Sodium taurocholate cotransporting polypeptide is a functional receptor for human hepatitis B and D...

7/30/2019 Sodium taurocholate cotransporting polypeptide is a functional receptor for human hepatitis B and D virus

1/28

elife.elifesciences.org

Yan et al. eLife 2012;1:e00049. DOI: 10.7554/eLife.00049 1 of 28

Sodium taurocholate cotransportingpolypeptide is a unctional receptor orhuman hepatitis B and D virusHuan Yan1,2, Guocai Zhong2, Guangwei Xu2, Wenhui He2,3, Zhiyi Jing2,Zhenchao Gao1,2, Yi Huang2,3, Yonghe Qi2, Bo Peng2, Haimin Wang2, Liran Fu2,3,Mei Song2,3, Pan Chen2,3, Wenqing Gao2, Bijie Ren2, Yinyan Sun2, Tao Cai2,Xiaoeng Feng2, Jianhua Sui2, Wenhui Li2*

1Graduate program in School o Lie Sciences, Peking University, Beijing, China;2National Institute o Biological Sciences, Beijing, China; 3Graduate program inChinese Academy o Medical Sciences and Peking Union Medical College,Beijing, China

Abstract Human hepatitis B virus (HBV) inection and HBV-related diseases remain a major publichealth problem. Individuals coinected with its satellite hepatitis D virus (HDV) have more severe

disease. Cellular entry o both viruses is mediated by HBV envelope proteins. The pre-S1 domain o

the large envelope protein is a key determinant or receptor(s) binding. However, the identity o the

receptor(s) is unknown. Here, by using near zero distance photo-cross-linking and tandem afnity

purifcation, we revealed that the receptor-binding region o pre-S1 specifcally interacts with

sodium taurocholate cotransporting polypeptide (NTCP), a multiple transmembrane transporter

predominantly expressed in the liver. Silencing NTCP inhibited HBV and HDV inection, while

exogenous NTCP expression rendered nonsusceptible hepatocarcinoma cells susceptible to these

viral inections. Moreover, replacing amino acids 157165 o nonunctional monkey NTCP with the

human counterpart conerred its ability in supporting both viral inections. Our results demonstratethat NTCP is a unctional receptor or HBV and HDV.

DOI: 10.7554/eLie.00049.001

IntroductionApproximately 2 billion people have been inected with human hepatitis B virus (HBV) worldwide.

Over 350 million people currently are chronically inected and are at high risk or progression to cir-

rhosis, liver ailure, or cancer. More than 50% o liver cancers worldwide are attributed to HBV inec

tion. HBV-related liver diseases remain a major public health problem, causing approximately 1 million

deaths per year. Individuals coinected with HBV and HDV are at greater risk or rapid progression and

severe disease (Lavanchy, 2004; Hughes et al., 2011). Despite its enormous medical and social rele

vance, progress in HBV research has been impeded by the lack o understanding o HBV entry bywhich the virus specically inects human liver cells. HBV is an enveloped virus containing a small

genome o 3.2 kb o partially double-stranded DNA encoding our overlapping reading rames. The

HBV envelope consists o the small (S), middle (M), and large (L) envelope proteins, which are multiple

transmembrane spanners sharing the same C-terminal domain corresponding to the S protein but di

ering at their N-terminal domains (Figure 1A) (Heermann et al., 1984; Seeger et al., 2007). HDV is

a small satellite RNA virus o HBV carrying all three HBV envelope proteins and can only propagate

when coexisting with HBV. The mechanism o viral entry o HDV is believed to be similar to that o HBV

and HDV has been used as a surrogate or studying HBV inection at the entry level ( Barrera et al.

2004; Sureau, 2006; Hughes et al., 2011). The L protein and integrity o S protein are critical or HBV

*For correspondence:liwenhui@

nibs.ac.cn

These authors contributed

equally to this work

Competing interests: The authors

have declared that no competing

interests exist

Funding:See page 25

Received: 11 July 2012

Accepted: 05 September 2012

Published: 13 November 2012

Reviewing editor: Zhijian J

Chen, UT Southwestern MedicalSchool, United States

Copyright Yan et al. This

article is distributed under the

terms o the Creative Commons

Attribution License, which

permits unrestricted use and

redistribution provided that the

original author and source are

credited.

RESEARCH ARTICLE
http://elife.elifesciences.org/http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.001mailto:[email protected]:[email protected]://creativecommons.org/licenses/by/3.0/http://creativecommons.org/licenses/by/3.0/http://creativecommons.org/licenses/by/3.0/http://creativecommons.org/licenses/by/3.0/mailto:[email protected]:[email protected]://dx.doi.org/10.7554/eLife.00049.001http://dx.doi.org/10.7554/eLife.00049http://creativecommons.org/licenses/by/3.0/http://www.elifesciences.org/the-journal/open-access/http://elife.elifesciences.org/


2/28

Biochemistry | Microbiology and infectious diseases


Research article

and HDV inections. The pre-S1 domain o the L protein is a key determinant or entry o both HBV and

HDV and is believed to mediate viral interaction with the cellular receptor(s) on hepatocytes (Gripon

et al., 1995; Le Seyec et al., 1999; Chouteau et al., 2001; Blanchet and Sureau, 2007; Le Du

et al., 2009). Although a number o HBV receptor candidates have been reported in the past, none

has been conrmed to be unctional in supporting viral inection (Glebe and Urban, 2007).

An N-terminal myristoylated peptide corresponding to amino acids (aa) 248 o the pre-S1 domain

o the L protein has been shown to eectively block both HBV and HDV inections o hepatocytes

through engaging an unknown cellular component, most likely a viral receptor (Barrera et al., 2005

Glebe et al., 2005; Gripon et al., 2005; Engelke et al., 2006; Schulze et al., 2010). In the curren

study, by using a synthetic modied peptide originating rom the native aa 248 lipopeptide (Myr-47/

WT) as a probe and employing a series o biochemical approaches and virological assays, we identied

and conrmed that sodium taurocholate cotransporting polypeptide (NTCP), a multiple transmem-brane transporter mainly expressed in the liver, interacts specically with the L proteins o HBV and

HDV and unctions as a common receptor or both viruses.

Results

Photoreactive ligand peptides or identication o interacting protein(s)o pre-S1 domain o L envelope proteinTo identiy the pre-S1 interacting molecule(s), we employed a photo-cross-linking approach using

a synthetic peptide derived rom the native pre-S1 peptide with particular residues replaced by

eLie digest Liver diseases related to the human hepatitis B virus (HBV) kill about 1 millionpeople every year, and more than 350 million people around the world are inected with the virus.

Some 15 million o these people are also inected with the hepatitis D virus (HDV), which is a satellite

virus o HBV, and this places them at an even higher risk o liver diseases, including cancer. The

viruses are known to enter liver cells by binding to receptors on their surace beore being enguled.

Both HBV and HDV have outer coats that consist o three kinds o envelope proteins, and a

region called the pre-S1 domain in one o them is known to have a central role in the interactionbetween the viruses and the receptors and, thereore, in inecting the cells. However, the identity o

the HBV receptor has remained a mystery. Now Yan et al. have identifed this receptor to be sodium

taurocholate cotransporting polypeptide. This protein, known as NTCP or short, is normally

involved in the circulation o bile acids in the body.

In addition to humans, only two species are known to be susceptible to inection by human HBV

and HDVchimpanzees and a small mammal known as the treeshrew. Yan et al. started by isolating

primary liver cells rom treeshrews, and then used a combination o advanced purifcation and mass

spectrometry analysis to show that the NTCP on the surace o the cells interacts with the pre-S1

domain in HBV.

The authors then perormed a series o gene knockdown experiments on liver cells o both

human and treeshrew origin: when the gene that codes or NTCP was silenced, HBV inection was

greatly reduced. Moreover, they were able to transect HepG2 cellswhich are widely used in

research into liver disease, but are not susceptible to HBV and HDV inectionwith NTCP romhumans and treeshrews to make them susceptible. Similarly, although monkeys are not susceptible

to HBV, replacing just fve amino acids in monkey NTCP with their human counterparts was enough

to make the monkey NTCP a unctional receptor or the viruses.

In the past, basic research into HBV and the development o antiviral therapeutics have both

been hindered by the lack o suitable in vitro inection systems and animal models. Now, the work

o Yan et al. means that it will be possible to use NTCP-complemented HepG2 cells or challenges

as diverse as undamental studies o basic viral entry/replication mechanisms and large-scale drug

screening. It is also possible that HBV and HDV inection might interere with some o the important

physiological unctions carried out by NTCP, so the latest work could also be o interest to medical

scientists working on other diseases related to these inections.DOI: 10.7554/eLie.00049.002
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.002http://dx.doi.org/10.7554/eLife.00049.002http://dx.doi.org/10.7554/eLife.00049


3/28



Research article

Figure 1. Developing photoreactive peptide ligands and an antibody or identiying pre-S1 binding partner(s) by zero distance cross-linking.

(A) Schematic diagram o HBV envelope proteins and N-terminal peptides o pre-S1 domain. Pre-S1 (2-47): 2-47th residues o the pre-S1 domain o the L

Figure 1. Continued on next page
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049


4/28



Research article

nonnatural amino acids (L-photo-leucine, L-2-amino-4,4-azi-pentanoic acid) (Figure 1A). L-photo

leucine contains a photoactivatable diazirine ring. Irradiation o ultraviolet (UV) light at 365 nm induces

a loss o nitrogen o the diazirine ring and yields a reactive carbene group with short hal-lie or cova

lent cross-linking at nearly zero distance (Suchanek et al., 2005). Primary hepatocytes isolated romtreeshrews (Tupaia belangeri), the only species susceptible to human HBV inection other than humans

and chimpanzees (Su et al., 1987; Walter et al., 1996; Glebe et al., 2003), were used as target cells

To maximize the eciency o photo-cross-linking, two residues (leu11 and phe14) in a region (aa 915

known to be critical or viral inection (Schulze et al., 2010) were chosen or substitution with L-photo

leucine. Leu11 is 100% conserved among HBV genotypes, and the 14th residue is a phenylalanine in

most genotypes but a leucine in some HBV strains o genotypes F and G. Changing phe14 to leucine

(F14L) did not signicantly aect the binding o HDV virion to primary Tupaia hepatocytes (PTHs

(Figure 1B). The activity o the synthesized peptide ligand Myr-47/WTb (or WTb hereater) contain

ing photo-leucines at positions 11 and 14 was also conrmed ( Figure 1C,D). WTb inhibited HDV

binding to PTHs with eciency comparable to Myr-47/WT that is comprised o all natural amino

acids (Figure 1A,C). A peptide Myr-47/N9Kb (or N9Kb hereater) similar to WTb but with an additiona

mutation at the ninth residue (N9K) did not block HDV binding to PTHs (Figure 1C). WTb but no

N9Kb inhibited viral inection o HBV and HDV on PTHs (Figure 1D). Both WTb and N9Kb peptideswere myristoylated at the N-terminus and conjugated with a biotin tag on a C-terminal lysine residue

(Figure 1A). N9Kb diers rom WTb by only one amino acid but completely lost these blocking activ

ities. Thus, N9Kb was used as a negative control or WTb. In addition, a monoclonal antibody (mAb

2D3, which specically recognizes an epitope adjacent to the critical receptor-binding region o the

peptides and shared by both WTb and N9Kb, was developed (Figure 1E).

Identication o NTCP as a specic binding protein o pre-S1The WTb or control N9Kb peptide at 200 nM was then applied to PTHs in culture and near zero distance

cross-linking was induced by UV irradiation. The cross-linked peptide and associated partners were

precipitated by streptavidin T1 beads and separated by SDSPAGE. Western blotting using 2D3 as a

probe revealed several bands including a major smeared band with apparent molecular weight o65 kDa

in the WTb but not N9Kb cross-linked sample. The 65-kDa band shited to 43 kDa upon treatmen

with the deglycosylation enzyme PNGase F (Figure 2A, let), indicating that it is highly N-glycosylatedThe WTb cross-linked protein apparently contained no intermolecular disulde bonds as it migrated

similarly under both nonreducing and reducing conditions (Figure 2A, right). The non-photoreactive

Myr-47/WT peptide but not its N9K mutant peptide eectively competed with WTb or cross-linking to

the 65-kDa band (Figure 2B). The cross-linked protein rom PTHs decreased in abundance rapidly ove

time during culture (Figure 2C). We also examined primary human hepatocytes (PHHs) in the cross-

linking experiments. Bands with slightly smaller molecular weights than those seen in the PTH cells

were also observed in PHHs (Figure 2D).

We then proceeded to identiy the target protein(s) using anity purication ollowed by mass spec-

trometry (MS) analysis. The purication procedure included three tandem steps ater photo-cross-linking

protein o HBV (S472 strain, genotype C). Residue numbering is based on genotype D. Asterisk indicates highly conserved residues among genotypes.

Epitope o mAb 2D3 was shaded in gray. (B) Eect o alterations o the critical N-terminal residues within pre-S1 region o L protein on HDV binding to

PTHs. Both wild-type (WT) and mutant HDV virions carry HBV envelope proteins. Mutant HDV carries point mutation as indicated in the pre-S1 region o

L protein. PTHs were incubated with HDV at 16C or 4 hr and ollowed by extensive wash; bound virions were quantied by qRT-PCR or virus genome

RNA copy, and the data are presented as percentage o virus binding, the binding o WT virus was set as 100%. (C) Myr-47/WTb bait peptide dose-

dependently inhibited HDV virion binding. The binding assay was perormed similarly as panel B except that PTHs were pre-incubated with indicated

peptides. (D) Inhibition o viral inection by the photoreactive peptides. Let: PTHs were pre-incubated with peptides at indicated concentrations at

37C or 1 hr and then inoculated with HDV virus. Viral inection was examined by measuring viral RNA in inected cells with qRT-PCR 6 dayspost-inection (dpi). Data are presented as percentage HDV inection. Right: peptides at indicated concentrations were added to PTHs beore HBV

inoculation. The cell culture medium was replenished every 2 days. Secreted viral antigen HBeAg was measured by ELISA on 6 dpi, and the data are

presented as percentage o that in the absence o peptides. (E) Antibody 2D3 recognizes residues 1933 o pre-S1. Peptide NC36 (aa 436 o pre-S1,

NLSVPNPLGFFPDHQLDPAFGANSNNPDWDFNP) conjugated with keyhole limpet hemocyanin (KLH) was the immunogen peptide or generating mouse

mAb 2D3. Binding activity o 2D3 with ull-length pre-S1 protein was measured by ELISA in the presence o competition peptides at indicated concentra-

tions. LD15 peptide compassing residues 1933 o pre-S1 inhibited 2D3 binding in a dose-dependent manner, indicating that 2D3 recognizes an epitope

within this region. HBV: hepatitis B virus; mAb: monoclonal antibody; HDV: hepatitis D virus; PTH: primary Tupaia hepatocytes; HBeAg: HBV e antigen.

DOI: 10.7554/eLie.00049.003

Figure 1. Continued


5/28



Research article

Figure 2. Identication o pre-S1 binding protein on primary hepatocytes with photoreactive peptide Myr-47/WTb.

(A) Let: Cultured PTHs at 2448 hr ater isolation and plating were photo-cross-linked with 200 nM Myr-47/WTb

(WTb) or Myr-47/N9Kb (N9Kb), ollowed by Streptavidin Dynal T1 beads precipitation and Western blot analysis usingmAb 2D3. The protein cross-linked by WTb is sensitive to PNGase F treatment and shited rom 65 to 43 kDa.

Right: WTb cross-linked samples were treated with 100 mM DTT and/or PNGase F as indicated and detected

similarly as in the let panel. (B) Non-photoreactive Myr-47/WT peptide (WT) but not its N9K mutant competed with

200 nM o WTb peptide or cross-linking with PTHs in a dose-dependent manner. (C) The abundance o the target

protein(s) in PTH cells decreased over time. PTHs on dierent days o in vitro culturing were photo-cross-linked with

200 nM WTb. The cross-linked samples were analyzed by Western blot. The two bands at 65 and 43 kDa were

due to incomplete deglycosylation by PNGase F. (D) WTb cross-linking with primary human hepatocytes (PHH).

Frozen PHH cells were thawed and plated 1 day beore cross-linking. With same procedure as in panel A, 200 nM

WTb but not N9Kb cross-linked with a glycoprotein o molecular weight at 60 kDa, which shited to 39 kDa upon

PNGase F treatment. (E) Purication o target protein(s) or MS analysis. PTHs photo-cross-linked with 200 nM o

WTb or N9Kb peptide were lysed, then the peptides and their cross-linked proteins were puried in tandem with

Streptavidin Dynal T1 beads, mAb 2D3 conjugated beads, and Streptavidin Dynal T1 beads in 1 RIPA buer.

Extensive wash was applied or each purication step. The samples were treated with or without PNGase F as

indicated prior to the last step o Streptavidin beads precipitation. The nal puried samples were subjected toSDS-PAGE ollowed by silver staining (let). Bracketed areas indicate the bands cut or MS analysis. Western blot

analysis (right) o the same cross-linked samples were perormed similarly as in panel A. The top 10 nonredundant

proteins identied in the 3 samples by MS analysis are listed in Figure 2Source data 1. The common protein hit

identied by MS analysis o the 65- and 43-kDa bands cut rom the WTb cross-linked sample was Tupaia NTCP

(tsNTCP), and the representative MS/MS spectra and parameters o the peptide hits are shown in Figure 2fgure

supplement 5. The control band cut rom N9Kb cross-linked sample did not generate any hits on any o these

peptides. (F) Predicted tsNTCP protein sequence. A 30-amino acid insertion unique to tsNTCP is underlined. Two

peptides identied by LC-MS/MS were highlighted in green. All lysine and arginine are highlighted in red to



6/28



Research article

capturing all biotin-labeled proteins with streptavidin T1 beads, sorting out the target protein(s) with

2D3 antibody anity beads, and then puriying with streptavidin T1 beads again to remove residua

molecules that were not covalently cross-linked with the bait peptide. The puried samples were sub

sequently subjected to SDS-PAGE ollowed by silver staining. Similar to the Western blotting results

with the 2D3 antibody, a 65-kDa protein band was visible by silver staining. The band was also

shited to 43 kDa upon PNGase F treatment (Figure 2E). Both the original 65-kDa and the shited

43-kDa bands were subsequently excised rom the gel and subjected to LTQ-Orbitrap Velos (Thermo

Fisher Scientic, MA. USA) MS analysis ater trypsin digestion. The tandem mass spectra were searched

against a Tupaia hepatocyte protein database, which we had established by deep sequencing o the

transcriptome (Figure 2fgure supplements 14). Two dierent tryptic peptide ragments, which were

identied rom both the 65-kDa and 43-kDa bands (Figure 2fgure supplement 5), matched to

a protein homolog o human NTCP. Tupaia NTCP (tsNTCP) shares 83.9% protein sequence identitywith its human counterpart and has an insertion o 30 aa near its C-terminus (Figure 2F). The peptide

(TEETIPGTLGNSTH) containing 4 aa o this insertion (underlined) was one o the two peptides identi

ed by the MS analysis at a high condence level (Figure 2fgure supplement 5). These data sug

gest that NTCP is the protein specically interacting with the WTb bait peptide.

Conrmation o NTCP as a specic binding protein o pre-S1We next cloned human and TupaiaNTCPsand validated the binding o the exogenously expressed

NTCPs with the WTb peptide and an N-terminal myristoylated pre-S1 peptide with native residues

Both human NTCP (hNTCP) and tsNTCP could be eciently cross-linked by WTb but not N9Kb when

expressed in 293T cells as shown by Western blotting with the anti-WTb antibody 2D3 as well as an

anti-C9 antibody recognizing the C-terminal C9 tag o the recombinant hNTCP and tsNTCP proteins

(Figure 3A). WTb but not the control N9Kb peptide bound to 293T cells expressing a green fuores

cent protein (GFP)-tagged tsNTCP (tsNTCP-EGFP) and co-localized with tsNTCP-EGFP on the celsurace. This binding was readily competed o by the ree Myr-47/WT peptide (Figure 3B). Moreover

a native pre-S1 peptide specically recognized the human hepatocellular carcinoma Huh-7 cell line

transected with hNTCP (Figure 3C). Consistently, Huh-7 cells transected with either tsNTCP o

hNTCPs had markedly increased HDV binding to the cells. The Myr-47/WT peptide readily com-

peted with binding o the wild-type HDV, whereas a noninectious mutant HDV virus bearing a single

N9K mutation in the pre-S1 domain o its L envelope protein ailed to bind either hNTCP- or tsNTCP

expressing Huh-7 cells (Figure 3D). Collectively, these data demonstrated a specic interaction

between NTCP and the pre-S1 domain o the L protein, which directly mediates the binding o HDV

virions to target cells.

indicate trypsin cleavage sites. Many o the potential tryptic peptides are not appropriate or LC-MS detection

because o unavorable size and/or hydrophobicity. PTH: primary Tupaia hepatocytes; MS: mass spectrometry.

DOI: 10.7554/eLie.00049.004

The ollowing source data and gure supplements are available or gure 2.

Source data 1. Top 10 nonredundant proteins identied in the three samples (Figure 2E) by MS analysis

DOI: 10.7554/eLie.00049.005

Figure supplement 1. Generation oTupaia hepatocytes proteome database rom Illumina deep sequencing-

determined transcriptome o PTHs.DOI: 10.7554/eLie.00049.006


determined transcriptome o PTHs.

DOI: 10.7554/eLie.00049.007



DOI: 10.7554/eLie.00049.008



DOI: 10.7554/eLie.00049.009

Figure supplement 5. Representative MS/MS spectra and parameters o the identied peptide hits.

DOI: 10.7554/eLie.00049.010

Figure 2. Continued
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.004http://dx.doi.org/10.7554/eLife.00049.005http://dx.doi.org/10.7554/eLife.00049.006http://dx.doi.org/10.7554/eLife.00049.007http://dx.doi.org/10.7554/eLife.00049.008http://dx.doi.org/10.7554/eLife.00049.009http://dx.doi.org/10.7554/eLife.00049.0010http://dx.doi.org/10.7554/eLife.00049.0010http://dx.doi.org/10.7554/eLife.00049.009http://dx.doi.org/10.7554/eLife.00049.008http://dx.doi.org/10.7554/eLife.00049.007http://dx.doi.org/10.7554/eLife.00049.006http://dx.doi.org/10.7554/eLife.00049.005http://dx.doi.org/10.7554/eLife.00049.004http://dx.doi.org/10.7554/eLife.00049


7/28



Research article

NTCP expression is required or HBV and HDV inectionTo test the requirement o endogenous expression o NTCP or HBV and HDV inection, we rst

examined the eect oNTCPgene silencing on viral inection o PTHs. PTHs were transected with

tsNTCP-specic or a control small interering RNA (siRNA) prior to viral inoculation. When tsNTCP

Figure 3. Binding o NTCP with N-terminal peptide o pre-S1 and HDV virions. (A) 293T cells transected with an expression vector or plasmid containing

cDNA o hNTCPor tsNTCPused with a C9 tag at its C-terminus were cross-linked with 200 nM Myr-47/WTb or Myr-N9Kb similarly as in Figure 2A at 24 hr

post-transection. Cross-linked protein samples were precipitated by Streptavidin Dynal beads ollowed by treatment with PNGase F as indicated, and

then analyzed by Western blotting using mAb 2D3 or anti-C9 tag antibody. (B) 293T cells transected with tsNTCP-EGFP or a control hSDC2-EGFP

(encoding human heparan sulate proteoglycan core protein used with EGFP at C-terminus) expression plasmid were incubated with WTb or N9Kb in the

presence or absence o 200 nM non-photoreactive Myr-47/WT as indicated. Bound peptides were probed with PE-streptavidin and the colocalization o

peptide and NTCP on cell surace was shown in the merged images. (C) FACS analysis o pre-S1 peptide binding with hNTCP transiently transected

Huh-7 cells. 24 hr post-transection with hNTCP or a control plasmid, the cells were stained with 200 nM FITC-pre-S1 (FITC-labeled lipopeptide

corresponding to the N-terminal 59-amino acid o pre-S1). The binding was analyzed by fow cytometry. (D) Huh-7 cells, ater 24 hr o transection

o indicated plasmids, were incubated with wild-type HDV or HDV with a N9K mutation on its L protein. Bound virions were quantied by qRT-PCR.

The result is presented as old changes o binding over the background virus binding to pcDNA6-transected cells. mAb: monoclonal antibody; tsNTCP:

Tupaia NTCP; NTCP: sodium taurocholate cotransporting polypeptide.

DOI: 10.7554/eLie.00049.011


8/28



Research article

mRNA level was reduced to 30% in tsNTCP siRNA-transected cells (Figure 4A, upper-let), tota

HDV RNA copies were markedly reduced in these cells comparing to those transected with contro

siRNA. We urther quantied the HDV genome and antigenome RNA copies using strand-specic

reverse transcription ollowed by quantitative real-time polymerase chain reaction (qPCR). The HDV

antigenome is a circular replication intermediate that is complementary to the genome. It is not

present in the inoculum and only appears in inected cells (Chen et al., 1986). As shown in Figure 4A

(upper-middle panel), both HDV genomic and antigenomic RNA copies were greatly reduced in cells

transected with tsNTCP-specic siRNA but not the control siRNA, indicating that tsNTCP is required

or de novo HDV inection. By contrast, lenti-VSV-G virus inection, or which viral entry is mediatedby glycoprotein protein G o VSV, was not aected in the tsNTCP- and siRNA-transected cells

(Figure 4A, upper-right). These data demonstrate that HDV viral entry requires NTCP. As HDV is

enveloped by HBV envelope proteins and can only inect target cells in a single round in the absence

o HBV, these data support that tsNTCP unctions at entry level or viral inection mediated by HBV

envelope proteins.

We then tested HBV inection on tsNTCP knockdown PTHs. Inection with HBV can be assessed by

measuring secreted viral antigens HBV S antigen (HBsAg) and HBV e antigen (HBeAg). HBV inocula

may contain residual HBsAg that can release and interere with the detection o newly synthesized

HBsAg during the rst ew days o inection. To dierentiate de novo HBsAg synthesis rom the con

taminating inoculum, we assayed HBsAg secretion over time rom days 6 to 12 ater inection with the

culture medium changed every 23 days. In addition, the kinetics o production o HBeAg with minima

or no residuals in the inoculum was also examined in the same time course experiment. As shown in

Figure 4A (lower-let), both HBsAg and HBeAg levels were markedly reduced by transection o

tsNTCP-specic but not a control siRNA at all three time points tested, demonstrating that tsNTCP

expression is required or bona de HBV inection. To conrm that tsNTCP unctions at the viral entry

level or HBV as it does or HDV, we tested AAV8-HBV virus inection on tsNTCP knockdown PTHs

AAV8-HBV is a recombinant adenovirus-associated virus containing a 1.05 overlength HBV genome

or which viral entry is mediated by AAV8 capsid instead o HBV envelope proteins. AAV8-HBV inec

tion o PTHs can nevertheless transduce the HBV genome into cells and lead to subsequent HBV vira

antigen expression. NTCP knockdown did not aect AAV8-HBV inection in PTHs, as shown by the

kinetics o HBeAg (Figure 4A, lower-right). This result shows that NTCP has no eect on post-entry

steps o HBV inection.

We next examined the eect o silencing human NTCP on HBV and HDV inections in human hep

atocytes. Human hepatoma cell line HepaRG is the only cell line known to date to be susceptible to

HBV and HDV inections upon dierentiation into a mixture o hepatocyte-like and biliary-like cells(Gripon et al., 2002). HepaRG dierentiation requires a lengthy cell culture procedure, including

maintaining undierentiated cells or 2 weeks beore induction, ollowed by induction with corticoids

and DMSO or another 24 weeks (Gripon et al., 2002). The NTCP mRNA level was low in HepaRG

cells beore induction when examined on days 5 and 10 ater initial plating, but increased dramatically

when the cells dierentiated ater induction (Figure 4B, upper-let). To examine i the acquired hNTCP

expression on dierentiated HepaRG cells is required or HDV and HBV inections, the cells were

transected with siRNAs targeting hNTCP. About 70% HDV inection was reduced by hNTCP knock-

down as indicated by decreased levels o HDV viral RNAs (Figure 4B, upper-right). Similarly, HBV

inection was also inhibited as indicated by signicantly reduced HBeAg at multiple time points

(Figure 4B, lower-let), as well as viral RNAs including the 3.5 kb RNA or HBV pre-C and pregenome

RNA (pgRNA) and HBV total RNA (Figure 4B, lower-right) quantied at the end o the experiment

We urther validated the critical role o hNTCP on HBV inection in PHHs, the natural host o the virus

Consistently, knockdown o hNTCP signicantly reduced HBV inection, which was correlated with theNTCP mRNA knockdown eciency. Both viral antigens and viral RNAs were decreased in cells trans

ected with hNTCP-specic siRNAs but not with the control siRNA (Figure 4C). Taken together, these

data demonstrate NTCP as a common key cellular receptor component necessary or HBV and HDV

inections o hepatocytes.

NTCP expression renders nonsusceptible hepatocarcinoma cellssusceptible to HDV and HBV inectionsWe then investigated the ability o NTCP to render nonsusceptible cells susceptible to viral inection

NTCP mRNA expression is low in human hepatocarcinoma cell lines that are not susceptible to HBV or


9/28



Research article

Figure 4. HDV and HBV inections o hepatocytes require NTCP. (A) Inections o HDV and HBV in PTHs were inhibited by tsNTCP knockdown. Freshly

isolated PTHs were transected with siRNAs against tsNTCP or a control siRNA. 3 days later, 1 105 PTHs were inoculated with HDV, HBV, or control



10/28



Research article

HDV inection. The levels o NTCP mRNA in Huh-7 and HepG2 cells were about 10,000 times lowe

than that in primary human and Tupaia hepatocytes (Figure 5A). We rst examined i NTCP expression

renders Huh-7 susceptible to HDV inection. Human NTCP-transected Huh-7 cells supported HDV

inection with an eciency comparable to that o PTHs; nearly 10% o cells were inected as shown by

staining o the HDV delta antigen that mainly locates in cell nuclei, whereas Huh-7 cells transected with

a vector plasmid allowed no HDV inection (Figure 5B). Moreover, the inection could be blocked by

known HBV entry inhibitors, such as pre-S1 lipopeptide and hepatitis B immune globulin (HBIG), dem

onstrating a genuine inection o HDV mediated by HBV envelope proteins on these cells (Figure 5C)

HDV RNAs, including antigenomic RNA that is only produced during HDV replication, rapidly increased

over time in the inected cells (Figure 5D). Moreover, the inection eciency correlated with both the

inoculation dose o HDV (Figure 5E) and the expression level o hNTCP (Figure 5F). HDV also inected

HepG2 cells transiently transected with hNTCP (Figure 5fgure supplement 1) as well as a cell line

established by G418 selection o HepG2 cells ater hNTCP transection, which expresses hNTCP stablyand could be readily stained by the FITC-pre-S1 peptide (Figure 5fgure supplements 2 and 3).

Although HDV is an accepted surrogate or HBV entry, we urther examined i exogenous expres

sion o NTCP directly renders nonsusceptible cells susceptible to HBV inection. HepG2 cells tran-

siently transected with either tsNTCP- or hNTCP-supported HBV inection as evidenced by continuous

secretion o HBeAg over time and accumulation o HBV replicative intermediate viral RNAs in the

inected cells. The entry inhibitor Myr-59 blocked the inection (Figure 6fgure supplement 1)

More ecient HBV inection was achieved on stable HepG2-hNTCP cells with about 510% o the cells

being inected as revealed by intracellular staining o HBsAg, whereas there was no HBV inection in

the parental HepG2 cells (Figure 6A). HBV inection in the HepG2-hNTCP stable cells was urthe

evidenced by the continuously increased production o HBeAg during the testing period. HBV entry

inhibitors, in particular the Myr-59 peptide and 17B9, eciently inhibited the inection ( Figure 6B)

The inection eciency as evidenced by HBV total and 3.5 kb RNA levels correlated with the inocula

tion dose. Moreover, the ormation o HBV covalently closed circular DNA (cccDNA), which is a replicative intermediate and transcriptional template or production o viral RNAs, was conrmed by Southern

blot analysis (Figure 6D). To urther demonstrate that the replicative intermediates o HBV were syn

thesized de novo in HepG2-hNTCP cells ater inection, we perormed additional time course experi

ments. HBV viral replicative intermediates, including cccDNA, the 3.5 kb HBV RNA, as well as the tota

HBV RNA in the HBV-inected HepG2-hNTCP cells were quantied at dierent time points. The

cccDNA became detectable at 24 hr post-inection. It markedly increased at day 3 post-inection and

maintained a relatively stable level or the rest o the time points examined, whereas the ormation o

HBV cccDNA was completely abolished i entry inhibitor Myr-59 was included with the initial virus

inoculation (Figure 6E). Consistently, HBV RNA levels o the 3.5 kb transcript and the total HBV

viruses AAV8-HBV and Lenti-VSV-G. For HDV and HBV, PTHs were inected at 500 and 100 genome equivalent copies per cell, respectively. The level o

HDV viral RNAs in inected cells was quantied by qRT-PCR on 6 dpi. Strand-specic primers were used to dierentiate the HDV genomic and anti-

genomic RNAs (see Materials and methods). For VSV-G control virus inection, recombinant lentivirus pseudotyped by VSV-G carrying a lucierase

reporter was inoculated to PTHs 3 days ater siRNA transection. The lucierase activity was assessed on 6 dpi. For HBV inection, the kinetics o secreted

viral antigens HBsAg and HBeAg were measured by ELISA. The medium was changed every 3 days. For AAV8-HBV inection, PTHs were inected with a

recombinant AAV8 carrying 1.05 overlength HBV genome. Secreted HBeAg was assessed on indicated days post-inection. The eect o tsNTCP

silencing in all viral inections was independently evaluated with a total o our siRNAs against tsNTCP (see Materials and methods). The data shown are

the result o a representative siRNA out o the our tested. (B) Dierentiated HepaRG cells express high level o NTCP mRNA and knockdown NTCP inthese cells inhibited HDV and HBV inections. HDV and HBV inection o siRNA-transected HepaRG cells was conducted similarly as in panel A. HDV

RNA levels in the inected cells were measured on 9 dpi. For HBV inection, secreted HBeAg was collected every 2 days as indicated and analyzed by

ELISA. The copy numbers o HBV total RNA and 3.5 kb RNA in the inected cells were measured at the end o the experiment, 10 dpi. (C) Knockdown

hNTCP in PHHs hampered HBV inection. Frozen PHHs were thawed and plated 1 day beore transecting with siRNAs against hNTCP or a control

siRNA. Similar to panels A and B, 3 days ater transection, PHHs were inoculated with 100 genome equivalent copies o HBV per cell, and the levels o

secreted HBeAg were determined at indicated dpi. HBV RNAs were quantied at the end o the experiment, 9 dpi. The knockdown eciency o siRNA

targeting tsNTCP or hNTCP shown in panels AC was determined by real time RT-PCR on day 4 ater transection. NTCP: sodium taurocholate cotrans-

porting polypeptide; HBV: hepatitis B virus; HDV: hepatitis D virus; PTH: primary Tupaia hepatocytes; tsNTCP: Tupaia NTCP; siRNA: small interering

RNA; dpi: days post-inection; hNTCP: human NTCP.

DOI: 10.7554/eLie.00049.012

Figure 4. Continued


11/28



Research article

Figure 5. NTCP expression coners Huh-7 susceptibility to HDV inection. (A) NTCP mRNA expression level in the

indicated cell lines and primary hepatocytes. The Huh-7 was used to normalize the relative expression levels in

other cells. (B) 1 105 Huh-7 cells were transected with 100 ng hNTCP/pcDNA6 or a vector control in 24-well plateand maintained in PMM, 24 hr ater transection, transected cells were inected with HDV at 500 genome equiva-

lent copies per cell. On 8 dpi, HDV delta antigen, which typically locates in nuclei, was stained with 4G5 antibody in

green, nuclei were stained with DAPI in blue. (C) Huh-7 cells transected with hNTCP were inected with HDV

similarly as in panel B in the presence or absence o HBV entry inhibitors: HBIG (hepatitis B immune globulin),

Myr-59, and anti-HBsAg mAb, 17B9. 4G5 was used as an antibody control. HDV RNA copies o inected cells were

quantied by real-time RT-PCR on 6 dpi. (D) Huh-7 cells transected with hNTCP were inected with HDV similarly as

in panel B. The HDV viral RNAs in inected cells at indicated time points were quantied by real-time RT-PCR.

(E) HDV inection with increasing multiplicities o genome equivalents (mge). With 100 ng hNTCP/pcDNA6, 1 105



12/28



Research article

transcripts in the inected HepG2-hNTCP cells gradually increased during rst several days o inection

and reached a steady level ater day 5 (Figure 6F). Together these data show that NTCP contributes

substantially to HBV inection.

We next compared the eciency o HBV inection in HepG2-hNTCP cells with that in PHHs. Asshown by intracellular staining o HBV core antigen (HBcAg) on day 8 post-inection, about 10%

HepG2-NTCP cells were inected at multiplicities o genome equivalents (mge) o 100, which is com

parable to the eciency o HBV inection o PHHs (Figure 6fgure supplement 2). In contrast to

PHHs, HepG2-NTCP cells propagate in cultures, thus the actual inection eciency o HepG2-NTCP

cells may be more likely than not underestimated by the observed end-point HBcAg staining. We also

compared the levels o secreted HBeAg and intracellular viral RNAs in these two types o cells inected

with three inoculation doses. The level o secreted HBeAg rom HepG2-NTCP appeared to be highe

than that in PHHs rom two donors, whereas the levels o viral RNAs per nanogram o total cell RNA

in both inected cell types are comparable (Figure 6fgure supplement 3). This may be partially

explained by their dierent abilities in propagation and supporting viral replication and protein expres

sion that would require more detailed studies.

Ecient HBV inection o PHHs or HepaRG cells in vitro normally requires high dose o virus inocu

lums, and only limited progeny viruses are produced ater inection (Gripon et al., 1988; Griponet al., 2002; Boehm et al., 2005). To assess viral particles released rom HBV-inected HepG2-NTCP

cells, we rst quantied viral DNA in the medium collected at dierent time points ater the inection

As indicated by drastic decline o viral DNA level on day 4 post-inection, the majority o residua

viruses rom inocula were removed by changing the medium and washing during the rst ew days o

inection. The levels o viral DNA in the media resulted rom the ongoing inection during days 413

post-inection were low (equivalent to 1% o input viral DNA copies) despite signicant amount o

HBeAg secretion during this period. Similarly, only low levels o viral DNA were detected in the

medium rom HBV-inected PHHs (Figure 6fgure supplement 4, right). It is reasonable to speculate

that some host actors that are lacking in cell cultures might be needed or ecient viral particles or

mation or releasing; or some cellular actors in cultures may hinder these processes during inection

The culture medium collected rom inected HepG2-NTCP cells was subsequently tested or inection

o PHHs. In line with the low HBV viral DNA level in the medium inoculum, very low number o intracel

lular HBV total RNA copies were detected in PHHs on day 13 post-inection (Figure 6fguresupplement 5), indicating that only very limited HBV inection might have occurred, which may be

attributed to the low multiplicity o inection.

Residues 157 to 165 o hNTCP are critical or pre-S1 binding and viralinectionsWe nally investigated the molecular determinants o NTCP or HBV and HDV inections. Crab-eating

monkey (Macaca fascicularis) NTCP (mkNTCP) shares high protein sequence identity with hNTCP

(96.3%) (Figure 7fgure supplement 1). However, mkNTCP neither supports HDV inection nor pre-

S1 peptide (Myr-59) binding (Figure 7A), consistent with the known narrow species specicity o the

Huh-7 cells were transected, as in panel B. Transected cells were inected with increasing mge o HDV as

indicated. HDV delta antigen was detected as in panel B on 8 dpi. (F) HDV inection o cells with increasing levels

o hNTCP. About 1 105 Huh-7 cells were transected with a vector pcDNA6 or hNTCP/pcDNA6 at indicated

amounts and cells were inoculated with 500 mge o HDV. HDV delta antigen was detected on 8 dpi as in panel B.

NTCP: sodium taurocholate cotransporting polypeptide; PMM: primary hepatocytes maintenance medium; HBV:

hepatitis B virus; HDV: hepatitis D virus; mAb: monoclonal antibody; HBsAg: HBV S antigen.

DOI: 10.7554/eLie.00049.013

The ollowing gure supplements are available or gure 5.Figure supplement 1. Inection o HDV on HepG2 cells expressing hNTCP.

DOI: 10.7554/eLie.00049.014

Figure supplement 2. Inection o HDV on HepG2 cells expressing hNTCP.

DOI: 10.7554/eLie.00049.015

Figure supplement 3. Inection o HDV on HepG2 cells expressing hNTCP.

DOI: 10.7554/eLie.00049.016

Figure 5. Continued
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.013http://dx.doi.org/10.7554/eLife.00049.014http://dx.doi.org/10.7554/eLife.00049.015http://dx.doi.org/10.7554/eLife.00049.016http://dx.doi.org/10.7554/eLife.00049.016http://dx.doi.org/10.7554/eLife.00049.015http://dx.doi.org/10.7554/eLife.00049.014http://dx.doi.org/10.7554/eLife.00049.013http://dx.doi.org/10.7554/eLife.00049


13/28



Research article

Figure 6. NTCP expression coners susceptibility to HBV inection. (A) Intracellular HBsAg expression in HBV-inected cells. HepG2-hNTCP stable cells or

parental HepG2 cells were inoculated with HBV at 100 mge. On 9 dpi, intracellular HBsAg o inected cells on coverslips was stained with antibody 17B9 in

green, and nuclei were stained with DAPI in blue. (B) Secreted HBeAg levels in the supernatants o HBV-inected cells. The cells were inected with HBV at

100 mge in the presence or absence o entry inhibitors as indicated. The medium was changed every 2 days. Secreted HBeAg was measured at 3, 5, 7 dpi;

each time point represents the level o newly synthesized HBeAg within every 2 days. ( C) HBV inection eciency is correlated with the viral inoculum dose.

With increasing dose o HBV, 2 105 HepG2-hNTCP cells were inected as indicated. HBV RNAs in inected cells was examined on 10 dpi with real-time

RT-PCR or the total HBV RNAs and the 3.5 kb transcripts. (D) Southern blot analysis o cccDNA. HepaG2-hNTCP cells or HepG2 parental cells were

inected with 100 mge o HBV in the presence or absence o Myr-59. HBV cccDNA was extracted rom 23 106 inected cells on 7 dpi. Hal o the

extracted DNA o each sample was subjected to a 1.3% agarose gel and analyzed by Southern blotting. A plasmid DNA marker or cccDNA (see Materials



14/28



Research article

and methods) at dierent concentrations rom 0.8 to 100 pg was included in the same gel. ( E)(F) Kinetic analysis o

HBV cccDNA and RNAs in HBV-inected HepG2-hNTCP cells. HBV cccDNA (in panel E) was quantied at indicated

time points post inection (see Materials and methods). A dotted line indicates the background amplication. HBV

RNA copies (in panel F) in the inected cells were measured at indicated time points, data o similarly inected

parental HepG2 cells on 9 dpi were also shown. NTCP: sodium taurocholate cotransporting polypeptide; HBV:

hepatitis B virus; HBsAg: HBV S antigen; HBeAg: HBV e antigen; mge: multiplicities o genome equivalents; dpi: days

post-inection; cccDNA: covalently closed circular DNA; hNTCP: human NTCP.

DOI: 10.7554/eLie.00049.017The ollowing gure supplements are available or gure 6.

Figure supplement 1. NTCP expression coners susceptibility to HBV inection.

DOI: 10.7554/eLie.00049.018

Figure supplement 2. Comparative HBV inection o PHHs and HepG2-NTCP cells.

DOI: 10.7554/eLie.00049.019

Figure supplement 3. Comparative HBV inection o PHHs and HepG2-NTCP cells.

DOI: 10.7554/eLie.00049.020

Figure supplement 4. Kinetics o HBV viral DNA in the culture medium o primary inection and reinection o the

released viruses on PHHs.

DOI: 10.7554/eLie.00049.021

Figure supplement 5. Kinetics o HBV viral DNA in the culture medium o primary inection and reinection o the

released viruses on PHHs.

DOI: 10.7554/eLie.00049.022

Figure 6. Continued

virus. We then made a series o human NTCP variants to cover all the dierent amino acids between

hNTCP and mkNTCP. In each variant, two or a ew residues were mutated to their mkNTCP counter-

parts (Figure 7fgure supplement 1). Whereas most mutations did not signicantly interere with

Myr-59 binding or HDV inection, alteration o ve residues o hNTCP between aa 157165 and its

monkey counterpart (rom KGIVISLVL to GRIILSLVP, distinct residues are underlined) completely abol

ished Myr-59 binding and the ability to support HDV inection. Remarkably, replacing the moti o aa

167156 in mkNTCP with the corresponding human residues converted mkNTCP to an ecient recep

tor or HDV inection (Figure 7A). All the NTCP variants tested were examined or NTCP expression

and comparable levels o cell surace expression were conrmed (Figure 7B). Similar to HDV, HBV

inection was also abolished on HepG2 cells expressing hNTCP carrying monkey-like mutations

GRIILSLVP, while mkNTCP-bearing human residues KGIVISLVL within the moti o aa 157165 largely

restored HBV inection (Figure 7C). These data show that residues between 157 and 165 o NTCP are

crucial or binding to the receptor-binding region o the pre-S1 domain o the L protein o HBV, and

critically contribute to NTCP-mediated HBV and HDV inections.

DiscussionIn this study, by employing a unique approach o tandem anity purication combined with MS analy

sis against a Tupaia hepatocyte proteome database established by deep sequencing, we revealed

that the liver bile acid transporter, NTCP, specically interacts with a key region in the pre-S1 domain

o the HBV envelope L protein. By perorming a series o virological analyses, we showed that silenc

ing NTCP expression markedly inhibited viral inection o HBV and HDV in Tupaia as well as human

hepatocytes. Exogenous expression o NTCP rendered nonsusceptible human hepatoma cells suscep

tible to the viral inections. The authentic viral inections in cells complemented with NTCP wereshown by the kinetic analyses o several markers o viral inections, in particular the quantication o

newly synthesized viral replicative intermediates. Moreover, the NTCP-rendered inections were

blocked by known entry inhibitors. NTCP residues 157 to 165 were identied to be critical or pre-S1

binding and viral inections. These data clearly demonstrate that NTCP is a unctional receptor or

both HBV and HDV.

Identication o cellular receptor(s) o HBV and HDV has been challenging. In our study, we utilized a

short peptide ligand, WTb, which was originated rom the known receptor-binding domain o the L pro-

tein (Barrera et al., 2005; Glebe et al., 2005; Gripon et al., 2005; Engelke et al., 2006; Schulze et al.

2010), but with specially designed properties suitable or photo-cross-linking and tandem purication
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.017http://dx.doi.org/10.7554/eLife.00049.018http://dx.doi.org/10.7554/eLife.00049.019http://dx.doi.org/10.7554/eLife.00049.020http://dx.doi.org/10.7554/eLife.00049.021http://dx.doi.org/10.7554/eLife.00049.022http://dx.doi.org/10.7554/eLife.00049.022http://dx.doi.org/10.7554/eLife.00049.021http://dx.doi.org/10.7554/eLife.00049.020http://dx.doi.org/10.7554/eLife.00049.019http://dx.doi.org/10.7554/eLife.00049.018http://dx.doi.org/10.7554/eLife.00049.017http://dx.doi.org/10.7554/eLife.00049


15/28



Research article

Figure 7. Identication o a critical region (aa 157165) o NTCP or pre-S1 binding and viral inections. (A) Pre-S1

binding and HDV inection on cells expressing wild-type or mutant NTCPs. Corresponding amino acids (one-letter



16/28



Research article

orm) at the mutated positions o NTCP are shown or hNTCP, crab-eating monkey NTCP (mkNTCP), and tsNTCP.

Huh-7 cells were transected with plasmids encoding tsNTCP, hNTCP, mkNTCP, or NTCP mutants as indicated.

The mutant NTCPs include hNTCP-bearing mutations o mkNTCP residues and mkNTCP-bearing mutations o

human residues at indicated positions. The transected cells were maintained in PMM or 24 hr and then either

stained with 200 nM FITC-pre-S1 or inected with 500 mge HDV. HDV delta antigen in inected cells was detected

with mAb 4G5 on 7 dpi. Replacing aa 157165 o mkNTCP with human counterpart rendered mkNTCP an ecient

receptor or pre-S1 binding and HDV inection. (B) All NTCP variants expressed comparable levels o NTCP. Huh-7

cells transected as in panel A were biotinylated 24 hr ater the transection, then lysed and analyzed or cell suraceNTCP expression (top), total NTCP expression (middle), and GAPDH (bottom), respectively. For cell surace

expression, cell lysates were pulled down with streptavidin T1 Dynabeads and subsequently examined by western

blot with mAb 1D4 recognizing a C9 tag at the C-terminus o each NTCP variant. For total NTCP expression, cell

lysates were directly subjected to SDS-PAGE, ollowed by Western blot analysis with 1D4. (C) Eects o NTCP

mutations on HBV inection. HepG2 cells were transected with plasmids encoding hNTCP, mkNTCP, or hNTCP

variants bearing the indicated monkey residues, or mkNTCP variants with the indicated human residues.

Transected cells were maintained in PMM or 24 hr, and subsequently inected with HBV at 100 mge. HBeAg and

HBV 3.5 kb RNA were assayed on 6 dpi. Similar to panel B, comparable NTCP surace expression levels in the

transected HepG2 cells were conrmed or all the NTCP variants tested (Figure 7fgure supplement 2). NTCP:

sodium taurocholate cotransporting polypeptide; HDV: hepatitis D virus; hNTCP: human NTCP; tsNTCP: Tupaia

NTCP; PMM: primary hepatocytes maintenance medium; mge: multiplicities o genome equivalents; mAb:

monoclonal antibody; HBV: hepatitis B virus; HBeAg: HBV e antigen; dpi: days post-inection.

DOI: 10.7554/eLie.00049.023

The ollowing gure supplements are available or gure 7.Figure supplement 1. Protein sequence alignment o human, monkey, and Tupaia NTCP. Residues dierent

between human and monkey are highlighted in red. Tupaia residues dierent rom humans are in blue. Residues

157165 are boxed.

DOI: 10.7554/eLie.00049.024

Figure supplement 2. Total and surace NTCP expression levels in the transected HepG2 cells. Transiently

transected HepG2 cells rom the same batch o transection as in Figure 7Cwere analyzed or total or cell surace

NTCP expression at 24 hr post-transection as described in Figure 7B.

DOI: 10.7554/eLie.00049.025

Figure 7. Continued

Two photo-leucines were incorporated into the critical receptor-binding region o WT b without interer

ing with its receptor-binding activity, which allowed highly specic zero distance cross-linking o itsdirect binding partner(s) but not other neighboring molecules. A biotin moiety o WTb acilitated puri

cation o the complex o WTb and its binding partner(s) by streptavidin beads. An mAb, 2D3, was

developed to recognize WTb on an epitope outside the receptor-binding site, serving as a highly spe

cic tool or detection as well as additional anity purication o the complex. Thus, the binding

partner(s) was rst cross-linked by WTb, and then puried by using streptavidin and 2D3 beads in tan

dem. The covalent interaction between the WTb ligand and its partner(s) enabled a purication pro-

cess under high-stringency conditions, and ecient isolation was achieved irrespective o the nature

o the binding partner(s) even i it is a membrane protein(s) with multiple transmembrane domains, like

NTCP identied here.

NTCP (Slc10a1) is the ounding member o the SLC10 amily o solute carrier proteins. It is a hepatic

Na+ bile acid symporter and is responsible or cotransportation o sodium and bile acids across cellula

membranes to maintain the enterohepatic circulation o bile acids (Hagenbuch and Meier, 1994

Stieger, 2011). NTCP is a multiple transmembrane glycoprotein presumed to span the cellular membrane up to 10 times with small extracellular loops (Mareninova et al., 2005; Hu et al., 2011). It is

mainly expressed in the liver (Stieger, 2011), consistent with the liver tropism o HBV and HDV. NTCP

localizes to the sinusoidal (basolateral) plasma membrane o hepatocytes (Stieger et al., 1994), a loca

tion that ts well with its receptor role or blood-borne HBV and HDV. Whereas HBV rst attaches to

hepatocytes mainly through heparan sulate (Schulze et al., 2007; Leistner et al., 2008), our data

demonstrate that the interaction between NTCP and L protein o HBV is highly specic, and NTCP is

crucial or productive viral entry o hepatocytes. Consistent with previous reports on primary cultures

o rat hepatocytes (Liang et al., 1993; Rippin et al., 2001), NTCP expression rapidly decreased ove

time in cultured PTHs ater isolation. This may at least partially explain the observations that primary
http://dx.doi.org/10.7554/eLife.00049http://dx.doi.org/10.7554/eLife.00049.023http://dx.doi.org/10.7554/eLife.00049.024http://dx.doi.org/10.7554/eLife.00049.025http://dx.doi.org/10.7554/eLife.00049.025http://dx.doi.org/10.7554/eLife.00049.024http://dx.doi.org/10.7554/eLife.00049.023http://dx.doi.org/10.7554/eLife.00049


17/28



Research article

hepatocytes typically remain susceptible to HBV inections in vitro or only a ew days ater isolation

rom liver tissues (Gripon et al., 1988; Seeger et al., 2007).

NTCP is unctionally conserved in mammalians, but protein sequences o NTCP vary among

species, which is likely to contribute to the narrow species tropism o viral inection. Strikingly, despite

the high level o protein sequence homology between human and monkey NTCP, the later did not

support HBV and HDV inection. Replacing a small moti o aa 157165 o mkNTCP with the corre-

sponding hNTCP residues converted mkNTCP to a receptor or pre-S1 binding as well as HDV and

HBV inection. Further studies are warranted to determine i and how NTCP contributes to the species

specicity o HDV and HBV inection in other species. It also remains to be determined i othermolecule(s) additional to NTCP contributes to the cellular entry o HBV and/or HDV as a coreceptor(s

or receptor component(s), and i other host actors such as the microenvironment or architecture o

hepatocytes in liver, or soluble blood components like those that have been shown to involve in inec

tions o other viruses (Shayakhmetov et al., 2005; Morizono et al., 2011), contribute to HBV and/o

HDV inection.

Expression and subcellular distribution o NTCP are precisely regulated under physiological condi-

tions. NTCP accounts or most, i not all, hepatic Na+-dependent bile acid transport (Stieger, 2011)

NTCP expression is low and inversely correlated with the degree o dedierentiation o cancer cells in

human hepatocellular carcinoma (Kullak-Ublick et al., 1997; Zollner et al., 2005) and the severity o

HBV-related liver cirrhosis (Lee and Kim, 2007). The newly discovered role o NTCP as an entry recep

tor or HBV and HDV raises interesting questions regarding its involvement in viral pathogenesis.

Identication o NTCP as a unctional receptor or HBV and HDV advances our understanding o thei

entry into host cells and may lead to new prevention and treatment strategies against these viruses

and related diseases.

Materials and methods

Isolation and culture o primary Tupaiahepatocytes (PTHs)Adult tree shrews (Tupaia belangeri chinensis) were housed in a Tupaia animal acility at the

National Institute o Biological Science, Beijing. All studies were perormed in accordance with

institutionally approved protocols and adherent to guidelines o the National Institute o Biologica

Sciences Guide or the care and use o laboratory animals. PTH cells were obtained rom anesthe

tized Tupaia (100150 g) with a two-step perusion method as previously described (Walter et al.

1996). Cell suspensions ater perusion were ltered through a 70-m cell strainer and centri

uged at 50 g or 3 min. The cell pellet containing PTHs was resuspended in plating mediumo Williams E medium supplemented with 10% FBS, 5 g/ml transerrin, 5 ng/ml sodium selenite

2 mM L-glutamine, 100 U/ml penicillin, and 100 g/ml streptomycin. The cells were then plated

on collagen-coated cell culture dishes or plates. 4 hr ater plating, medium were changed to pri

mary hepatocytes maintenance medium (PMM), that is, Williams E medium supplemented with

5 g/ml transerrin, 10 ng/ml EGF, 3 g/ml insulin, 2 mM L-glutamine, 18 g/ml hydrocortisone

40 ng/ml dexamethasone, 5 ng/ml sodium selenite, 2% DMSO, 100 U/ml penicillin, and 100 g/m

streptomycin. Cells were maintained in 5% CO2 humidied incubator at 37C with regular medium

change every 23 days.

Primary human hepatocytes (PHHs)PHHs were purchased rom Becton Dickinson (United States) or Shanghai RILD Inc. (Shanghai, China)

The cells were cultured similarly as PTHs using the same plating medium and maintaining PMM

medium as described above.

Cell linesHuman embryonic kidney cell lines 293 and 293T, human cervix carcinoma cell line Hela, and human

hepatocellular carcinoma cell line HepG2 were rom American Type Culture Collection (ATCC); human

hepatocellular carcinoma cell lines Huh-7, SMMC-7721 (SMMC), and Bel-7404 (BEL) were rom the

Cell Bank o Type Culture Collection, Chinese Academy o Sciences. The cells were cultured with

Dulbeccos Modication o Eagles Medium (DMEM; Invitrogen, United States) supplemented with

10% etal bovine serum, 100 U/ml penicillin, and 100 g/ml streptomycin at 37C in 5% CO2 humidied

incubator except otherwise indicated. HepaRG cells were purchased rom Biopredic Internationa


18/28



Research article

(Rennes, France) and were cultured ollowing the product manual. Dierentiated HepaRG cells were

obtained ollowing a two-step procedure as described by Gripon et al. (2002).

Viruses

HDVA plasmid containing a head to tail trimer o 1.0 HDV cDNA o a genotype I virus (Genebank acces

sion number: AF425644.1) under the control o a CMV promoter was constructed with de novo syn-

thesized HDV cDNA or the production o HDV RNPs. A pUC18 plasmid containing nucleotide

24311990 o HBV (Genotype D, Genebank accession number: U95551.1), or the same plasmid bearing mutation generated by site-directed mutagenesis, was used or expressing HBV envelope proteins

under the control o endogenous HBV promoter. HDV virions were produced by transection o the

plasmids in Huh-7 as previously described by Sureau et al. (1992).

HBVHBV genotype B virus was obtained by ultracentriugation o plasma rom an HBV chronic carrier with

written consent. HBV genotype D virus was produced by transection o Huh-7 cells with a plasmid

containing 1.05 copies o HBV genome under the control o a CMV promoter similarly as previously

described by Blanchet and Sureau (2006). The Genebank accession numbers or the viruses are

JX978431 and U95501.1, respectively.

AAV8-HBV recombinants

Recombinant adeno-associated virus 8 (AAV8) carrying 1.05 copies o HBV genome was producedsimilarly as previously described (Xiao et al., 1998) by cotransection o 293 cells with plasmids o

AAV8 packaging, 1.05 overlength HBV genome (genotype D) and adenovirus helper.

Lenti-VSV-GAn HIV-1 genome-based lentivirus pseudotyped by glycoprotein o vesicular stomatitis virus and carrying

a refy lucierase reporter gene was produced by cotransection o 293T cells with plasmids or VSV-G

expression, HIV genome packaging, and lucierase reporter, respectively, as described (Sui et al., 2005)

Virus-related experiments were conducted in a BSL-2 acility at the National Institute o Biologica

Sciences, Beijing.

Peptides and antibodiesPeptides with nonnatural amino acid L-2-amino-4,4-azi-pentanoic acid(L-photo-leucine) were syn

thesized by American Peptide Company Inc. (United States). Other peptides corresponding to theN-terminal o pre-S1 domain o HBV L protein (genotype C, strain S472, GeneBank EU554535.1)

were synthesized by SunLight Peptides (Beijing, China). Mouse monoclonal antibodies (mAb) 2D3

1C10, and 4G5 were generated in the laboratory; all are o IgG1 isotype. 2D3 specically recognizes

the 1933 amino acids o the pre-S1 domain o HBV L protein; 1C10 recognizes HBcAg; 4G5 target

HDV delta antigen. 17B9, a mouse mAb-specic to HBV S protein, was provided by Dr. Lin Jiang

China National Biotec Group. Hepatitis B immune globulin (HBIG) was rom the National Institutes o

Food and Drug control, Beijing, China. 2D3 magnetic beads were prepared by covalently cross-linking

2D3 to Dynabeads M-270 Epoxy ollowing manuacturers instructions. Secondary antibodies o

immunofuorescence staining and Western blot were purchased rom Invitrogen or Sigma-Aldrich

(United States).

ELISA kits and other reagents

ELISA kits or HBsAg and HBeAg measurement were purchased rom Wantai Pharm Inc. (Beijing,China). SYBR Premix Ex Taq quantitative real-time PCR kit and Reverse Transcriptase (RT) kit were rom

Takara Inc. (Beijing, China). Streptavidin-coupled magnetic beads (Dynabeads MyOne Streptavidin T1

and magnetic beads coated in glycidyl ether (Epoxy) groups (Dynabeads M-270 Epoxy) were pur-

chased rom Invitrogen. Other reagents were purchased rom New England Biolabs (United States)

Lie Technologies (United States), or Sigma-Aldrich.

Assays or HBV viral antigens rom supernatant o inected cellsHBV viral antigens HBsAg and HBeAg were examined using 50 l supernatants with commercial ELISA

Kits (Wantai Pharmacy, Beijing, China) ollowing manuacturers instructions. In most cases, HBsAg


19/28



Research article

level was normalized with WHO HBsAg reerence serum (kindly provided by Dr. Zhenglun Liang rom

the National Institutes or Food and Drug control, Beijing, China) and presented as international units

per milliliter.

Quantication o HDV total RNA (genome equivalent) copies and HBVgenome equivalent copies

HDVViral RNA was isolated with Trizol reagent ollowing manuacturers instructions. Total RNA was

reverse transcribed into cDNA with random primers (PrimeScript RT kit; Takara) and 2 l o thecDNA was used or real-time PCR assay. Primers or quantiying HDV total RNA or genome equivalent

copies are complemented with the delta antigen coding region o HDV RNA genome: orward

primer HDV-1184F, 5-TCTTCCTCGGTCAACCTCTT-3, and backward primer HDV-1307R

5-ACAAGGAGAGGCAGGATCAC-3 .

HBVViral DNA was isolated by standard genomic DNA isolation method. The DNA was quantied using spe

cic primers: 5-GAGTGTGGATTCGCACTCC-3 (orward) and 5-GAGGCGAGGGAGTTCTTCT-3

(backward) by real-time PCR. The viral genome equivalent copies were calculated based on a standard

curve generated with known copy numbers. Real-time PCRs were perormed using SYBR Premix

Ex Taq kit on an ABI Fast 7500 real-time system instrument (Applied Biosystems, United States).

HDV binding and inhibition assaysWith 5 107 copies o genome equivalent HDV, 1 105 o target cells were incubated at 16C or 4 h

in the presence o 4% PEG8000, ollowed by extensive wash with cold PBS or our times. The cells

were then lysed directly with Trizol reagent and ollowed by reverse transcription. RNA copy numbers

o viral genome and internal control glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA

were determined by real-time PCR. For HDV binding inhibition assay, peptides were pre-incubated

with target cells at 16C or 1 hr beore incubating with the virus; antibodies against viral envelope

protein were pre-incubated with viruses beore adding to target cells.

HDV and HBV inection and inhibition assaysViral inections o HDV and HBV were conducted in 48-well plates at multiplicities o genome equiva

lents o 500 and 100, respectively. Normally, 5 107 copies o genome equivalent HDV or 1 10

copies o genome equivalent HBV were inoculated in the presence or absence o entry inhibitors with

1 105 cells and incubated or 16 hr except otherwise indicated. Cells were then washed with mediumor three times and maintained in PMM medium with medium change every 23 days. For HDV inec

tion, 4% PEG 8000 was present during the 16 hours viral inoculation period similarly as described by

Barrera et al. (2004). Viral inection at dierent time points was analyzed by measuring viral DNA/

RNAs and viral antigen expression. Quantitative real time RT-PCR was used to quantiy HDV tota

RNAs, strand-specic real time RT-PCR to determine copies o HDV genome and antigenome RNA

(see below). For HBV viral inection on HepaRG cells and PHH, 4% PEG 8000 was present during

the inoculation period as previously described by Gripon et al. (Gripon et al., 1993; Gripon et al.

2002; Schulze et al., 2007). Viral inection o PTH was conducted in the absence o PEG8000. Culture

medium was changed every 23 days. Secreted HBsAg and/or HBeAg were determined with com-

mercial ELISA kits. Real-time PCR, with or without a prior reverse transcription step, was used o

quantication o HBV-specic 3.5 kb pre-C and pregenomic RNA, total HBV sub-genomic RNA, and

HBV cccDNA copies.

Quantication o HDV genome and antigenome RNAs bystrand-specic real time RT-PCRThe strand-specic qRT-PCR was perormed as previously described by Freitas et al. (2012). Briefy

the genomic and antigenomic RNAs were reverse transcribed separately with strand-specic primers

into cDNAs: primer HDV398R (5-CGCTTCGGTCTCCTCTAACT-3) or genomic RNA; primer HDV288F

(5-GCAGACAAATCACCTCCAGA-3) or antigenomic RNA. The reverse transcribed cDNAs o genmo-

mic or antigenmoic HDV RNAs were used as templates or real-time PCR using the HDV398R and

HDV288F primer pair. TaqMan probe was 5 FAM-AGAGCTCTGACGCGCGAGGAGTAAGC-TAMRA 3

Real-time PCR assays were conducted with an ABI Fast 7500 real-time PCR instrument.


20/28



Research article

Quantication o HBV-specic RNAsTotal RNA rom HBV-inected cells was isolated with Trizol reagent (Invitrogen). About 400 ng tota

RNA was reverse transcribed into cDNA with PrimeScript RT kit (Takara) in a 10 l reaction. cDNA

derived rom 20 ng total RNA was used as template or real-time PCR amplication. In a separate real

time PCR reaction, 20 ng o total RNA was directly used as template to assess the possible HBV vira

DNA contamination in the RNA preparation. Primers (HBV2270F: 5-GAGTGTGGATTCGCACTCC-3

and (HBV2392R: 5-GAGGCGAGGGAGTTCTTCT-3) were used or HBV 3.5 kb transcripts

(HBV1805F: 5-TCACCAGCACCATGCAAC-3) and (HBV1896R: 5-AAGCCACCCAAGGCACAG-3

were or total HBV-specic transcripts. Amplication o 123-bp ragment or 3.5 kb transcripts and

92-bp product or total HBV-specic transcripts were both conducted by denaturation at 95C or 30 s,

ollowed by 40 cycles o 95C denaturation or 3 s, and 60C annealing/elongation or 30 s. Real-time

PCR was perormed using SYBR Premix Ex Taq kit on an ABI Fast 7500 real-time system instrument.

Real-time PCR using either set o the primers generated highly specic amplication product. HBV

RNA copy numbers were deduced rom a standard curve generated rom known nucleic acid

quantities. Then the HBV RNA copy number per nanogram RNA in the inected cell cultures was

calculated by subtracting the background amplication noise derived rom the viral DNA contam

ination in the RNA preparation rom that cDNA amplication. The signal-to-noise ratio or HBV tota

transcripts is usually 50, and or 3.5 kb transcripts 20. The real-time PCR detection limits o

total HBV-specic transcripts and 3.5 kb transcripts are 0.5 and 3.5 copies per nanogram cellula

total RNA, respectively.

Southern blot analysis o HBV covalently closed circular DNA (cccDNA)HBV cccDNA Southern blot was conducted ollowing a similar procedure as described by Summers et al

(1990)with modications. Briefy, to selectively extract HBV cccDNA, inected hepG2-NTCP cells in

6-cm dishes were lysed with 1 ml lysis buer at 37C or 60 min, ollowed by addition o 0.25 ml o

2.5 M KCl and incubation at 4C overnight. The lysis buer was not supplemented with proteinase K

containing 50 mM TrisHCl, pH 7.4, 10 mM EDTA, 150 mM NaCl, 1% SDS. The lysate was then claried

by centriugation at 12,000 g or 30 min at 4C and extracted with phenol and phenol:chloroorm

DNA was precipitated with equal volume o isopropanol in the presence o 20 g glycogen (Roche)

and nally dissolved in TE buer. The prepared DNA sample was then treated with plasmid-sae

adenosine triphosphate (ATP)-dependent deoxyribonuclease DNase (Epicentre Technologies) ollow

ing manuacturers instructions. For Southern blotting, the plasmid-sae DNase-treated DNA was

separated on a 1.3% agarose gel and then transerred to a nylon membrane (Hybond-N+; Amersham

using a standard neutral transer procedure. A 3280-bp plasmid constructed by inserting a 588-bpHBV DNA ragment (rom 1805 to 2392, genotype D, Southern blot probe) into a 2692 bp pMD18T

vector (Takara) was also run on the same agarose gel to serve as the molecular marker or cccDNA in

Southern blot analysis. The plasmid is o similar size o HBV genome and was mainly in supercoiled

orm; thereore it runs at similar size as HBV cccDNA in agarose gel. The nylon membrane was hybrid-

ized with a [-32P] dCTP-labeled HBV probe (genotype D HBV DNA ragment rom 1805 to 2392

prepared by random primer DNA labeling kit (Ver.2.0; Takara). Hybridization was carried out in 7 ml o

Perect Hyb Plus Hybridization Buer (Sigma) with 1 hr pre-hybridization, ollowed by overnight

hybridization at 67C. The membrane was then washed once with 2 SSC/0.1% SDS, 1 SSC/0.1%

SDS, and 0.5 SSC/0.1% SDS at 67C or 20 min, respectively. Finally, the membrane was subjected to

autoradiographic exposure.

Quantication o HBV cccDNA by qPCR

HBV cccDNA (double-stranded DNA without nick and gap) was quantied by real-time PCR usinga protocol as previously described (Bowden et al., 2004; Werle-Lapostolle et al., 2004) with

modications. In particular, specic primers or cccDNA detection reported by Glebe et al. (2003

(ccc-1582F: 5-TGCACTTCGCTTCACCT-3; ccc-2316R: 5-GACCACCAAATGCCCCT-3) were

validated and used or quantiying copy numbers o cccDNA using real-time PCR. Viral DNAs othe

than cccDNA, including single-stranded and relaxed circular DNAs, were degraded prior to ampli-

cation by treatment o the DNA templates with plasmid-sae adenosine triphosphate (ATP)

dependent deoxyribonuclease DNase (Epicentre Technologies). In brie, HBV-inected cells were

lysed or 4 hr at 65C in lysis buer (50 mM TrisHCl, pH 8.0, 50 mM EDTA, 100 mM NaCl, 1% SDS)

supplemented with proteinase K (200 g/ml) and ollowed by phenol-chloroorm extraction. A tota


21/28



Research article

o 250 ng o the extracted DNA was digested with 510 units plasmid-sae DNase in a 50 l volume

or 8 hr at 37C ollowed by DNase inactivation at 70C or 30 min. 2 l o the 50 l reaction was

then added to 20 l o a real-time PCR reaction. Amplication o 735 bp cccDNA product was

conducted by denaturation at 95C or 5 min, ollowed by 45 cycles o denaturation at 95C or

30 s, 62C annealing or 25 s, and 72C elongation or 45 s. HBV cccDNA copy numbers were

calculated with a standard curve rom plasmid with known nucleic acid quantities. The detection

limit or cccDNA is 10 copies cccDNA per reaction (equivalent to 10 ng o total cell lysate DNA)

Real-time PCR was perormed with SYBR Premix Ex Taq kit on an ABI Fast 7500 real-time system

instrument.

PTH cDNA library construction, deep sequencing oTupaiatranscriptome, and bioinormatics analysis o Illumina deepsequencing-determined transcriptomePrimary Tupaia hepatocytes were isolated as described above. PTH mRNA was puried rom 10 g o

total RNA using Oligo-dT magnetic beads. The mRNA was ragmented into small pieces by incubation

with divalent cations at 94C or exactly 5 min. The rst strand cDNA was synthesized using random

primers and SuperScript II reverse transcriptase (Invitrogen) with ragmented mRNAs. RNA template

was then removed by RNase H, and double-stranded cDNA was prepared with DNA polymerase I

cDNA with blunt ends was created by T4 DNA polymerase and Klenow DNA polymerase and an A

base was subsequently added to the 3 end o the blunt phosphorylated DNA ragments by Klenow

ragment (3 to 5 exo minus). The cDNA was then ligated with adapters and then ran on a 2%

agarose gel. The ragments with a size range rom 200 25 bp were puried, ollowed by amplication using the manuacturers primers. The PCR products were then puried using QIAquick PCR

purication Kit (Qiagen), quantied and diluted or cluster generation and deep sequencing. The

72-cycle pair-end sequencing was perormed with Sequencing Kits (Version 5) on an Illumina Genome

Analyzer IIx (Illumina, San Diego, United States). Illumina CASAVA pipeline v1.8.1 was used or

sequence extraction and ltering.

Bioinormatics analysis o Illumina deep sequencing-determinedtranscriptome o primary Tupaiahepatocytes

De novo reconstruction o transcriptome rom cDNA library deepsequencing dataTotal 253,919,616-pair 72 nt sequences with 36.6G base rom the sequencing results o the hepato-

cyte cDNA library described above were ed to Trinity (Grabherr et al., 2011) r20110519 using paiend RNA-seq protocol, with which 209,063 transcripts o average length o 1421 nt (minimum 300 nt

maximum 21,043 nt, and scaold N50 o 3674 nt) were generated.

Protein sequences identifcation and annotationFollowing assembly, GENSCAN (Burge and Karlin, 1997) was used with deault parameters to

identiy coding sequences and the encoded protein sequences o these transcripts. A total o

91,479 protein sequences were identied. Each chosen protein sequence was rst annotated with

its corresponding blastp (Camacho et al., 2009) matches rom NCBI human protein sequences

Those not annotated in the rst step were then submitted or similar annotation process with

UniprotKB human proteome and NCBI nonredundant protein sequence database. Protein

sequences that were not annotated by previous steps were submitted or annotation with their

corresponding transcripts. Protein sequences with their corresponding transcripts that can be

annotated by the blastx (Camacho et al., 2009) hits o NCBI human protein sequences, UniProtKBhuman proteome or NCBI nonredundant protein sequence database were annotated with these

hits rom transcripts.

Generation oTupaia hepatocyte protein sequences databaseAll identied protein sequences were included in the hepatocyte protein sequences database.

The protein sequences were labeled with corresponding unctional annotation results. Any identi

ed protein sequences that were not successully annotated are labeled with Uncharacterized

protein. Total 50,951 annotated and 40,528 uncharacterized protein sequences generated rom

the Tupaia hepatocytes transcriptome were combined into the database. A numeric ID (gi) was


22/28



Research article

generated or each protein sequence in the database. The corresponding cDNA sequences were

deposited to NCBI Transcriptome Shotgun Assembly (TSA) database o GenBank with accession

number rom JU120276 to JU170736 ater removal o cDNAs shorter than 200 bp and a vector

sequence.

Analysis o protein classes in Tupaia hepatocytesPANTHER database (Mi et al., 2005) was used to determine the protein class distribution o

annotated transcripts and protein sequences in primary Tupaia hepatocytes (PTHs) generated in

this study and transcriptome rom primary human hepatocytes (PHHs) reported by Hart et al(2010).

Photo-cross-linking with peptide ligand and tandem purication o thetarget molecule(s)L-photo-leucine-bearing wild-type bait peptide (WTb) or control bait peptide (N9Kb) was dissolved

in DMSO in dark. L-photo-leucine contains a photoactivatable diazirine ring, irradiation o UV light

at 365 nm induces a loss o nitrogen o the diazirine ring, and yields a reactive carbene group with

short hal-lie or covalent cross-linking at nearly zero distance. For tandem purication, WTb or N9K

at indicated concentrations was applied to 1 107 hepatocytes plated on collagen-coated dishes

Cells were cross-linked by UV irradiation and then washed to remove residual ree peptides and

subsequently lysed with 1 ml radioimmunoprecipitation assay (RIPA, pH 7.4) buer containing 20 mM

Tris, 150 mM NaCl, 0.1% SDS; 0.5% sodium deoxycholate, 1% NP40, and 1 protease inhibitor cocktail (Roche). The cell lysates were precipitated with 100 l streptavidin T1 magnetic beads, eluted

with 50 l nonreducing SDS-PAGE loading buer and then diluted with cold RIPA buer to a na

volume o 1 ml and was precipitated with 100 l (1 108) 2D3-conjugated M-270 dynabeads, then

eluted with 100 l nonreducing loading buer. The elute, with or without PNGase F treatment, was

diluted to 1 ml with RIPA buer and was precipitated again with 100 l streptavidin T1 magnetic

beads, ollowed by extensive washing, and nally eluted by boiling 5 min with 20 l SDS-PAGE load

ing buer. The samples were then analyzed with 12% SDSPAGE and silver staining. For photo

cross-linking analysis o primary cells, cell lines, or NTCP- or control plasmid-transected Huh-7 or

293T cells, WTb or N9Kb bait peptide in the presence or absence o competing peptide was applied

to 2 106 cells, photo-cross-linking was conducted similarly as described above. The cross-linked

samples were precipitated with streptavidin T1 magnetic beads and separated by SDS-PAGE, and

ollowed by Western blotting with mAb 2D3 (recognizing bait peptides) or mAb 1D4 agains

C-terminal tag C9.

LC-MS/MS and data analysisSilver stained gel bands were cut, ollowed by in-gel reduction, alkylation, and trypsin digestion

as previously described (Shevchenko et al., 2006). Digested peptide mixtures containing 0.1%

ormic acid were loaded onto a 4 cm, 75-m inner diameter used silica capillary column packed

with 10-m YMC C18 material (YMC, Kyoto, Japan). Ater desalting, sampl

Sodium taurocholate cotransporting polypeptide is a functional receptor for human hepatitis B and D...

Documents

Transcript of Sodium taurocholate cotransporting polypeptide is a functional receptor for human hepatitis B and D...