 |
INTRODUCTION |
The ICP34.5 protein of herpes simplex virus type 1 (HSV-1)1 is important for
promoting viral neurovirulence. Deletion mutants lacking ICP34.5 fail
to multiply in the brain and cannot cause encephalitis after
intracerebral inoculation of mice (1, 2) or is unable to replicate in
dorsal root ganglia after footpad inoculation (3). The requirement for
the ICP34.5 protein for HSV-1 replication is cell type- and cell
state-dependent (2-7). Deletion mutants can trigger a
total shut-off of protein synthesis mediated by protein kinase R in
response to HSV infection in nonpermissive cells, such as neuronal
cells (6, 8). ICP34.5 can reactivate protein synthesis by binding and
activating the PP1 cellular protein phosphatase. PP1 reactivates
eukaryotic initiation factor 2
by removing the phosphate attached by
protein kinase R (9, 10). ICP34.5 can also bind to proliferating cell
nuclear antigen (11). Viruses lacking the ICP34.5 gene also show cell
type-dependent deficits in virion maturation and egress
(12). Previous studies in our laboratory correlated differences in
ICP34.5 sequence with the ability to cause neuroinvasive disease and
also differences in tissue culture behaviors, such as plaque size,
efficiency of glycoprotein processing, and viral release, for members
of two families of HSV-1 strains (13).
The ICP34.5 gene of HSV-1 is located in the inverted repeats of the
unique long (UL) region of the HSV-1 genome and therefore is diploid
(Fig. 1) (14). The coding region of the ICP34.5 gene is 0.8-1.0 kb
long with a remarkably high guanosine/cytosine content (>80%) (13,
15). The ICP34.5 protein consists of three regions: an N terminus with
a variable number of arginines in an arginine-rich cluster; a bridge
region in the center of the protein consisting of a variable number of
Pro-Ala-Thr (PAT) repeats; and a C-terminal sequence, homologous to the
mammalian GADD34 (growth arrest and DNA damage) protein and murine Myd116
(myeloid differentiation primary response)
protein (13, 15-17). The GADD34-like sequence at the carboxyl terminus
of the ICP34.5 protein contains a PP1 binding motif, which can activate
PP1 to dephosphorylate eukaryotic initiation factor 2 (10, 18).
HSV-1 replication of viruses that lack the entire ICP34.5 or
lack or contain specific mutations in the GADD34-like sequences is much
less efficient or is inhibited in neuronal cells, including a human
neuroblastoma cell line (SK-N-SH), and some noncycling primary cells
(2-4, 6). Compensatory mutation(s) can rescue virus replication in
neuronal cell lines (19-23).
The functions associated with ICP34.5 suggest that ICP34.5 should be
present in different parts of the infected cell. Proliferating cell
nuclear antigen is located in the nucleus (reviewed in Ref. 24),
whereas PP1 is present in the nucleus (25, 26) and cytoplasm (27).
Interestingly, protein kinase R is also present in both the cytoplasm
and the nucleus, where it is specifically localized to the nucleolus in
association with ribosomes (28, 29). A previous study using
immunofluorescence microscopy suggested that ICP34.5 is present only in
the cytoplasm of infected cells (30), but in other studies, ICP34.5 was
associated with cytoplasmic and nuclear fractions of infected cells
(17, 31). The presence of the Arg-rich cluster at the N terminus of
ICP34.5, which resembles a putative nuclear and nucleolar targeting
sequence and RNA binding sequence in the human proliferation-associated
protein p120 (32), suggests that the protein may also be present and
have activity in the nucleus and nucleolus. Similar Arg-rich sequences
are present at the N terminus of the I14L protein of African swine
fever virus (33), the ORF2 protein of porcine circovirus type 2 (34)
and human immunodeficiency virus REV (35, 36) and act as
nucleolar targeting sequences for these proteins.
In this study, the influence of the N-terminal arginine-rich region and
the PAT repeat region of ICP34.5 on the cellular localization of
ICP34.5 was evaluated by comparing the properties of natural variants
of ICP34.5 from two families of HSV-1 strains. Although the protein
sequence of the C-terminal region of ICP34.5 and most of the rest of
the protein is highly conserved, a major difference between these
strains is the number of PAT repeats in the middle region and the
number of arginines at the N terminus of ICP34.5 (Fig. 1) (13, 15, 37).
SP7 and LP5 are intrastrain variants isolated from a neonate with
disseminated, fatal HSV-1 disease, including encephalitis. Both SP7 and
LP5 have the potential to cause neuroinvasive disease in mice, but SP7
is more neuroinvasive than LP5 (13). KOS321 and KOS79 were obtained
from lip lesions of the same individual on different occasions. KOS321
is highly passaged and completely attenuated for disease (13), and
KOS79 is a low passage strain that can cause neuroinvasive disease
(38).
Due to the low expression of the ICP34.5 protein in the natural HSV-1
infection and the difficulty in the development of effective and
specific antibodies against ICP34.5 (30, 37), an alternative approach
was taken. The natural variants of the ICP34.5 gene from SP7, LP5,
KOS321, and KOS79 were amplified by PCR and cloned into mammalian
expression vectors to generate fusion proteins with a reporter peptide
(c-Myc or hrGFP) at the C terminus. Studies of the expression and cell
distribution of these recombinant proteins upon transfection into Vero
cells and SK-N-SH neuroblastoma cells demonstrated that the N-terminal
Arg-rich cluster was necessary and sufficient for determining ICP34.5
localization in the cell; however, the length of the PAT repeats in the
middle region of the ICP34.5 variants is the ultimate determinant of
the cellular location of the protein. In addition, PP1 colocalized with
the ICP34.5 variant in the cells expressing ICP34.5. These studies show
that the peptide sequence-determined cellular localization of ICP34.5
is a determinant in the function of the protein and hence may influence
the tissue culture behavior and the virulence of HSV-1.
| |
EXPERIMENTAL PROCEDURES |
Cells and Virus
Vero cells were grown in medium 199 supplemented with 5% fetal
calf serum, penicillin (100 IU/ml), streptomycin (100 µg/ml), and
2.25 mM NaHCO3 at 37 °C. SK-N-SH cells were
grown in Eagle's minimal essential medium supplemented with 10% fetal
calf serum, 1.0 mM sodium pyruvate, and the same amount of
antibiotics and NaHCO3 as described for Vero cells.
The isolation and characteristics of the clinical strains, SP7 and LP5
(13), and KOS79 (38) have been described. KOS321 is a plaque-purified
and highly passaged strain provided by T. Holland (Wayne State
University School of Medicine) (39). The ICP34.5 deletion mutant
(d34.5) and the parental McKrae strain were provided by
G. C. Perng (Cedars Sinai Medical Center) (40).
Plasmids
Recombinant ICP34.5-myc Cloned from KOS321, KOS79, SP7, and
LP5--
HSV-1 uses the latter of two start codons (ATG) as its start
codon (15). To truly express the ICP34.5 gene in fusion with c-Myc and
His6, two PCR primers were synthesized in which the first start codon (ATG) and the stop codon (TAA) of the ICP34.5 gene
were mutated, respectively, as indicated by the underlines: MU34
(5'-CTGCACGCACTTGCTTGCCT-3') and 34MU
(5'-CTCGGGTGTAACGATAGACC-3') (Sigma, Genosys). PCR
amplification was performed using DNA purified from SP7-, LP5-, KOS79-,
and KOS321-infected cells (Wizard Genomic DNA purification kit;
Promega). Due to the high guanosine/cytosine content of the sequence,
PCR was performed using the MasterAmpTM PCR optimization
kit (premixture KN; Epicentre) and tfl polymerase (Epicentre) as
previously described (13). The PCR-amplified products were ligated into
the PCRTM2.1 vector (Invitrogen) with T4 DNA ligase
(Promega). Each of the insertions was examined by restriction digestion
and sequencing (M. Detweiler, Roswell Park Cancer Institute, University
of Buffalo) to check on the fidelity of the PCR and cloning
amplification procedures. The cloned PCR DNA fragment was cleaved from
the PCRTM2.1 vector with EcoRI and ligated into
the pcDNA3.1 myc-His B(
) vector (Invitrogen), which was
linearized with EcoRI. The purified recombinant DNAs were
named kos321m, sp7a, lp5a, and kos79a, representing the ICP34.5
variants cloned from the respective HSV-1 strains. These recombinant
DNAs were sequenced to make sure that the genes encoding the
myc-His6 epitopes are in the reading frame with the inserted ICP34.5.
Recombinant ICP34.5-hrGFP Genes (kos321hrGFP, sp7hrGFP, lp5hrGFP,
and kos79hrGFP)--
PCR-amplified ICP34.5 variants cloned into the
PCRTM2.1 vector were cut with BamHI and
XhoI. The small DNA fragment, including the ICP34.5 gene,
was purified and ligated with the hrGFP fragment cleaved from the hrGFP
vector (VitalityTM, Stratagene) by the same enzymes. The
recombinant DNAs were confirmed by sequencing.
N Terminus-truncated ICP34.5-myc Fusion Genes (tkos321 and tsp7)
without the Sequence Encoding the Arg-rich Cluster--
A new PCR
primer (T34, 5'-CCATGGGCGCGGTCCCAACC-3') was synthesized to
initiate the sequence downstream of the N-terminal Arg-rich cluster
from a new start codon (underlined in the T34 sequence). Truncated
ICP34.5 genes were amplified by PCR using the new T34 primer and the
34MU primer (described above) and cloned into the pcDNA3.1
myc-His vector. The truncated genes derived from kos321m and
sp7a are named tkos321 and tsp7, respectively. The mutant ICP34.5-myc gene was confirmed by sequencing.
Recombinant Genes with Sequences Encoding the ICP34.5 N-terminal
Arg-rich Cluster Attached to the 5'-End of the hrGFP Gene
(kos321nhrGFP, sp7nhrGFP, and kos79nhrGFP)--
PCR was performed
using two new primers (AR5, 5'-TGTGCTGGATATCTGCAGAATTCGGCT-3'; ar3,
AGGCCTCGAGCTGTGCGGTTGGGACCG-3'), which incorporate restriction enzyme
sites for EcoRI and XhoI, respectively. Plasmid
DNA variants including kos321m, sp7a, and kos79a were used as
templates. The amplified products encompassed a 5' noncoding region and
a leading sequence encoding the ICP34.5 N-terminal Arg-rich cluster.
The products were cut with EcoRI and XhoI and ligated with an hrGFP fragment cleaved from the hrGFP vector with the
same enzymes. The recombinant DNAs were named for the ICP34.5 variants,
which provided the Arg-rich cluster and were sequenced to make sure
that the insertion was present and in the reading frame with hrGFP.
Hybrid ICP34.5 Genes (7nKb and Kn7b)--
Recombinant DNAs
kos321m and sp7a were cut with NotI to produce two
fragments. The small fragment, which includes the N-terminal Arg-rich
cluster from one plasmid, was ligated with the large fragment
containing the remaining segment of the ICP34.5 gene from the other
plasmid to generate hybrid recombinant ICP34.5 DNA. The 7nKb has the
first 44 amino acids from SP7 and a KOS321-like backbone, whereas the
Kn7b has the first 43 amino acids from KOS321 and an SP7-like backbone.
The new constructs were also sequenced.
Transient Expression of ICP34.5 Fusion Genes
Recombinant DNA was transfected into Vero cells or SK-N-SH
cells using LipofectAMINE PLUS (Invitrogen). After 2 days, cells were
either lysed and analyzed by Western blot or processed for immunocytochemistry.
Complementation Study
SK-N-SH cells were transfected with the ICP34.5-Myc DNA
variants. After the cells reached confluence (3-4 days), the cells were grown with M199 containing 0.5% fetal calf serum for another 2-3
days and then infected by a null ICP34.5 mutant of HSV-1
(d34.5) (40) at a multiplicity of infection of 2. After 2 days, the supernatant and the infected cells were collected separately
after centrifugation (12,000 × g, 2 min). After
freeze, thaw, and sonication, viral plaque assays were performed on
Vero cells.
Antibodies
A specific antibody to ICP34.5 was developed against a peptide
(AQSQVTSTPNSEPVVC) near the N terminus of the ICP34.5 protein (QCB,
Hopkinton, MA). The peptide was attached to keyhole limpet hemocyanin,
ovalbumin, and then tetanus toxoid, and rabbits were immunized four
times with each immunogen. Sera were collected after each procedure.
The antisera were purified by affinity chromatography on a
peptide-agarose column. Monoclonal anti-Myc, anti-Myc-horseradish peroxidase (HRP), and goat anti-mouse TRITC (TRITC-GAM) antiserum were
purchased from Invitrogen. The polyclonal anti-PP1 (FL-18) antibody
(Santa Cruz Biotechnology, Inc., Santa Cruz, CA) used in this study was
generated to a peptide present in most forms of PP1.
Western Blotting
Cells were lysed by radioimmune precipitation assay lysis buffer
(100 mM NaCl, 10 mM Tris, pH 7.4, 1% Triton
X-100, 1% sodium deoxycholate, 0.1% SDS) in the presence of a
protease inhibitor mixture (Sigma) and DNase I (100 µg/ml; Sigma) at
48 h after transfection. Samples were mixed with an SDS-PAGE gel
loading buffer (2% SDS, 80 mM Tris, pH 6.8, 32% glycerol,
0.001% bromphenol blue, 5% 
-mercaptoethanol) and boiled. The
lysates were separated by electrophoresis on 10-20% SDS-PAGE gels and
transferred to polyvinylidene difluoride membranes. The membranes were
blotted with anti-Myc-HRP or anti-peptide antibody followed by goat
anti-rabbit conjugated with HRP. The signals were examined by
chemiluminescence using ECL (Amersham Biosciences), and the x-ray films
were digitized with an HP scanner to produce the final images.
Immunohistochemistry
Vero cells, transfected with the ICP34.5 expression vector, were
seeded onto coverslips and cultured overnight. These cells were then
fixed with cold acetone for 10 min. After the samples dried, they were
blocked by normal mouse serum and then incubated with anti-Myc-HRP for
1 h at 37 °C. After washing with PBS, these cells were stained
by immersing in the disclosing reaction mixture, including
diaminobenzidine tetrahydrochloride and hydrogen peroxide, for 7 min.
Samples were further dehydrated by immersing sequentially in a series
of graded alcohols (70, 95, and 100%) and then in xylene and mounted
with Permount. Photomicrographs were obtained using an Olympus B-max 60.
Indirect Immunofluorescence
Transfected cells were fixed in 4% paraformaldehyde for 15 min
and permeabilized in 0.2% Triton X-100 for 5 min. After incubation with 0.1 M glycine and 1% bovine serum albumin, the cells
were incubated with anti-Myc monoclonal antibody (Invitrogen) in the presence of 1% bovine serum albumin for 1 h. After washing, these fixed cells were incubated with FITC-GAM or TRITC-GAM for another 1 h. The cells were then treated with equilibration buffer and mounted using the SlowFade-light antifade kit (Molecular Probes, Inc.,
Eugene, OR) and viewed by fluorescence microscopy (Olympus AX70) or
confocal microscopy (Olympus IX70). The digital images were obtained
and analyzed by Spot Advance software (Diagnostic instruments, Inc.).
For colocalization studies, organelle-specific dyes such as propidium
iodide for the nucleus, or DiOC6(3) for the ER (41)
were used at concentrations of 5 µg/ml and 2.5 µg/ml, respectively,
during the incubation with the secondary antibodies.
| |
RESULTS |
Strain Differences in ICP34.5 Gene/Protein Sequence--
DNA
encoding ICP34.5 was obtained from each of the HSV-1 strains by PCR
amplification using primers from the ends of the coding sequence. The
PCR products from each variant were sequenced, and the DNA sequences
were aligned and converted to peptide sequences using the multiple
alignment program and six frame translation programs (Sequence
Launcher, BCM) (42). The ICP34.5 genes from SP7, LP5, KOS79, and KOS321
strains had extensive homology except for major sequence differences in
the number of repeats of CCC GCG ACC, encoding PAT, in the middle
region of the gene and the CGC repeats at the N terminus encoding a
string of Arg (e.g. RRRRHRGPRRPR for LP5) (Fig.
1) (13, 15). The sequence of the SP7
ICP34.5 used in this and subsequent studies differs from that reported in Bower et al. (13) in that the Gly-Glu-Gly-Ala (GEGA) at
positions 153-156 (with respect to the LP5 sequence) is present. As a
result, the only difference between the SP7 and the LP5 ICP34.5 is the number of the PAT repeats, with 18 and 22, respectively. The numbers of
PAT repeats in the ICP34.5 from LP5 and SP7 are larger than for other
HSV-1 strains that have been reported (15, 37). The ICP34.5 from LP5
and SP7 have 8 Arg residues in the N-terminal Arg cluster. The ICP34.5
protein from KOS79, another low passage strain, has 11 PAT repeats and
nine Arg residues, while the KOS321 ICP34.5 has only three PAT repeats
and seven Arg residues. Single amino acid polymorphisms are noted
between SP7, LP5, and KOS321 at positions 140, 158, and 227, with
respect to the LP5 sequence (13). The KOS79 ICP34.5 has the following
residues: Ala140, Glu158, and
Ala227. In addition, the KOS79 ICP34.5 has Gly instead of
Ser at position 54, Glu instead of Ala at position 161, insertion of
Asp between positions 81 and 82, and deletion of Leu and Arg between
positions 148 and 149. The C-terminal GADD34-like sequence is identical for all of the HSV-1 strains.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1.
Location of the ICP34.5 gene in the HSV-1
genome and structural comparison of several natural variants of the
ICP34.5 protein. Top line, unique long (UL)
and unique short sequences (US) of the HSV-1 genome flanked
by the inverted repeats (boxes). The small dots with
arrows below the top line indicate the
locations of the ICP34.5 gene (diploid) in the viral genome and the
direction of its transcripts. The expanded representions show the three
regions of the ICP34.5 protein: an N terminus with an
Arg-rich cluster, a central bridge region with PAT repeats, and a
C-terminal GADD34-like region. The numbers of Arg in the Arg-rich
cluster and the PAT repeats are indicated for each variant.
|
|
The PCR products of the ICP34.5 variants were cloned into the
pcDNA3.1 myc-His expression vector (Invitrogen) with
c-myc and His6 at the C terminus. Each of the
new vectors was sequenced to confirm the insertion of the proper
sequences. Vero cells were transfected with the same amount of
recombinant DNA for each of the variants, and after 2 days, protein
samples were obtained from the same number of cells and analyzed by
Western blot. Polypeptides with the same mobility were detected from
cells transfected with ICP34.5-myc DNA variants with either
an anti-Myc monoclonal antibody or a polyclonal antibody developed
against a peptide (AQSQVTSTPNSEPVVC) near the N terminus of the ICP34.5
protein. Detection of identical polypeptides for the fusion protein
with antibodies raised against epitopes at the N and C termini (c-Myc
epitope) indicate that the complete recombinant proteins were expressed
(Fig. 2). Although useful, the
anti-peptide antibody was too weak for use in subsequent studies. The
expressed ICP34.5-Myc variants differed in mobility with apparent
molecular masses ranging from 40 to 60 kDa. The ICP34.5 from the
LP5 and KOS321 strains were the largest and smallest, respectively,
while the ICP34.5 from SP7 was slightly smaller than from LP5. The
KOS79 ICP34.5 had a medium size in comparison with the proteins from
LP5 and KOS321. Minor bands with faster mobility, which may correspond
to degradation products, were also observed, especially for the KOS321
ICP34.5. The apparent molecular masses of these ICP34.5-Myc
polypeptides are consistently larger than the predicted molecular mass
that was calculated from their gene sequences (LP5-Myc, 36.1 kDa;
SP7-Myc, 35 kDa; KOS79-Myc, 33.2 kDa; KOS321-Myc, 30.8 kDa) but are
consistent with the differences in their genetic sequence (see Fig. 1).
This may be due to either extensive posttranslational modification of
the ICP34.5-Myc protein or more likely a structural conformation of the
protein (possibly due to the PAT repeats), which affects its mobility
in an SDS-PAGE electrophoresis.
View larger version (88K):
[in this window]
[in a new window]
|
Fig. 2.
Expression of ICP34.5-Myc proteins. Vero
cells were transfected with ICP34.5 variant recombinant DNAs and grown
for 48 h. Cells were lysed in radioimmune precipitation assay
buffer under reducing conditions. The lysates were separated on a
10-20% gradient SDS-PAGE gel and transferred to a polyvinylidene
difluoride membrane. The membrane was cut into half, probed with the
indicated antibodies, and visualized by chemiluminescence.
A, scanned image of a blot probed with anti-Myc conjugated
with HRP. B, scanned image of a duplicate sample probed with
anti-ICP34.5 peptide antiserum followed by goat anti-rabbit-HRP.
Molecular mass standards are indicated. kos321m,
sp7a, lp5a and kos79a represent the
proteins expressed from the ICP34.5-myc DNA variants from
the respective HSV-1 strains (KOS321, SP7, LP5, and KOS79).
|
|
Recombinant ICP34.5-Myc Complements Replication of a Null ICP34.5
Virus in SK-N-SH Cells--
ICP34.5 is essential for viral replication
in neuronal cells (1, 2, 6), and replication of a null ICP34.5 mutant virus (d34.5) (40) was greatly reduced in confluent SK-N-SH cells (a neuroblastoma cell line) when the cells were grown in low
concentrations of serum (data not shown). SK-N-SH cells transfected with recombinant ICP34.5-myc DNA variants were infected with
d34.5 virus, and at 48 h postinfection the supernatants
and cell pellets were harvested separately. Production of virus was
assayed using permissive Vero cells. Only a small fraction of cells
successfully expressed ICP34.5-Myc proteins upon transfection of the
SK-N-SH cells. Nevertheless, the presence of KOS321 ICP34.5-Myc
increased viral production by at least 15-fold in separate experiments, while the SP7 ICP34.5-Myc augmented the replication of the null virus
by at least 7-fold compared with cells transfected with an
antisense-KOS321 ICP34.5 sequence (Fig.
3). The enhancement of d34.5
replication by the presence of ICP34.5-Myc, although statistically
significant, was not as large as anticipated. This may be due to an
increased level of permissiveness of this SK-N-SH neuroblastoma cell
line or the lower transfection efficiency of SK-N-SH cells than
expected based on results with Vero cells. The ability to enhance
replication of the null virus in the SK-N-SH cells indicates that the
transfected gene and its protein are active and that the c-Myc and
His6 fused at the C terminus of the ICP34.5 do not inhibit
the function of the protein.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 3.
Complementation of the replication of a null
ICP34.5 HSV-1 (d34.5) in SK-N-SH cells. SK-N-SH
cells were transfected with ICP34.5-Myc variant recombinant DNAs or the
antisense DNA to the KOS321 ICP34.5 gene, and 72 h later they were
infected with d34.5 for 48 h. Total virus production
was quantitated by plaque assay on Vero cells. A representative
experiment is shown. The titration of each sample was examined in
triplicate, and the error bar represents the
S.E.M. Results for ICP34.5-Myc fusion proteins from KOS321 and SP7
lacking the 18 N-terminal amino acids are indicated as
tkos321 and tsp7.
|
|
Cellular Localization of ICP34.5
Variants--
Immunohistochemistry and immunofluorescence microscopy
were performed to delineate the intracellular locations of the variants of ICP34.5. Immunohistochemistry showed a difference in the cellular localization of the SP7 and KOS321 variants of ICP34.5-Myc. The SP7
ICP34.5-Myc was localized to regions of the cytoplasm close to the
nucleus, but unlike SP7 and previous reports (30), the KOS321
ICP34.5-Myc localized predominantly to the nucleus (data not shown).
The differences in cellular localization of the ICP34.5-Myc variants
could be discerned better by immunofluorescence microscopy. ICP34.5-Myc
was detected with a monoclonal antibody against c-Myc followed by a
secondary antibody conjugated with FITC (green). Propidium iodide
(red), which binds specifically to DNA, was used to stain the nucleus.
The SP7 ICP34.5-Myc concentrated in the cytoplasm, especially in
perinuclear regions, while in a small fraction of the cells, very low
amounts of the SP7 ICP34.5-Myc were also found in the nucleus (Fig.
4, A and C). The
localization is more clearly shown in the images in which the FITC and
propidium iodide staining are merged. For cells expressing the SP7
ICP34.5-Myc, the FITC fluorescence and propidium iodide signal did not
colocalize, indicating that a nuclear presence of the SP7 variant is
not detectable by this technique (Fig. 4A). In contrast, the
KOS321 ICP34.5-Myc was almost exclusively in the nucleus (Fig. 4,
A and C) with the propidium iodide signal and
FITC signal colocalized in the nucleus to give a yellow color. Further
analysis by confocal microscopy confirmed that the expression of the
KOS321 ICP34.5-Myc was limited to the nucleus (Fig. 4C). The
KOS321 ICP34.5-Myc fluorescence appeared to cover the nucleus (Fig. 4,
C and D) in many of the cells, but in some cells
the fluorescence appeared as an intranuclear ring (Fig. 4C)
or was localized to specific regions of the nucleus or as dots within
the nucleus (data not shown).
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 4.
Intracellular localization of the ICP34.5-Myc
variants examined by indirect immunofluorescence. Vero or SK-N-SH
cells transfected with ICP34.5-Myc recombinant DNAs were incubated for
48 h, fixed, reacted with anti-Myc monoclonal antibody followed by
FITC or TRITC-GAM, and viewed with fluorescence illumination. The
images were captured by a digital camera using SPOT Advance software.
A, expression of kos321m or sp7a ICP34.5-Myc variants in
Vero cells. ICP34.5-Myc was visualized by FITC-GAM (green).
Nuclei of the cells were also counterstained with propidium
iodide (PI) (red), and the image was merged with
the FITC image. B, expression of lp5a or kos79a ICP34.5-Myc
variants in Vero cells. ICP34.5-Myc was visualized by FITC-GAM
(green). C, expression of kos321m, kos79a, or
sp7a ICP34.5-Myc variants in Vero or SK-N-SH cells were visualized with
FITC-GAM and viewed with confocal microscopy. D, expression
of kos321m or sp7a ICP34.5-Myc variants in Vero cells. ICP34.5-Myc was
visualized by TRITC-GAM (red). ER of cells was
counterstained with DiOC6(3) (green) and merged
with TRITC signals.
|
|
To determine whether the cytoplasmic SP7 ICP34.5-Myc was in the ER,
another colocalization study was performed in which the ICP34.5-Myc was
visualized with anti-Myc and TRITC-GAM (red) and the ER was stained
with DiOC6(3) (green) (41). The signal from DiOC6(3) was limited to areas surrounding the nucleus,
indicative of the ER network. Some of the SP7 ICP34.5-Myc appeared to
colocalize with DiOC6(3) upon merging of the images, as
indicated by the yellow color, but much of the SP7 ICP34.5-Myc was
present in other areas of the cytoplasm (Fig. 4D).
The localization of the KOS79 and LP5 ICP34.5-Myc variants in the cell
was also examined. The KOS79 ICP34.5-Myc was present in both the
cytoplasm and specific regions of the nucleus (Fig. 4B). The
fluorescent signal for cells expressing the LP5 variant of ICP34.5-Myc
was seen almost exclusively in the cytoplasm. The cellular distribution
of the SP7, LP5, KOS79, and KOS321 ICP34.5-Myc variants in SK-N-SH
cells was similar to that in Vero cells (Fig. 4C and data
not shown).
As an alternative method to study the cellular localization of the
ICP34.5 variants, the distributions of hrGFP fusion proteins of the
ICP34.5 variants were studied. Expression of hrGFP resulted in a
diffuse fluorescence throughout the cell with complete coverage of the
nucleus. The cellular distribution of the ICP34.5-hrGFP variants
resembled that of the ICP34.5-Myc variants such that the cell
distribution correlated with the number of PAT repeats in the proteins
(Fig. 5, Table
I). The LP5 ICP34.5-hrGFP variant, with
the largest number of PAT repeats (n = 22), was
restricted to the cytoplasm of 97% of the cells, with nuclear and
cytoplasmic distribution in the remaining 3%. In cells expressing the
SP7 ICP34.5-hrGFP, with 18 PAT repeats, the distribution was 88 and 12%, respectively. No cells were observed expressing only nuclear fluorescence for the LP5 or SP7 ICP34.5-hrGFP variants. The KOS79 ICP34.5-hrGFP has 11 PAT repeats, and the fluorescence was located in
both the cytoplasm and in restricted areas within the nucleus. For the
KOS79 ICP34.5-hrGFP, 66% of the ICP34.5-expressing cells had GFP in
both nuclear and cytoplasmic compartments, 28% of the cells had
expression only in the nucleus, and the remaining 6% of the cells
expressed GFP only in the cytoplasm. The KOS321 ICP34.5-hrGFP, with the
smallest number of PAT repeats (n = 3), was
predominantly in the nucleus, with 90% of the cells expressing only
nuclear fluorescence and the other 10% of the cells with the GFP
signal in both the nucleus and cytoplasm. The nuclear
immunofluorescence for all of the variants of GFP-ICP34.5 was localized
to discrete regions. The distribution of the different ICP34.5-hrGFP
variants was not changed by infection of the cells with HSV-1 SP7 virus (data not shown). These findings indicate that ICP34.5 variants with a
long PAT region are more likely to be absent from the nucleus.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 5.
The cellular localization of ICP34.5-hrGFP
fusion protein. Vero cells transfected with variants of
ICP34.5-hrGFP were visualized by fluorescence. The solid
red arrows indicate cells with cytoplasmic
fluorescence. The red arrowheads indicate cells
with signals predominantly in discrete regions of the nucleus. The
hollow red arrows point to cells with
signals in both cellular compartments.
|
|
Role of the N-terminal Arg-rich Cluster for Cellular Localization
of the ICP34.5 Proteins--
To determine the role of the Arg-rich
clusters for the cellular localization of ICP34.5, constructs lacking
the N-terminal 18 amino acid residues (including the Arg-rich cluster)
were prepared for ICP34.5 from KOS321 and SP7. The expression and size
of the truncated proteins (tkos321 and tsp7) were confirmed by Western blot analysis. Both truncated proteins exhibited a faster mobility than
their parental proteins (Fig.
6A). The truncated ICP34.5-Myc variants were able to complement the replication of the
d34.5 virus in SK-N-SH cells to levels similar to the intact
ICP34.5-Myc variants (Fig. 3), indicating that the protein was
functional.
View larger version (37K):
[in this window]
[in a new window]
|
Fig. 6.
Expression and intracellular localization of
the N terminus truncated ICP34.5-Myc. Vero cells were transfected
with tkos321 or tsp7 ICP34.5-myc DNA and prepared for
Western blot and indirect immunofluorescence. A, a scanned
image of a Western blot of an extract from transfected Vero cells
probed with anti-Myc-HRP and visualized by chemiluminescence. Molecular
mass standards are indicated at the right. B,
immunofluorescence image of transfected Vero cells using FITC-GAM as
described in the legend to Fig. 4.
|
|
Cells expressing the truncated KOS321 ICP34.5-Myc protein exhibited a
diffuse fluorescence throughout the cell. The cytoplasmic distribution
of the tsp7 protein was more diffuse than for SP7, and unlike tkos321,
it excluded the nucleus (Fig. 6B). These results indicate
that the N terminus, which includes the Arg-rich sequence, is necessary
for the characteristic cellular distribution of the KOS321 and SP7
variants of ICP34.5. The exclusion of tsp7 from the nucleus indicates
that other sequences in the SP7 ICP34.5 also influence the distribution
of the SP7 variant.
Whether the N-terminal Arg-rich cluster is sufficient to determine the
cellular distribution of ICP34.5 was tested by comparing the cellular
distributions of hrGFP and fusion proteins of the Arg-rich clusters
from three different ICP34.5 variants (KOS321, SP7, and KOS79) inserted
at the N terminus of the hrGFP. These clusters have 7, 8, or 9 Arg
residues, respectively. The unmodified hrGFP diffused throughout the
cell, including the nucleus. The fluorescent signal for hrGFP modified
by the N-terminal attachment of Arg-rich clusters from any of the
ICP34.5 variants was distributed within the cytoplasm and to discrete
regions of the nucleus, resembling the nucleolus (Fig.
7). The cellular targeting of hrGFP was
the same regardless of the number (n = 7, 8, or 9) of
Arg residues in the cluster (data not shown). This experiment indicates
that the N-terminal Arg-rich cluster from ICP34.5 directs proteins to
the cytoplasm and discrete regions of the nucleus, but it is not the
sole determinant for the cellular localization of the ICP34.5
variants.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 7.
Intracellular localization of fusion proteins
with the Arg-rich cluster of ICP34.5 attached to the N terminus of
hrGFP. Vero cells transfected with hrGFP or sp7nhrGFP were fixed
and then analyzed by fluorescence microscopy. The arrows
point to cells with immunofluorescence in both the cytoplasm and
discrete regions of the nucleus.
|
|
The Localization of the ICP34.5 Protein Is Influenced by the Number
of PAT Repeats--
To confirm that the number of PAT repeats in an
ICP34.5 variant determines the cellular localization of the protein,
hybrid ICP34.5-Myc proteins were constructed in which the N termini of the SP7 and the KOS321 ICP34.5 variants were exchanged at amino acid 44 of the SP7 variant of ICP34.5. The 7nKb ICP34.5-Myc was prepared with
an SP7-like N-terminal Arg-rich cluster (8 Arg residues) and a KOS321
backbone (three PAT repeats in the bridge region). The Kn7b ICP34.5-Myc
was prepared with a KOS321-like N-terminal Arg-rich cluster (7 Arg
residues) and an SP7 backbone (18 PAT repeats). The Kn7b ICP34.5-Myc
was distributed primarily within the cytoplasm, especially to
perinuclear regions, and in some cells a small amount was present in
specific regions of the nucleus, similar to the SP7 ICP34.5-Myc. In
contrast, the 7nKb ICP34.5-Myc was in the nucleus, similar to the
KOS321 ICP34.5-Myc (Fig. 8). This
indicates that although the N-terminal Arg-rich clusters provide a
cellular localization signal for the cytoplasm and discrete regions of
the nucleus, the final location of ICP34.5 is determined by other
regions of the protein. Since the major difference in ICP34.5 structure
between the different variants is the number of the PAT repeats, it is
most likely that the PAT repeats are the determinant of the final
cellular location of ICP34.5.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 8.
Intracellular localization of SP7/KOS321
hybrid ICP34.5-Myc proteins. Vero cells transfected with 7nKb or
Kn7b DNA were analyzed by immunofluorescence using FITC-anti-Myc and
PI.
|
|
ICP34.5 Directs the Cellular Localization of PP1--
ICP34.5
binds to PP1 through the GADD34-like C terminus of the protein (9, 10,
18). The intracellular location of PP1 was examined by
immunofluorescence using a polyclonal anti-PP1 antibody, which
recognizes most forms of PP1, and visualized with FITC-goat anti-rabbit
(green). ICP34.5-Myc was visualized with TRITC-GAM (red) in cells
transfected with the SP7 or KOS321 variants. In cells that did not
express the ICP34.5 protein, the PP1 exhibited a diffuse intracellular
fluorescence that included the nucleus. In contrast, the distribution
of PP1 fluorescence was mainly in the cytoplasm and colocalized with
the fluorescence of ICP34.5 in cells expressing the SP7 ICP34.5-Myc
variant. Some nuclear fluorescence of PP1 was also present and may be
due to PP1 isoforms that do not bind to the SP7 ICP34.5. Similarly, the
distribution of PP1 in cells expressing the KOS321 ICP34.5-Myc variant
was concentrated in the nucleus, and for some cells, the PP1 and
ICP34.5 were in discrete regions of the nucleus and, to a minor extent, in the cytoplasm (Fig. 9).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 9.
Intracellular localization of PP1 in cells
transfected with ICP34.5-Myc variants. Vero cells transfected with
SP7 or KOS321 ICP34.5-myc recombinant DNAs were incubated
for 48 h. The cells were fixed, reacted with anti-PP1 followed by
FITC-goat anti-rabbit and anti-Myc monoclonal antibody followed by
TRITC-GAM, and viewed with fluorescence illumination.
|
|
| |
DISCUSSION |
The functions attributed to ICP34.5 have been associated with the
C-terminal GADD34-like sequences (9-11, 18, 43). By comparing the
cellular distribution of the natural variants of ICP34.5 expressed as
fusion proteins with C-terminal c-Myc or hrGFP, we show that the
ICP34.5 variants are localized to different cellular compartments and
the Arg-rich cluster at the N terminus and the PAT repeats in the
middle-bridge region of the protein work together to manipulate the
cellular distribution of the ICP34.5.
Although the properties of the fusion proteins and their expression
differ somewhat from the native protein during a natural infection, the
possible influence of the additional sequences on the properties of the
ICP34.5 is discounted by the following: both the c-Myc-His6
and hrGFP fusion proteins exhibited similar cell distributions for each
of the recombinant ICP34.5 variant proteins; the fusion proteins were
also shown to be functional by their ability to enhance the replication
of an ICP34.5 null virus (d34.5) (40) in SK-N-SH
neuroblastoma cells; and the cellular distribution of the different
ICP34.5 variant-hrGFP fusion proteins was not changed by infection of
the cells with HSV-1.
The Arg-rich cluster at the N terminus of ICP34.5 (e.g.
RRRRHRGPRRPR, for LP5) acts as a cellular localization signal to direct the ICP34.5 into the cytoplasm, especially to perinuclear regions, and
to specific regions of the nucleus, resembling the nucleolus. Without
the Arg-rich cluster, the KOS321 variant of ICP34.5 (tkos321) exhibits
a diffuse cellular distribution unlike the distribution of any of the
variant ICP34.5 proteins. Similarly, the lack of an N-terminal Arg-rich
cluster for the p32 isoform of the human I-mfa domain-containing
protein deprives this protein of a nucleolar targeting mechanism (44).
The ability of the Arg-rich cluster to direct hrGFP to specific
cytoplasmic, nuclear, and discrete nuclear sites resembling the
nucleolus indicates that the sequence can be sufficient to direct
ICP34.5 to these cellular regions. The Arg-rich clusters with 7, 8, or
9 arginines, corresponding to the ICP34.5 variants, were equivalent in
their ability to target hrGFP. Arg-rich clusters, like that present in
ICP34.5, act as nuclear and nucleolar localization sequences for other
proteins, including the ORF2 protein in porcine circovirus type 2 (34), the catalytic subunit of HSV-1 DNA polymerase (45), the p14ARF (alternative reading frame) tumor suppressor (46), nucleolar p120 protein (32), and the I14L protein of African swine fever virus
(33), a protein that resembles ICP34.5 in several ways. For the
REV protein of human immunodeficiency virus, an arginine-rich cluster serves as both a nuclear/nucleolar targeting sequence and an
RNA binding sequence (35, 36). Although the reason for ICP34.5 to be in
the nucleus and the cytoplasm is not known, it is interesting that
protein kinase R (28, 29) and PP1, the ligand of ICP34.5 (24-26), are
present in the nucleus and/or nucleolus in addition to the cytoplasm.
The presence of an Arg-rich cluster appears to be important for wild
type levels of neurovirulence in mouse models of infection (1, 47)
despite the ability of the ICP34.5 lacking the N-terminal Arg-rich
cluster (tICP34.5) to promote the growth of a deletion mutant in
neuroblastoma cells (Ref. 16; Fig. 3). The position of the Arg-rich
cluster with respect to the N terminus of ICP34.5 also appears to be
important, since a mutant virus with a new epitope (16 amino acids)
added to the N terminus of ICP34.5 was 10 times less neurovirulent (1),
whereas another mutant virus with an 8-amino acid deletion downstream
of the cluster maintained its neurovirulence and the ability to
replicate in neuronal cells (2).
Stark differences in the distribution of the ICP34.5 variants
correlated best with the number of PAT repeats. The LP5 and KOS321
variants of ICP34.5 have the longest (n = 22) and the
shortest (n = 3) PAT repeats, respectively, reported
for HSV-1, and they show the greatest contrast in cellular
distribution. The KOS321 ICP34.5 was present almost exclusively in the
nucleus, while the LP5 ICP34.5 variant was limited almost exclusively
to the cytoplasm. Similar to the LP5 variant, the SP7 ICP34.5, with 18 PAT repeats, was primarily in the cytoplasm, but unlike LP5, there was
a small amount of ICP34.5 present in nuclei. The tsp7, lacking the
N-terminal Arg region, but with the 18 PAT repeats, was restricted from
the nucleus. The KOS79 ICP34.5, with an intermediate number of PAT repeats (n = 11), distributed to both the cytoplasm and
to discrete regions of the nucleus, as if to have characteristics of
both the KOS321 and LP5 variants. Interestingly, most variants of
ICP34.5 have more than five PAT repeats (15, 37). The PAT region of ICP34.5 modulates the targeting ability of the Arg-rich cluster at the
N terminus of ICP34.5 and, as a result, determines where the protein
goes in the cell.
The longer PAT repeat regions, as for SP7 and LP5, are likely to form
structures consisting of a rigid collagen-like helix with 9 amino acids
per turn and with the threonine residues facing the outside surface (as
suggested by a computer model (SpartanPro; Wave Function, Inc.)). Such
an ICP34.5 protein might adopt a dumbbell-like shape with the N- and
C-terminal globular regions bridged by the rod formed by PAT repeats.
The prediction of an extended structure is supported by the slower
SDS-PAGE mobility for the ICP34.5 protein variants than would be
expected from their molecular weights (Fig. 2). The proline-directed
structure of the PAT repeats would not be extensively disrupted even
upon boiling in SDS. The structure formed by a long sequence of PAT
repeats, as present in the SP7 and LP5 variants, may prevent the
nuclear entry of the ICP34.5, whereas an intermediate number of PAT
repeats may hinder but not prevent the nuclear presence of ICP34.5, as
seen for the KOS79 variant with its 11 PAT repeats. The combination of
the targeting signal from the N-terminal arginine cluster and the
modulating effect of the PAT region appear to be the primary
determinants of the amount of ICP34.5 localized to the nucleus.
The strain-dependent distributions of ICP34.5
were also reflected in the distribution of its primary ligand, PP1. The
PP1 in cells expressing the KOS321 variant of ICP34.5 was concentrated to discrete sites in the nucleus, the sites where the ICP34.5 was
localized, rather than the disperse distribution of PP1 of the
nontransfected cells. Similarly, the PP1 in cells expressing the SP7
variant of ICP34.5 was concentrated in the cytoplasm rather than the
nucleus. The ICP34.5 variants, acting as cellular PP1 binding proteins,
either concentrate or deliver the PP1 to specific sites in the cell,
either nuclear or cytoplasmic. Cellular proteins, like PNUT (25), which
binds to PP1 and localizes it to the nucleus, also determine the
nuclear distribution of PP1. Binding of ICP34.5 to PP1 may deliver the
PP1 to specific sites in the cell in support of important viral function(s).
The number of PAT repeats in the bridge region of ICP34.5 appears to be
the structural determinant that delineates the distribution of ICP34.5
and PP1 within the cell. As a result, the greatest differences in PP1
distribution and its functions would be expected between HSV strains
with ICP34.5 variants containing extreme numbers of PAT repeats, such
as for KOS321 and LP5 or SP7. This difference may be accentuated during
natural infection because the levels of ICP34.5 are low.
PP1 isoforms have been attributed with functions other than the
dephosphorylation and reactivation of the eukaryotic initiation factor
2 protein after it is phosphorylated in response to HSV infection or
interferon action (43). For example, in yeast, PP1 is essential for
membrane fusion processes that are involved in vesicular transport from
the ER to the Golgi apparatus (48), and in Xenopus extracts,
inhibition of PP1 increased formation of ER networks (27). These
actions of PP1 may be relevant to our previous studies in which we
correlated the number of PAT repeats in ICP34.5 for a particular HSV-1
strain with efficiency of glycoprotein processing and virion release
(13). These processes involve vesicular transport. The KOS321, SLP5,
and SLP10 strains are readily released from infected cells and have
only three PAT repeats in ICP34.5. In contrast, the processing of
glycoproteins of the SP7 strain of HSV-1 (the parent strain of SLP5 and
SLP10), which has 18 PAT repeats, is limited to steps that occur in the ER, and the virus is more cell-associated. The neuroinvasive disease potential of these strains also correlates with the number of PAT
repeats such that the KOS321-like viruses are completely attenuated and
the viruses with larger numbers of PAT repeats are virulent (13, 38).
These correlations suggest a connection between the sequence-determined
cellular distribution of ICP34.5; its interaction with its target
proteins, PP1, possibly proliferating cell nuclear antigen, or other
proteins; and the influence of ICP34.5 on virus replication and pathogenesis.
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Northeastern Ohio
Universities College of Medicine, 4209 State Route 44, Box 95, Rootstown, OH 44272. Tel.: 330-325-6134; Fax: 330-325-5914; E-mail: ksr@neoucom.edu.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
