key: cord-0294783-1g0isp98
authors: Dent, Matthew; Hamorsky, Krystal; Vausselin, Thibaut; Dubuisson, Jean; Miyata, Yoshinari; Morikawa, Yoshio; Matoba, Nobuyuki
title: Safety and Efficacy of Avaren-Fc Lectibody Targeting HCV High-Mannose Glycans in a Human Liver Chimeric Mouse Model
date: 2020-04-24
journal: bioRxiv
DOI: 10.1101/2020.04.22.056754
sha: ac7815740ea4bf31331fc7c34fc9f9bed9b60b67
doc_id: 294783
cord_uid: 1g0isp98

Infection with hepatitis C virus (HCV) remains to be a major cause of morbidity and mortality worldwide despite the recent advent of highly effective direct-acting antivirals. The envelope glycoproteins of HCV are heavily glycosylated with a high proportion of high-mannose glycans (HMGs), which serve as a shield against neutralizing antibodies and assist in the interaction with cell-entry receptors. However, currently there is no approved therapeutic targeting this potentially druggable biomarker. Here, we investigated the therapeutic potential of the lectibody Avaren-Fc (AvFc), a HMG-binding lectin-Fc fusion protein. In vitro assays showed AvFc’s capacity to neutralize cell culture-derived HCV in a genotype independent manner with IC50 values in the low nanomolar range. A histidine buffer-based AvFc formulation was developed for in vivo studies using the PXB human liver chimeric mouse model. Systemic administration of AvFc was well tolerated; after 11 consecutive doses every other day at 25 mg/kg, there were no significant changes in body or liver weights, nor any impact noted in blood human albumin levels or serum alanine aminotransferase activity. Gross necropsy and liver pathology further confirmed the lack of discernible toxicity. This treatment regimen successfully prevented genotype 1a HCV infection in all animals, while an AvFc mutant lacking HMG binding activity failed to block the infection. These results suggest that targeting envelope HMGs is a promising therapeutic approach against HCV infection. In particular, AvFc may provide a safe and efficacious means to prevent recurrent infection upon liver transplantation in HCV-related end-stage liver disease patients.

histidine buffer or PBS was then concentrated to 10 mg/mL and incubated at 4°C or room 128 temperature. Absorbance at 280 nm and 600 nm was measured immediately after concentration 129 and then again after 16 and 72 h. A 280 was measured after centrifugation of precipitate.

A pharmacokinetic profile for AvFc was generated following a single 25 mg/kg i.p. 133 injection in C57bl/6 mice (The Jackson Laboratory, BarHarbor, ME, USA; 8-week-old males 134 and females; n=4 per time point) and sampling blood at 0.5, 1, 2, 4, 8, 12, 24 and 48 h post 135 injection. The concentration of AvFc was then measured using an HIV gp120-coated ELISA. 136 Briefly, a recombinant gp120 (HIV CM235, NIH AIDS Reagent Program) was coated overnight 137 at 0.3 μg/mL followed by blocking with 5% dry milk-PBST. Serum samples at varying dilutions 138 were incubated for 2 h followed by detection by a goat anti-human Fc-HRP secondary antibody 139 (ThermoFisher Scientific). The plasma concentration of AvFc was calculated by interpolating 140 from a standard curve. PK parameters were calculated using the PKSolver Microsoft Excel Add-141 on (35).

Toxicological analysis and HCV challenge study in PXB-mice 144 The mouse model of toxicological analysis and HCV infection and toxicological analysis 145 was performed in PXB-mice® (cDNA-uPA wild/+ /SCID, cDNA-uPA wild/+ : B6;129SvEv-Plau, 146 SCID: C.B-17/Icr-scid/scid Jcl). These mice contain transplanted human hepatocytes with a 147 replacement index of greater than 70% as determined by blood human albumin (h-Alb) 148 measurements prior to virus inoculation (25). Mice were separated into 3 treatment groups:

AvFc lec-(25 mg/kg, n=5) for 11 doses, or AvFc (25 mg/kg, n=5 each) for 8 or 11 doses.

Treatments were co-administered i.p. with virus inoculation (5 x 10 5 copies/kg) on day 0 with a 151 genotype 1a strain (PBC002) and every other day thereafter. The general conditions and body 152 weights of the animals were monitored every other day, while serum HCV RNA and blood h-Alb 153 were measured every 7 days by RT-PCR and latex agglutination immunonephelometry (LZ Test 154 "Eiken" U-ALB, Eiken Chemical Co., Ltd.) respectively. Serum alanine aminotransferase 1 (ALT) levels were determined either using a Fujifilm DRI-CHEM NX500sV clinical chemistry 156 instrument or by ELISA (Institute of Immunology Co., Ltd., Tokyo, Japan). At the study 157 termination on day 35, animals were euthanized and subject to gross necropsy and general 158 health. Blood was also drawn via cardiac puncture and used for ALT, HCV RNA and h-Alb 159 analyses. Multiple comparisons between groups at each time point were conducted and corrected using the 175 Tukey method with the threshold of significance set at p = 0.05. Liver:body weight ratios were 176 compared using one-way ANOVA. 180 Building on our previous observation that AvFc has affinity to a recombinant HCV E2 181 envelope protein (24), we first examined whether AvFc inhibits HCV infection in vitro using 182 multiple genotypes of cell culture-produced virus (HCVcc) or pseudotyped virus (HCVpp). 

Previously, we found that AvFc has limited solubility in phosphate-buffered saline (PBS) 194 at concentrations above 1 mg/mL (unpublished observation). In order to facilitate in vivo studies 195 we screened for an optimal liquid formulation for systemic administration that can impart 196 improved stability and solubility to AvFc at higher concentrations. Initial buffer screening 197 revealed that AvFc is prone to degradation at and below pH 6.5, suggesting that AvFc is not 198 stable in acidic pH conditions ( Figure S1 ). Further pre-formulation studies led us to identify an 199 optimal buffer composed of 30 mM histidine, pH 7.0, 100 mM sucrose and 100 mM NaCl. Figure 1 ). Consequently, these results suggested that administration of the drug every other day 216 (Q2D) might be sufficient to keep the virus under control in a murine HCV challenge model. 217 We then assessed the safety of Q2D administration of AvFc in PXB-mice®. To 218 effectively discern potential toxicity associated with AvFc's HMG-binding activity, we included 219 an AvFc variant lacking HMG-binding activity as a control (AvFc lec-; Figure S2 ). PXB mice 220 received either the vehicle (the histidine buffer described above) Q2D for 11 total doses, AvFc at 221 25 mg/kg Q2D for a total of 8 or 11 doses, or AvFc lecat 25 mg/kg Q2D for 11 total doses. As 222 shown in Figure 4A -C, no significant differences in either body weights, blood h-Alb levels or weight were seen ( Figure 4D ). These results indicate that AvFc, formulated in the histidine 225 buffer, is well tolerated in the immunocompromised mice engrafted with human hepatocytes.

Histopathology was performed to evaluate any potential toxicity to the human liver grafts 227 due to AvFc administration (Table 2 and Figure 5 ). In the human hepatocyte area, slight to 228 moderate (score 2 to 3 in Table 2 Figure 5F ). 235 Collectively, it was concluded that there was no treatment-related adverse effect in the liver 236 tissue. The full pathology report may be found in the Supplementary Information. 

The mechanism of HCV neutralization by AvFc is likely through binding to HMGs on 258 the E1/E2 envelope protein dimer, which blocks their interaction with host cell receptors and 259 viral entry. Unlike HIV envelope glycoproteins, whose glycan content can vary widely between 260 strains, the number and position of glycosylation sites on E1/E2 are highly conserved, indicating 261 their critical role in HCV's infectious processes (37). The notion that AvFc functions as an entry 262 inhibitor is supported by the facts that the lectibody has affinity to the E2 protein (24) The present study also demonstrated that AvFc therapy is well tolerated in mice and 275 human hepatocytes, as Q2D i.p. administration of 25 mg/kg of AvFc up to 11 doses did not show 276 any obvious toxicity in PXB mice by gross necropsy or histopathology of engrafted human 277 hepatocytes, nor did it result in significant changes in body weight, h-Alb, or ALT levels ( Figure   278 3, 4). We hypothesize that the lack of any significant toxicity is attributable to AvFc's unique 279 HMG-binding mechanism, whereby it requires multivalent interaction with several HMGs in 280 proximity to exhibit high affinity binding to a glycoprotein target. In line with this hypothesis, prolonged exposure to a carbohydrate-binding agent like AvFc, may result in significant decrease in viral fitness by decreasing E1/E2 incorporation into HCV particles or increased 317 susceptibility to humoral immunity due to breach in the glycan shield (37, 51). Our results 318 provide a foundation to test the above hypotheses and feasibility of the HMG-targeting anti-HCV 319 strategy.

In conclusion, the present study provided an important proof of concept for the 321 therapeutic potential of AvFc against HCV infection via targeting envelope HMGs. In particular, 322 the lectibody may provide a safe and efficacious means to prevent recurrent infection upon liver 323 transplantation in HCV-related end-stage liver disease patients. Table 2 ).

Histiocytic brown pigmentation in the Glisson's sheath is noted only in this mouse. Fatty change,  macrovesicular  2  3  3  3  3  3  3  3  3  3  3  3  3  3  3 Infiltrate, inflammatory cell, around vacuolated hepatocyte 

Structural biology of

AvFc (25 mg/kg), 11 doses AvFc lec-(25 mg/kg), 11 doses