(Received for publication, October 28, 1996, and in revised form, December 20, 1996)
From the § Department of Molecular Genetics and
Microbiology, School of Medicine, State University of New York,
Stony Brook, New York 11794 and Max-Delbrück-Centrum
für Molekulare Medizin, Robert-Rössle-Straße 10, 13122 Berlin, Germany
The cellular receptors for poliovirus
(PVR) are glycoproteins belonging to the immunoglobulin superfamily.
Functional receptors for poliovirus are only expressed by primates;
known rodent homologues lack the ability to bind virus due to amino
acid differences. Human poliovirus infections are targeted to the
gastrointestinal tract and, rarely, to motor neurons in the central
nervous system. Available evidence suggests that poliovirus uses only
one cellular receptor, implying that the tissue tropism of poliovirus
is likely to be related to the expression of the human PVR (hPVR).
However, low levels of expression of hPVR-specific mRNAs can be
detected in many human tissues other than the apparent target cells.
The nonpathogenic function of hPVR is unknown. For a study of the transcriptional control of hPVR expression, we have isolated and characterized the promoter of the hPVR gene. Deletion
analysis defined an approximately 280 base pair minimal promoter
fragment that: 1) lacks TATA- and CAAT-like elements, 2) is
distinguished by a high GC content, and 3) promotes transcription at
multiple start sites. The pattern of activity caused by transfection of serial 5- and 3
-promoter deletions is almost identical in HEp2, HeLa,
COS-1, and mouse L929 cells, indicating a similar transcriptional regulation of the hPVR promoter in these cell lines.
However, on transfection of Raji cells, a Burkitt's lymphoma cell line harboring a transcriptionally inactive hPVR gene, all
promoter reporter constructs tested exerted only residual activity.
These results suggest that the cis-element(s) governing
cell type-specific hPVR expression resides in the minimal promoter
region. We also report the sequences of the promoters of two monkey
homologues to hPVR (AGM
1 and
AGM
2). Transcripts encoding the monkey poliovirus receptors originate from a region analogous to that identified for hPVR
transcripts.
The human receptors for poliovirus
(hPVR)1 are cell surface proteins
possessing three immunoglobulin-like domains, designated V-C-C (1, 2;
reviewed in Ref. 3). Analyses of cDNAs and of genomic DNA suggested
that hPVR-related proteins are expressed in the forms of four splice
variants: two membrane-bound (hPVR and hPVR
) and two secreted
(hPVR
and hPVR
) polypeptides (Refs. 1 and 2 and Fig. 1). The
membrane-bound receptors hPVR
and hPVR
differ only in the
sequences of their cytoplasmic C-terminal domains, and they are highly
glycosylated (4, 5). Moreover, receptor hPVR
is phosphorylated on
one or more serine residues of its cytoplasmic domain (6). The nature
of the secreted splice variants (hPVR
and hPVR
) has been largely
inferred only from analyses of their corresponding mRNA species
(1).
Humans are the only natural hosts for poliovirus. Monkeys, however, can
be experimentally infected, because they express receptors homologous
to hPVR. Koike et al. (7) have shown that, in contrast to
hPVR, monkey poliovirus receptors (referred to in the following as
mPVRs) of African green monkey cells are encoded by two related genes
AGM1 and AGM
2 (African green monkey
receptor). The predicted gene products mPVR
1 and mPVR
1 (for
AGM
1) and mPVR
2 (for AGM
2) are integral
membrane glycoproteins that were shown to be functional poliovirus
receptors (7). Secreted splice variants of AGM
1 and
AGM
2 have not been detected (7).
A gene (MPH, mouse poliovirus receptor homologue) homologous to the human and monkey poliovirus receptor genes exists also in mouse cells (8, 9). Similarly to hPVRs and mPVRs, the receptor homologues encoded by MPH belong to the immunoglobulin superfamily possessing the putative domain structure V-C-C (reviewed in Ref. 3). The MPH gene product, however, is unable to function as a poliovirus receptor due to differences in its amino acid sequence in comparison with the human and monkey receptors. This is true particularly in the case of the V domain, the structure of hPVR that has been shown to bind the virus (9-12).
Recently, two new human genes, PRR1 and PRR2 (poliovirus receptor-related genes) have been deduced from cDNA sequences that can be predicted to encode proteins related to hPVR (13, 14). Therefore, it appears that a gene family encoding PVR-like molecules may exist in humans similar to families of Ig-like CEA (15) and PSG genes (16). Interestingly, the latter two gene families map to human chromosome 19q, as does the hPVR gene (1, 17-19).
The cellular function(s) of hPVR and the hPVR-related proteins in humans, monkeys, and mice is obscure. Early studies have led to the prediction that the observed narrow tissue tropism of poliovirus would mirror the distribution of hPVR expression in human tissues (20). However, results of Northern blot analyses have indicated that transcripts related to hPVR sequences can be detected in nearly all human tissues that have been analyzed (1, 2; see discussions in Refs. 21 and 22). Since human monocytes express hPVR (23), the apparent ubiquity of hPVR transcripts in human tissues may be explained, at least in part, by contaminating blood cells in tissue samples. It is also possible that the probes used to detect hPVR-related transcripts cross-reacted with transcripts of PRR1 and PRR2.
In our efforts to characterize cellular function(s) for hPVR and to gain insight into the pathology of poliovirus infection in humans, we have isolated and characterized the promoters of the hPVR and AGM genes. Structural as well as functional analyses indicate that the expression of the hPVR gene is controlled by a TATA and CAAT box-deficient promoter. Our results suggest that cis-acting elements of the hPVR core promoter and trans-acting factors available in cell lines of primate and mouse origin generally support hPVR expression, a result in concurrence with the expression of the hPVR gene in transgenic mice (24, 25). However, the fine tuning of promoter activity of the hPVR gene may differ slightly depending on the cell type investigated. Interestingly, hPVR is not expressed in a variety of hematopoetic cell lines, including the Burkitt's lymphoma cell line Raji. We provide evidence that the hPVR core promoter is inactive in these cells, indicating that this sequence harbors information for the control of basal and cell type-specific expression of the hPVR gene.
Isolation of hPVR Genomic DNA
A chromosome 19-specific genomic library was purchased from the American Type Culture Collection (ATCC 57766; library name LL19NL01). The bacterial strain used was LE392 (ATCC 33572).
Screening the Genomic Library2.5 × 106
plaque-forming units were screened with a 32P-labeled
Bss HII-SacI fragment (153 bp) isolated from the
5-NTR of cDNA clone H20A (2). Positive plaques were picked, and
1500 plaque-forming units each were plated for a second screen,
resulting in the identification of single positive plaques.
109
competent L392 bacteria were infected with 107
plaque-forming units of -phage. These suspensions (70 ml) were
incubated at 37 °C until lysis of the bacteria was almost complete.
The infection was stopped by adding 0.5 ml of chloroform. Bacterial debris was removed by centrifugation, and RNase A and DNase I were
added to the supernatant to final concentrations of 50 and 10 µg/ml,
respectively. Following an incubation at 37 °C for 1.5 h the
phages were pelleted at 27,000 rpm for 2 h at 4 °C using a
Sorvall SW 41 rotor. The pellet was resuspended in 500 µl of TE
buffer (10 mM Tris/HCl, 1 mM EDTA, pH 8.0),
mixed with 5 µl of 10% SDS and 5 µl of 0.5 M EDTA, pH
8.0, and incubated for 15 min at 65 °C. Following a phenol
extraction the DNA was precipitated with ethanol.
Cloning and Restriction Analysis of Genomic DNA
The 5.7-kbp BamHI fragment harboring the first exon
and the upstream genomic region (see Fig. 2) was cloned into the
BamHI site of vector Bluescript KS (Stratagene), yielding
plasmid p5.7/7. To generate subclones this fragment was mapped with
respect to the restriction enzymes ApaI, ClaI,
HindIII, KpnI, NotI, SalI, XbaI, SmaI, EcoRI, PstI,
SpeI, and XhoI. No sites were found for ClaI, HindIII, KpnI, NotI,
SalI, and XbaI. The 2.9-kbp SacI
fragment (see Fig. 2; one SacI site is located in the
vector) was subcloned into Bluescript KS
, yielding pSac2.9/2.
Bal 31 Deletions and Sequencing
Ten µg of plasmid pSac2.9/2 were linearized with
HindIII and precipitated with ethanol. The DNA pellet was
dissolved in 340 µl of water, mixed with 85 µl of 5 × buffer
(3 M NaCl, 30 mM CaCl2, 60 mM MgCl2, 100 mM Tris/HCl, 5 mM EDTA, pH 8.0), and prewarmed at 37 °C before 15 units
of Bal 31 (Biolabs) were added. Each minute 20 µl were removed from
the reaction and added to an ice-cold mixture of 24 µl of water, 1 µl of 1 M EGTA, pH 7.0, and 5 µl of 3 M
sodium acetate, pH 5.2. After 10 min at 65 °C the samples were
ethanol-precipitated, and the DNA was digested with BamHI and XbaI. Gel-purified DNA fragments (100 ng) were cloned
into a SmaI-BamHI-cut Bluescript KS vector.
Colonies were analyzed according to standard procedures. Sequencing
reactions were performed using the Pharmacia Biotech Inc. T7 sequencing
kit.
RNA Isolation, Northern Blotting, and Hybridization
Seven ml of lysis buffer (4 M guanidinium thiocyanate, 25 mM sodium citrate, pH 7.0, 0.5% N-lauroylsarcosine, 5% 2-mercaptoethanol) were added to 5 × 107 cells, and the viscous solution was homogenized by pipetting. The lysate was loaded onto a cushion of 5.7 M CsCl and spun for 18 h at 36,000 rpm in an SW 41 rotor. The RNA pellet was dissolved in 500 µl of buffer (10 mM EDTA, pH 7.0, 0.5% N-lauroylsarcosine, 5% 2-mercaptoethanol, 5% phenol) and precipitated twice with ethanol.
Northern BlottingFifty µg of total RNA were dissolved in 50 µl of denaturing buffer (2.2 M formaldehyde, 50% formamide, 10% 10 × electrophoresis buffer (200 mM MOPS, 50 mM sodium acetate, 10 mM EDTA, pH 7.0)), heated to 65 °C for 15 min, mixed with 5 µl of 10 × FF (10% Ficoll 400, 0.2% bromphenol blue), electrophoresed through a 1.2% agarose gel (containing 2.2 M formaldehyde), and blotted onto nitrocellulose.
For hybridization the filter was incubated overnight with hybridization buffer (50% formamide, 50 mM PIPES, pH 6.5, 5 × SSC (1 × SSC = 0.15 M NaCl, 0.015 M sodium citrate, pH 7.0), 0.1% Denhardt's solution, 400 µg/ml yeast tRNA) at 42 °C before radiolabeled probe was added. Incubation at 42 °C was continued for another 24 h. The filter then washed with 2 × SSC, 0.1% SDS twice for 20 min. at room temperature followed by two additional washes with 0.1 × SSC, 0.1% SDS for 20 min each. The filter was exposed to an x-ray film using an intensifying screen.
Radioactive Labeling
For DNA labeling the Life Technologies, Inc. random-primed DNA labeling kit and [32P]dCTP (Amersham Corp.) were used according to the instructions of the manufacturers.
End labeling of oligodeoxynucleotides was done by incubating 5 pmol of
an oligonucleotide in a total volume of 25 µl with 50 µCi of
[-32P]ATP (Amersham) and 3 units of polynucleotide
kinase (Biolabs) for 15 min at 37 °C. The labeled samples were
desalted using a Sephadex G25 column (Pharmacia).
PCR
PCR was done using a Perkin-Elmer or Biometra thermocycler.
Conditions for the generation of deletion fragments were: 10 ng of
plasmid template, 50 pmol of each primer, 5 µl of 10 × buffer, 1 µl of 10 mM dNTP mix, 5 µl of formamide, and 0.5 µl
of Taq polymerase (2.5 units, Boehringer Mannheim) in a
total reaction volume of 50 µl. PCR was started with five cycles of 1 min at 94 °C, 30 s at 37 °C, and 1 min at 72 °C, followed
by 25 cycles of 1 min at 94 °C, 30 s at 55 °C, and 1 min at
72 °C. For genomic DNA (100 ng) as template the probe composition
was essentially the same, but 2 µl of 10 mM dNTPs and 1 unit of vent polymerase (Biolabs) were used. PCR was done for 35 cycles. For cloning the PCR products were cut with appropriate
restriction enzymes and gel purified. The oligonucleotides used were
(restriction sites are underlined): S20,
5-GCGCCGGGAGAGACCTGC-3
; S22,
5
-CC
TAGCGTCAGTAGCAGCGG-3
; 1729, 5
-CCAC
CAGCCG-3
; 3745, 5
-
GGCC
GAGAAACCACT-3
; 3912, 5
-GCCACTCGGCAACCGCG-3
;
4410, 5
-GGG
GGCTCGGGCCTTGCCAGTTGCTCC-3
; 4411, 5
-GGG
GCGGCGGCCTTGTGAGTTGCC-3
; 4413, 5
-GGG
GGTTCCGGGCCTTGCAGTTGGTCCGG-3
; 4438, 5
-GGC
GAGAAACCAC-3
; 4463, 5
-CG
GTACACGGTGAGGTTAACAGTCAGCAGC-3
; 4526, 5
-GCC
GCTGCGCAGGTCTCTCC-3
; 4527, 5
-GCC
GGAGTTGGCGGCCAG-3
; 4528, 5
-GCC
GGAGGGAAGGGGAATA-3
; 4529, 5
-
GCC
GCTCGCTCTGCCGCGG-3
; 4530, 5
-GGC
CTGGCCGCCAACTCC-3
; 4531, 5
-GGC
GAGAGACCTGCGCAGC-3
; 4532, 5
-GGC
GCCGCCTCTTCTAGTG-3
; and 4533, 5
-GGC
GATCTGTCCCATCACG-3
.
Primer pairs used to generate the promoter and promoter deletion
constructs were: 4438 and 4410 (H), 4533 and 4410 (A), 4532 and 4410 (B), 4531 and 4410 (C), 4530 and 4410 (D), 4438 and 4529 (E), 4438 and
4528 (F), 4438 and 4527 (G20), 4438 and 4526 (G17), 4532 and 4529 (BE),
4438 and 4411 (AGM2), and 4438 and 4413 (AGM
1).
RACE
Rapid amplification of cDNA ends (5-RACE) was done
according to the protocol described by Frohman (26) with the following modifications.
Partially purified mRNA from HEp2 and JA-1 cells was prepared using a Quickprep mRNA purification kit (Pharmacia). One µg of RNA in 12 µl of water containing 50 fmol of primer (primer 4463 or 1729, see below) was heated to 80 °C and then allowed to cool to 40 °C. After a 2-h incubation period at 40 °C, buffer, dithiothreitol, dNTP, RNAsin, and Superscript reverse transcriptase were added according to the manufacturer's instructions (Life Technologies). The reaction mixture was incubated at 40 °C for 90 min.
First PCRGSP1 was primer 1729. Five cycles with 45 s at 94 °C, 3 min at 48 °C, and 45 s at 72 °C were done, followed by 30 cycles with 30 s at 94 °C, 45 s at 50 °C, and 45 s at 72 °C. The annealing temperature was increased by 0.3 °C/cycle.
Second PCRGSP2 was primer 4529. Thirty cycles with 30 s at 94 °C, 45 s at 50 °C, and 45 s at 72 °C were done. The increment for the annealing temperature was 0.2 °C/cycle.
RACE products were cut with BglII-HinDIII and
cloned into Bluescript KS and BamHI-HinDIII,
and colonies were sequenced with T3 primer.
Cloning of Reporter Plasmids
The promoter fragments generated by PCR (see above) were cloned into the NheI-BglII site of the pGL2-Basic vector (Promega), and the vector and insert transitions were sequenced. The H reporter construct was cut with SacI, and the insert was replaced by the 2.9-kbp SacI fragment of p5.7/7, thus generating the large promoter/reporter vector (promoter insert of approximately 3000 bp).
Cell Culture
HeLa R19, HEp2 (a human laryngeal cell line), Hep G2 (a human liver cell line), COS-7, COS-1 (African green monkey kidney cell lines), SK-N-SH (a human neuroblastoma-derived cell line), 293 (a human kidney cell line), RD (a human rhabdomyosarcoma cell line), and mouse L cells were grown in Dulbecco's modified Eagle's medium DMEM with 8% bovine calf serum. L735, CEM (human T leukemia cell lines), HL60 (a human promyelocytic cell line), and Raji (a Burkitt's lymphoma (BL) cell line) cells as well as all other BL cell lines (F126, JBL2, JI, LY66, LY67, LY91, BL40, BL60, BL70, BL72, BL90, and BL99) were grown in RPMI 1640 medium with 8% fetal bovine serum. JA-1 cells (a mouse L cell line harboring the hPVR gene) were maintained in Dulbecco's modified Eagle's medium with 8% bovine calf serum and hypoxanthine/aminopterin/ thymidine.
Transfection and Harvest of Cells for CAT Enzyme-linked Immunosorbent Assay and Luciferase Assay
HEp2, COS-1, Hep G2, HeLa, and L cells were transfected by the calcium phosphate procedure. 5-20 µg of test plasmid, 2 µg of pSV2CAT (in combination with luciferase vectors only), and 62.5 µl of 2 M CaCl2 in a total volume of 500 µl were combined dropwise with 500 µl of ice-cold 2 × Hank's balanced salt solution. The precipitate was added to the cells. Four hours later the medium was removed, and a solution of 20% glycerol in Hank's balanced salt solution was added. Following a 3-min incubation at 37 °C, 10 ml of medium was added, and the supernatant was removed again and replaced by fresh medium with serum.
Raji cells were electroporated. 107 cells/transfection were spun, and the cell pellet was resuspended in medium without serum. The cell suspension was combined with 5-20 µg of test plasmid and 2 µg of pSV2CAT. Electroporation was done at 340 V and 1180 microfarads with a Life Technologies electroporator (electrode gap, 0.4 cm). Cells were allowed to sit for 10 min at room temperature and were then transferred with fresh medium and serum into culture flasks.
All transfected cells were harvested 36 h after transfection, and cell extracts (usually 400 µl) were made using the reporter lysis buffer from Promega.
CAT activity was measured with a CAT enzyme-linked immunosorbent assay kit (Boehringer Mannheim), whereas luciferase activity was determined using the luciferase assay system from Promega.
The hPVR gene maps to chromosome 19 (19q13.1-13.2; Refs. 1 and 17-19). Using cDNA sequences and cosmid clones, Koike et al. (1) have determined the intron and exon structure of the hPVR gene (Figs. 1A and 2A). They found that the first coding exon hybridized to a 5.7-kbp genomic BamHI fragment. However, neither the start site of transcription nor the promoter of the gene has been characterized.
To isolate the putative promoter region, a human chromosome 19-specific
genomic library was screened for -phages harboring the 5
-region of
the hPVR gene (see "Materials and Methods"). The 5
-NTR of cDNA
clone H20B maps to the center of the 5.7-kbp BamHI fragment
(Fig. 2B, open box; H20B contains
the most extended 5
-sequence information). The
BamHI-SacI fragment was likely to contain part or
all of the promoter of the hPVR gene, for which reason it
was subcloned and its nucleotide sequence was determined. Appropriately, this BamHI-SacI fragment was found
to represent the 5
-end of a cosmid insert containing a functional
hPVR gene (not shown). This cosmid has been used previously
by us to generate the mouse cell line JA-1 expressing functional
hPVR
and hPVR
isotypes (4). The BamHI-SacI
fragment was cloned into a CAT reporter plasmid and found to exert
promoter activity.2 As discussed in greater
detail below, the promoter region lacked TATA- and CAAT-like
elements.
Northern blot analyses were carried
out to determine the level of transcriptional expression and to
estimate the size of hPVR mRNAs in various cell lines. As can be
seen in Fig. 3, hybridization of RNAs of different human
and monkey cell lines (HEp 2, HeLa, RD, SK-N-SH, and COS7) with a probe
recognizing all splice variants of the hPVR gene yielded a
single band in all cases. A similar signal was obtained with RNAs of
the human cell lines HL60 and 293, whereas, interestingly, RNAs of a
variety of hematopoetic cell lines (CEM, L735, F126, 549, JI, LY66,
LY67, LY91, BL40, BL60, BL64, BL70, BL72, BL90, BL99, and Raji) gave no
detectable signals in Northern blots (data not shown). The intensity of
the hPVR signal from cells known to be permissive for poliovirus
replication suggests rather low levels of hPVR-related mRNA. In
contrast to the blots using human and monkey cell line RNAs,
experiments carried out using RNA obtained from JA-1 cells yielded a
very intense signal, showing some bands below and above the authentic
hPVR band. JA-1 cells are mouse cells stably transformed with the
hPVR gene (see above); they harbor approximately 100 copies
of the hPVR gene per cell. This is most likely the reason
for the apparent huge overproduction of hPVR RNA-specific signal.
Indeed, on short exposure of the blot, only a single band was seen in
JA-1 cells (Fig. 3), migrating to the same position as that from HeLa
cells (data not shown). We conclude that hPVR mRNAs are
approximately 3.5 kilobases in size.
The mRNAs for hPVR,
, and
share the same 3
-NTR whereas
that of hPVR
is different (Fig. 1B). Analyses with probes
specific for the detection of the isoforms hPVR
,
, and
or for
hPVR
revealed differences in the steady state levels of these two
classes of mRNAs. Whereas hybridization with the former probe
resulted in band intensities identical to those seen in Fig. 3, signals for the 3.5-kilobase-long hPVR
mRNA were observed only on
prolonged exposure of the Northern blots, regardless of the cell lines
used (data not shown). However, the ratio of
,
, and
versus
messages is approximately the same in all cell
lines tested, including JA-1.
Initial
experiments to map the start sites of hPVR mRNA by RNase protection
or primer extension yielded inconsistent results in that putative start
sites were spread over a relatively large area of the promoter (data
not shown). Therefore, we used the 5-RACE technique to determine the
hPVR mRNA 5
-ends (26; see "Materials and Methods"). Sequence
analysis of HEp2 RACE clones revealed the presence of multiple start
sites for transcription within HEp2 cells (Table I and
Fig. 4). Seventeen of 22 events mapped to either of
three start sites within a 20-bp region (Fig. 4, closed
circles). Minor start sites were suggested by a few clones mapping
to the 3
-vicinity of the major start sites (Fig. 4, open
circles). We then performed the same analysis using RNA isolated
from the JA-1 cell line. It was of interest to us to examine the
pattern of transcriptional start sites in JA-1 cells, in which the
hPVR gene is transcribed in the mouse cell context, to
determine whether it was similar to one we had mapped in HEp2 cells. As
listed in Table I, 10 of 25 RACE clones obtained from hPVR mRNA of
JA-1 cells revealed starts sites at the same locations as found in HEp2
cells. These data suggest that the 5
-ends of hPVR transcripts isolated
from JA-1 cells were slightly more heterogeneous with respect to the
start sites mapped in the HEp2 cell line (data not shown).
|
We had reported earlier the isolation of two cDNA clones (H20A and
H20B) with 5-NTRs longer than those determined by the 5
-RACE method
(see Fig. 4 for the location H20B). Therefore, we used RT-PCR to
determine whether transcriptional initiation events took place in the
region upstream of the major start sites detected by 5
-RACE (see
"Materials and Methods"). Primer C (Fig. 4, arrow)
yielded a product with mRNAs of HEp2, HeLa, COS-1, and JA-1 cells,
whereas primer B did not (data not shown). The start sites detected by
RT-PCR upstream of the major initiation sites seemed to be used at a
low frequency, because the uncloned HEp2 and JA-1 RACE pools could not
be cleaved to any noticeable extent with restriction enzyme
EaeI (see Fig. 4).
Inspection of the DNA sequence upstream of the start site cluster (Fig. 4) did not reveal TATA- or CAAT-like elements that are typical signals for precise initiation of transcription (27). Moreover, the sequences in the vicinity of the transcriptional start sites were found to have a much higher than average GC content. Many genes with these characteristics often harbor multiple start sites of transcription (28, 29). Therefore, our observation that the hPVR gene possesses multiple transcriptional initiation sites would be consistent with these previous investigations.
Location of the hPVR PromoterTo determine the sequences that
are essential for promoter activity, an XhoI-ATG fragment
(Fig. 5, fragment H) was generated by PCR and
used for further analysis. Earlier results suggested that sequences
upstream of the XhoI site are dispensable for basic promoter
activity (CAT reporter assays; data not shown). By PCR, a number of 5-
and 3
-deletions were introduced into the XhoI-ATG fragment,
and the resulting DNAs (Fig. 5, 5
-deletions, A-D;
3
-deletions, E, F, G20, and G17) were cloned
into a luciferase reporter vector. The results of transfections of the
deletion constructs into HEp2, HeLa, COS-1, and L929 cells are shown in
Fig. 6A. It is apparent that the activity
pattern caused by the deletions is very similar for the four cell lines
tested. This indicates that the hPVR promoter activity is
controlled in a similar fashion in cells of human, monkey, and mouse
origin. A significant decrease of activity was found as result of
deletions from B to C or from E to F. Thus, a minimal DNA fragment
possessing core promoter activity under the conditions of the
experiment is the B to E sequence. Accordingly, the BE segment (Fig. 5,
from
343 to
58) was cloned into the luciferase expression vector
and was shown to possess full activity in the different cell lines
(Fig. 6A). The 5
-end of the BE segment is located some 150 bp upstream of the major transcriptional start site cluster, whereas
the 3
-end maps to 56 bp upstream of the translation initiator AUG
codon. Taken together, these data suggest that sequences important for
basic promoter activity reside in the region of the transcriptional
start site cluster as well as within the region encoding the 5
-NTR of
hPVR mRNA.
It is known that certain cell lines of the hematopoetic lineage cannot be infected with poliovirus (30, 31). Most likely, these cells do not express hPVR polypeptides. As indicated before, we have tested 13 BL cell lines, representing all three classes of translocation types (32) for the presence of hPVR RNA by Northern blotting. None of these cells showed detectable levels of a signal indicative of hPVR mRNA. Raji cells, one of the BL cell lines tested by Northern blotting, was further analyzed for hPVR expression by fluorescence-activated cell sorting analysis. Using hPVR-specific monoclonal antibodies, we were unable to detect hPVR at the surface of these cells, as expected (data not shown). On the other hand, Southern blot analysis has shown that the hPVR gene is not deleted in Raji cells (data not shown). Therefore, the lack of hPVR expression in Raji cells may be due to an inactive hPVR promoter. To test this hypothesis, we have transfected the H and BE promoter constructs into Raji cells. As can be seen in Fig. 6B, Raji cells were unable to support transcription from the exogenously transfected H (and BE, data not shown) promoter construct, a result in agreement with the lack of endogenous hPVR mRNA in these cells.
As seen in Fig. 6B, expression of luciferase under the control of the H fragment promoter is relatively low in COS-1 cells. This may be explained by the fact that the expression of luciferase from the positive control vector is driven by the SV40 promoter and enhancer, which is particularly strong in these cells.
Cloning of the Monkey PVR PromotersIn contrast to humans,
monkeys possess at least two genes encoding PVRs (AGM1
and AGM
2) (7). On the basis of the extent of the sequence
divergence between the two genes, it is likely that a duplication of
the monkey PVR gene occurred after species differentiation (7).
Sequence homologies suggest that the hPVR proteins are more closely
related to the corresponding protein encoded by AGM
1
(90.2% identity) than to those encoded by AGM
2 (86.4%
identity) (7). The nonpathological function of any of these proteins is
yet to be determined, but it is possible that mPVR
1 and mPVR
2 may
have different functions or that the expression of these proteins may
be regulated via their promoters in a different manner. These
considerations prompted us to isolate the monkey promoters.
Oligodeoxynucleotides located in the 5-coding region of mPVR
1 and
mPVR
2 mRNAs, derived from the published sequence (7), and a
5
-oligonucleotide spanning the XhoI site (see Fig.
2B) were used to PCR amplify sequences corresponding to the
human H fragment of COS-1 genomic DNA. As a control, a similar
experiment was performed with HEp2 genomic DNA, yielding the H fragment
(Fig. 5). The HEp2 PCR product was sequenced, and no differences were found when compared with the sequence of the original
-human DNA
insert. The promoters of both the AGM
1 and
AGM
2 genes were sequenced (Fig. 4, alignment to
hPVR). AGM
1 clones showed heterogeneity at
three positions (designated Y, Y, and W), a finding that was confirmed
by a second, independent round of PCR amplification and sequencing. A
triplet, CTT, at position
552 of AGM
1 was also found in
one of six clones of AGM
2. It is apparent that the degree
of conservation between the promoter sequences is approximately the
same as found for the coding regions of the corresponding mRNAs.
The AGM1 and AGM
2 fragments generated by
PCR were cloned into the luciferase reporter vector and found to exert
promoter activity to an extent similar to the human counterpart (data
not shown). Since RT-PCR of COS-1 mRNA, using the same primers as for human mRNAs (see above), suggested a transcriptional start site
region in COS-1 cells similar to that found in human cell lines, we
assume that the transcriptional control exerted by the monkey promoters
is similar to that of their human counterpart. This conclusion,
however, must await confirmation by 5
- and 3
-deletion experiments of
the two monkey promoters.
It should be noted that the region upstream of the 5-NTRs is
particularly well conserved among all three promoters (Fig. 4), whereas
the region preceding the major start sites harbors a hot spot of
sequence divergence between the monkey and human promoters. In light of
the finding that the upstream sequences are largely dispensable for
basic hPVR promoter function, this was unexpected. However,
the potential importance of these sequences for fine tuning of promoter
activity was already suggested by the 5
-deletion analyses; the A and B
deletions caused an increase of activity when transfected into primate
cells and a slight but constant decrease on transfection into mouse L
cells (see Fig. 6).
Typical infections of poliovirus in humans, initiated by ingestion of the virus, are restricted to cells of gastrointestinal organs. Only rarely (1% of all infections) does the virus reach the central nervous system, where it destroys motor neurons with great preferences, thereby causing a disease syndrome called poliomyelitis. Despite this apparent tropism, mRNA encoding hPVR can be detected in small quantities in many human tissues (reviewed in Refs. 21 and 22). The control mechanism(s) governing poliovirus tissue tropism is not understood. It may be related to patterns of expression of hPVR both at transcriptional (22) and translational (33) levels, as it may also involve cellular factors such as CD44 (34).
The current study has been undertaken to dissect the promoter region of the hPVR gene. Understanding the regulation of expression of hPVR proteins may not only be crucial to understanding the unique poliovirus pathogenicity, it may also shed light on the nonpathogenic function of these proteins belonging to a new human gene family.
All subtypes of mRNAs encoding hPVRs are found in low quantities in
human tissue culture cells. Additionally, hPVR mRNA is much less
abundant than hPVR
-type mRNAs in all cases tested (data not
shown). Whether the low steady state level of the hPVR
message is
due to less frequent splicing events or is the result of decreased hPVR
mRNA stability (2) remains to be determined. Unexpectedly, this is not reflected by the amount of the corresponding proteins observed in various cell lines (10). The reason for the unequal expression of the hPVR mRNAs, on the one hand, and the roughly similar abundance of the polypeptides, on the other hand, is not known.
Using the RACE procedure, we have found that transcription of hPVR mRNAs initiates at multiple sites with a prevalence at two or three sites (Fig. 4, closed circles). This result is likely to be correct because: 1) two independent experiments using two different cell lines (HEp2 and JA-1) revealed similar results; 2) all three start sites match to some degree the consensus sequence for eukaryotic transcriptional initiation sites (CAG/TT) (27, 35); and 3) a modified RACE procedure (26) was also used to exclude that the results were derived from degraded RNA or products of incomplete reverse transcription. In this modified procedure, a RNA oligonucleotide is ligated to decapped mRNA molecules before reverse transcription is performed such that only full-length cDNA molecules can be amplified by subsequent PCR amplification. Moreover, degraded RNA is subjected to a dephosphorylation reaction before decapping of mRNAs is performed.
Minor start sites were observed by RT-PCR experiments, suggesting that
a minority of transcripts originate from the region extending
approximately to position 280 (Fig. 4). A start site window similar
to that of hPVR transcripts was also found for transcripts of mPVR
1
and mPVR
2 mRNAs by using the same 5
- and 3
-primers as for hPVR
in the RT-PCR (data not shown).
Since the hPVR gene maps to chromosome 19, we have used a
corresponding -phage library to search for sequences upstream of the
first hPVR exon. A fragment of the hPVR gene,
approximately 2.9 kilobases in length, was identified as containing
promoter activity (Fig. 2B). Sequence analysis of this
fragment and successive 5
-deletions suggested that the promoter
activity resides downstream of the XhoI site (data not
shown). Accordingly, the remaining sequence of the hPVR gene
(XhoI to initiating AUG) was subjected to an intensive
analysis revealing a minimal promoter fragment of approximately 280 bp
that contained the transcriptional start site cluster.
The characteristics of the hPVR promoter (short minimal DNA sequence, relatively CG-rich, and lack of TATA- or CAAT-like sequences) are common to a variety of genes, such as the human phenylalanine hydroxylase (36), murine complement component C4 (37), mouse thymidylate synthetase (38), mouse mb-1 (28), mouse fibroblast growth factor receptor 1 (39), and human pregnancy-specific protein (16) genes and the human RET proto-oncogene (40). It is noteworthy that most of these genes show cell type and/or developmentally regulated expression.
Several cell lines derived from hematopoetic tissues, BL lines included, have been reported to be resistant to poliovirus infection (30, 31). Northern blot analyses for hPVR mRNA in RNAs of 12 BL cell lines were negative (data not shown). RNAs of BL40 and Raji cells did not yield a signal even by the RT-PCR method. On the other hand, transfection of poliovirus RNA into Raji cells leads to poliovirus replication, a result eliminating the possibility of a cell internal block to viral replication.3 Therefore, we conclude that resistance of these cells to poliovirus replication is determined by the failure to express a suitable receptor. Since an exogeneously transfected hPVR promoter is inactive in Raji cells (see Fig. 6B), the apparent lack of hPVR expression in these cells is likely to be due to transcriptional repression.
Preliminary footprint analyses, using HeLa cell extracts, revealed three areas that suggested protection (for location see Fig. 4; data not shown). A very similar footprint pattern was obtained when the experiment was carried out with nuclear extracts of mouse L cells (data not shown), a result suggesting that factors controlling hPVR promoter activity and transcriptional start site selection are similar in human and mouse cells. This was not unexpected, since mice transgenic for the hPVR gene are not only sensitive to poliovirus infection, but when inoculated intravenously or intracerebrally, they produce a disease syndrome nearly identical to human poliomyelitis (24, 25, 41, 42).
As determined by computer analysis (43), possible candidates for
factors binding to the hPVR core promoter region are PuF, AP2, and GCF factors. AP2 and GCF binding sites have been observed in
promoters related to the structure of the hPVR promoter,
such as the human ATP synthase subunit (44), rat SSTR1 (45), rat
SSTR4 (46), human ret proto-oncogene promoters (40), and mouse
fibroblast growth factor receptor 1 (39). Gel mobility shift analyses
needed to further characterize the binding activities to the
hPVR promoter are in progress. It will be particularly interesting to test whether the expression of the hPVR gene
is developmentally regulated or modulated in specific differentiation pathways, since, besides the AP2 sites, two hypothetical PEA3 sites are
apparent in the hPVR promoter sequence. PEA3 motifs have
been observed in promoters of developmentally regulated genes (for
example, see Ref. 47).
We have also isolated and analyzed the promoter regions of the AGM genes, the monkey homologues to hPVR. These promoters are highly homologous to that of the hPVR gene with the exception of the region upstream of footprint 3. The significance of this difference is not known.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) X94226[GenBank] for hPVR, X94227[GenBank] for AGM1, and X94228[GenBank] for
AGM
2.
We thank Patrick Hearing and Birgit Woelker for their generous help in carrying out the footprint analyses, Patrick Hearing, Winship Herr, and Rolf Knippers for discussions, M. Frohman for technical advice concerning the RACE strategies, and Fred Lahser for proofreading the manuscript.