Rat Hybrid Complement System Is Controlled by Rat Factor H

The C3 humanized mice have spontaneous complement activation potentially due to the failure of mouse factor H to control the human-mouse hybrid complement system [13]. To examine whether rat factor H could regulate the human-rat hybrid complement system, we did an alternative pathway-mediated hemolytic assay using human C3-supplemented C3 KO rat serum in the presence of 0–3.4 μm of purified rat or human factor H. We found that both the rat and human factor H inhibited the hybrid complement-mediated hemolysis in a concentration-dependent manner, even though human factor H worked ∼5-fold better than rat factor H (Fig. 1b). These results indicate that C3 humanized rats should not have the same severe issues reported in the C3 humanized mice [13].

Complement Activity of the Human C3-Supplemented C3 KO Rat Serum Is Inhibited by Compstatin

The above-described results suggested that this hybrid complement system could be used to evaluate human C3-specific complement inhibitors. To test this hypothesis, we supplemented C3 KO rat serum with 50 μg/mL of purified human C3 protein, then measured its hemolytic activity in the presence of 0, 10, or 50 μm of compstatin, an established primate-specific C3 inhibitor that could previously only be tested in human or nonhuman primates [12]. The result showed that compstatin inhibited the human-rat hybrid complement system-mediated hemolysis in a concentration-dependent manner (Fig. 1c).

Human C3 Restores C3 KO Rat Complement Activity in vivo

Encouraged by the positive results from the above in vitro experiments, we injected purified human C3 protein (500 μg) into a C3 KO rat through the tail vein and then evaluated its complement activity in vivo using a hemolytic model described in our previous work [17, 19]. The result showed that in contrast to our previous observation that rabbit red blood cells (Erabb) were protected in the C3 KO rats, the infused Erabb were destroyed rapidly in the C3 KO rats supplemented with human C3, as indicated by the increased levels of free hemoglobin in the plasma (Fig. 1d). This result thus far provided a strong rationale to develop a C3 humanized rat, which could be valuable to evaluate the treatment efficacies and pharmacodynamics/pharmacokinetics of human C3-specific inhibitors in vivo.

Development of a C3 Humanized Rat by Knocking in Human C3 cDNA into Rat C3 Gene Using CRISPR/Cas9

We designed a vector containing the human C3 cDNA flanked by homologous arms of the rat C3 gene so that the human C3 cDNA can be knocked into the exon 1 of the rat C3 gene while knocking out the rat C3 at the same time. We identified a founder rat with the correct KI event by PCR and sequencing. This founder rat was bred with a WT rat to generate F1 rats with germline-transmission of the targeted rat C3 gene, and then the F1s were bred with each other to generate homozygous hC3 KI rats, as determined by PCR genotyping.

hC3 KI Rats Express Human but Not Rat C3 and Have No Detectable Spontaneous C3 Activation

To demonstrate that the novel hC3 KI rats express human but not rat C3, we first tested for human C3 cDNA transcripts by RT-PCR using RNA prepared from livers, spleen, and retina of WT and hC3 KI rats after complete DNase I digestion to remove any contaminating genomic DNA. These experiments show that human C3 transcripts are only detectable in the tissues from hC3 KI rats but not WT rats (Fig. 2a–c).

Fig. 2.

Human C3 transcripts are detectable in multiple tissues from the hC3 KI rats. Total RNAs were isolated from the liver (a), spleen (b), and retina (c) tissues of male human C3 knock-in (hC3 KI) rats and DNase I-treated. Human C3 transcripts were amplified by PCR using cDNAs reverse-transcribed from these total RNAs as templates (cDNA). The same RNA samples after DNase I treatment but without reverse transcription were used as negative controls (RNA).

Fig. 2.

Human C3 transcripts are detectable in multiple tissues from the hC3 KI rats. Total RNAs were isolated from the liver (a), spleen (b), and retina (c) tissues of male human C3 knock-in (hC3 KI) rats and DNase I-treated. Human C3 transcripts were amplified by PCR using cDNAs reverse-transcribed from these total RNAs as templates (cDNA). The same RNA samples after DNase I treatment but without reverse transcription were used as negative controls (RNA).

Close modal

To confirm these RNA-level results at the protein level, we probed purified human and rat C3 proteins using a polyclonal anti-human C3 antibody in Western blots. These experiments show that the antibody recognizes the beta chains of both human and rat C3 (Fig. 3a). Consistent with the predicted molecular weights based on the amino acid sequences, the rat C3 beta chain is slightly smaller than the human C3 beta chain, making it feasible to distinguish rat C3 from human C3 in this assay.

Fig. 3.

hC3 KI rats expressed human but not rat C3 and has no detectable spontaneous C3 activation. a 50 ng purified human C3 (hC3) or rat C3 (rC3) were boiled and loaded onto an SDS-PAGE gel, then detected by a polyclonal goat anti-human C3 IgG. The rat C3 beta chain showed a smaller molecular weight than its human counterpart. Using the same Western blot assay, human C3 proteins were detected in samples only from normal human serum (human) and hC3 KI rats (hC3KI). Rat C3 proteins were only detected in samples from WT rats (WT). Human C3-depleted serum (hC3-dpl) and C3 KO rat plasma (C3KO) were used as negative controls for plasma samples. b Human C3 levels in the plasma of both human C3 knock-in (hC3 KI) male (M) and female (F) rats were examined using a commercially available ELISA kit that specifically recognizes human C3 but not rat C3, per the manufacturer’s instruction. WT rat plasma (WT) was diluted at the same condition as hC3 KI rats and served as the negative control. Normal human plasma (human) was diluted for an additional 1:50 fold as compared to hC3 KI plasma and served as the positive control. c Human C3 levels in the vitreous and aqueous humor samples of male WT and hC3 KI male rats were detected using the same human C3 ELISA kit. d C3a levels in hC3 KI male rats were evaluated using a cytometric bead array-based assay. Cobra venom factor (CVF)-activated hC3 KI rat serum was used as positive controls. **p < 0.01.

Fig. 3.

hC3 KI rats expressed human but not rat C3 and has no detectable spontaneous C3 activation. a 50 ng purified human C3 (hC3) or rat C3 (rC3) were boiled and loaded onto an SDS-PAGE gel, then detected by a polyclonal goat anti-human C3 IgG. The rat C3 beta chain showed a smaller molecular weight than its human counterpart. Using the same Western blot assay, human C3 proteins were detected in samples only from normal human serum (human) and hC3 KI rats (hC3KI). Rat C3 proteins were only detected in samples from WT rats (WT). Human C3-depleted serum (hC3-dpl) and C3 KO rat plasma (C3KO) were used as negative controls for plasma samples. b Human C3 levels in the plasma of both human C3 knock-in (hC3 KI) male (M) and female (F) rats were examined using a commercially available ELISA kit that specifically recognizes human C3 but not rat C3, per the manufacturer’s instruction. WT rat plasma (WT) was diluted at the same condition as hC3 KI rats and served as the negative control. Normal human plasma (human) was diluted for an additional 1:50 fold as compared to hC3 KI plasma and served as the positive control. c Human C3 levels in the vitreous and aqueous humor samples of male WT and hC3 KI male rats were detected using the same human C3 ELISA kit. d C3a levels in hC3 KI male rats were evaluated using a cytometric bead array-based assay. Cobra venom factor (CVF)-activated hC3 KI rat serum was used as positive controls. **p < 0.01.

Close modal

We repeated the assay with plasma samples from hC3 KI rats and WT rats, using C3 KO rat plasma, normal human plasma, and C3-depleted human serum as controls. These Western blots showed that while no C3 protein was detectable in either C3 KO rat plasma or C3-depleted human serum, as expected, human C3 protein was detectable in normal human plasma and hC3 KI rat plasma, while rat C3 protein was only detectable in WT rat plasma (Fig. 3a).

We then measured human C3 concentrations in the hC3 KI rat plasma using an established ELISA. These studies show that while the male KI rats had 50 ± 9 μg/mL of human C3 in the plasma, the females had only 1.4 ± 0.3 μg/mL (Fig. 3b). In light of the low human C3 levels in the female KI rats, we only studied male KI rats in subsequent experiments.

Given the interests of local C3-targeted therapies for ocular diseases, we also collected vitreous and aqueous humor samples from male WT and hC3 KI rats, then measured their human C3 levels using the same ELISA kit. These studies showed that again, while there was no detectable human C3 in these local ocular fluids from the WT rats, human C3 concentrations in the vitreous and aqueous humors from male C3 KI rats were 215 ± 26 and 182 ± 62 ng/mL, respectively (Fig. 3c). hC3 KI rats appear to be healthy with no detectable spontaneous systemic C3 activation and no detectable sign of renal dysfunction at 11 weeks of age.

Human C3 KI mice showed spontaneous C3 activation, developed renal problems before 11 weeks of age, and have a median survival age of 16 weeks [13]. The hC3 KI rats appear normal and healthy with the oldest being over 52 weeks old to date. To evaluate spontaneous C3 activation in hC3 KI rats, we measured their human C3a levels in the serum before and after incubation with complement activator cobra venom factor (CVF) using a cytometric bead array-based assay. These measurements show that while significant levels of human C3a are detectable after CVF incubation with hC3 KI serum, there is no detectable C3a in the same plasma samples from the naïve hC3 KI rats (Fig. 2d). When urine samples from 11-week-old male C3 KI rats were analyzed for urine albumin, nitrogen, and creatine, there was no sign of renal dysfunction at all (Table 1).

Table 1.

Biochemical analyses of urine samples from WT and human C3 KI rats

SampleGenderAge, weeksSample typeUrea nitrogen, mg/dLCreatinine, mg/dLMicroalbumin, mg/L
WT 10 Rat urine 2,422 91 <12 
WT 10 Rat urine 3,151 113 <12 
hC3 KI 11 Rat urine 2,438 107 <12 
hC3 KI 11 Rat urine 1,129 44 <12 
hC3 KI 11 Rat urine 3,027 131 <12 
hC3 KI 11 Rat urine 532 19 <12 
hC3 KI 11 Rat urine 1,180 49 <12 
SampleGenderAge, weeksSample typeUrea nitrogen, mg/dLCreatinine, mg/dLMicroalbumin, mg/L
WT 10 Rat urine 2,422 91 <12 
WT 10 Rat urine 3,151 113 <12 
hC3 KI 11 Rat urine 2,438 107 <12 
hC3 KI 11 Rat urine 1,129 44 <12 
hC3 KI 11 Rat urine 3,027 131 <12 
hC3 KI 11 Rat urine 532 19 <12 
hC3 KI 11 Rat urine 1,180 49 <12 

Biochemical analyses of rat urine samples. Urine samples from WT and hC3 KI rats were collected and analyzed for urine levels of albumin, nitrogen, and creatinine, using an automatic biochemical analyzer. No signs of renal dysfunction were detected.

hC3 KI Rat Serum Possesses Complement Activities

To demonstrate that the developed hC3 KI rats possess complement activity, we carried out both classical and alternative pathway-mediated hemolysis assays using different concentrations of their serum samples together with WT rat serum, C3 KO rat serum, and normal human serum as controls. These studies showed that at 5% and 30% concentrations tested, hC3 KI rats exhibit reduced classical pathway-mediated hemolytic activity; however, at 80% serum concentration, hC3 KI rat serum exhibits comparable classical pathway-mediated hemolytic activity as WT rat serum and normal human serum. In the alternative pathway assays, hC3 KI rats exhibit ∼30% lower hemolytic activities than the two controls (Fig. 4a, b).

Fig. 4.

hC3 KI rats process complement activity in vitro and in vivo that can be inhibited by compstatin. a 5-80% hC3 KI male rat serum or WT rat serum (WT rat) or normal human serum (human) were incubated with antibody-sensitized sheep red blood cells (EShA) at 37°C for 20 min in GVB++ buffer. The hemolysis% was determined by OD414 readings. b 5–80% hC3 KI male rat serum or WT rat serum or normal human serum or C3 KO rat serum were incubated with Erabb for 1 h in Mg-EGTA GVB buffer. The hemolysis was evaluated by OD414 readings. c 0–5 µM compstatin were incubated with Erabb mixed with 50% hC3 KI male rat serum, or 20% WT rat serum, or 10% normal human serum in Mg-EGTA GVB buffer at 37°C for 1 h. The hemolysis% was determined by OD414 readings. d 2 × 109 Erabb were injected intravenously into each hC3 KI male rats, with or without immediate treatment of 20 μm compstatin (Comp). The rats were euthanized after 60 min, and blood was collected by cardiac puncture. Levels of released hemoglobin in serum were evaluated via measuring OD414. The baseline reading of rat serum was evaluated by collecting WT naïve rat serum (without Erabb infusion) via cardiac puncture after euthanasia (naïve WT). *p < 0.05, **p < 0.01; ns, no significance.

Fig. 4.

hC3 KI rats process complement activity in vitro and in vivo that can be inhibited by compstatin. a 5-80% hC3 KI male rat serum or WT rat serum (WT rat) or normal human serum (human) were incubated with antibody-sensitized sheep red blood cells (EShA) at 37°C for 20 min in GVB++ buffer. The hemolysis% was determined by OD414 readings. b 5–80% hC3 KI male rat serum or WT rat serum or normal human serum or C3 KO rat serum were incubated with Erabb for 1 h in Mg-EGTA GVB buffer. The hemolysis was evaluated by OD414 readings. c 0–5 µM compstatin were incubated with Erabb mixed with 50% hC3 KI male rat serum, or 20% WT rat serum, or 10% normal human serum in Mg-EGTA GVB buffer at 37°C for 1 h. The hemolysis% was determined by OD414 readings. d 2 × 109 Erabb were injected intravenously into each hC3 KI male rats, with or without immediate treatment of 20 μm compstatin (Comp). The rats were euthanized after 60 min, and blood was collected by cardiac puncture. Levels of released hemoglobin in serum were evaluated via measuring OD414. The baseline reading of rat serum was evaluated by collecting WT naïve rat serum (without Erabb infusion) via cardiac puncture after euthanasia (naïve WT). *p < 0.05, **p < 0.01; ns, no significance.

Close modal

Compstatin Treatment Inhibits Complement-Mediated Hemolysis within the hC3 KI Rats in vivo

To demonstrate that the hC3 KI rats can be used for evaluating human C3 inhibitors in vivo, we tested compstatin in a complement alternative pathway-mediated hemolysis assay in vitro using hC3 KI rat serum, normal human serum, or WT rat serum as the source of complement. This assay showed that compstatin inhibited complement-mediated hemolysis in a concentration-dependent manner when hC3 KI rat serum and normal human serum were used but had no effect on rat complement-mediated hemolysis, as expected (Fig. 4c).

These positive in vitro results served as the rationale for conducting an in vivo test with an established complement-mediated hemolysis model by infusing rabbit red blood cells (Erabb) into hC3 KI rats, then treating them with either 2 mg/kg of compstatin or the same volumes of vehicle (PBS). We bled the rats 60 min later and evaluated the extent of hemolysis by measuring levels of free hemoglobin in the plasma. We found that while the mock-treated hC3 KI rats infused with Erabb showed elevated levels of free plasma hemoglobin as compared with naïve WT rats, a clear sign of intravascular hemolysis, compstatin-treated hC3 KI rats with the same Erabb infusion exhibited comparable plasma levels of free hemoglobin to those in the naïve WT rats (Fig. 4d).

In this report, using the available C3 KO rats, we found that human C3 supplementation restored the hemolytic activity of C3 KO rat serum and that rat factor H controlled the new human-rat hybrid complement system. We then developed a novel C3 humanized rat by knocking in human C3 cDNA into the first exon of rat C3 gene with CRISPR/Cas9 technology. In this way, the endogenous rat C3 promoter/enhancers are expected to govern human C3 expression while the rat C3 expression is ablated. We found that the resulting rats were healthy without spontaneous C3 activation and expressed human but not rat C3. These rats also possessed complement hemolytic activities that can be inhibited both in vitro and in vivo by compstatin, a C3 inhibitor known to be specific for primate C3. These data suggest that the hC3 KI rats can serve as an in vivo tool for the development of new C3 inhibitors instead of nonhuman primates before a clinical trial.

Animal models are critical for the development of new drugs. Before a clinical trial, drug candidates are usually required to be tested in animals to demonstrate their in vivo treatment efficacy, in addition to other studies. However, efficacy studies require that the drug candidates cross-react with their target counterparts in animals. In the case of C3 inhibitors, because most, if not all, of the known C3 inhibitors are primate C3-specific and do not cross-react with C3 in other animal models, nonhuman primates are the only option to test these drug candidates in vivo before a clinical trial. But drug development studies involving nonhuman primates have many serious issues such as cost, time, and ethical concerns, making the task extremely difficult and hindering the development of novel C3 inhibitors.

Developing a humanized mouse by genetically replacing the mouse target gene with its human counterpart has been employed to evaluate drug candidates specific only for human targets with some success. But in the case of complement humanized animals, it has been reported that C5 humanized mice are essentially C5 KO mice [20], potentially due to the reason that mouse C5 convertases cannot cleave human C5, indicating that human C5 is not compatible with the mouse complement system. A C3 humanized mouse model was recently developed by genetically manipulating mouse embryonic stem cells to replace the mouse C3 gene with its human counterpart [13]. However, chronic and spontaneous alternative pathway of complement activation occurred in these mice, and consequently, the C3 humanized mice developed severe renal complications at young age and died prematurely. The usefulness of these C3 humanized mice is thus limited.

Rats are physiologically closer to humans than mice in general and thus are better for modeling human diseases. The rat complement system is also much closer to its human counterpart than the mouse complement system and has comparable hemolytic activities as the human complement system [15]. Despite these advantages, due to rat genome editing technology limitations prior to the advent of CRISPR/Cas9 technology, rat complement models were scarce with only one spontaneously occurring C6-deficient rat model commonly used in complement-related studies [21]. With the recent development of CRISPR/Cas9 technology, the rat genome can be efficiently manipulated at a reasonable cost. Indeed, taking advantage of this technology, rats deficient in factor B [22], an essential component of the alternative pathway of complement activation, and in CD59 [23], a critical cell surface complement inhibitor, were developed. We also successfully developed the first C3 KO rats using the CRISPR/Cas9 technology [16, 17] and used them to study the role of complement in chemotherapy-induced periphery neuropathy and the potential therapeutic values of C3 in diseases involving different complement activation pathways.

Our previously developed C3 KO rats gave us a unique opportunity to examine whether human C3 is compatible with the rat complement system. By supplementing purified human C3 into C3 KO rat serum in vitro and C3 KO rat serum in vivo, we demonstrated that human C3 is compatible with the rat complement system (Fig. 1a). More importantly, using purified rat factor H, and human C3-supplemented C3 KO rat serum, we found that rat factor H inhibited the human-rat hybrid complement-mediated hemolysis in a concentration-dependent manner, but with a reduced efficacy compared with human factor H (Fig. 1b). However, possibly due to the low expression levels of human C3 in the hC3 KI rats, native rat factor H appeared to be sufficient to control the potential spontaneous complement activation in vivo even with its reduced efficacy.

Indeed, our novel hC3 KI rats are healthy and fertile, as predicted by the studies using C3 KO rats and purified human C3. Human C3 transcripts can be detected in multiple tissues besides livers such as spleen and the retina (Fig. 2). Ocular fluids from the C3 KI rats also contain human C3, suggesting that these rats could be used to test local human C3-targeted therapeutics for ocular diseases in which complement is integrally involved in the pathogenesis, e.g., AMD and diabetic retinopathy. But human C3 plasma levels in these KI rats are low, especially in female rats, as measured by ELISA (Fig. 3b). These lower-than-expected serum C3 levels in the hC3 KI rats were not due to spontaneous complement activation-mediated C3 consumption because we found that rat factor H was able to control the activation of this hybrid complement system (Fig. 1b), and more importantly, we did not detect any elevated serum human C3a levels in the naïve hC3 KI rats (Fig. 3d). Therefore, the low serum levels of human C3 in these rats were likely due to the human C3 cDNA used in the expression construct, as previous studies by others showed that the use of cDNA could result in low expression levels in vivo. For example, attempts to express human apolipoprotein B (APOB) from cDNA-based transgenes regulated by exogenous promoters in transgenic mice were inefficient and only one of 29 transgenic founders expressed detectable levels of APOB [24]. We expected that by introducing the human C3 DNA into the context of the complete rat genome in the future might improve the expression levels in rats. The unexpected low levels of hC3 in female rats may provide an opportunity to study hC3 regulatory elements that control gene expression. Nevertheless, the complement hemolytic activities of the male hC3 KI rat serum were comparable to those of normal humans or WT male rats when tested at a higher concentration (80% serum) (Fig. 4a, b). When rabbit red blood cells were infused into the male hC3 KI rats, complement-mediated intravascular hemolysis occurred rapidly, and it was controlled by compstatin, the C3 inhibitor known to be selective for primate C3 (Fig. 4d). All these data suggest that the male hC3 KI rats are suitable for evaluating novel primate (human) C3 inhibitors in vivo.

In summary, using our previously developed C3 KO rats, we found that human C3 was compatible with the rat complement system. We thus developed an hC3 KI rat that expressed human but not rat C3. These hC3 KI rats were healthy without detectable spontaneous complement activation and renal dysfunction as reported in the C3 humanized mice [13]. They also possessed complement hemolytic activities both in vitro and in vivo. More importantly, compstatin, a primate-specific C3 inhibitor, efficiently reduced intravascular hemolysis in these rats in a model of complement-mediated hemolysis. All these data, taken together, demonstrated the successful development of the new C3 humanized rat and provided the proof-of-concept that the male hC3 KI rats can be used for evaluating novel C3 inhibitors in vivo as potential therapeutics.

All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committees of the University of Michigan and Cleveland Clinic (Approval #2398&2557).

The authors declare no conflicts of interest.

This project is supported in part by the National Institute of Health grant EY032458 (F.L.).

J.Y.C., L.Z., M.Y., and E.D.H. carried out the experiments, analyzed the data, and edited the manuscript; Z.F. and T.L.S. designed the experiments, analyzed the data, and edited the manuscript; F.L. conceptualized the study, designed the experiments, analyzed the results, and prepared the manuscript.

All the data generated or analyzed during the study are included in this article. Further inquiries can be directed to the corresponding author.

1.
Merle
NS
,
Church
SE
,
Fremeaux-Bacchi
V
,
Roumenina
LT
.
Complement system part i: molecular mechanisms of activation and regulation
.
Front Immunol
.
2015
;
6
:
262
.
2.
Rosse
WF
.
Dr Ham's test revisited [editorial]
.
Blood
.
1991
;
78
(
3
):
547
50
.
3.
Klein
RJ
,
Zeiss
C
,
Chew
EY
,
Tsai
JY
,
Sackler
RS
,
Haynes
C
et al
.
Complement factor h polymorphism in age-related macular degeneration
.
Science
.
2005
;
308
(
5720
):
385
9
.
4.
Patriquin
CJ
,
Kuo
KHM
.
Eculizumab and beyond: the past, present, and future of complement therapeutics
.
Transfus Med Rev
.
2019
;
33
(
4
):
256
65
.
5.
Rother
RP
,
Rollins
SA
,
Mojcik
CF
,
Brodsky
RA
,
Bell
L
.
Discovery and development of the complement inhibitor eculizumab for the treatment of paroxysmal nocturnal hemoglobinuria
.
Nat Biotechnol
.
2007
;
25
(
11
):
1256
64
.
6.
Hill
A
,
Rother
RP
,
Arnold
L
,
Kelly
R
,
Cullen
MJ
,
Richards
SJ
et al
.
Eculizumab prevents intravascular hemolysis in patients with paroxysmal nocturnal hemoglobinuria and unmasks low-level extravascular hemolysis occurring through c3 opsonization
.
Haematologica
.
2010
;
95
(
4
):
567
73
.
7.
Hillmen
P
,
Young
NS
,
Schubert
J
,
Brodsky
RA
,
Socié
G
,
Muus
P
et al
.
The complement inhibitor eculizumab in paroxysmal nocturnal hemoglobinuria
.
N Engl J Med
.
2006
;
355
(
12
):
1233
43
.
8.
Harder
MJ
,
Höchsmann
B
,
Dopler
A
,
Anliker
M
,
Weinstock
C
,
Skerra
A
et al
.
Different levels of incomplete terminal pathway inhibition by eculizumab and the clinical response of pnh patients
.
Front Immunol
.
2019
;
10
:
1639
.
9.
Baas
I
,
Delvasto-Nuñez
L
,
Ligthart
P
,
Brouwer
C
,
Folman
C
,
Reis
ES
et al
.
Complement c3 inhibition by compstatin cp40 prevents intra- and extravascular hemolysis of red blood cells
.
Haematologica
.
2020
;
105
(
2
):
e57
e60
.
10.
Hillmen
P
,
Szer
J
,
Weitz
I
,
Röth
A
,
Höchsmann
B
,
Panse
J
et al
.
Pegcetacoplan versus eculizumab in paroxysmal nocturnal hemoglobinuria
.
N Engl J Med
.
2021
;
384
(
11
):
1028
37
.
11.
Liao
DS
,
Metlapally
R
,
Joshi
P
.
Pegcetacoplan treatment for geographic atrophy due to age-related macular degeneration: a plain language summary of the filly study
.
Immunotherapy
.
2022
;
14
(
13
):
995
1006
.
12.
Sahu
A
,
Morikis
D
,
Lambris
JD
.
Compstatin, a peptide inhibitor of complement, exhibits species-specific binding to complement component c3
.
Mol Immunol
.
2003
;
39
(
10
):
557
66
.
13.
Devalaraja-Narashimha
K
,
Meagher
K
,
Luo
Y
,
Huang
C
,
Kaplan
T
,
Muthuswamy
A
et al
.
Humanized c3 mouse: a novel accelerated model of c3 glomerulopathy
.
J Am Soc Nephrol
.
2021
;
32
(
1
):
99
114
.
14.
Ferreira
VP
,
Pangburn
MK
,
Cortés
C
.
Complement control protein factor h: the good, the bad, and the inadequate
.
Mol Immunol
.
2010
;
47
(
13
):
2187
97
.
15.
Ong
GL
,
Mattes
MJ
.
Mouse strains with typical mammalian levels of complement activity
.
J Immunol Methods
.
1989
125
1–2
147
58
.
16.
Xu
J
,
Zhang
L
,
Xie
M
,
Li
Y
,
Huang
P
,
Saunders
TL
et al
.
Role of complement in a rat model of paclitaxel-induced peripheral neuropathy
.
J Immunol
.
2018
;
200
(
12
):
4094
101
.
17.
Zhang
L
,
Dai
Y
,
Huang
P
,
Saunders
TL
,
Fox
DA
,
Xu
J
et al
.
Absence of complement component 3 does not prevent classical pathway-mediated hemolysis
.
Blood Adv
.
2019
;
3
(
12
):
1808
14
.
18.
Filipiak
WE
,
Saunders
TL
.
Advances in transgenic rat production
.
Transgenic Res
.
2006
;
15
(
6
):
673
86
.
19.
Lin
K
,
Zhang
L
,
Kong
M
,
Yang
M
,
Chen
Y
,
Poptic
E
et al
.
Development of an anti-human complement c6 monoclonal antibody that inhibits the assembly of membrane attack complexes
.
Blood Adv
.
2020
;
4
(
9
):
2049
57
.
20.
Latuszek
A
,
Hu
Y
,
Cao
J
,
Wiegand
SJ
,
Yancopoulos
GD
.
Evaluation of complement activity in wild type, c5 knockout and humanized mice for drug discovery
.
Invest Ophthalmol Vis Sci
.
2014
;
55
:
1324
–.
21.
Leenaerts
PL
,
Stad
RK
,
Hall
BM
,
Van Damme
BJ
,
Vanrenterghem
Y
,
Daha
MR
.
Hereditary c6 deficiency in a strain of pvg/c rats
.
Clin Exp Immunol
.
1994
;
97
(
3
):
478
82
.
22.
Coan
PM
,
Barrier
M
,
Alfazema
N
,
Carter
RN
,
Marion de Procé
S
,
Dopico
XC
et al
.
Complement factor b is a determinant of both metabolic and cardiovascular features of metabolic syndrome
.
Hypertension
.
2017
;
70
(
3
):
624
33
.
23.
Yao
X
,
Verkman
AS
.
Marked central nervous system pathology in cd59 knockout rats following passive transfer of neuromyelitis optica immunoglobulin g
.
Acta Neuropathol Commun
.
2017
;
5
(
1
):
15
.
24.
Chiesa
G
,
Johnson
DF
,
Yao
Z
,
Innerarity
TL
,
Mahley
RW
,
Young
SG
et al
.
Expression of human apolipoprotein b100 in transgenic mice. Editing of human apolipoprotein b100 mrna
.
J Biol Chem
.
1993
;
268
(
32
):
23747
50
.