Modification of Three Amino Acids in Sodium Taurocholate Cotransporting Polypeptide Renders Mice Susceptible to Infection with Hepatitis D Virus In Vivo

Abstract

Sodium taurocholate cotransporting polypeptide (NTCP) was identified as a functional receptor for hepatitis D virus (HDV) and its helper hepatitis B virus (HBV). In cultured cell lines, HDV infection through mouse NTCP is restricted by residues 84 to 87 of the receptor. This study shows that mice with these three amino acids altered their corresponding human residues (H84R, T86K, and S87N) in endogenous mouse NTCP support de novo HDV infection in vivo HDV infection was documented by the presence of replicative forms of HDV RNA and HDV proteins in liver cells at day 6 after viral inoculation. Monoclonal antibody specifically binding to the motif centered on K86 in NTCP partially inhibited HDV infection. These studies demonstrated specific interaction between the receptor and the viral envelopes in vivo and established a novel mouse model with minimal genetic manipulation for studying HDV infection. The model will also be useful for evaluating entry inhibitors against HDV and its helper HBV.

Importance: NTCP was identified as a functional receptor for both HDV and HBV in cell cultures. We recently showed that neonatal C57BL/6 transgenic (Tg) mice exogenously expressing human NTCP (hNTCP-Tg) in liver support transient HDV infection. In this study, we introduced alterations of three amino acids in the endogenous NTCP of FVB mice through genome editing. The mice with the humanized NTCP residues (H84R, T86K, and S87N) are susceptible to HDV infection, and the infection can be established in both neonatal and adult mice with this editing. We also developed a monoclonal antibody specifically targeting the region of NTCP centered on lysine residue 86, and it can differentiate the modified mouse NTCP from that of the wild type and partially inhibited HDV infection. These studies shed new light on NTCP-mediated HDV infection in vivo, and the NTCP-modified mice provide a useful animal model for studying HDV infection and evaluating antivirals against the infection.

FIG 1

Generation of mouse lines with modification of aa 84 to 87 of NTCP. (A) Schematic diagram of mouse Ntcp editing by TALEN. In mouse zygotes, the first exon in the Ntcp gene locus was specifically targeted by TALEN protein pairs, and DNA double-strand breaks (DSB) were introduced by the dimerized FokI nucleases. Homologous recombination was triggered with the donor homologous arm harboring the designed 5-bp editing which could introduce three amino acid modifications between residues 84 and 87 of the mouse NTCP. (B) Examining the Ntcp TALEN activity in vitro. Schematic of the Surveyor nuclease assay for TALEN cleavage efficiency is shown (top). A DNA double-strand break mediated by TALEN cleavage introduces mutations by nonhomologous end joining (NHEJ). The mutations in mouse Ntcp were amplified by genomic DNA PCR, and then the amplicons were reannealed to form base pair mismatches which were digested by Surveyor nuclease. TALEN cleavage efficiency was calculated based on the fraction of cleaved DNA. The result of a gel Surveyor nuclease assay of two TALEN pairs targeting Ntcp in Hep1-6 mouse hepatocytes 48 h after transfection is shown (bottom). Cutting ratios calculated by Quantity One software are indicated at the bottom of each lane. In lane 2 for the cells transfected with the TALEN pair b (see the legend of panel A for details), about 5.9% of the 500-bp amplicon was digested into the 400-bp fragment (arrow). (C) Sequencing results indicating the correct 5-bp editing from the tail genomic DNA of the founder mouse. Sequence alignment (top) and the chromatogram (bottom) are shown for the founder mouse with the designed base pair modification. (D) Summary of founder mice with the correct 5-bp mutation or other deletion/mutation. Genomic DNA was isolated from the FVB and ICR mouse tails and sequenced for the targeted genome region. (E) Tail genomic DNA genotyping of the offspring of the founder mice with the 5-bp editing in Ntcp. Genotyping primer pairs for the wild-type and edited Ntcp are shown in the schematic diagram (top). The amplicons for Ntcp from the mouse with indicated genotypes are shown (bottom). WT, wild type.

FIG 2

Expression in the liver of mouse NTCP with the modification of aa 84 to 87. (A) Quantifying mRNA of the mouse Ntcp with or without the 5-bp editing by qRT-PCR. The two qRT-PCR primer pairs specific for the cDNA from the edited or wild-type mouse Ntcp mRNA are shown in the left panel. The mRNA levels of the edited Ntcp as well as wild-type Ntcp from the homozygotes (n = 5) or heterozygotes (n = 5) are shown (left). mRNA copy numbers from 20 ng of liver RNA are presented. Bars indicate the median of each group. Mice were born in the same litter, and their parents were heterozygotes with the edited Ntcp. (B) Mouse MAb 18D1 specifically bound an epitope harboring aa 84 to 87 of human NTCP. Human (hNTCP) and mouse (mNTCP) NTCPs with residues 84 to 87 swapped with the indicated counterparts were expressed in 293T cells; 18D1 (10 μg/ml) and FITC-labeled anti-mouse secondary antibody were used for the staining. (C) Liver sections from an Ntcp-edited FVB mouse were stained with 18D1 labeled by CF555 (red) (left). The mouse was homozygous for the humanizing mutations 84R, 86K, and 87N. Liver sections from a wild-type FVB mouse were used as a negative control. Liver sections from the wild-type or edited mice were also stained by a MAb 36C1M (green; right); 36C1M recognizes an epitope outside the region of aa 84 to 87 of NTCP to confirm the expression of mouse NTCP regardless of the gene editing. Nuclei were stained in blue.

FIG 3

HDV infection of NTCP-modified neonatal and adult mice in vivo. (A) The schematic diagram for viral inoculation and infection detection. FVB mice were inoculated with HDV by intraperitoneal (i.p.) injection on day 9 after birth (DOB 9) and sacrificed on day 15 after birth (DOB 15) (top). Homozygotes (n = 7), heterozygotes (n = 7), and wild-type (n = 5) littermates were inoculated with 2 × 10¹⁰ GEq of HDV. HDV RNA levels were assessed by qRT-PCR from the liver of challenged mice; the levels of Ntcp mRNA containing the editing were also determined. The RNA copy numbers from 20 ng of liver RNA are presented (bottom). Data points from the same mouse are shown in the same color and pattern. Bars indicate the median of each group. Mice were born in the same litter, and their parents were heterozygotes of the edited Ntcp. (B) HDV RNA was undetectable in the spleen or pancreas of the challenged mice. The detection limit for HDV RNA was 10 to 100 copies per 20 ng of liver total RNA. Ntcp mRNA containing the editing was also determined by qRT-PCR, and the levels were below or close to the detection limit of the assay (∼100 copies per 20 ng of liver total RNA). (C) Northern blot analysis for HDV genome and antigenome RNA from the challenged mouse livers. Results of RNA samples of individual mice from the experiment shown in panel A are shown. Two micrograms of total RNA was loaded for analysis. Mouse GAPDH RNA was used as an RNA loading control. RNA from HepG2-NTCP cells with or without HDV infection was used as the positive (Pos) or negative (Neg) control, respectively. (D) Detection of HDV RNA editing. The 349-bp PCR amplicons from the cDNA reverse transcribed from the HDV RNA of the indicated mouse liver from the experiment shown in panel A were digested with the restriction enzyme NcoI. NcoI cutting yielded a band of 269 bp only from HDV products that underwent RNA editing, which converted the sequence from CCATAG to CCATGG. The PCR amplicons and the edited products are shown by silver staining. (E) Immunofluorescence staining of HDV delta antigens in the liver of an infected mouse. Photos taken of liver sections of a homozygous FVB mouse challenged with 2 × 10¹⁰ GEq of HDV at day 9 after birth and sacrificed at day 15 after birth are shown. The edited mouse NTCP was stained by MAb 18D1 (red); HDV delta antigens were stained by MAb 4G5 (green); nuclei are shown in blue. (F) HDV infection of adult FVB and ICR mice with modified NTCP. FVB and ICR with homozygous editing of residues 84 to 87 were inoculated with 1 × 10¹⁰ GEq of HDV/g weight at 70 days of age and sacrificed at day 6 post-viral inoculation (top). RNA copy numbers of HDV and the mRNA of the edited mouse Ntcp in the liver were determined by qRT-PCR. HDV RNA and edited Ntcp mRNA values from male (triangle) and female (circle) mice are shown. Data points from the same mouse are shown in the same color and pattern. Bars indicate the median of each group.

FIG 4

Analysis of MAb 18D1 for neutralization of HDV infection. (A) MAb 18D1 inhibited HDV infection on HepG2-NTCP cells. HepG2 cells stably expressing NTCP were infected by HDV with the indicated concentration of 18D1. HDV RNAs were quantified by qRT-PCR assay on day 6 postinfection. The result is presented as the percentage of HDV RNA relative to the level of the control (PBS) group. Anti-pre-S1 MAb 2D3 was used as a positive control for inhibition. (B) 18D1 partially inhibited HDV infection in vivo. 18D1 (60 mg/kg; n = 19), isotype-matched control antibody (mouse IgG2 [mIgG2]; 60 mg/kg, n = 6), or PBS (n = 14) was intraperitoneally injected 30 min prior to the viral challenge; HDV was inoculated at a dose of 2 × 10¹⁰ GEq, and the infection was evaluated on day 6 post-viral infection. The level of infection is presented as the percentage of HDV RNA relative to the average level of the PBS group. Results from four independent experiments with a total of 39 mice are shown collectively in the box-and-whiskers plot. Center lines show the medians; box limits indicate the 10th and 90th percentiles. Statistical significance was calculated by a Mann-Whitney-Wilcoxon test (two-sided).