A potent truncated form of human soluble CR1 is protective in a mouse model of renal ischemia–reperfusion injury

The complement system is a potent mediator of ischemia–reperfusion injury (IRI), which detrimentally affects the function and survival of transplanted kidneys. Human complement receptor 1 (HuCR1) is an integral membrane protein that inhibits complement activation by blocking the convertases that activate C3 and C5. We have previously reported that CSL040, a truncated form of recombinant soluble HuCR1 (sHuCR1), has enhanced complement inhibitory activity and improved pharmacokinetic properties compared to the parent molecule. Here, we compared the capacity of CSL040 and full-length sHuCR1 to suppress complement-mediated organ damage in a mouse model of warm renal IRI. Mice were treated with two doses of CSL040 or sHuCR1, given 1 h prior to 22 min unilateral renal ischemia and again 3 h later. 24 h after reperfusion, mice treated with CSL040 were protected against warm renal IRI in a dose-dependent manner, with the highest dose of 60 mg/kg significantly reducing renal dysfunction, tubular injury, complement activation, endothelial damage, and leukocyte infiltration. In contrast, treatment with sHuCR1 at a molar equivalent dose to 60 mg/kg CSL040 did not confer significant protection. Our results identify CSL040 as a promising therapeutic candidate to attenuate renal IRI and demonstrate its superior efficacy over full-length sHuCR1 in vivo.

deposition of the complement split products C4d and C3d 15 . In mice, a deficiency in complement components (C3, C5, C6, or Factor B) or C3a and/or C5a receptors protected against renal IRI 13,16,17 , whereas mice deficient in complement regulators were more susceptible 18,19 . Therefore, complement activation at multiple stages contributes to renal IRI. However, despite several attempts in recent years 9,[20][21][22] , the effective clinical therapy of renal IRI by attenuating complement activation has not been reported.
Human CR1 (HuCR1), also known as the C3b/C4b receptor, is a multi-functional polymorphic glycoprotein expressed on the surface of erythrocytes, monocytes, neutrophils, B cells, some T cells, follicular dendritic cells, and glomerular podocytes 23 . The complement regulatory activity of HuCR1 is contained within its 1971-amino acid extracellular domain, which is composed of four long homologous repeats (LHR-A through -D) with a total of 30 short consensus repeats (SCRs) 24 . A soluble form of HuCR1 containing the extracellular domain but lacking the transmembrane and cytoplasmic domains is present in the circulation at extremely low concentrations, most likely as a by-product of proteolytic shedding 25 . Both membrane-bound and soluble CR1 regulate complement activity by accelerating the decay of the C3 and C5 convertases and acting as a cofactor for the serine protease factor I, which degrades C3b and C4b 26 . A recombinant form of soluble HuCR1 (sHuCR1) known as TP10 has previously been tested for its efficacy against complement-mediated pathologies. Studies using animal models demonstrated that treatment with TP10 had beneficial effects in arthritis 27 , hyperacute rejection 28 , IRI of small intestine 29 and skeletal muscle 30 , tissue injury and burns 31 . However, TP10 has a relatively short half-life 32,33 and failed to meet primary end points in clinical trials 33,34 . Recently, we have developed CSL040, a sialylated truncated variant of sHuCR1 lacking the LHR-D domain. CSL040 exhibits higher complement regulatory activity and an extended half-life compared with sHuCR1 35 , and has shown benefit in a model of antibody-mediated glomerulonephritis in mice 35 . These data led us to further investigate the efficacy of CSL040 in a mouse model of warm renal IRI.

Results
Complement activation in renal IRI. In our mouse model of unilateral warm renal IRI, the right kidney is removed, the left renal pedicle is clamped for 22 min, and kidney injury and function are assessed 24 h post-reperfusion. The first set of experiments focused on validation of the complement pathway as a target in this model. For this purpose, kidneys were examined for the deposition of complement activation markers by immunofluorescence staining/confocal microscopy, and serum was collected to measure consumption of complement components. Compared to sham controls, mice undergoing kidney IRI exhibited significant deposition of C3d, C4d and C9 on the basement membrane of tubular epithelium and on vascular and glomerular endothelium (Fig. 1A). C3d is a marker for complement activation by any of the three pathways, while C4d indicates engagement of the classical and/or lectin pathways, and C9 indicates formation of the MAC, the final step of complement activation. To further investigate which complement pathways were activated, kidneys were stained for the pathway-specific components C1q (classical), MBL (lectin) and factor Bb (alternative). Deposition of all 3 markers was evident in IRI kidneys (Fig. 1B). Serum collected from the mice at 24 h was analysed for the activity of each pathway. Lectin and alternative pathway activities were significantly reduced in IRI versus sham mice ( Fig. 1D-E; both p < 0.001), indicating consumption of components in those pathways. Classical pathway activity was also reduced in IRI mice, although this did not reach statistical significance ( Fig. 1C; p = 0.09). Together these data suggest activation of the lectin and alternative complement pathways, and to a lesser extent of the classical complement pathway, in the warm renal IRI model. Treatment with CSL040 reduced IR-induced renal injury. IRI/vehicle-treated mice showed significantly elevated serum creatinine and urea (both p < 0.001 versus sham) at 24 h, indicating a significant loss of renal function ( Fig. 2A-C). Treatment with two doses of 15, 30 or 60 mg/kg CSL040 protected against loss of renal function in a dose-dependent manner, reaching statistical significance with the highest dose (p < 0.05 versus IRI/vehicle) ( Fig. 2A-C). Histological analysis of kidneys from the IRI/vehicle group revealed extensive tubular injury (p < 0.001 versus sham) with necrosis, cell swelling, dilatation, cast formation, and overall disruption of renal architecture. Consistent with the functional readouts, CSL040 treatment dose-dependently reduced tubular injury, with significance at the highest dose of 60 mg/kg (p < 0.05 versus IRI/vehicle) (Fig. 2D,E). Treatment with two doses of 85.2 mg/kg full-length sHuCR1 (an equimolar dose to 60 mg/kg CSL040) did not confer significant protection against IR-induced renal dysfunction or tubular injury ( Fig. 2B-E). CSL040 at the highest dose significantly reduced the number of cleaved caspase-3 positive apoptotic cells in the kidney, whereas fulllength sHuCR1 did not ( Supplementary Fig. 1).
A similar pattern was observed for the pathway-specific markers C1q, MBL and factor Bb. Mice from the IRI/vehicle group showed significant intrarenal deposition of all 3 markers compared to sham mice ( Fig. 3D-F; all p < 0.001), confirming the activation of the classical, lectin and alternative pathways in this model. CSL040 at 60 mg/kg significantly reduced C1q (p < 0.01), and at 30 mg/kg for MBL (30 mg/kg: p < 0.05, 60 mg/kg: p < 0.01) and factor Bb (30 mg/kg: p < 0.01, 60 mg/kg: p < 0.001); full-length sHuCR1 also caused a reduction, but this did not reach statistical significance for any of the markers (Fig. 3D-F www.nature.com/scientificreports/ Complement activation was assessed at the systemic level by measuring circulating C3b and C5a. In line with the histological analyses, serum C3b and C5a were significantly increased in IRI/vehicle mice compared to sham mice (Fig. 4A,B; both p < 0.001), and significantly reduced by treatment with 60 mg/kg CSL040 (Fig. 4A,B; both p < 0.01 versus IRI/vehicle). Full-length sHuCR1 at an equimolar dose did not significantly reduce C3b or C5a in serum (Fig. 4A,B). Treatment with CSL040 attenuated renal IR-induced endothelial cell activation and glycocalyx shedding. Renal endothelial cell activation and glycocalyx degradation as a consequence of kidney IRI contribute to inflammation and loss of renal architecture. Activation of the renal endothelium was assessed by measuring tissue VCAM-1 expression. Expression of VCAM-1 was significantly upregulated in IRI/vehicle mice compared to sham mice (Fig. 5A,B; p < 0.001). Treatment with 60 mg/kg CSL040 significantly reduced VCAM-1 expression (Fig. 5A,B; p < 0.001 versus IRI/vehicle). The equimolar dose of full-length sHuCR1 also reduced VCAM-1 expression, but this was not statistically significant (Fig. 5A,B).
Plasma syndecan-1 and hyaluronan levels were measured to assess shedding of the endothelial glycocalyx. Both markers were significantly increased in IRI/vehicle mice compared to sham mice (Fig. 5C,D; both p < 0.001). Treatment with 60 mg/kg CSL040 significantly reduced the levels of circulating syndecan-1 and hyaluronan compared to IRI/vehicle (Fig. 5C,D; both p < 0.05 versus IRI/vehicle). Full-length sHuCR1 treatment had no effect on the levels of plasma syndecan-1 and hyaluronan (Fig. 5C,D). Together, these data suggest that inhibition of complement by CSL040 reduced IR-induced endothelial activation and glycocalyx degradation, thereby preserving endothelial integrity.      www.nature.com/scientificreports/ rescence staining/confocal microscopy. Renal IRI was associated with recruitment of both cell types to the tubular interstitium at the corticomedullary junction (Fig. 6A,B) which was significant for both populations compared to sham (Fig. 6C,D; both p < 0.001 IRI/vehicle versus sham). Treatment with CSL040 dose-dependently reduced the number of infiltrating neutrophils and macrophages following IR, reaching statistical significance with the highest dose of 60 mg/kg (Fig. 6C,D; both p < 0.01 versus IRI/vehicle). As observed for many of the other readouts, the equimolar dose of full-length sHuCR1 did not significantly affect macrophage or neutrophil recruitment/infiltration (Fig. 6C,D).

Discussion
Transplant-related IRI is an unavoidable inflammatory response that happens in all transplanted organs during the process of organ retrieval and reperfusion in the recipient. In kidney transplantation, IRI is the predominant mechanism leading to organ dysfunction or DGF, which is associated with an increased risk of chronic rejection, and worse long-term graft function and survival 20,37,38 . To date, no approved treatments for IRI or DGF are available. There are abundant pre-clinical and clinical data demonstrating that unregulated complement activation contributes to the pathogenesis of a wide range of conditions including renal IRI and DGF 20,37-39 . It has been shown in animal models that deficiency in complement components or treatment with complement inhibitors resulted in significantly reduced renal IRI 7,9,16,17,40,41 . In this study, we identified CSL040, a truncated form of recombinant sHuCR1 with enhanced inhibitory activity and pharmacokinetic properties 35 , as a promising therapeutic candidate to attenuate complement activation and ameliorate renal IRI. It is known that all complement activation pathways can be involved in the pathogenesis of renal IRI, DGF, and transplant rejection, but the precise mechanisms and the relative contribution of the individual pathways www.nature.com/scientificreports/ has been a subject of debate. Each pathway uses different triggers under various pathogenic conditions such as infection or ischemia. Determining which pathway(s) initiates the complement cascade in IRI may inform the development of more specific therapeutic strategies. Initial studies using C4-or factor B-deficient mice indicated a predominant role for the alternative pathway and lack of involvement of the classical and lectin pathways 42,43 . In contrast, the lectin pathway was implicated in a study using mice deficient in MBL-A and MBL-C 44 . There is also evidence to suggest that all 3 pathways are activated by renal IR in large animals including swine and humans 45,46 .
To clarify which of the pathways was activated in our model, we stained for C1q, MBL, and factor Bb in injured kidneys. Deposition of all 3 markers was observed and was most pronounced for MBL and factor Bb. In line with this, analysis of complement functional activity in serum revealed that renal IRI induced higher consumption of the components of the lectin and alternative pathways than of the classical pathway. The injured kidneys also showed intense deposition of C3d, C4d, and C9, indicating full activation of complement. In renal IRI, complement deposition occurs on the outer surface of the proximal tubules early after reperfusion 9,16 . Proximal tubular epithelial cells generate local C3 and express only low levels of complement regulatory proteins 13,47 , helping to explain the vulnerability of the kidney, particularly tubular cells, to complement attack following IRI. It has been suggested that intrarenal C3 synthesis may play a pathogenic role during kidney transplant rejection episodes 48 . A study of renal transplant patients showed a correlation between the level of C3 gene expression in donor kidneys at the time of implantation and the decline of graft function after transplantation 49 . C3 activation, the convergence point of all three complement pathways, is the central stage to generate nearly all complement effector products, including C3b. The formation of C3b is a prerequisite for the generation of C3 and C5 convertases and the formation of MAC, and its tissue deposition amplifies the The numbers of neutrophil and macrophage infiltrate were expressed as count per high-power field (HPF), and quantitatively measured by image J analysis. Significance was tested using Mann-Whitney U test ( ### p < 0.001 for sham vs. IRI/vehicle), and One-way ANOVA Kruskal-Wallis and Dunn multiple comparisons test (**p < 0.01 for IRI/vehicle vs. all treatments). The data shown are mean ± SEM (n = 7-8 per group).  50 . Altogether, the published and present data suggest that multiple complement pathways are activated during reperfusion in renal IRI, and its severity depends critically on both early activation steps during reperfusion and on local C3 upregulation. Upstream complement inhibition at the level of C3/C3b or the C3 convertase has therefore been considered a promising approach. Currently no C3-targeted inhibitors are in clinical use for IRI and DGF, however, several promising candidates have emerged in recent years 9,51 . sHuCR1 has been investigated as a treatment for IRI of various organs including the kidney 31,52-54 , because it blunts the impact of C3 by accelerating the decay of the C3 and C5 convertases and promoting the factor I-mediated inactivation of C3b and C4b. TP10, a recombinant sHuCR1 produced in CHO cells, protected rat renal allografts from complement-mediated inflammatory injury 32 . TP10 has demonstrated modest efficacy in a clinical trial in which 59 patients undergoing lung transplantation received TP10 or placebo before reperfusion 34 . Although early postoperative deaths and the incidence of rejection episodes were not significantly different, the proportion of patients extubated at 24 h was significantly higher in the TP10 group 34 . This study confirmed the potential for sHuCR1 in the treatment of IRI. In the current study, we show that two doses of 60 mg/kg of CSL040, a sialylated and truncated form of sHuCR1 with enhanced complement inhibitory activity and an extended half-life 35 , effectively blocked all 3 pathways of complement activation and ameliorated renal IRI in mice. Treatment with an equimolar amount of full-length sHuCR1 with a similar level of sialylation to CSL040 35 did not significantly reduce complement activation or renal dysfunction. It is important to note that the same sHuCR1 material was used in this and our earlier functional comparison with CSL040 35 , ruling out the possibility that the observed difference in efficacy was due to batch-to-batch variation in protein quality and activity. We conclude that CSL040 provides significant benefits over full-length sHuCR1 in protecting the kidney against ischemic injury. Activation of complement can directly affect the function of the endothelium by inducing glycocalyx shedding, release of proinflammatory cytokines and chemokines, and upregulation of adhesion molecule expression. Endothelial dysfunction is concomitant with permanent loss of up to 50% of peritubular capillaries, interstitial fibrosis and tubular atrophy 55,56 . We have previously shown the rapid activation of the endothelium and degradation of the endothelial glycocalyx after renal IRI 10 . Disruption of the glycocalyx results in increased vascular permeability, interstitial edema, and loss of endothelial integrity and function 57 . In addition, breakdown components of the glycocalyx (syndecan-1, hyaluronan fragments, heparan sulfate, and biglycan) have been shown to activate innate and adaptive immunity 58 . In this study, we found that treatment with CSL040 attenuated endothelial activation and glycocalyx shedding. These results provide further evidence that complement activation is a critical effector mechanism of post-ischemic renal inflammation and injury. Treatment with full-length sHuCR1 showed no significant reduction in endothelial injury, consistent with its lower potency compared to CSL040. The activation of complement promotes tissue infiltration by leukocytes via chemokine expression and enhancement of adhesion molecule interactions and transmigration across the impaired vascular barrier 59 . The infiltrating leukocytes produce proinflammatory cytokines and oxygen radicals and release proteases and myeloperoxidase, aggravating injury 60 . As previously reported 9,10 , we found that IR induced infiltration of neutrophils and macrophages, particularly in the corticomedullary junction of the kidney. Treatment with CSL040 significantly inhibited their recruitment, confirming that complement activation is critical in this aspect of the injury process. In a separate study, we have demonstrated that interfering with neutrophil activation and homing, by antibody blockade of the G-CSF receptor (G-CSFR), is protective in our model of warm renal IRI 36 . Future studies will examine the possibility that CSL040 treatment and G-CSFR blockade, which target different mechanisms, may have complementary or synergistic protective effects.
In conclusion, we have demonstrated that CSL040 effectively protects against renal IRI, significantly reducing renal dysfunction, tubular injury, complement activation and deposition, endothelial activation and damage, and neutrophil and macrophage infiltration. This study identifies CSL040 as a potential therapeutic candidate to attenuate IRI and thereby improve graft function and survival in kidney and/or other solid organ transplants. Warm renal IRI. 10-12-week-old male C57BL/6 mice were subjected to warm renal IRI as described previously 10 . Briefly, mice were anesthetized by intraperitoneal (i.p.) administration of ketamine (100 mg/kg) and xylazine (15 mg/kg) and core body temperature was kept at a constant 37 °C by placing a heat pad beneath the animal. Prior to surgery, 0.1 mg/kg of the analgesic buprenorphine was administered subcutaneously (s.c). Using midline abdominal incision, the kidneys were exposed and the renal pedicles were bluntly dissected. After right nephrectomy, the left renal pedicle was occluded with a microvascular clamp (Roboz, Rockville, MD) for 22 min at 37 °C in a temperature-controlled chamber. After clamp removal, the kidney was observed for the colour change indicative of even blood reflow. The surgical wound was then sutured, and mice received 8 ml/kg warm saline into the abdominal cavity prior to wound closure. Mice were recovered on a heat pad at 37 °C with atipamezole (0.25 mg/kg, s.c. injection) to hasten recovery from anaesthesia. 24 h after reperfusion, the mice were anesthetized and exsanguinated, and blood and kidney samples were obtained. Sham-operated mice had identical surgical procedures except that the left renal pedicle was not clamped. www.nature.com/scientificreports/ A dose escalation study was performed using two doses of CLS040 in the range 15 to 60 mg/kg (in a volume of 200-220 μL), 85.2 mg/kg of sHuCR1 (molar equivalent dose to 60 mg/kg CSL040) or Phosphate Buffered Saline (PBS, vehicle control) were administered via i.p. injection 1 h prior to ischemia and 3 h after the first injection. CSL040 and sHuCR1 with similar glycoprofiles were generated by CSL Ltd. (Parkville, VIC, Australia) as described previously 35 . Experimental groups (n = 8/group) were as follows: (1)  Assessment of renal function. Renal function was assessed by measuring serum creatinine using a kinetic colorimetric assay and analysed on a COBAS INTEGRA 400 plus analyzer (Roche, Castle Hill, NSW, Australia). In addition, serum urea was measured using Urea Assay Kit STA-382 (Cell Biolabs, San Diego, CA) as per the manufacturer's instructions.

Animals
Histopathology. Formaldehyde (10%)-fixed and paraffin-embedded kidney biopsies were cut in 4-μm-thick sections and stained with hematoxylin-eosin (H&E) or periodic acid-Schiff (PAS) according to standard protocols. PAS sections were used first to locate injured or necrotic areas, and H&E sections were subsequently used to score renal tubular injury. Injury was determined by calculating the number of injured or necrotic renal tubules as a percentage of total renal tubules within each section. Every sample was evaluated in a blinded manner by a trained pathologist for the following histological changes as findings of tubular injury: tubular epithelial cell degeneration or sloughing, tubular dilatation, cast formation, loss of brush borders, and thickening of tubular basement membranes. For each sample, twelve randomly selected corticomedullary fields (four from each of the upper, mid, and lower poles; 400× magnification) were assessed, and the percentage injury was averaged.
Enzyme-linked immunosorbent assays for complement pathway activity, C3b, C5a, and hyaluronan. Serum samples collected 24 h after reperfusion were assessed for functional complement pathway activities using the following ELISA kits from Hycult Biotech (Uden, The Netherlands): mouse classical (HIT420), lectin (HIT421), and alternative pathway (HIT422). Plasma samples were analysed for C3b using the mouse C3b ELISA kit (HK216, Hycult Biotech) and for C5a using the Mouse Complement Activation Component C5a DuoSet ELISA kit (DY2150; R&D Systems, Minneapolis, MN) as per the manufacturer's instructions. Plasma samples were also analysed for the glycocalyx components syndecan-1 using the mouse CD138 ELISA kit (PromoCell, Heidelberg, Germany) and hyaluronan using the DY3614 ELISA kit (R&D Systems) as per the manufacturer's instructions. OD was measured at 450 nm using a FLUOstar Omega microplate reader (BMG Labtech, Offenburg, Germany).