J. Mol. Biol. (1997) 272, 484±492
The Active Domain of the Herpes Simplex Virus
Protein ICP47: A Potent Inhibitor of the Transporter
Associated with Antigen Processing (TAP)
Lars Neumann1, Wolfgang Kraas2, Stephan Uebel1, GuÈnther Jung2
and Robert TampeÂ1,3*
1
Max-Planck-Institut fuÈr
Biochemie, Am Klopferspitz
18a, D-82152, Martinsried
Germany
2
Institut fuÈr Organische
Chemie der UniversitaÈt
TuÈbingen, Auf der Morgenstelle
18, D-72076, TuÈbingen
Germany
3
Lehrstuhl fuÈr Biophysik E22
Technische UniversitaÈt
MuÈnchen, D-85747 Garching
Germany
The herpes simplex virus type 1 (HSV-1) protein ICP47 binds speci®cally
to the transporter associated with antigen processing (TAP), thereby
blocking peptide-binding and translocation by TAP and subsequent loading of peptides onto MHC class I molecules in the endoplasmic reticulum. In consequence, HSV-infected cells are masked for immune
recognition by cytotoxic T-lymphocytes. To investigate the molecular
details of this, so far, unique transporter-inhibitor interaction, the active
domain and critical amino acid residues were identi®ed by using short
overlapping fragments and systematic deletions of the viral inhibitor. A
fragment of 32 amino acid residues, ICP47(3-34), was found to be the
minimal region harboring an activity to inhibit peptide-binding to TAP
comparable to the action of the full-length protein and therefore representing the active domain. Further N or C-terminal truncations cause an
abrupt loss in activity. Within the identi®ed active domain, various
mutants and chimeras of ICP47 derived from HSV-1 and HSV-2 helped
to identify amino acid residues critical for TAP inhibition. On the basis of
these results, therapeutic drugs could be designed that are applicable in
treatment of allograft rejection or in novel vaccination strategies against
HSV, restoring the ability of the immune system to recognize HSVinfected cells.
# 1997 Academic Press Limited
*Corresponding author
Keywords: ABC transporter; membrane protein; multi-protein complex;
vaccination; virus persistence
Introduction
Cytotoxic T-lymphocytes recognize and effectively eliminate virus-infected cells. This requires
presentation of viral fragments on the surface of
the infected cell in association with major histocompatibility complex (MHC) class I molecules
Abbreviations used: ABC, ATP-binding cassette;
Boc, butyloxycarbonyl; DMF, N,N-dimethylformamide;
ER, endoplasmic reticulum; Fmoc,
¯uorenylmethoxycarbonyl; HPLC, high-pressure liquid
chromatography; HSV, herpes simplex virus; HSV-1,
HSV type 1; HSV-2, HSV type 2; ICP47-1, ICP47
derived from HSV-1; ICP47-2, ICP47 derived from
HSV-2; mAU, milli absorbance units; MHC, major
histocompatibility complex; Pmc, 2,2,5,7,8pentamethylchroman-6-sulphonyl; TAP, transporter
associated with antigen processing; tBu, tert-butyl; TFA,
tri¯uoroacetic acid; Trt, trityl; wt, wild-type. The singleletter code is used for protein and peptide sequences.
0022±2836/97/390484±09 $25.00/0/mb971282
(Townsend & Bodmer, 1989). Protein fragments are
generated in the cytosol and nucleus mainly by the
proteasome complex and are translocated into the
endoplasmic reticulum (ER) by the MHC-encoded
transporter associated with antigen processing
(TAP). After synchronized, chaperone-assisted
assembly and loading onto MHC class I molecules,
kinetically stable peptide-MHC complexes are
transported to the cell surface, where they are
monitored by cytotoxic T-lymphocytes for their
antigenic cargo (for reviews, see Cresswell &
Howard, 1997; Koopmann et al., 1997; Tampe et al.,
1997).
TAP1 and TAP2 are members of the protein
superfamily of ATP-binding cassette (ABC) transporters, which are characterized by highly conserved Walker A/B motifs and the C-loop
signature (for review, see Higgins, 1992). They
form a heterodimeric membrane complex localized
in the ER-membrane (Kelly et al., 1992; Kleijmeer
et al., 1992), where each subunit contains a trans# 1997 Academic Press Limited
485
Active Domain of ICP47
membrane and a nucleotide-binding domain. TAP
mediates peptide transport into the ER lumen
(Neefjes et al., 1993; Shepherd et al., 1993;
Androlewicz et al., 1993; Meyer et al., 1994) and its
function can be dissected into an ATP-independent
peptide-binding step and an ATP-dependent translocation step (van Endert et al., 1994). ATP binding
itself does not require the presence of peptides
(MuÈler et al., 1994). The peptide-binding step has
been identi®ed to be critical for peptide selection in
the subsequent translocation step (Androlewicz &
Cresswell, 1994; Uebel et al., 1995; van Endert et al.,
1995; Uebel et al., 1997).
Due to the effective pathways of antigen processing and presentation, persistent viruses have been
forced to evolve strategies to escape immune surveillance (for review, see Hill & Ploegh, 1995). The
immediate-early protein ICP47 (IE12) encoded by
herpes simplex virus type 1 (HSV-1) was identi®ed
to be responsible for the down-regulation of the
MHC class I surface expression in human ®broblasts (York et al., 1994). By speci®cally binding to
the TAP-complex, ICP47 blocks TAP-mediated
peptide transport into the ER (FruÈh et al., 1995; Hill
et al., 1995). On the molecular level, we and others
have demonstrated that ICP47 inhibits peptidebinding to TAP, thereby blocking the ®rst and essential step in the translocation pathway (Ahn et al.,
1996; Tomazin et al., 1996). The interaction of ICP47
with the TAP complex is highly species-speci®c,
since ICP47 has a 100-fold higher af®nity for human
than for murine TAP (Ahn et al., 1996). Herpes simplex virus type 2 (HSV-2) encodes a related protein
(ICP47-2), which has 65% sequence identity with
ICP47-1 within the ®rst 52 to 54 residues (Whitton &
Clements, 1984; McGeoch et al., 1985; S. KohlstaÈtt &
U. H. Koszinowski, unpublished results). Apart
from these variants of ICP47 encoded by different
virus strains, no sequence similarity to any other
protein was found in the data base.
In addition to its unique primary structure,
ICP47 shows very interesting structural aspects. It
appears to be loosely folded (random-coiled conformation) in aqueous solution, thereby limiting
approaches to solve the structure at high resolution
by solution NMR or X-ray diffraction (Beinert et al.,
1997). We found that ICP47 interacts speci®cally
with lipid membranes. Interestingly, this membrane association results in profound conformational changes and a more a-helical structure of
the viral inhibitor is formed. These results are consistent with predictions of a-helical regions within
the N-terminal half of the ICP47 molecule. Fluorescence analysis revealed that the tryptophan residue at position 3, which is part of a predicted
a-helix, participates in the membrane-ICP47
interaction (Beinert et al., 1997).
To identify the active region and essential amino
acid residues of the ICP47-TAP interaction, we
tested a variety of different deleted or mutated
ICP47 molecules, including ICP47 chimeras of
HSV-1 and HSV-2, for their ability to block peptidebinding to the human TAP complex. ICP47(3-34)
was identi®ed to be the shortest region that reveals
nearly the same activity as the full-length protein.
Therefore, this 32mer fragment represents the active
domain of the viral inhibitor. Within this interacting
domain, several charged amino acid residues were
found to be essential for inhibition of TAP.
Results
Overlapping 25mer fragments of ICP47 fail to
interact with TAP
Peptide-binding was identi®ed to be a prerequisite step in the translocation mechanism of the
TAP transporter. HSV escapes immune surveillance by blocking this peptide-binding step, thereby preventing loading of MHC class I molecules.
ICP47 reveals a higher af®nity to human TAP (Kd
50 nM) than all peptides analyzed so far, thereby
being most effective in blocking peptide-binding to
TAP (Ahn et al., 1996). Thus, we were interested in
the mechanism underlying this high af®nity and in
whether ICP47 contains small fragments that are
able to block TAP function. In general, human
TAP preferentially binds peptides of eight to 16
amino acid residues (van Endert et al., 1994). In
order to bypass binding due to a given peptide
length, we synthesized overlapping fragments of
25 amino acid residues, which cover the entire
sequence of the viral inhibitor, and determined
their activity using TAP inhibition assays. Interestingly, none of these fragments was able to inhibit
peptide-binding to TAP (Figure 1a). In addition,
we analyzed various combinations of these fragments. Although fragments 1-25, 27-51, and 53-77
nearly cover the entire ICP47 molecule, none of
these combinations can complement the function
of the full-length protein (data not shown). The
residual potential of these 25mer fragments to
interfere with peptide-binding to TAP might be due
to the fact that they still are somehow poor substrates for TAP.
The failure of these peptides to restore ICP47
function might indicate that multiple contact sites
embodied by a larger fragment are required. In
order to identify these contact sites, we used the
25mer fragments to disturb the interaction between
ICP47 and TAP. For this analysis, inhibition of peptide-binding by full-length ICP47 was adjusted to
50%, while, in addition, a large molar excess of
different fragments was present. As shown in
Figure 1b, none of these fragments, or a combination of these (not shown), was able to interfere
with the interaction of ICP47 and TAP. In summary,
the failure of these short fragments to block TAP
function directly or to affect the ICP47-TAP interaction might indicate that the structural entity of a
larger domain is required for the activity of ICP47.
Identification of the active domain of ICP47
In order to identify the essential region required
for TAP inhibition, we synthesized N and C-term-
486
Active Domain of ICP47
decrease in activity, as shown in Figure 2c. However, full activity was found for fragments 1 ± 40 to
1 ± 34. Fragments shorter than ICP47(1 ± 31) fully
lost their ability to block peptide-binding to TAP.
These results were con®rmed by the af®nity constants determined for these ICP47 fragments
(Table 1). From this set of data, we conclude
that the ®rst 34 amino acid residues of ICP47 are
essential and suf®cient for TAP inhibition.
In analogy, we performed N-terminal truncations of ICP47. Here, only the ®rst and second
amino acid residues were found to be dispensable.
Further truncations of tryptophan and alanine at
positions 3 and 4 resulted in loss of activity, indicating that these residues are critical for ICP47 function (Figure 2d). These results are con®rmed by the
af®nity constants determined for various fragments
(Table 1). Combining the data from the C and
N-terminal deletions, we ®nally synthesized fragment ICP47(3± 34) and determined its af®nity constant, Kdˆ 89 nM (Table 1). The minor decrease in
af®nity of ICP47(3 ± 34) as compared to full-length
ICP47 might be due to a cumulative effect of the
slightly reduced af®nities of fragment ICP47(3± 53)
and ICP47(1 ±34) and appears not to be very drastic in contrast to the loss of activity observed for
shorter fragments. Therefore, ICP47(3 ± 34) was
identi®ed to be the active domain of the viral
inhibitor for blocking TAP function.
Figure 1. The 25mer, overlapping fragments of ICP47
fail to complement the ICP47-TAP interaction. (a) Inhibition of speci®c peptide-binding (100 nM radiolabeled
RRYQKSTEL) to human TAP was assayed in the presence of a twofold (open bars) or a 40-fold molar excess
(®lled bars) of full-length ICP47 or overlapping 25mer
fragments of ICP47. (b) TAP inhibition assays were performed in the presence of 50 nM full-length ICP47 and
5 mM 25mer fragments. The broken line indicates the
inhibition of peptide-binding by full-length ICP47 in the
absence of fragments.
inally truncated fragments of ICP47. The purity
and activity of these fragments were analyzed by
reverse phase chromatography and TAP inhibition
assays, respectively (Figure 2). Starting from the C
terminus, deletions revealed that fragment 1-53
still has the same activity as the full-length protein.
By Scatchard analysis as well as by competition
assays, the af®nity constant of ICP47(1 ±53) was
determined to be 50(12) nM for human TAP
(Figure 2b), which is, within the range of error,
identical with that of the full-length protein (Kd
50 nM) (Ahn et al., 1996; Tomazin et al., 1996).
Moreover, ICP47(1-53) has a 50-fold lower af®nity
for murine TAP than for human TAP (Figure 2b).
Thus, we conclude that deletion of one-third of the
protein at the C terminus affects neither the
activity nor the species-speci®city of the viral TAPinhibitor. Further deletions of the C terminus
beyond fragment 1-53 initially resulted in a slight
Critical amino acid residues within the
active domain
In order to identify critical amino acid residues
within the active domain, we persued two working
hypotheses. The ®rst was based on structural
aspects predicting a helix-turn-helix motif, whereas
the second was derived from a pattern of charged
amino acid residues that are highly conserved
between the active domain of ICP47-1 and ICP47-2
(Table 2). The ®rst presumption seems particularly
attractive, since truncation of the N-terminal predicted a-helix causes a complete loss of activity for
ICP47. To see whether a glycine-proline turn motif
is essential for ICP47 function, ICP47 mutants were
generated that have the turn motif (G18S/P19S) or
a segment between both helical regions (HSpH)
replaced by serine or glycine residues (Table 2A).
As shown in Figure 3, the mutant G18S/P19S has
nearly the same activity in inhibition of peptidebinding to TAP as the wild-type fragment, demonstrating that neither the turn motif nor these two
amino acid residues are essential for ICP47 function. In contrast, a complete loss of activity was
observed for the mutant HSpH, indicating that the
regions of predicted a-helicity alone are not suf®cient to retain activity. This result led us to the
second working hypothesis, based on charged
amino acid residues located between the predicted
a-helical regions. For technical reasons, during
peptide synthesis, we have chosen the active fragment ICP47(1± 32), whereby charged residues
at positions 20, 24, 31, and 32 were replaced by
Active Domain of ICP47
487
Table 1. Af®nity constants of ICP47 fragments for
human TAP.
Af®nity constants were determined by TAP inhibition assays.
Human TAP-containing microsomes were incubated at 4 C for
45 minutes with 100 nM radiolabeled RRYQKSTEL in the presence of various concentrations of the ICP47 fragments or nonlabeled RRYQKSTEL. The Kd values of the ICP47 variants were
determined by comparing the inhibition of peptide-binding
caused by the ICP47 fragments with that of the non-labeled
RRYQKSTEL. The Kd value of RRYQKSTEL is known to be 145
nM. Data were derived from at least four measurements (error:
SD).
Figure 2. Inhibition of peptide-binding to TAP by truncated ICP47 fragments. (a) The purity of the ICP47 fragments was analyzed by reversed phase HPLC. ICP47(153) (broken line) and ICP47(1-34) (continuous line) was
eluted from a C2/C18 column by using an acetonitrile
gradient (0.1% TFA) and monitored at 216 nm. (b) Using
a TAP inhibition assay the blocking of speci®c peptidebinding (100 nM radiolabeled RRYQKSTEL) to human
TAP (triangles) or murine TAP (circles) was analyzed in
the presence of various concentrations of ICP47(1-53).
The reporter peptide has the same af®nity for human
and murine TAP. (c) and (d) To identify the active
domain of ICP47, the inhibition of peptide-binding to
human TAP was assayed in the presence of a tenfold
molar excess of full-length, C-terminally (c), or Nterminally truncated ICP47 (d). Data were obtained by
duplicate or triplicate measurements (error bars: SD).
488
Active Domain of ICP47
Table 2. Primary and predicted secondary structure of
ICP47 encoded by HSV-1 or HSV-2.
A, Secondary structures are predicted by MacMatch 1.1d (University of California, San Francisco). Similar results were
obtained by using the BCM Protein Scondary Structure Prediction Program (NNSSP), the PredictProtein Service (EMBL, Protein Design Group, or the Genetic Computer Group (GCG,
Madison, Wisconsin). a or t indicate the a-helix or turn motifs,
respectively. The ®rst helical region is well determined (sources
from 8 to 9), whereas the second is less de®ned (sources from 5
to 8). For ICP47(3-45), mutated residues are underlined and
highlighted by superscription. B, Sequence comparison of fulllength ICP47 derived from HSV-1 and the partial sequence of
ICP47-2 as found in the data base (Whitton & Clements, 1984).
glycine (Table 2). These mutants were tested in
TAP inhibition assays (Figure 3). The single
mutants or the double mutant R20G/D24G
showed a reduced ability to block peptide-binding
to TAP as compared to the wild-type ICP47(1± 32).
The triple mutant D24G/K31G/R32G lost its
activity almost completely. These results demonstrate that the examined charged residues within
the active domain are critical for ICP47 function.
Chimeras of ICP47 derived from HSV-1
and HSV-2
In general, chimeras constructed between two
homologous proteins harboring distinct activities
are helpful tools to identify critical amino acid residues or essential regions involved in the function
of a given protein. As shown in Table 2B, the ®rst
52 to 54 residues covering the active region are
highly conserved between ICP47 encoded by HSV1 and HSV-2. Interestingly, we found that ICP47
fragments derived from HSV-2 show a signi®cantly reduced activity in blocking peptide-binding
to TAP in comparison to ICP47-1. Using TAP inhibition assays, the af®nity constant of ICP47(1 ±53)-2
was determined as 2.7(0.8) mM, revealing a 50fold lower af®nity than for ICP47(1 ± 53)-1 (Table 1).
We have constructed chimeras of ICP47(1± 32),
Figure 3. Critical amino acid residues within the active
domain. Inhibition of speci®c peptide-binding (100 nM
radiolabeled RRYQKSTEL) to human TAP was
measured in the presence of various mutants of
ICP47(3-45) or ICP47(1-32). Data were obtained by
duplicate or triplicate measurements (error bars: SD).
The sequences of these mutants are summarized in
Table 2A.
which cover most of the divergent residues of the
homologous region. Two different approaches
were followed. First, chimeras containing only
small regions in the active ICP47(1-32)-1 replaced
by the type 2 sequence were analyzed (Figure 4a).
The insertion of the HSV-2-encoded sequence MPS
at positions 10 to 12 causes a drastic reduction of
its activity, whereas exchanges of the other divergent amino acids lead to only minor effects. Substitutions of the sequence KTT and CA at positions 6
to 8 and 26± 27 resulted in a slightly enhanced
activity. In a second approach, chimeras were generated in which the type 1 sequence was transformed step-by-step into an ICP47-2 (Figure 4b).
Chimeras that possess the HSV-2 sequence up to
position 10 were as active as the wild-type fragment, whereas the insertion of proline at position
11 causes the drastic loss of its ability to inhibit
TAP function. These data are in line with the
results of the ®rst approach. Combining both sets
of data, the proline residue at position 11 was
identi®ed to be destructive for the activity of
ICP47. This conclusion was independently con®rmed by the mutant F11P of ICP47(1 ±32)-1, which
has a signi®cantly lower activity (data not shown).
Interestingly, this proline residue is located within
an a-helical region predicted for ICP47-1. Thus, the
loss in activity caused by the proline substitution
in ICP47-2 might correlate with disrupting the
a-helix in this region. As ICP47(1-32) underwent
further metamorphosis, only slight variations in
the low activity are observed, indicating that the
rest of the divergent residues between ICP47-1 and
ICP47-2 play only a minor role in TAP inhibition.
Active Domain of ICP47
Figure 4. Chimeras of ICP47(1-32) derived from HSV-1
and HSV-2. Inhibition of speci®c peptide-binding (100
nM radiolabeled RRYQKSTEL) to human TAP was
measured in the presence of a 25-fold molar excess of
ICP47(1-32). Chimeras were constructed by replacing
small regions of the type 1 sequence by the type 2
sequence (a) or by successive substitution of ICP47-1 by
ICP47-2 starting from the N terminus (b). The sequence
derived from HSV-2 is highlighted in black boxes. Data
were obtained by duplicate or triplicate measurements
(error bars: SD).
Discussion
The herpes simplex virus protein ICP47 blocks
peptide transport into the ER by preventing peptide-binding to the TAP complex. As a consequence, virus-infected cells escape immune
recognition by cytotoxic T lymphocytes. Using
overlapping or stepwise truncated fragments of the
viral TAP-inhibitor, a 32mer peptide, ICP47(3± 34),
was identi®ed as the active domain, revealing
nearly the same activity as the full-length protein
in blocking TAP function. Further truncations at
the N or C terminus result in a rapid loss of
activity. Furthermore, we have identi®ed critical
charged residues within the active domain that are
highly conserved between ICP47 encoded by HSV1 and HSV-2. Substitution of aspartate, lysine and
arginine with glycine at positions 24, 31 and 32
causes a complete loss of activity. By contrast, similar modi®cations at positions 18 and 19 or most of
the divergent residues between the ICP47 sequence
489
of HSVI-1 and HSV-2 had no, or only a minor,
effect on the activity of ICP47. Because peptides
require free N and C termini for their binding to
TAP (Androlewicz & Cresswell, 1994; Uebel et al.,
1997), one might speculate that the basic and acidic
side-chains of ICP47, which are separated by a distance resembling that between peptides ideal for
recognition by TAP, mimic the free terminal carboxy or amino group. However, fragment 14 ± 38
alone, containing the conserved, charged residues,
but lacking the ®rst predicted helical region, is not
suf®cient to block TAP function.
Beside these sequence characteristics, ICP47 has
some interesting structural properties. The viral
TAP inhibitor appears to be loosely folded (random-coiled conformation) in aqueous solution.
However, in the presence of lipid membranes,
drastic conformational changes were observed, and
upon binding to negatively charged membranes,
an a-helical conformation was induced. Interestingly, full-length ICP47 and ICP47(1± 53) behaved
nearly identically, indicating that the active
domain might be involved in structural reorganization upon binding to the membrane and the TAPcomplex. Evidence was provided that the tryptophan residue interacts directly with the membrane
(Beinert et al., 1997). These experimental results are
in agreement with secondary structure predictions,
proposing two a-helical regions (residues 3 to 13
and 35 to 45) within the ®rst N-terminal half of the
protein (Table 2A). Interestingly, the tryptophan
and alanine residues at positions 3 and 4, which
promote helix formation, are essential for ICP47
function. Vice versa, insertion of proline at position
11, which is known to break a-helices, causes a
drastic loss of activity. Taken together, these data
point to the signi®cance of the predicted secondary
structural elements. C-terminal truncations and
various mutations indicate that the second predicted a-helical region and the turn-motif is
dispensable.
In conclusion, we propose a model for the
active domain of ICP47 that consists of three sections. The ®rst segment is formed by a predicted
a-helical region (residues 3 to 13). Here, reduction
of the probability of helix formation by truncation
or proline insertion correlates with a loss in
activity. Furthermore, it was demonstrated that
this segment, particularly the tryptophan residue
at position 3, interacts with the membrane
(Beinert et al., 1997). This region might serve as a
membrane anchor to increase the local concentration of the TAP inhibitor, as has been observed
for various peptide hormone receptors (Schwyzer,
1995). The second segment (residues 14 to 34) is
supposed to interact with the substrate binding
site via a pattern of charged amino acid sidechains. Interestingly, these charged amino acid
residues are conserved between ICP47 encoded
by HSV-1 and HSV-2. For speci®c interaction,
these residues must be optimally arranged, possibly due to the help of the ®rst segment, since
the ®rst and the second segment are active only
490
in combination. This combination, a 32mer fragment, forms the active domain in respect to
blocking TAP function. Interestingly, considering
that particularly the conserved charged residues
are important for interacting with human TAP,
the sequence variations found between ICP47-1
and ICP47-2 within that region do not drastically
affect the activity of the viral inhibitor. The third
segment, residues 35 to 88, is not required for
TAP inhibition. Considering that due to evolutionary pressure viruses have been forced to
condense multiple functions to a limited genome
size, it seems attractive to speculate whether this
apparent redundant region has a second, and so
far unknown function. Furthermore, in this
respect, it is surprising that the most conserved
region (residues 35 to 52) with only two divergent residues is not needed for TAP inhibition
(Table 2B).
During the refereeing procedure of this paper,
Galocha et al. (1997) reported also on the characterization of the active region of ICP47. Their conclusions are largely in agreement with our results,
although different assay systems were used. However, in contrast to our study, ICP47 derived from
HSV-2 (strain HG52) was found to be as active as
ICP47-1. Most interestingly, ICP47 of HSV-2 differed by two amino acid residues (P11F and S12L)
from the published, partial ICP47 sequence of
HSV-2 strain HG52 (Whitton & Clements, 1984) as
well as from the ICP47 sequence of HSV-2 strain
186 (S. KohlstaÈtt & U. H. Koszinowski, unpublished results), the last two being identical and
given in Table 2. Because of the exchange of two
amino acid residues and four base-pairs, the difference can hardly be explained by a PCR error
during sequencing. Thus, it has to be clari®ed why
both ICP47-2 sequences from HSV-2 strain HG52
submitted by Whitton & Clements (1984) and
Galocha et al. (1997) differ by these amino acid residues. The difference con®rms our conclusion concerning the proline residue at position 11 to be
responsible for the reduced activity of the ICP47-2
protein as examined in this work. In addition, this
highlights exciting epidemic aspects of HSV-2.
Mutations in various HSV strains could result in
variants of ICP47 that harbor different potentials to
block TAP function. Thus, altered pathogenicity of
HSV strains may be related to variants of ICP47.
Here, we have identi®ed the active domain and
critical residues involved in this unique inhibitortransporter interaction. Based on these results, it
may be possible in the future to design peptidometrics based on ICP47(3 ± 34) that are speci®c and
highly potent inhibitors of the TAP transporter. By
increasing the stability against proteolysis and by
solving the targeting aspects of these ICP47 derivatives, therapeutic drugs could be developed that
act as effective immune suppressors in the treatment of allograft rejection or MHC class I linked
diseases. In addition, novel virus-based gene shuttles can be engineered, encoding the DNA of the
active domain and therefore able to protect trans-
Active Domain of ICP47
formed cells from lysis by virus-speci®c T-lymphocytes. Finally, novel vaccines could be developed
that speci®cally interfere with the unique TAPICP47 interaction and thereby restore the ability of
the immune system to eliminate the persistent
virus from the human body.
Materials and Methods
Preparation of TAP-containing insect
cell microsomes
The heterologous expression of human and murine
TAP in insect cells has been reported (Meyer et al., 1994;
Ahn et al., 1996). TAP-containing microsomes were isolated by a combination of differential and sucrose density-gradient centrifugation (Meyer et al., 1994; Uebel
et al., 1995). The microsomal membranes were resuspended in phosphate-buffered saline, 1 mM 1,4-dithioDL-threitol, snap-frozen in liquid nitrogen and stored at
ÿ80 C.
Isolation of full-length ICP47
Full-length ICP47 was expressed in Escherichia coli as a
6 His fusion protein (His-ICP47) and puri®ed by metal
af®nity chromatography and reversed phase HPLC as
described (Ahn et al., 1996). Purity of the protein was
analysed by SDS-PAGE and reversed phase HPLC.
Synthesis of variants of ICP47
Synthetic fragments of ICP47 were prepared by solid
phase synthesis on a multiple peptide synthesizer (SMPS
350, Zinsser Analytic, Frankfurt; Software Syro, MultiSynTech, Bochum) according to Fmoc/tBu-strategy
(Jung & Beck-Sickinger, 1992). Peptides were synthesized
on p-benzyloxybenzyl alcohol resins preloaded with
Fmoc-amino acids. Fmoc-protected amino acids were
activated with diisopropylcarbodiimide/1-hydroxybenzotriazol in DMF and coupled in tenfold excess for 90
minutes The N-terminal protecting group was cleaved
by piperidine/DMF (1:1, v/v). Side-chains of the amino
acids were protected as follows: D(tBu), E(tBu), T(tBu),
T(tBu), Y(tBu), N(Trt), Q(Trt), K(Boc), R(Pmc), H(Boc),
C(Trt), W(Boc). Peptides were cleaved from the resin
using Reagent K (1 ml) containing 82.5% TFA, 5% (w/v)
phenol, 5% (w/v) thioanisol, 2.5% (w/v) ethane dithiol
and 5% (w/v) water for three hours at room temperature. After precipitation and washing (three times) with
cold diethyl ether, the compounds were lyophilized from
tBu (4:1, v/v). All peptides were puri®ed by reversed
phase HPLC using a C2/C18-column (mRPC C2/C18 SC
2.1/10, Pharmacia) or by size exclusion chromatography
(Superdex Peptide PC 3.2/30, Pharmacia). Purity of the
ICP47 fragments was >95%. Identity of all ICP47 fragments was veri®ed by mass spectroscopy. Peptide concentrations were determined using the MicroBCA
protein assay (Pierce).
TAP inhibition assay
TAP-containing microsomes were diluted with ice-cold
assay buffer (phosphate-buffered saline with 1 mg/ml
dialyzed bovine serum albumin, 1 mM 1,4-dithio-DLthreitol, 5 mM MgCl2, 0.05 % polyethylene glycol 6000)
to a ®nal protein concentration of 25 mg/ml. The protein
concentration was determined using the MicroBCA pro-
Active Domain of ICP47
tein assay (Pierce). The suspension was homogenized by
drawing through a 23-gauge needle. From this suspension, 150 ml was incubated with radiolabeled RRYQKSTEL (100 nM) in the presence of a molar excess of the
corresponding ICP47 molecule. In order to reach equilibrium for all ICP47 variants, samples were incubated for
45 minutes on ice. During that period peptides or ICP47
molcules are not modi®ed or degraded. Subsequently,
350 ml of cold assay buffer was added and the microsomes were pelleted by centrifugation (12,000 g for eight
minutes at 4 C). After washing with 500 ml of cold assay
buffer, microsome-associated radioactivity was quanti®ed by g-counting. The amount of bound peptide was
background corrected for non-speci®c binding in the presence of a 400-fold molar excess of unlabeled peptide,
which corresponds also to 100% inhibition of peptidebinding to TAP. To determine the dissociation constant
of ICP47 derivatives (Kd) the inhibitor concentration was
variied over three orders of magnitude and the data
were ®tted by a competition function as described
(Uebel et al., 1995). The activity of the ICP47 variants
was measured by at least two independent assays. Competition by unlabeled RRYQKSTEL (Kd ˆ 145 nM) served
as an internal reference. Peptides were radiolabeled as
described (Meyer et al., 1994).
Acknowledgments
We thank Dr Christoph Eckerskorn for mass spectrometry analysis and Stefanie Urlinger, Titia Plantinga,
Anja Jestel and Kurt Pawlitschko for helpful discussions.
We are grateful to Dr Sybille KohlstaÈtt and Dr Ullrich H.
Koszinowski for sharing the unpublished ICP47-2
sequence. This work was supported by the Deutsche Forschungsgemeinschaft (DFG).
References
Ahn, K., Meyer, T. H., Uebel, S., SempeÂ, P., Djaballah,
H., Yang, Y., Peterson, P. A., FruÈh, K. & TampeÂ, R.
(1996). Molecular mechanism and species-speci®city
of TAP inhibition by herpes-simplex virus protein
ICP47. EMBO J. 15, 3247± 3255.
Androlewicz, M. J. & Cresswell, P. (1994). Human transporters associated with antigen processing possess
a promiscuous peptide-binding site. Immunity, 1,
7 ±14.
Androlewicz, M. J., Anderson, K. S. & Cresswell, P.
(1993). Evidence that transporter associated with
antigen processing translocate a major histocompatibility complex class I-binding peptide into the
endoplasmic reticulum in an ATP-dependent
manner. Proc. Natl Acad. Sci. USA, 90, 9130± 9134.
Beinert, D., Neumann, L., Uebel, S. & TampeÂ, R. (1997).
Structure of the TAP-inhibitior ICP47 induced by
membrane association. Biochemistry, 36, 4694± 4700.
Cresswell, P. & Howard, J. C. (1997). Antigen
recognition. Curr. Opin. Immunol. 9, 71 ± 74.
FruÈh, K., Ahn, K., Djaballah, H., SempÂ, P., van Endert,
P. M., TampeÂ, R., Peterson, P. A. & Yang, Y. (1995).
A viral inhibitor of peptide transporters for antigen
presentation. Nature, 375, 415± 418.
Galocha, B., Hill, A., Barnett, B. C., Dolan, A.,
Raimondi, A., Cook, R. F., Brunner, Y., McGeoch,
D. J. & Ploegh, H. L. (1997). The active site of
ICP47, a herpes simplex virus-encoded inhibitor of
491
the major histocompatibility complex (MHC)encoded peptide transporter associated with antigen
processing (TAP), mops to the NH2-terminal
35 residues. J. Exp. Med. 185, 1565± 1572.
Higgins, C. F. (1992). ABC transporters: from microorganisms to man. Annu. Rev. Cell Biol, 8, 67± 113.
Hill, A. & Ploegh, H. (1995). Getting the inside-out ± the
transporter associated with antigen-processing
(TAP) and the presentation of viral-antigen. Proc.
Natl Acad. Sci. USA, 92, 341 ± 343.
Hill, A., Jugovic, P., York, I., Russ, G., Bennink, J.,
Yewdell, J., Ploegh, H. & Johnson, D. (1995). Herpes
simplex virus turns off the TAP to evade host
immunity. Nature, 375, 411± 415.
Jung, G. & Beck-Sickinger, A. G. (1992). Multiple peptide synthesis methods and their applications.
Angew. Chem. 104, 375 ± 391.
Kelly, A. P., Powis, S. H., Kerr, L.-A., Mockridge, I.,
Elliott, T., Bastin, J., Uchanska-Ziegler, B., Ziegler,
A., Trowsdale, J. & Townsend, A. (1992). Assembly
and function of the two ABC transporter proteins
encoded in the human major histocompatibility
complex. Nature, 355, 641± 644.
Kleijmeer, M., Kelly, A., Geuze, H. J., Slot, J. W.,
Townsend, A. & Trowsdale, J. (1992). Location of
MHC-encoded transporters in the endoplasmic reticulum and cis-golgi. Nature, 357, 342±344.
Koopmann, J. O., HaÈmmerling, G. J. & Momburg, F.
(1997). Generation, intracellular-transport and loading of peptides associated with MHC class-I
molecules. Curr. Opin. Immunol. 9, 80 ± 88.
McGeoch, D. J., Dolan, A., Donald, S. & Rixon, F. J.
(1985). Sequence determination and genetic content
of the short unique region in the genome of herpes
simplex virus type 1. J. Mol. Biol. 181, 1 ± 13.
Meyer, T. H., van Endert, P. M., Uebel, S., Ehring, B. &
TampeÂ, R. (1994). Functional expression and puri®cation of the ABC transporter complex-associated
with antigen-processing (TAP) in insect cells. FEBS
Letters, 351, 443 ±447.
MuÈller, K. M., Ebensperger, C. & TampeÂ, R. (1994).
Nucleotide binding to the hydrophilic C-terminal
domain of the transporter associated with antigen
processing (TAP). J. Biol. Chem. 269, 14032± 14037.
Neefjes, J. J., Momburg, F. & HaÈmmerling, G. J. (1993).
Selective and ATP-dependent translocation of peptides by the MHC-encoded transporter. Science, 261,
769 ± 771.
Schwyzer, R. (1995). 100 years lock-and-key concept: are
peptide keys shaped and guided to their receptors
by the target cell membrane?. Biopolymers, 37, 5 ± 16.
Shepherd, J. C., Schumacher, T. N., Ashton-Rickardt,
P. G., Imaeda, S., Ploegh, H. L., Janeway, C. A. J. &
Tonegawa, S. (1993). TAP1-dependent peptide
translocation in vitro is ATP dependent and peptide
selective. Cell, 74, 577± 584.
TampeÂ, R., Urlinger, S., Pawlitschko, K. & Uebel, S.
(1997). The transporter associated with antigen processing (TAP). In Unusual Secretory Pathways: From
Bacteria to Man (Kuchler, K., Rubartelli, A. &
Holland, B., eds), pp. 115± 136, Springer, New York.
Tomazin, R., Hill, A. B., Jugovic, P., York, I., van Endert,
P., Ploegh, H. L., Andrews, D. W. & Johnson, D. C.
(1996). Stable binding of the herpes simplex virus
ICP47 protein to the peptide-binding-site of TAP.
EMBO J. 15, 3256± 3266.
Townsend, A. & Bodmer, H. (1989). Antigen recognition
by class I-resticed T lymphocytes. Annu. Rev. Immunol. 7, 601 ±624.
492
Active Domain of ICP47
Uebel, S., Meyer, T. H., Kraas, W., Kienle, S., Jung, G.,
WiesmuÈller, K. H. & TampeÂ, R. (1995). Requirements for peptide-binding to the human transporter
associated with antigen-processing revealed by peptide scans and complex peptide libraries. J. Biol.
Chem. 270, 18512± 18516.
Uebel, S., Kraas, W., Kienle, S., WiesmuÈller, K.-H.,
Jung, G. & TampeÂ, R. (1997). Recognition principle
of the TAP-transporter disclosed by combinatorial
peptide libraries. Proc. Natl. Acad. Sci. USA, 94,
8976± 8981.
van Endert, P. M., TampeÂ, R., Meyer, T. H., Tisch, R.,
Bach, J. F. & McDevitt, H. O. (1994). A sequential
model for peptide-binding and transport by the
transporters associated with antigen processing.
Immunity, 1, 491± 500.
van Endert, P. M., Riganelli, D., Greco, G., Fleischhauer,
K., Sidney, J., Sette, A. & Bach, J. F. (1995). The peptide-binding motif for the human transporter associated with antigen-processing. J. Exp. Med. 182,
1883± 1895.
Whitton, J. L. & Clements, J. B. (1984). The junctions
between the repetitive and the short unique
sequences of the herpes simplex virus genome are
determined by the polypeptide-coding regions of
two spliced immediate-early mRNAs. J. Gen. Virol.
65, 451± 466.
York, I. A., Roop, C., Andrews, D. W., Riddell, S. R.,
Graham, F. L. & Johnson, D. C. (1994). A cytosolic herpes simplex virus protein inhibits antigen
presentation to CD8‡ T lymphocytes. Cell, 77,
525± 35.
Edited by I. B. Holland
(Received 25 April 1997; received in revised form 15 July 1997; accepted 16 July 1997)