Gene Therapy (2000) 7, 750–758
 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
www.nature.com/gt
NONVIRAL TRANSFER TECHNOLOGY
RESEARCH ARTICLE
Novel cell permeable motif derived from the
PreS2-domain of hepatitis-B virus surface antigens
S Oess1,2 and E Hildt2,3
1
Max-Planck-Institut für Biochemie, AG Virusforschung, Martinsried; 2Institut für Experimentelle Onkologie und
Therapieforschung der Technischen Universität München; and 3Robert Koch-Institut, Berlin, Germany
Efficient transfer of proteins or nucleic acids across cellular
membranes is a major problem in cell biology. Recently the
existence of a fusogenic sequence was predicted in the junction area of the PreS2- and S-domain of the hepatitis-B virus
surface antigens. We have identified cell permeability as a
novel property of the PreS2-domain. Cell permeability of
PreS2 is not restricted to hepatocytes. PreS2 translocates
in an energy-independent manner into cells and is evenly
distributed over the cytosol. Detailed analysis revealed that
cell-permeability is mediated by an amphipatic ␣-helix
between amino acids 41 and 52 of PreS2. Destruction of this
translocation motif (PreS2-TLM) abolishes cell permeability.
PreS2-TLM per se can act as a shuttle for peptides and
functional proteins (such as EGFP). This permits the highly
specific modulation of intracellular signal transduction by
transfer of peptides competing protein–protein interactions
as demonstrated by specific inhibition of TNF␣-dependent
activation of c-Raf-1 kinase. Moreover in vivo functionality
was demonstrated by PreS2-TLM-dependent protein transfer into primary bone marrow cells and into the liver. The
amphipatic motif is conserved between the different hepatitis-B virus subtypes, and the surface proteins of avian and
rodent hepadnaviruses exhibit similar amphipatic peptide
sequences. In respect to hepatitis-B virus-infection, the
PreS2-TLM could represent the postulated fusion peptide
and play a crucial role in the internalization of the viral
particle. Gene Therapy (2000) 7, 750–758.
Keywords: cell permeable peptides; hepatitis-B virus; PreS2; signal transduction
Introduction
Recently a few peptide motifs have been identified conferring the unexpected property of cell permeability to
proteins. These motifs are derived from the third helix
of the antennapedia homeodomain (pAntp),1 the human
immunodeficiency virus (HIV) protein Tat,2 the Kaposi
fibroblast growth factor (K-FGF)3 and the herpes simplex
virus type-1 (HSV-1) structural protein VP22.4,5 While in
the case of pAntp 16 predominantly basic amino acids,
which adopt an amphipatic a-helix, mediate the cell permeability,1 an unstructured cluster of nine amino acids is
responsible for the cell permeability of the HIV-Tat protein. The cell permeability mediating domain derived
from K-FGF consists of 12–16 hydrophobic amino acids.
Within VP22 no clearly defined cell permeable peptide
motif has been identified so far.2–4 For none of these cell
permeable proteins the translocation mechanism is completely understood but an unspecific interaction with
membrane phospholipids, eventually resulting in
inverted micelle formation and subsequent release of the
peptides into the cytosol has been postulated.6,7
Based on the observation that cells non-permissive for
hepatitis-B virus (HBV) infection could be infected after
partial proteolytic cleavage of the viral particle, the exist-
Correspondence: E Hildt, Robert-Koch-Institut, Am Nordufer 20, D13353 Berlin, Germany
Received 20 December 1999; accepted 5 January 2000
ence of a fusogenic peptide in the PreS2-domain of the
viral surface proteins has been postulated.8,9 HBV is the
smallest human pathogenic DNA virus and belongs to
the family of hepadnaviridae which infect hepatocytes
with high specificity. The gene encoding for the hepatitisB virus surface proteins consists of a single open reading
frame, which is divided by three in frame AUG-start
codons into the preS1-, the preS2- and the S-region. By
initiation at each of the three AUG-start codons, the large
(LHBs), the middle (MHBs) and the small hepatitis-B
virus surface protein (SHBs) are synthesized. In addition
to their role as structural proteins of the viral envelope,
a transcriptional activator function has been ascribed to
the PreS2 domain. Prerequisite for the activator function
is the cytoplasmic orientation of the PreS2-domain, which
can be observed in the case of LHBs10–12 and of the Cterminally truncated MHBs proteins (MHBst).13,14 The
cytoplasmic PreS2 domain is a binding partner and activator of the protein kinase C (PKC). The activation of
PKC is mediated to the nucleus via the c-Raf-1/MAP2kinase signal transduction pathway resulting, for
example, in the activation of the transcription factors
AP-1 and NF-␬B.15
Here we describe cell permeability as a novel and
unexpected property of the PreS2 domain. The PreS2
domain harbors a short, amphipatic ␣-helical peptide,
which confers cell permeability to PreS proteins and
unrelated proteins when fused to it.
Cell permeable peptide from HBV
S Oess and E Hildt
Results
The PreS2 domain is a cell permeable protein
In order to study the cell permeability of the PreS2
domain, highly purified PreS2 protein was isolated from
an eubacterial expression system as a fusion protein harboring a N-terminal hexa-his-tag and a C-terminal extension of eight amino acids derived from the S-region
(Table 1). This protein, also termed MHBst63, belongs to
the family of PreS2 activators and exhibits transcriptional
activator function.15
For analysis of the cell permeability of the PreS2
domain HeLa (Figure 1a, b) and 293 cells (Figure 1d, e)
were grown for 30 and 120 min in the presence of 2.5 ␮m
PreS2 protein in the culture medium. After extensive
washing of the cells for removal of unspecifically bound
protein the internalization was investigated by immunofluorescence microscopy of the fixed cells using a PreS2specific antiserum. The immunofluorescence shows that
the PreS2 protein is present in 100% of the cultured cells,
demonstrating a high translocation efficiency. The
internalized PreS2 protein is homogeneously distributed
over the cytoplasmic compartment, while the nucleus
remains unstained. There is no indication for an enrichment in subcellular structures (vesicles) involved in
endocytosis.
To confirm the cytosolic localization of the PreS2 protein, cell fractionation experiments of HepG2 cells grown
for 120 min in the presence or absence of PreS2 protein
were performed (Figure 2). The cytosolic fraction was
analyzed by Western blotting using a PreS2-specific antiserum. The Western blot confirms the presence of PreS2
in the cytosolic fraction.
To prove unequivocally the presence of PreS2 in the
cytosolic compartment and to investigate whether the
membrane translocation affects its above-mentioned activator function, electrophoretic mobility shift assays
(EMSA) were carried out. The activator function of PreS2
depends on its cytoplasmic localization and can be monitored by the induction of AP-1 or NF-␬B.
Activation of AP-1 and NF-␬B was determined using
nuclear extracts derived from 293 cells exposed to
exogenous PreS2 protein (Figure 3). The EMSA shows
that after addition to the culture medium the PreS2 protein triggers the significant induction of AP-1 and NF-␬B
reflecting the translocation of PreS2 into the cytoplasm.
These results indicate that (1) the PreS2 domain is a
novel cell permeable protein, which (2) translocates with
high efficiency (3) into the cytoplasmic compartment,
and that (4) the cell permeability is not restricted to
hepatocytes.
751
PreS2-mediated membrane translocation is a passive,
energy independent process
The most frequently facilitated way for proteins to enter
living cells is that of receptor-mediated endocytosis. This
process is energy dependent and completely abolished
when the cells are incubated at 4°C.16 To investigate
whether PreS2-mediated cell permeability requires
energy, we tested the translocation ability of the complete
PreS1-PreS2 domain (169 amino acids in length, Table 1)
at 37°C and 4°C into HepG2 and 293 cells. At the same
time this allowed us to get an idea about a possible size
limitation for efficient membrane transport. For a detailed
analysis of the intracellular localization, pure cytosol was
prepared using a potter homogenisor to avoid disruption
of subcellular organelles and subsequent differential
ultracentrifugation of the homogenate. Western blot
analysis of the cytosolic fraction derived from HepG2 and
293 cells incubated at 4°C or 37°C cells (Figure 4) showed
Table 1 Schematic presentation of recombinant proteins and synthetic peptides used in this study
The shaded box resembles the amphipatic ␣-helix between aa 42–51 of PreS2 or PreS2-TLM, respectively, and the sequence is given in the
one letter amino acid code.
Gene Therapy
Cell permeable peptide from HBV
S Oess and E Hildt
752
a
d
b
c
f
« e
g
i
« h
Figure 1 PreS2 cell permeability is not restricted to hepatocytes. (a–i) Immunofluorescence of HeLa cells (top panels a–c) after 30 min incubation and
293 cells (middle panels d–f) after 120 min incubation (a) without exogenous peptide; (d) with unrelated peptide (2.5 ␮m); (b,e) PreS2 (2.5 ␮m); (c,f)
PreS2–3S-mutant (2.5 ␮m); (g–i) Immunofluorescence of HepG2 cells after 60 min incubation (g) without exogenous peptide at 37°C; (h) with 2 ␮m
PreS1-PreS2 at 37°C and (i) with 2 ␮m PreS1-PreS2 at 4°C.
that the cell permeability of the PreS1-PreS2 protein is not
affected by low temperatures and confirmed the cytosolic
localization. This was corroborated by indirect immunofluorescence of HepG2 cells (Figure 1g–i) showing that
the entire PreS1-PreS2 domain is indeed cell permeable
at 37°C, as well as at 4°C. These results demonstrate that
a protein of approximately 20 kDa is transported
efficiently and that translocation is a passive process, not
requiring energy. This argues against classic receptormediated endocytosis.
Gene Therapy
An amphipatic ␣-helix mediates cell permeability of the
PreS2-domain
Having established the ability of the PreS2 domain to
translocate through biological membranes, we strived to
identify the motif within PreS2 responsible for cell permeability. The amphipatic character of cell permeable
peptide motifs has repeatedly been described as a
requirement for their ability to translocate via biological
membranes.1,6 In the case of PreS2, previous mutagenesis
studies and CD spectroscopy revealed the existence of an
Cell permeable peptide from HBV
S Oess and E Hildt
753
Figure 2 The PreS2 domain translocates into the cytosol. Western blot
analysis of HepG2 cytosolic lysates. Cells were incubated with PreS2
(2 ␮m) at 37°C for 120 min and the lysates tested for the presence of
PreS2 using a PreS2-specific primary antibody (HBV 25–19). The same
results were obtained with a primary antibody directed against the Nterminal hexa-his-tag (data not shown). Similar results were obtained
using 293 cells (data not shown).
amphipatic ␣-helix between amino acids 41 and 52 of the
PreS2 domain (PLSSIFSRIGDP; subtype ayw).13 Moreover, this amphipatic ␣-helix has been shown to mediate
dimerization of PreS2. Replacement of three amino acid
residues within the amphipatic ␣-helix by serine (PreS2–
3S-mutant; PSSSSSSRIGDP; Table 1) resulted in a loss of
the ␣-helical structure and lead to ␤-sheet conformation
of the sequence of interest, thus considerably reducing
the oligomerization tendency of PreS2.13
In order to analyze the relevance of this amphipatic ␣helical motif for the membrane translocation of PreS2, we
compared the cell permeability of wtPreS2 protein and
the PreS2–3S mutant into HeLa and 293 cells. The internalization of the proteins was determined by immunofluorescence microscopy (Figure 1) and Western blot
analysis of cellular lysate (Figure 5). The immunofluorescence microscopy as well as the Western blot analysis
show that cell permeability of the PreS2–3S mutant
(Figure 1c, f) was drastically reduced as compared with
wt-PreS2 (Figure 1b, e). These results demonstrate that
the amphipatic ␣-helix between amino acids 41 and 52 is
a prerequisite for the cell permeability of the PreS2
domain. Hence it is termed PreS2-translocation-motif
(PreS2-TLM or TLM) in the following.
The PreS2-TLM alone is sufficient to mediate cell
permeability and allows highly specific modulation of
signal transduction
In order to investigate whether the PreS2-TLM is sufficient to mediate the observed cell permeability and to
test the potential of PreS2-TLM to act as a carrier for other
peptides we designed a synthetic fusion peptide con-
Figure 3 Internalized PreS2 triggers the activation of intracellular signal
transduction pathways (a) EMSA using NF-kB consensus oligonucleotides in nuclear extracts of 293 cells after 60 min and 210 min incubation
with PreS2 (2 ␮m). TPA stimulated cells served as positive control
unstimulated cells as negative control. Specificity of the interaction of NF␬B with its consensus motif was confirmed by competition with unlabelled
oligonucleotides with increasing concentration (from left to right: 5×, 10×,
25×, 50× that of 32P-labelled oligonucleotides). (b) EMSA using AP-1
consensus oligonucleotides in nuclear extracts of 293 cells after 10 min
and 210 min incubation with PreS2 (2 ␮m). TPA-stimulated cells served
as positive control, unstimulated cells as negative control.
Figure 4 PreS2-mediated translocation across the plasma membrane into
the cytoplasm is an energy-independent process. Western blot analysis of
the cytosolic fraction of HepG2 and 293 cells after 10 min and 30 min
incubation with PreS1-PreS2 at concentrations of 2 ␮m at 37°C and 4°C.
Western blot analysis was carried out using a primary antibody mixture
(Q19–10, F124, HBV 116). The upper band in the left panel correlates
with the PreS1-PreS2 dimer.
sisting of PreS2-TLM linked to a C-terminal PLAP motif
derived from the cytoplasmic domain of TNF-RI
(Table 1). In a previous investigation,17 we observed that
this PLAP motif mediates the interaction of TNF-RI with
the C-terminal SH3 domain of the adapter molecule
Grb2. This Grb2/TNF-RI-interaction is a prerequisite for
the TNF␣-dependent activation of c-Raf-1 kinase.17
Gene Therapy
Cell permeable peptide from HBV
S Oess and E Hildt
754
Figure 5 The cell permeability of the PreS2-domain is mediated by an
amphipatic a-helix. Western blot analysis of cellular lysates of HepG2 cells
after 10 min (lanes 2 and 3) and 30 min incubation (lanes 4 and 5) with
PreS2 (lanes 2 and 4) or PreS2–3S-mutant (lanes 3 and 5) at concentrations of 2 ␮m. Untreated HepG2 cells served as negative control (lane
1). Western blot analysis was carried out using a primary antibody
directed against the hexa-his-tag. Identical concentrations of the PreS2
and PreS2–3S-mutant gave equivalent signals in Western blot analysis
(data not shown). The same result was obtained using a PreS2-specific
primary antibody (data not shown).
To determine the effect of the fusion peptide on c-Raf1 kinase activity, HeLa cells were grown in the absence
or presence of the fusion peptide and stimulated with
TNF␣. The activity of c-Raf-1 kinase was determined by
immunocomplex assays as described under Materials
and methods. The immunocomplex assay shows that the
fusion peptide abolishes the TNF␣-dependent activation
of c-Raf-1 kinase (Figure 6). To confirm the specificity of
this effect additional controls were performed. The presence of a mutated fusion peptide harboring a KLAP motif
instead of the PLAP motif did not significantly affect the
TNF␣-dependent activation of c-Raf-1 kinase. The EGFdependent activation of c-Raf-1 kinase is mediated by the
interaction of the middle SH2 domain of Grb2 with a
phosphotyrosine residue of the EGF receptor. The EGFdependent activation of c-Raf-1 kinase was not significantly impaired by the presence of the cell permeable
peptides.
These results indicate (1) that the PreS2-TLM per se is
sufficient to translocate across membranes and (2) demonstrate its potential to act as a carrier for other peptides
permitting a specific modulation of intracellular signal
transduction.
PreS2-TLM can act as a carrier for functional proteins
into mammalian cell lines, primary cells and plant tissue
culture
In order to test the general translocation capacity of the
PreS2-TLM we tested the membrane translocation of
highly-purified 6H-TLM-EGFP fusion protein (27 kDa),
isolated from an E. coli expression system. For control
experiments a hexa-his-tagged EGFP-protein lacking the
PreS2-TLM was used. The fluorescence microscopy of 293
cells incubated for 30 min with 6H-TLM-EGFP
(Figure 7a) or with 6H-EGFP (Figure 7b) shows a significant fluorescence of the intracellular compartment of the
cells incubated with 6H-TLM-EGFP, whereas the cells
incubated with the 6H-EGFP, which lacks the translocation motif, remain unstained. These results demonstrate that the 12 amino acid sequence PreS2-TLM per se
is sufficient to confer cell permeability to a functional
protein of 27 kDa in vitro.
To investigate the capacity of the novel PreS2-TLM to
mediate a protein transfer into primary cells, primary
mouse bone marrow cells were incubated in the presence
or absence of 6H-TLM-EGFP for 60 min. The fluorescence
microscopy (Figure 7c, d) shows that 6H-TLM-EGFP
translocated into different species of the primary bone
marrow cells, indicating that the PreS2-TLM mediated
cell permeability is not restricted to immortalized cell
lines.
In view of the fact that PreS2-TLM mediated cell permeability is not targeted to certain cell types or species,
we wanted to investigate whether membrane translocation resembles a general mechanism which might also
occur in plant cells. For this purpose we tested the cell
permeability of wtPreS2 and the PreS2-TLM-deficient
mutant PreS2–3S into callus cultures of Lupinus albus.
Since complete lysis of the cells required considerable
dilution, hexa-his-tagged proteins were enriched by passing the lysates over a Ni-NTA-affinity column; specifically bound proteins were eluted by imidazole as
described under Materials and methods. The eluates
were analyzed by Western blotting using a PreS2-specific
antiserum. The Western blot (Figure 7e) shows that only
in case of the eluate obtained from plant cells incubated
with wtPreS2, PreS2-specific proteins were detectable,
whereas no specific signal was obtained in case of the
cells incubated with the PreS2–3S mutant. This confirms
the specificity of the observed internalization.
These results indicate that (1) the novel PreS2-TLM is
able to act as a carrier for functional proteins and (2) that
the carrier function is not restricted to immortalized
mammalian cells.
Figure 6 The PreS2-TLM can act as a carrier for other peptides modulating intracellular signal transduction pathways. Immunocomplex assay of cRaf-1 kinase activity in HeLa cells after treatment with cell permeable PreS2-TLM peptides and TNF␣ or EGF, respectively. Induction of c-Raf-1 kinase
activity after TNF␣ (lane 2) and EGF (lane 3) treatment is abolished by addition of PreS2-TLM-PLAP (lanes 4 and 5), but not by PreS2-TLM-KLAP
(lanes 6 and 7) to TNF␣-stimulated cells, while EGF stimulated c-Raf-1 kinase activity is detectable after addition of either peptide (lanes 8 and 9).
Gene Therapy
Cell permeable peptide from HBV
S Oess and E Hildt
755
a
a
b
b
c
d
e
Figure 7 The carrier function of the PreS2-TLM is not restricted to mammalian cells. (a-e) Fluorescence microscopy of 293 cells after 30 min incubation with PreS2-TLM-EGFP (panel a) and EGFP (panel b) at concentrations of 2 ␮m. Fluorescence microscopy of primary mouse bone marrow
cells after 60 min incubation with PreS2-TLM-EGFP, 2 ␮m (panel c)
and control without exogenous protein (panel d). Western blot of affinity
chromatography-concentrated lysates of lupinus albus callus cultures after
180 min incubation with PreS2 (left) or PreS2–3S (right), respectively
(panel e), at concentrations of 2 ␮m. Western blot was carried out using
a PreS2-specific primary antibody (HBV25–19).
PreS2-TLM mediates cell-permeability in vivo
In the experiments described above it was observed that
the PreS2-TLM is able to translocate over the plasma
membrane of cultured and primary cells in vitro with
very high efficiency. The resulting question was whether
the PreS2-TLM also permits the in vivo transfer into intact
organs. For this purpose wt-PreS2 and the PreS2–3S
mutants which lack a functional PreS2-TLM, were
infused into the portal vein of anesthetized rats and the
livers thereafter were removed for immunohistochemical
analysis using a PreS2-specific antiserum. The immunofluorescence microscopy (Figure 8) of liver sections
derived from rats infused with wtPreS2-protein
(Figure 8b) shows a cytoplasmic staining of various cell
types. The staining is not restricted to the endothelium
cells or cells surrounding the capillaries. The wtPreS2
protein is detectable in cell layers distant from the capillaries. The staining pattern of the sections derived from
rats infused with the PreS2–3S mutant (Figure 8a) shows
only an unspecific fluorescence of structures also stained
in livers derived from buffer-treated animals (data not
shown) and of extracellular structures.
These results demonstrate the functionality of the
PreS2-TLM in vivo and show that the PreS2-TLM enables
the internalized protein to spread from cell to cell.
Figure 8 PreS2-TLM dependent membrane translocation in vivo. Immunofluorescence of rat liver sections after infusion of PreS2 (a) or PreS2–
3S-mutant (b) at concentrations of 10 ␮m through the portal vein in vivo.
Indirect immunofluorescence was carried out using a primary antibody
directed against the PreS2-domain.
Discussion
In this study we describe the identification of a novel cell
permeable peptide which allows the transport of biologically active compounds across the plasma membrane into
cells. To date, several members of the heterogeneous family of cell permeable proteins have been described: the
third helix of the antennapedia homeodomain (pAntp),
the HIV-Tat protein, the K-FGF and the HSV-1 structural
protein VP22.5,18–24 While in the case of pAntp, an amphipatic ␣-helix of 16 amino acids mediates the cell permeability,1 and an unstructured cluster of nine basic
amino acids is responsible for the cell permeability of the
HIV-1 Tat protein,2 in case of VP224 and of K-FGF3 the
underlying structural motif is unknown.
The cell permeability of the PreS2-domain can be
clearly ascribed to an amphipatic ␣-helical peptide at the
C-terminus (aa 41–52). The novel cell permeable motif
derived from the PreS2 surface antigens of HBV
described in this study shares some striking features with
pAntp. Both proteins are transported over biological
membranes in an energy-independent manner, not
involving classical endocytosis. Like pAntp, PreS2 shows
a tendency to form dimers and tetramers in membrane
mimetic environments and the amphipatic ␣-helix
between amino acids 41 and 52 of PreS2 is involved in
dimerisation.13,25 The PreS2–3S mutant, where the amphipatic ␣-helical structure is replaced by a ␤-sheet conformation, exhibits a reduced oligomerisation tendency13
and no significant cell permeability.
This novel PreS2-derived translocation motif (PreS2Gene Therapy
Cell permeable peptide from HBV
S Oess and E Hildt
756
Figure 9 Amphipatic profiles are conserved in the PreS-region of all
hepadnaviruses. (a) Conservation of the amphipatic character of aa 42–51
of PreS2 in different HBV-subtypes. Schematic presentation of the distribution of hydrophilic and hydrophobic amino acids of the PreS2-TLM in
the subtype ayw (1) (blue) and other subtypes ayw (2) (green), adr (1),
adr (2), ayr (red) and adw and adw2 (grey). Y-axis: hydopathy-values
according to Kyte and Doolittle;28 positive values resemble hydrophobic,
negative values hydrophilic amino acid side chains; X-axis: 12 amino acids
of the PreS2-TLM of ayw (1) and the corresponding sequences in other
subtypes with the N-teminal Proline in position 1. (b) Amphipatic patterns and localisation of similar motifs in different avian and rodent
hepadnaviruses. Comparison of the hydropathy profiles of the PreS2-TLM
(dark blue) with sequences of the PreS-region of DHBV3 (red), HHBV
(green), WHV (light blue) and GSHV (yellow). Y-axis: hydopathy values
according to Ref. 28: positive values resemble hydrophobic, negative values
hydrophilic amino acid side chains; X-axis: 12 amino acids of the PreS2TLM of ayw (1) and the sequences with similar hydrophathy profiles with
their position within the PreS or PreS2-region in other hepadnaviruses.
TLM) per se is sufficient to act as a carrier for other
peptides/proteins across cellular membranes as demonstrated by the specific modulation of intracellular signal
transduction pathways. The PreS2-dependent activation
of AP-1 or NF-␬B requires the cytoplasmic localization of
the PreS2-domain.13,14 Moreover, the specific inhibition of
the TNF␣-dependent activation of c-Raf-1 kinase by a
PreS2-TLM-PLAP fusion peptide competing the
Grb2/TNF-RI interaction, which occurs in the cytosolic
compartment,17 confirms that the PreS2-TLM enables the
translocation across the membrane. The specificity and
effectiveness of these effects is demonstrated by the fact
that low concentrations of 2 ␮m protein or peptide added
to the culture medium are sufficient to exhibit these intracellular effects, whereas the respective mutants fail to
induce these effects.
PreS2- or PreS1-PreS2 proteins are found in the cytoplasm most probably due to their interaction with cytoplasmic binding partners like hsp90 or PKC15 (EH,
unpublished results). In contrast to the cytosolic localization of translocated PreS2- and PreS1-PreS2 domains, in
previous investigations we have observed an even distribution of the PreS2 domain over cytosol and nucleus
when expressed in mammalian and insect cells.13,25 Possibly this is due to the expression of large amounts of protein in excess to the cytosolic binding partners of PreS2
and reflects the capacity of PreS2 to enter the nucleus
when present in sufficient concentrations.
In contrast to the VP22-dependent translocation into
the nucleus,4 the PreS2-TLM-mediated transfer displays
no preference for any subcellular structures. The fusion
of PreS2-TLM to a protein harboring a nuclear localization signal results in a nucleoplasmic transport (EH,
unpublished results). The complete PreS1-PreS2 protein
is found in the cytoplasm. The small size of the PreS2TLM of 12 aa, its high transfer efficiency (about 100% of
the cells are affected), its low immunogenicity, the lack
of any activator function of the TLM, its spreading
capacity, and its defined structure function relation could
make it a helpful tool for the transfer of biomolecules into
cells or tissues.
In case of the antennapedia homeobox protein and the
HIV-Tat protein, it has been established that their cell
permeability is crucial for their biological function. This
leads to the question of the physiological role of PreS2TLM mediated cell permeability for the life cycle of HBV.
MHBs is not a crucial component of the viral envelope26
Gene Therapy
a
b
and it is not yet known whether the PreS2-dependent
transcriptional activator function is relevant for viral replication. Within the PreS2-domain a high conservation
can be observed for the PreS2-TLM between the different
HBV subtypes. Within the PreS2-TLM, seven amino acids
are conserved between the different subtypes and,
assuming an a-helical structure as in the case of subtype
ayw, a similar pattern of hydrophilic and hydrophobic
patches and hence conservation of the amphipatic character can be seen (Figure 9a). Very similar amphipatic profiles are found in the PreS domains of all investigated
hepadnaviruses, although the sequence homologies are
Cell permeable peptide from HBV
S Oess and E Hildt
weak (Figure 9b). Therefore we postulate, that the amphipatic character rather than the amino acid sequence itself
is conserved between the different hepadnaviridae. The
high degree of structural conservation of the PreS2-TLM
between the hepadnaviridae suggests an essential role for
the viral life cycle.
In the case of HBV, the PreS2-TLM is located in the
region of the previously postulated fusion peptide,8,9
which becomes exposed after partial proteolytic digestion
of the viral particle and seems to enable the infection of
non-permissive cells by these pre-digested viral particles.
We conclude that the PreS2-TLM is a good candidate for
the fusion peptide and that it is exposed to the viral surface following partial proteolysis of the viral surface proteins. It is tempting to speculate that during the HBV
infection, the liver cell specifically bound virus particle
is cleaved by a (receptor-immanent) proteolytic activity,
which results in a conformational change of the surface
proteins. This could result in the exposure of the PreS2TLM to the viral surface, permitting the internalization
of the viral particle from the cell surface.
Materials and methods
Recombinant production and purification of PreS2proteins
This was carried out as described previously.13,25
Recombinant production and purification of PreS2-TLMEGFP fusion proteins
The coding sequence for the PreS2-TLM was introduced
by a forward primer upstream to the coding sequence for
EGFP and amplified by PCR utilizing the plasmid
pEGFP-N1 (Clontech, Palo Alto, CA, USA). The PCR product was ligated in frame to a 5′-coding sequence for a
N-terminal hexa-his-tag into the vector pQe8 (Qiagen,
Hilden, Germany). Recombinant production and purification was carried out as described previously,13,25 with
the exception that sample application for affinity chromatography was carried out at pH 6.3 instead of 8.0.
Synthetic peptides
Synthesis of the peptides was performed as a step-wise
solid phase peptide synthesis method and purified by C18
reverse phase HPLC. Molecular weights of purified
peptides were verified by mass spectrometry.
Cell culture and protein incubation procedure
Mammalian cells (HepG2, 293, HeLa) were grown in Dulbecco’s modified Eagle’s medium (DMEM), 10% fetal calf
serum (FCS), penicillin/streptomycin, Na-pyruvate and
with the exception of immunofluorescence experiments,
cells were serum deprived for 16 h before commencing
the experiments. Recombinant PreS2-proteins were
diluted in 5 mm Na-acetate pH 5.5 in culture medium
(final pH 7.0) to final concentrations of 2 ␮m (for Western
blot and EMSA) or 2.5 ␮m (for immunofluorescence).
Recombinant PreS1-PreS2 proteins, EGFP fusion proteins
and synthetic peptides were diluted in DMEM, 0.5% FCS,
penicillin/streptomycin, Na-pyruvate to final concentrations of 2 ␮m (for Western blot and EMSA or immunocomplex
assays,
respectively)
or
2.5 ␮m
(for
immunofluorescence). Incubation was carried out for the
time period indicated in the results section at 37°C if not
stated otherwise.
Preparation of primary bone marrow cells
The bone marrow was prepared from the femur of 4month-old female mice. The cells were grown in DMEM,
20% FCS. Peptide incubation was performed as
described above.
757
Cell permeability studies in callus cultures of Lupinus
albus
Protein incubation was carried out at 16°C at concentrations of 2 ␮m. After 180 min incubation, cells were harvested, washed extensively and powdered under liquid
nitrogen. The cell powder was immediately suspended
in guanidine 6M, Tris 10 mm, Na-phosphate 100 mm, ␤mercaptoethanol 10 mm, pH 8.0 and lysed completely by
sonification. The lysate was cleared by centrifugation and
the supernatant subjected to affinity chromatography on
a Ni2+-NTA-column (Qiagen). Affinity chromatography
was carried out as described in Ref. 13 and eluates tested
for the presence of recombinant proteins by Western
blot analysis.
In vivo protein transfer
Portal vein infusion was carried out on anesthetized male
rats. After introduction of a cannula into the portal vein,
avoiding inhibition of the blood flow through the liver,
10 ␮m PreS2 or PreS2–3S, respectively, were infused
slowly over a period of 5 min (concentration is calculated
as peptide per plasma volume passing the liver during
the perfusion period of 10 min). Five minutes after termination of the peptide infusion, animals were killed, the
liver removed and immediately frozen in tissue freezing
medium (Jung, Berlin, Germany).
Cell fractionation and Western blot analysis
After protein incubation cells were harvested and
washed five times with cold phosphate buffered saline
(PBS) to remove excess protein. Cells were lysed by sonification and the lysate clarified by centrifugation. For cell
fractionation the cells were lysed by homogenization
using a Potter Elvjehm (Berlin, Germany) homogenisor.
Unbroken cells, lysosomes and nuclei were removed by
centrifugation at 10 000 g for 20 min. For preparation of
pure cytosolic fraction the supernatant was centrifuged
at 400 000 g for 18 min using a TLA100.2 rotor (Beckman,
Fullerton, USA). The lysate/cytosolic fraction were separated by SDS-PAGE. Western blot analysis was carried out
using a hexa-his-tag specific primary antibody (Qiagen)
or a primary antibody directed against PreS2 (HBV 25–
19, directed against amino acids 17–29 of PreS2; Q19–10
and F124, directed against the aminoterminal region of
PreS2) or a PreS1-specific antibody (HBV 116). As secondary anti body, horseradish peroxidase conjugated antimouse Ig F(ab′)2 fragment (Amersham) was used.
Electrophoretic mobility shift assays (EMSA)
EMSAs were carried out using double stranded AP-1 and
NF-␬B consensus oligonucleotides (Promega, Madison,
WI, USA) which were phosphorylated using polynucleotide kinase (Promega) and ␥-32P-ATP (Amersham-Pharmacia, Uppsala, Sweden). Preparation of nucleic extracts
was carried out according to the protocol of Dignam et
al.27 0.5 ng (50 000 c.p.m./ng) radiolabelled oligonucleotide were incubated with 10 ␮g of nuclear extracts, 2 ␮g
pdIdC, 20 ␮g acetylated bovine serum albumin and incubated for 30 min at room temperature. Gel electroGene Therapy
Cell permeable peptide from HBV
S Oess and E Hildt
758
phoresis was carried out on a non-denaturing, 4% polyacrylamide gel and the dry gel exposed to an
autoradiography film. As a positive control, cells were
stimulated with TPA (20 ng/ml) for 15 min and treated
accordingly.
Fluorescence microscopy
Following protein incubation of cells in vitro, indirect
immunofluorescence was carried out after washing the
cells with cold PBS three times and ethanol fixation. In
the case of PreS2 and PreS2–3S-mutant incubations a primary antibody directed against PreS2 (HBV 25–19) and
a Cy3-conjugated anti-mouse Fc (gamma) fragment
(Jackson ImmunoResearch, West Grove, PA, USA) were
used. PreS1-PreS2 was detected using a mixture of Q19–
10, F124 and HBV116. Following portal vein infusion,
cryocut sections of 5 ␮m were prepared, fixed in ethanol
and treated accordingly. Detection of EGFP fluorescence
was carried out on unfixed cells.
c-Raf-1 kinase assay
Serum-deprived HeLa cells were incubated with synthetic peptides for 2 h as described above and stimulated
with 100 units/ml TNF␣ or 5 ng/ml EGF, respectively,
for 15 min at 37°C. Together with the cytokines, 2 ␮m
synthetic peptides were added to avoid alteration of peptide concentrations. Activity of c-Raf-1 kinase was
determined by immunocomplex assay using recombinant
MEK (Santa Cruz Biotech, Santa Cruz, CA, USA) and
␥32P-ATP (Amersham) as substrates. Immunocomplex
assays were carried out as described in Ref. 17.
Acknowledgements
Antibodies were kindly provided by Dr L Mimms,
Chicago (HBV 25–19), Professor W Gerlich, Gie␤en (Q19–
10) and Dr A Budkowska, Paris (F124). We thank Dr C
Plank, Institute for Experimental Oncology, Munich, for
peptide synthesis, Professor M Wink, Department for
Pharmaceutical Biology, Heidelberg, for callus cultures of
lupinus albus and his kind invitation to carry out the
experiments in his laboratory and Dr B Munz, Stanford,
for helpful discussion. We are indebted to Professor PH
Hofschneider, Martinsried, for his continuous generous
support. This work was supported by the Deutsche Stiftung für Krebsforschung and the MMW (Munich).
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
References
1 Derossi D, Joliot AH, Chassaing G, Prochiantz A. The third helix
of the antennapedia homeodomain translocates through biological membranes. J Biol Chem 1994; 269: 10444–10450.
2 Vivès E, Brodin P, Lebleu B. A truncated HIV-1 Tat protein basic
domain rapidly translocates through the plasma membrane and
accumulates in the cell nucleus. J Biol Chem 1997; 272: 16010–
16017.
3 Rojas M, Donahue JP, Tan Z, Lin Y-Z. Genetic engineering of
proteins with cell membrane permeability. Nature Biotechnol
1998; 16: 370–375.
4 Elliot G, O’Hare P. Intercellular trafficking and protein delivery
by a herpesvirus structural protein. Cell 1997; 88: 223–233.
5 Phelan A, Elliott G, O’Hare P. Intercellular delivery of functional
Gene Therapy
24
25
26
27
28
p53 by the herpesvirus protein VP22. Nature Biotechnol 1998; 16:
440–443.
Derossi D et al. Cell internalization of the third helix of the
antennapedia homeodomain is receptor-independent. J Biol
Chem 1996; 271: 18188–18193.
Berlose J-P et al. Conformational and associate behaviours of the
third helix of antennapedia homeodomain in membrane
mimetic environments. Eur J Biochem 1996; 242: 372–386.
Gerlich WH, Lu X, Heermann KH. Studies on the attachment
and penetration of hepatits B virus. J Hepatol 1993; 17 (Suppl. 3):
S10–S14.
Lu X, Block TM, Gerlich WH. Protease-induced infectivity of
hepatitis-B virus for a human hepatoblastome cell line. J Virol
1996; 70: 2277–2285.
Prange R, Streek R. Novel transmembrane topology of the hepatitis B virus envelop proteins. EMBO J 1995; 14: 247–256.
Bruss V, Lu R, Thomssen R, Gerlich W. Post translational alterations in the transmembrane topology of the hepatitis-B virus
large envelope protein. EMBO J 1994; 13: 2273–2279.
Ostapchuk P, Hearing P, Ganem D. A dramatic shift in the
transmembrane topology of a viral envelope glycoprotein
accompanies hepatitis B viral morphogenesis. EMBO J 1994; 13:
1048–1057.
Hildt E, Urban S, Hofschneider PH. Characterization of essential
domains for the functionality of the MHBst transcriptional activator and identification of a minimal MHBst transactivator.
Oncogene 1995; 11: 2055–2066.
Hildt E, Saher G, Bruss V, Hofschneider PH. The hepatitis-B
virus surface protein (LHBs) is a transcriptional activator.
Virology 1996; 225: 235–239.
Hildt E, Hofschneider PH. The PreS2-activators of the hepatitisB virus: activators of tumour promotor pathways. Rec Res Cancer
Res, 1998; 154: 315–329.
Pastan IH, Willingham MC. Journey to the center of the cell:
role of the receptosome. Science 1981; 214: 504–509
Hildt E, Oess S. Identification of Grb2 as a novel binding partner
of tumor necrosis factor (TNF) receptor I. J Exp Med 1999; 189:
1707–1714.
Williams EJ et al. Selective inhibition of growth factor-stimulated
mitogenesis by a cell-permeable Grb2-binding peptide. J Biol
Chem 1997; 272: 22349–22354.
Schutze-Redelmeier M-P et al. Introduction of exogenous antigens into the MHC class I processing and presentation pathway
by drosophila antennapedia homeodomain primes cytotoxic T
cells in vivo. J Immunol 1996; 157: 650–655.
Lin YZ et al. Inhibition of nuclear translocation of transcription
factor NF-kB by a synthetic peptide containing a cell membranepermeable motif and nuclear localisation sequence. J Biol Chem
1995; 270: 14255–14258.
Allinquant B et al. Downregulation of amyloid precursor protein
inhibits neurite outgrowth in vitro. J Cell Biol 1995; 128: 919–927.
Troy CM et al. Downregulation of Cu/Zn superoxide dismutase
leads to cell death via the nitric oxide-peroxynitrite pathway. J
Neurosci 1996; 16: 253–261.
Troy CM et al. The contrasting role of ICE family proteases and
interleukin-1beta in apoptosis induced by trophic factor withdrawal and by copper/zinc superoxide dismutase downregulation. Proc Natl Acad Sci USA 1996; 93: 5635–5640.
Pooga M et al. Cell penetrating PNA constructs regulate galanin
receptor levels and modify pain transmission in vivo. Nature
Biotechnol 1998; 16: 857–861.
Hildt E et al. ER-localisation and functional expression of HBV
transactivator MHBst. Oncogene 1993; 8: 3359–3367.
Bruss V, Ganem D. The role of envelope proteins in hepatitisB virus assembly. Proc Natl Acad Sci USA 1991; 88: 1059–1063.
Dignam JP, LeBovitz RM, Roeder RG. Accurate transcription
initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acid Res 1983; 11: 1475–1489.
Kyte J, Doolittle RP. A simple method for displaying the hydropathic character of a protein. J Mol Biol 1982; 157: 105–132.