Evolution of Human and Non-human Primate CC Chemokine Receptor 5 Gene and mRNA
POTENTIAL ROLES FOR HAPLOTYPE AND mRNA DIVERSITY,
DIFFERENTIAL HAPLOTYPE-SPECIFIC TRANSCRIPTIONAL ACTIVITY, AND ALTERED
TRANSCRIPTION FACTOR BINDING TO POLYMORPHIC NUCLEOTIDES IN THE
PATHOGENESIS OF HIV-1 AND SIMIAN IMMUNODEFICIENCY VIRUS*,
Srinivas
Mummidi
§,
Mike
Bamshad§¶,
Seema S.
Ahuja
,
Enrique
Gonzalez,
Pablo M.
Feuillet,
Kazi
Begum,
M. Cristina
Galvis,
Vannessa
Kostecki,
Anthony J.
Valente,
Krishna K.
Murthy**,
Luis
Haro,
Matthew J.
Dolan§§,
Jonathan S.
Allan**, and
Sunil K.
Ahuja¶¶
From the Departments of Medicine,
University of Texas Health Science Center at San Antonio and South
Texas Veterans Health Care System, Audie L. Murphy Division,
San Antonio, Texas 78229, the ¶ Department of Pediatrics,
Eccles Institute of Human Genetics, University of Utah,
Salt Lake City, Utah 84112, the ** Department of Virology and
Immunology, Southwest Foundation for Biomedical Research,
San Antonio, Texas 78228, the Department
of Life Sciences, University of Texas, San Antonio, Texas 78249, and
the §§ Infectious Diseases Service, Wilford Hall
Medical Center, Lackland Air Force Base,
San Antonio, Texas 78236
Received for publication, July 22, 1999, and in revised form, January 10, 2000
 |
ABSTRACT |
Polymorphisms in CC chemokine receptor 5 (CCR5),
the major coreceptor of human immunodeficiency virus 1 (HIV-1) and
simian immunodeficiency virus (SIV), have a major influence on HIV-1 transmission and disease progression. The effects of these
polymorphisms may, in part, account for the differential pathogenesis
of HIV-1 (immunosuppression) and SIV (natural resistance) in humans and non-human primates, respectively. Thus, understanding the genetic basis
underlying species-specific responses to HIV-1 and SIV could reveal new
anti-HIV-1 therapeutic strategies for humans. To this end, we compared
CCR5 structure/evolution and regulation among humans, apes,
Old World Monkeys, and New World Monkeys. The evolution of the
CCR5 cis-regulatory region versus the open
reading frame as well as among different domains of the open reading
frame differed from one another. CCR5 cis-regulatory region
sequence variation in humans was substantially higher than anticipated.
Based on this variation, CCR5 haplotypes could be organized
into seven evolutionarily distinct human haplogroups (HH) that we
designated HHA, -B, -C, -D, -E, -F, and -G. HHA haplotypes were defined
as ancestral to all other haplotypes by comparison to the
CCR5 haplotypes of non-human primates. Different human and
non-human primate CCR5 haplotypes were associated with
differential transcriptional regulation, and various polymorphisms
resulted in modified DNA-nuclear protein interactions, including
altered binding of members of the NF-
B family of transcription
factors. We identified novel CCR5 untranslated mRNA sequences that
were conserved in human and non-human primates. In some primates,
mutations at exon-intron boundaries caused loss of expression of
selected CCR5 mRNA isoforms or production of novel mRNA
isoforms. Collectively, these findings suggest that the response to
HIV-1 and SIV infection in primates may have been driven, in part, by
evolution of the elements controlling CCR5 transcription and translation.
| |
INTRODUCTION |
Simian immunodeficiency viruses
(SIVs)1 comprise
a large and genetically diverse group of lentiviruses that originated
in sub-Saharan Africa (1-3). SIVs isolated from chimpanzees and
mangabeys are very similar to human immunodeficiency virus (HIV)-1 and
HIV-2, respectively (3-7). This similarity suggests that HIVs arose via cross-species transmission from non-human primate viral reservoirs. Yet, despite their common ancestry and close similarity, HIVs and SIVs
differ significantly with regard to clinical disease and pathogenesis.
Human infection with HIVs results in a progressive immunodeficiency
syndrome, whereas African apes and monkeys (e.g. African
Green monkeys (AGM)) infected with SIV exhibit no evidence of disease
(7-9). These differences in pathogenicity may be due, in part, to
primate species-specific variation in the genes controlling the host
response or expression of host HIV/SIV entry factors (10). Thus,
understanding the evolution and factors that control the expression of
these genes in primates will be an important step towards identifying
the molecular mechanisms underlying the response of primates to
infections with SIVs and HIVs. In turn, this may illuminate potential
strategies that could be used to mitigate or prevent infection with
HIV-1.
An important host genetic determinant of HIV-1 pathogenesis is
polymorphisms in the open reading frame (ORF) and
cis-regulatory region of CC chemokine receptor 5 (CCR5), a
major coreceptor for the entry of HIV and SIV (reviewed in Refs. 10 and
11). These polymorphisms could potentially influence cell-surface
density of CCR5 and thus have an impact on HIV/SIV pathogenesis. For
example, homozygosity for a 32-bp deletion in CCR5 ORF leads to loss of surface expression and profound resistance against HIV-1 infection (12). Similarly, a 24-bp deletion in the CCR5 ORF that was discovered in selected non-human primates might influence SIV pathogenesis (13).
More recently, we and others (14-17) have shown that polymorphisms in
the cis-regulatory region of human CCR5 are also
associated with altered rates of disease progression and transmission.
Thus, due to this close interaction with the lentiviral life cycle, CCR5 is an excellent candidate for exploring whether differences in its
gene/RNA structure and regulation account, in part, for the
differential pathogenesis of HIV and SIV. Additional interest in
analyzing host-specific differences in CCR5 regulation at the gene and
RNA level is spurred by the observation that despite having nearly
identical CCR5 coding regions, Asian macaques (e.g. Pig-tailed macaques (PTM)) but not AGMs infected experimentally with
SIV become symptomatic (reviewed in Refs. 1, 18, and 19). Given that
the differential pathogenicity of AGMs and PTMs to SIV is very
reminiscent of the differential pathogenicity of HIV and SIV, studies
aimed at dissecting the genetic basis for this differential
pathogenicity might provide additional insights into the varied
susceptibility to HIV-1 infection among humans.
The gene and RNA structure of human CCR5 is complex. We have
demonstrated that alternative splicing in the 5'-untranslated regions
(UTR) of CCR5 generates several distinct mRNA isoforms that are
under the control of at least two distinct promoters (20). Furthermore,
the 5'-UTR of CCR5 is encompassed within these
cis-regulatory regions that contain several single
nucleotide polymorphisms (SNPs) associated with altered rates of HIV-1
disease progression (14-17). Thus, polymorphisms in the noncoding
region of CCR5 could influence not only
cis-trans interactions that impact on gene
expression but also CCR5 mRNA stability and/or the efficiency of translation.
Given the multiple levels at which CCR5 expression could be regulated,
we performed a comprehensive analysis of the ORF, mRNA structure, and
transcriptional regulatory units of CCR5 relative to four important
events in human evolution (21): the divergence of humans from
chimpanzees 6 million years ago, from the Orang-Utan lineage 15 million
years ago, from the cercopithecoids (Old World monkeys (OWM)) ~35
million years ago, and from New World monkeys (NWM) 50 million years
ago. Results from these analyses enabled us to build the evolutionary
framework needed to define the relationships among human
CCR5 haplotypes that influence HIV-1 pathogenesis. This
evolutionary framework facilitated greatly our ability to determine the
influence of CCR5 haplotypes on HIV-1 transmission and
disease progression
(16).2
Additionally, we were able to test directly the hypothesis that polymorphisms in the human and non-human primate
cis-regulatory region of CCR5 confer differences
in transcriptional efficiencies and/or interact with different
trans-acting factors or to the same transcription factor but
with varying avidity.
| |
EXPERIMENTAL PROCEDURES |
Primate CCR5 ORFs--
The CCR5 ORF was PCR-amplified with
primers that flanked the human CCR5 ORF (gcg gcc gct tat gca cag ggt
gga aca ag (forward) and tct aga cca ctt gag tcc gtg tca (reverse)),
cloned, and sequenced on both strands from the following species
(common names are in parentheses): Chlorocebus
(Cercopithecus) aethiops sabaeus (sabaeus monkey);
Hylobates agilis unko (Agile gibbon); Callithrix
jacchus (marmoset); Ateles geoffroyi (black-handed
spider monkey); and Lagothrix lagothricha (woolly monkey).
In addition, the following primate sequences available in
GenBankTM (common names and GenBankTM accession
numbers are in parentheses) were used to construct the CCR5 ORF network
(Fig. 1a): Homo sapiens (human, X91492); Pan troglodytes (chimpanzee, AF035214); Gorilla
gorilla (gorilla, AF005659); Pongo pygmaeus
(Orang-Utan, AF075446); Hylobates leucogenys (white-cheeked
gibbon, AF075451); Cercocebus torquatus atys (sooty
mangabey, AF051905); Cercocebus aterrimus (black mangabey,
AF081578); Cercocebus galeritus (Tana River mangabey, AF035215); Macaca fascicularis (crab-eating macaque,
AF005660); Macaca mulatta (Rhesus macaque; AF005662);
Macaca assamensis (Assamese macaque, AF075449); Macaca
arctoides (stump-tailed macaque, AF075450); Macaca
nemestrina (pig-tailed macaque; AF105282) Papio hamadryas
hamadryas (baboon, AF005658); Papio hamadryas anubis
(Olive baboon, AF023452); Colobus guereza (black and
white colobus, AF141639); Cercopithecus neglectus (De
Brazza's guenon, AF035218); Cercopithecus nictitans
(greater spot-nosed guenon, AF035219); Cercopithecus lhoesti
(l'Hoest's monkey; AF081579); Cercopithecus cephus
(mustached guenon; AF035217); Cercopithecus ascanius
(red-tailed guenon, AF035216); Pygathrix nemeaus
(red-shanked douc langur, AF075448); Pygathrix bieti (black
snub-nosed monkey, AF075445); Rhinopithecus avunculus (Tonkin snub-nosed monkey, AF075447); Rhinopithecus
roxellana (Golden snub-nosed monkey, AF075444);
Trachypithecus francoisi (Francois langur, AF075442);
Erythrocebus patas (Patas monkey, AF035220);
Chlorocebus aethiops (grivet; AB015944); Chlorocebus aethiops sabaeus (AF035221); Chlorocebus aethiops
pygerythrus (vervet; AF035222); Chlorocebus aethiops
tantalus (tantalus monkey; AF081577); and Presbytis
phayrei (Phayre's leaf monkey, AF075443). All PCR amplifications
were performed with the following conditions: initial denaturation of
the DNA template for 90 s, followed by 30 cycles of a profile
consisting of 94 °C for 10 s, 55 °C for 30 s, and
72 °C for 1 min. To reduce the potential for errors, a proofreading
polymerase, Pfu (Stratagene, La Jolla, CA), was included
with the Taq DNA polymerase (Life Technologies, Inc.) at a
15:1 ratio during the PCR amplification.
CCR5 Numbering System, RNA and Promoter Nomenclature--
The
CCR5 numbering system that we used previously (15, 16, 20)
was based on the sequences deposited in GenBankTM
(accession numbers AF031236 and AF031237) and considered the
first nucleotide of the 5'-most UTR sequence as +1. However, since we
had identified additional 5'-UTR sequences (Fig. 2), and to maintain
uniformity in the numbering systems used by different investigators, it
was proposed at the CCR5-AIDS
symposium3 to designate the
first nucleotide of the CCR5 translational start site as +1 and the
nucleotide immediately upstream as 
1 (22). It was also proposed at
this meeting to change the numbering system for CCR5 noncoding exons
that we had identified previously (20). Exons 2 and 3 are two noncoding
exons that are not interrupted by an intron and are now designated as
exons 2A and 2B; exon 4 is now designated as exon 3. Finally, to
standardize the nomenclature of the CCR5 promoters that we had
identified previously (20), in this paper we have designated the
downstream CCR5 promoter as promoter 1 and the upstream promoter as
promoter 2. A similar nomenclature has been used by other investigators
for genes that have multiple promoters (23, 24).
Primate CCR5 cis-Regulatory Region--
The region corresponding
to human CCR5 2761 to 1835 was PCR-amplified, cloned,
and sequenced on both strands from the following primates (the number
of different members of the given non-human species that were sequenced
are shown in parentheses): P. troglodytes (n = 4); G. gorilla; P. pygmaeus; P. hamadryas
anubis (n = 3); M. mulatta
(n = 2); M. fascicularis; M. nemestrina; C. torquatus torquatus (red-capped
mangabey); C. galeritus chrysogaster (gold-bellied mangabey); C. guereza; C. guereza kikuyuensis
(Kikuyu colobus); Cercopithecus petaurista (lesser
spot-nosed guenon); C. neglectus; C. diana (Diana
guenon); C. l'hoesti; C. (Miopithecus) talapoin (Talapoin); C. (Erythrocebus) patas; C. aethiops (grivet;
n = 3); C. sabaeus (n = 8);
C. pygerythrus (n = 3); Presbytis
(Trachypithecus) francoisi; Saguinus oedipus
(cotton-topped tamarin); C. jacchus; Aotus
trivirgatus (owl monkey); A. geoffroyi; and L. lagotricha. A single allele per non-human primate was sequenced.
For Homo sapiens, 60 alleles were sequenced, and they were
derived from individuals with different genotypes, including those who
were homozygous or heterozygous for 2733A or 2733G, 2135C or
2135T, 1835T or 1835C (15). The CCR5 promoter region
from non-human primates was PCR-amplified using the following primers:
cat aaa gaa cct gaa ctt gac c (forward) and tag aat ttc taa tat aaa att cta tta aca tac tcg tga acc aca aac ggt cta (reverse). The PCR protocol
used to amplify the human and non-human primate CCR5 cis-regions and the ORF was identical. Notably, several alleles from a given species had identical sequences, and repeat
amplification/sequencing of the identical allele gave identical results.
Genotype Analysis of Non-human Primates--
Genotyping methods
for the CCR5 polymorphisms were described previously (15,
16).
5'-RACE and Reverse Transcription-PCR (RT-PCR)--
Total RNA
from human and non-human primate peripheral blood mononuclear cells
(PBMC) and human leukocyte subsets was extracted using Trizol reagent
(Life Technologies, Inc.). 5'-Rapid amplification of cDNA ends
(RACE) was performed on a human leukocyte cDNA library (CLONTECH, Palo Alto, CA) using an exon 2B-specific
primer (ggg aac gga tgt ctc agc tct tct) according to the
manufacturer's protocols. For RT-PCR, RNA was reverse-transcribed
using a CCR5 exon 3-specific oligonucleotide (acc aaa gat gaa cac cag
tga gta gag), and the resulting cDNA was amplified using a forward
primer derived from the newly identified sequence of CCR5 exon 1 (tgt
ctt ctc agc tct gct gac; Fig. 2) and a reverse primer derived from CCR5
exon 3 (gct ccg atg tat aat aat tga tgt). The specificity of the
products obtained from the PCR was confirmed further by performing a
nested PCR. The sequences of the primers used in the nested PCR were aat act tga gat ttt cag atg (forward) and aga ttg gac ttg aca ctt gat
aat cca t (reverse). All RT-PCR reactions were run with a negative
control that included no cDNA template.
Promoter Analysis--
To study the differences between the
CCR5 promoter activity of sabaeus AGM and that of humans, we
constructed a series of firefly luciferase-sabaeus (S1 to S5) and human
(H1 to H5) CCR5 promoter constructs in the promoterless
pGL3Basic vector (Promega, Madison, WI). A single sabaeus haplotype and
a haplotype representative of CCR5 human haplogroup (HH)-A
(16) were used to construct the reporter plasmids. The constructs were
transfected into human embryonic kidney (HEK), human erythroleukemia
(K562), and AGM kidney (COS) cell lines and tested for luciferase
reporter activity, as described previously (20). To study differences
in promoter activity exhibited by the cis-regulatory regions
of human CCR5 haplotypes, the genomic region spanning 2761
to 1814 was PCR-amplified from haplotypes corresponding to HHA, HHC,
HHE, HHF, or HHG (16) and cloned into the pGL3Basic vector.
Transfection into K562 and Jurkat cell lines and dual luciferase assays
were as described previously (20). For all promoter activity analyses,
at least two different plasmid preparations were used, and the DNA in
each plasmid preparation was quantified spectrophotometrically twice. The Wilcoxon signed-ranks test was used to compare the mean luciferase activity between homologous sabaeus and human promoter constructs. Statistical analysis to determine the differences in the mean luciferase activity among human CCR5 promoter alleles was by
one-way analysis of variance followed by the Scheffe's post hoc test.
Electrophoretic Mobility Shift Assay (EMSA)--
All cell lines
were obtained from American Type Culture Collection and were maintained
as described previously (20). Nuclear extracts were prepared from K562,
THP-1 (human monocyte), Jurkat (human T-cells), and COS cell lines,
according to standard protocols (25). Nuclear proteins from
PHA-stimulated human PBMCs were extracted according to the procedure of
Schreiber et al. (26). Prior to nuclear protein extraction,
human PBMCs were stimulated for 3 h with phorbol myristic acid (25 ng/ml) and ionomycin (1 µg/ml). PBMCs from SIV-uninfected sabaeus AGM
and vervet AGM were stimulated with 10 µg/ml PHA for 2 days and were
maintained in culture for 11 days in the presence of 10% human IL-2
(Advanced Biotechnologies Inc., Columbia, MD) before extracting the
nuclear proteins. Nuclear extracts from similarly PHA/IL-2-treated
human PBMCs served as controls for the EMSA experiments that used
sabaeus/vervet PHA-blast nuclear extracts. The nucleotide sequences of
the oligonucleotides used in EMSAs are shown in Table
I. For competition experiments, unlabeled
competitor oligonucleotides were incubated with the nuclear extracts
for 10 min on ice prior to adding the labeled probe. The
specificity of the binding reactions was confirmed by using nonspecific
double-stranded oligonucleotide competitors. For supershift
experiments, the indicated antibodies (Santa Cruz Biotechnology, Santa
Cruz, CA) were incubated with the nuclear extracts and the radiolabeled
probes for 45 min at room temperature prior to electrophoretic
resolution of the complexes. Densitometry analysis of the EMSA
gels was performed with the NIH Image software package (version
1.61).
View this table:
[in this window]
[in a new window]
|
Table I
Nucleotide sequences of the oligonucleotides used in EMSAs
Only the sequence of the sense oligonucleotides is shown. The
nucleotide differences between the oligonucleotides used to determine
differential nuclear factor binding at human CCR5 SNPs are
in capital letters and are in parentheses (see "Experimental
Procedures").
|
|
Phylogenetic Analysis--
Sequences were aligned using the
SEQUENCHER software package. Descriptive statistics were obtained using
ARLEQUIN software.4 Mean
nucleotide diversity within populations was estimated using the
equation, 
= (n/n 1)
xixjij; where
n is the number of DNA sequences examined,
xi and xj are the population
frequencies of the ith and jth type of DNA
sequences, and ij is the proportion of nucleotides that
differ between the ith and jth types of DNA
sequence. Genetic distances between sequences were estimated using
DNADIST of the PHYLIP software package using Kimura's two-parameter
model.5 The transition to
transversion ratio was varied from 2:1 to 10,000:1 but had no
substantial impact on the results. Distances between populations were
estimated from distances between individuals using NEIDIST (27).
Relationships between lineages and/or populations were depicted as
neighbor-joining networks (28), using NEIGHBOR. Inferred branch lengths
with negative values were converted to branches of length zero.
Parsimony networks were constructed using DNAPARS. The robustness of
branches was assessed by using 100 bootstrap data sets obtained using
SEQBOOT. Neighbor-joining and parsimony trees were produced using mouse
sequence as an outgroup, and CONSENSE was used to find the consensus
tree. Networks were visualized using TREETOOL. Estimates of the rates
of nonsynonymous (dN) and synonymous (dS) substitutions for all
pairwise comparisons were calculated using the method of Nei and
Gojobori (29) as implemented in the PAML package (30).
| |
RESULTS |
Molecular Evolution of the CCR5 ORF in Primates--
Comparison of
the complete CCR5 ORF from 37 different primates revealed that the
nucleotide sequence and amino acid identity of CCR5 were highly
conserved (alignment is shown in Supplemental Figs. 1 and 2; species
list is provided under "Experimental Procedures"). Of the variable
sites, 184 were SNPs, including 148 transitions and 55 transversions.
To quantify the amount of sequence variation found among primates, we
estimated the nucleotide diversity of each group. Nucleotide diversity
is equivalent to the number of nucleotide differences per site between
a random pair of chromosomes drawn from a population. For all primates,
the mean nucleotide diversity of the CCR5 ORF was 0.025 (~1 variant
in every 40 bp). Levels of total nucleotide diversity differed
substantially among hominoids (humans and apes), OWM, and NWM.
Nucleotide diversity in hominoids (0.009) and OWM (0.0013) was
substantially lower than that found within the total primate group,
whereas nucleotide diversity in NWM was the highest of all primate
groups (0.032).
In coding regions, mutation and selection are expected to have
different effects on nonsynonymous (dN) and synonymous (dS) nucleotide
substitutions. Consequently, comparisons of the rate of dN to dS
substitutions (dN/dS) can be utilized to explore molecular sequence
evolution (31). The neutral theory predicts that despite varying
mutation rates between lineages, dN/dS should remain constant among
lineages. Thus, variation of dN/dS among lineages is considered evidence against neutrality, whereas dN/dS ratios >1.0 are evidence for positive selection (32).
Pairwise maximum likelihood estimates of dN/dS among primate
CCR5 ORFs were consistently <1.0. However, estimation of
dN/dS for each of the domains of CCR5 (e.g. NH2
terminus, extracellular loops, and intracellular tail) revealed an
interesting trend. Pairwise estimates of dN/dS among hominoids and NWM,
for the sequence encoding the NH2 terminus, were
consistently >1.0. Pairwise estimates of dN/dS for the second
extracellular loop among hominoids and two of three NWM (spider or
woolly monkey) were >1.0 (dN/dS between hominoids and marmosets was
always <1.0). These findings suggested that the effects of natural
selection might vary among specific domains of CCR5 ORF.
Moreover, these results indicated that substitutions in the
NH2 terminus and second extracellular loop may underlie a
selective response to the pathogens after the NWM and Catarrhine split.
This was consistent with the finding that the bulk of polymorphisms in
the human and non-human CCR5 ORF have been found primarily in the NH2 terminus, and only a few are found elsewhere
(33, 34).
Phylogenetic reconstruction of the genetic affinities among hominoids,
OWM, and NWM demonstrated that NWM were substantially more divergent
from either hominoids or OWM (Fig.
1a). That is the genetic
distance between NWM and hominoids (0.054) or NWM and OWM (0.055) was
much larger than the genetic distance between hominoids and OWM
(0.014). These findings were consistent with estimates of genetic
divergence among these groups based upon analysis of morphological and
neutral genetic markers (35). Thus, it is noteworthy that despite the
different roles that CCR5 may have played in mediating responses to
pathogens (e.g. SIV and HIV-1) among African OWM, Asian OWM,
and hominoids, the pattern of variation observed among primate CCR5
ORFs is similar to that observed for neutral markers. Overall these
data suggest that the expression/function of CCR5 among OWM and
hominoids is more likely to be controlled by factors that regulate CCR5
transcription, mRNA processing, and/or translation than selection
for different ORF variants. For this reason, we explored the nature of
variation in the mRNA structure and cis-regulatory
region of CCR5 in NWM, OWM, and hominoids.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 1.
Phylogenetic relationship of and genetic
variation in the coding and a cis-regulatory region of
primate CCR5. a, phylogenetic networks
for primate CCR5 OR5 and promoter 1 (downstream promoter, old
nomenclature (20)). The higher primates included in the analysis are
shown in red, OWM in blue, and NWM in
purple. The number of alleles from a particular species that
was included in the analysis is shown in parentheses. The
bootstrap supports for each branch is denoted at the branch point
(bootstrapping is a commonly used procedure for estimating the
statistical significance of individual branches within a network).
Estimates of bootstrap support for both phylogenetic networks were
obtained using the homologous mouse sequences to root the consensus
trees. The sequence alignment of the CCR5 ORF sequences and
cis-regulatory region used in these analyses is shown in
Supplemental Figs. 1 and 3, respectively. b, comparison of a
CCR5 cis-regulatory region of primates (2761 to 1835).
The region corresponding to human CCR5 2761 to 1835 was
sequenced from genomic DNA of 60 humans and 43 non-human primates. The
sequence from 2762 to 2903 was obtained by sequencing human and
non-human primate CCR5 mRNA. The complete sequence alignment of the
non-human primates is shown in Supplemental Fig. 3. The numbering
system is based on the first nucleotide of the CCR5 ORF as +1. The
sequence of two representative great apes and OWMs and a single NWM is
shown. Dots, sequences not shown; dashes,
deletions/gaps found (relative to human); asterisks, regions
upstream of CCR5 2761 that were not determined. 2761
denotes +1 in the old numbering system (20). Arrows
demarcate the boundaries of human CCR5 exons 1, 2A, and 2B and introns
1 and 2. 2761 and 2903 correspond to the 5'-most end of the longest
human CCR5 mRNA isoform characterized previously (20) and in this
study (see Fig. 2), respectively. Exon sequences are in
uppercase and represent the CCR5 mRNA identified in this
study from humans, chimpanzee, rhesus, and sabaeus monkey. Intron
sequence is in lowercase. Orang-Utan and woolly monkey CCR5
mRNA structure was not determined, and hence we cannot predict the
exon-intron boundaries; the exon-intron boundaries shown are based on
the sequence homology of the splice acceptor-donor sites between these
two species and humans. The seven common polymorphic nucleotides
identified in the CCR5 cis-regulatory region spanning from
2761 to 1835 are indicated on the top row in blue (human
polymorphisms: 2733, 2554, 2459, 2135, 2132, 2086, and
1835). The sequence corresponding to these seven polymorphic sites in
CCR5 human haplogroup A (HHA) and non-human primates is
shown in red. Green residues indicate where chimpanzee
sequence differs from humans. The gaps (relative to human
sequence) present in the vast majority of OWMs (gaps 1-6)
and the dinucleotide gap present in chimpanzees are boxed.
Note that the nucleotide sequence differences between humans and
Orang-Utan, rhesus, sabaeus, or woolly monkey are not highlighted. The
polymorphic nucleotides/gaps (relative to human sequence) in
chimpanzees and OWMs that are present between 2903 and 2761 are not
shown. c-i, intra- and interspecies polymorphisms in
non-human primates. The unique CCR5 haplotypes identified in
four chimpanzees, three baboons, three vervets, two rhesus macaques,
three grivets, one cynomolgus macaque, one PTM, and eight sabaeus AGMs
are shown. Dots, sequences not shown; dashes,
gaps found. The numbering system is identical to that in b.
Exons and intron sequences are in upper and
lowercase, respectively. The polymorphisms that are found at
the same position in several species are in green
(e.g. the polymorphism at a position corresponding to human
CCR5 2719 is common to rhesus and sabaeus). Intra-species
polymorphisms are shown in red.
|
|
Evolution of a cis-Regulatory Region of CCR5 in Non-human
Primates--
Alignment of the nucleotide sequence of a
cis-regulatory regions of CCR5 from non-human
primates revealed high sequence conservation (Fig. 1b and
Supplemental Fig. 3). Nevertheless, substantial intra- and interspecies
sequence variation was observed (Fig. 1, c-i). Compared
with the human sequence, one gap was required to align the sequence of
the chimpanzee CCR5 cis-regulatory region, and six gaps were
inserted to align the OWM sequences. No gaps were required to align the
gorilla and Orang-Utan CCR5 promoter sequences; however,
several gaps were required to align the 44 primate sequences (Supplemental Fig. 3).
Compared with the CCR5 ORF, a cis-regulatory
region of CCR5 demonstrated similar levels of nucleotide
sequence diversity. Of the polymorphic sites, 237 were SNPs, including
177 transitions and 68 transversions. For all primates, the mean
nucleotide diversity of the cis-regulatory region of
CCR5 was 0.022, which is approximately 1 variant in every 45 bp. Mean nucleotide diversity was 0.007, 0.007, and 0.028 in hominoids,
OWM, and NWM, respectively. The cis-regulatory regions of
CCR5 in chimpanzee and human differed at 41 sites, including
8 fixed and 33 variable sites, suggesting that the CCR5
cis-regulatory regions of chimpanzees and humans were considerably
more different from one another than their CCR5 ORFs.
Genetic distances estimated from the cis-regulatory region
of CCR5 of hominoids, OWM, and NWM indicated that hominoids
were nearly equally divergent from OWM and NWM (Fig. 1a),
i.e. the genetic distance between hominoids and OWM (0.058)
was comparable to the genetic distance between hominoids and NWM
(0.067). This was in contrast to the closer affinity of hominoids and
OWM, as estimated from analysis of the CCR5 ORF. In other
words, the genetic distance between OWM and NWM was similar regardless
of whether the CCR5 ORF or cis-regulatory regions
were compared, whereas the genetic distances between OWM and hominoids
estimated from the cis-regulatory region of CCR5
was 4 times greater than estimates of the genetic distance from the
CCR5 ORF. These data suggested that the CCR5
cis-regulatory region of hominoids was substantially more
divergent from OWM than is the CCR5 ORF. This underscores the potential role that different evolutionary forces (e.g.
natural selection) may have played in shaping the genetic variation of the cis-regulatory region of CCR5 in hominoids
versus OWM.
CCR5 mRNA Splicing Patterns in Primates: mRNA Isoform
Diversity--
We designed experiments to investigate whether the
complex mRNA structure of CCR5 (20) is conserved across different
primate species. Furthermore, because the cis-regulatory
region of CCR5 encompasses the 5'-UTR region of human CCR5
(Fig. 2a and b), we determined whether the genetic variation we observed in this region in
non-human primates would influence the nature of the mRNA isoforms produced in different primates.
View larger version (45K):
[in this window]
[in a new window]
|
Fig. 2.
Conservation of a complex CCR5 mRNA
structure in primates. a, genomic organization of human
CCR5. Numbered boxes, exons; slashed lines, gaps;
Promoter 1 and Promoter 2 indicate the previously
designated downstream (Pd) and upstream (Pu) promoter, respectively
(20). 5'-RACE extended the previously described exon 1 by an additional
142 bp (2762 to 2903 which is equivalent to 1 to 142 of
GenBankTM accession numbers AF031237 and AF031237).
Arrows under the open boxes indicate
the location of the sense and antisense primers used in RT-PCR
experiments. b, novel CCR5 mRNA sequence (2903 to
2762) and genomic sequence upstream of human CCR5 exon 1. Exon and
intron/ cis-regulatory sequences are in uppercase
and lowercase, respectively. The novel mRNA sequences
identified in this paper are highlighted in red, and the
previously identified exon 1 mRNA sequences (20) are in black
uppercase. Arrows indicate the start of exon 1 and
intron 1. Putative transcription factor-binding sites upstream of exon
1 are overlined with a straight arrow.
c, schema of alternative CCR5 mRNA splicing
patterns in primates. mRNA isoforms that contain exon 1 are
arbitrarily designated as full-length transcripts (CCR5A and CCR5B),
and those lacking it are designated as truncated transcripts and are
not assigned a name (20). Tick marks indicate exon
boundaries, and the angled dashed lines indicate the
splicing patterns. gt and ag indicate the splice
donor and acceptor sites, respectively. The mutations in the splice
acceptor site are denoted in red. These mutations prevent
the production of the CCR5 mRNA isoforms that are marked by an X
(e.g. CCR5B was not found in cynomologous, rhesus or sabaeus
AGM). In a sabaeus monkey, a CCR5 mRNA species was found whose
structure includes a truncated exon 1 (lacks the region corresponding
to human CCR5 2752 to 2705) and exons 2A, 2B, and 3. The
non-consensus splice donor site of the truncated sabaeus exon 1 is
shown in green. d, expression of novel exon
1 sequence in primates. An ethidium bromide-stained gel of the RT-PCR
products is shown. The sources for the RNA used in RT-PCR are indicated
on the top, and abbreviations used are as
follows: P, human peripheral blood mononuclear cells;
L, human lymphocytes; AC, human activated CD4+
cells; and SAB, peripheral blood mononuclear cells from a
sabaeus AGM. The location of the final set of primers used in the
RT-PCR is shown in a. The length of the products amplified
from human cells is indicated on the left. In sabaeus,
the isoform homologous to CCR5A is shorter because of a deletion (48 nucleotides) in the 3' end of the region corresponding to human CCR5
exon 1.
|
|
We have shown previously that two "full-length" CCR5 mRNA
transcripts (CCR5A and CCR5B) arise by alternative splicing (20) (Fig.
2c). Several "truncated" transcripts can also originate in either exon 2A (old nomenclature exon 2) or exon 2B (old
nomenclature exon 3) of CCR5 (Fig. 2c). By using 5'-RACE on
a human leukocyte cDNA library, we extended the known CCR5 mRNA
sequence common to both CCR5A and CCR5B by 142 additional nucleotides
(2903 to 2762; Fig. 2, a and b). Analysis of
the genomic DNA sequence immediately upstream of the newly identified
mRNA sequence in exon 1 indicated the presence of several potential
transcription factor binding sites (Fig. 2b). This new exon
1 sequence was subsequently found in different human leukocyte subsets
as well as in mononuclear cells of several non-human primate species
(Fig. 2, c and d, and data not shown).
Comparison of the genomic DNA sequence extending from exon 1 to exon 2B
among non-human primates (Fig. 1b and Supplemental Fig. 3)
and RNA transcripts in mononuclear cells derived from chimpanzees,
rhesus macaque, cynomolgus macaque, and sabaeus AGM revealed the
following (Fig. 2, c and d). The exon-intron
splice donor and acceptor sites were conserved between humans and
Orang-Utan, gorilla, langur, and NWM. Mutations in the exon-intron
splice acceptor donor sites lead to loss of expression of selective
CCR5 mRNA isoforms in different non-human primates. Alternatively, usage of a non-canonical splice donor site in exon 1 of sabaeus resulted in the expression of a novel mRNA isoform. Despite these differences, it appears that the overall mRNA structure of CCR5 has
been conserved for at least 35 million years, suggesting that the
retention of this complicated RNA organization may have afforded a
selective advantage.
Evolution of a cis-Regulatory Region of CCR5 in
Humans--
Sequence analysis of the cis-regulatory region
(2761 to 1835) of 60 human CCR5 alleles revealed a total
of 32 variable sites that define 27 unique human haplotypes (Fig.
3). An additional unique CCR5
haplotype was found by sequencing a genomic clone (GenBankTM accession number AF009962; deposited by Dr.
N. L. Michael's group). Sequencing of the homologous region from
the 43 non-human primates (Fig. 1b and Supplemental Fig. 3)
and genotypic data from 40 additional non-human primates, including 23 chimpanzees, enabled us to define the CCR5 haplotype
ancestral to humans. That is we determined the polarity (the
ancestral-descendant relationship) of each nucleotide variant in the
cis-regulatory region of human CCR5. In previous
studies (15, 20), we found seven common polymorphic sites in the region
between CCR5 2761 to 1835 (Fig. 3a). 2733A,
2554G, 2459G, 2135T, 2132C, 2086A, and 1835C represented
the ancestral state for these variable sites in human CCR5
(Figs. 1b and 3c). The nucleotide identity at
each of these positions was invariant among great apes (except Gorilla
which had a CCR5 2132T) and OWM (Fig.
1b). This ancestral CCR5 haplotype was used to
root a phylogenetic network depicting the possible evolutionary
relationships among unique human CCR5 haplotypes (Fig.
3b).
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 3.
CCR5 gene map and phylogenetic
network of CCR5 haplotypes and haplogroups.
a, schema of CCR2 and CCR5 loci on
chromosome 3 (not to scale; hatched marks denote gaps). The
CCR5 exons (open boxes) and two introns are
numbered; the ORF is in exon 3. Downward-pointing
arrows indicate the common polymorphisms found in CCR5
ORF, CCR2 ORF, and in a cis-regulatory region
spanning from CCR5 2761 to 1835. The arrow
above the gene map denotes the CCR5 promoter 1, previously referred as downstream promoter (20). b, a
phylogenetic tree depicting the relationships among the seven
CCR5 human haplogroups (HHA-HHG). A chimpanzee
CCR5 haplotype was used as an outgroup. The sequences of 28 unique CCR5 haplotypes were used to generate the
phylogenetic tree. Each haplotype was assigned a number (1-28), which
is displayed at the tips of the branches. The CCR5
haplotypes that have a common evolutionary history are clustered
together and are boxed. Each cluster of CCR5
haplotypes therefore defined a unique CCR5 haplogroup, and
all haplotypes within a haplogroup share several distinct genetic
features. The CCR5 cis-regulatory polymorphism(s) that
define a haplogroup and the bootstrap support for each branch are
denoted at the branch point. The subset of haplotypes within HHF and
HHG that are in linkage disequilibrium with the CCR2-64I and
CCR5- 32 polymorphisms, respectively, are indicated by a
suffix following their identification number. The CCR2-64I and
CCR5-32 polymorphisms were genotyped as described
previously (15, 16). c, a schematic representation of the
nucleotide sequences of the unique human CCR5 haplotypes
(2761 to 1835). The sequences of human CCR5 haplotypes
were compared with those found in the homologous region of chimpanzee
CCR5. The numbers at the bottom of the
panel correspond to human CCR5 sequence. The sequences found
at the corresponding nucleotide positions in chimpanzee CCR5
are shown and are in red. Dashes represent gaps, and
dots denote identity between human and chimpanzee
CCR5 sequences for the indicated nucleotide position. Each
row is numbered serially (1-28) and represents the sequence
for the 28 haplotypes displayed in the phylogenetic tree.
CCR5 SNPs common to several human haplotypes are
boxed, whereas those that are unique to individual
haplotypes are not boxed and are in black.
Polymorphisms unique to chimpanzee CCR5 are not shown.
CCR5 haplotypes that form a haplogroup are
bracketed. d, classification of CCR5
human haplogroups. Each haplotype within a haplogroup is characterized
by the constellation of invariant polymorphisms indicated but differ
from each other by additional SNPs. This cassette of nucleotide
sequences is designated by a 7-letter SNP signature motif. The
sequences within a SNP signature motif that are common to those found
in the ancestral CCR5 haplotype, designated as HHA, are in
red. HHF*2 and HHG*2 designate the subset of haplotypes
within HHF and HHG that are in linkage disequilibrium with the CCR2-64I
and CCR5-32 polymorphisms, respectively. The 7-letter SNP
signature motifs for HHF*2 and HHG*2 have the prefix, 64I, and the
suffix, 32, respectively. The sequence for the haplotype
representing HHG*2 is derived from a CCR5 genomic DNA clone
(GenBankTM accession number AF009962). The HHB haplotype
was found by genotyping over 2,000 individuals (16) and confirmed by
sequencing. e, a model illustrating the potential evolution
of human CCR5 haplogroups. HHB, HHC, and HHD differ from HHA
by having a 2554T mutation. However, unlike HHC or HHD, HHB is not
mutated at either CCR5 2132 or 2086. HHB may therefore
be ancestral to HHC and HHD. HHG*1 and HHF*1 are likely to be ancestral
to HHG*2 and HHF*2, respectively.
|
|
A phylogenetic network of unique CCR5 haplotypes provided
the evolutionary framework for defining seven biologically distinct clusters of haplotypes that we designated as CCR5 human
haplogroups (HH)-A, -B, -C, -D, -E, -F, and -G (Fig. 3,
b-d). HHA represented the ancestral CCR5
haplogroup. The haplogroups, HHC through HHG, were defined by at least
one SNP, i.e. SNPs 2086G, 2132T, 1835T, and 2733G
distinguish CCR5 HHC, HHD, HHF, and HHG, respectively. HHB
haplotypes had a 2554T mutation but lacked the 2132T and 2086G
SNPs. An HHB haplotype is likely to be ancestral to HHC and HHD (Fig.
3e). SNPs 2459A and 2135C were in complete linkage disequilibrium. Haplotypes with 2459A and 2135C but lacking 2733G
and 1835T defined HHE. The polymorphisms CCR5 2733G, 32, 1835T, and CCR2-64I defined the haplotypes that are
descendants of ancestral haplotypes in HHE (Fig. 3e). The
CCR2-64I and CCR5-32 polymorphisms were found only on
CCR5 haplotypes in haplogroups F (HHF*2) and G (HHG*2),
respectively. To assess the robustness of each of the branches that
define, in part, human CCR5 haplogroups, a bootstrap
analysis was performed. Each branch was observed in 60% or more of the
networks generated (Fig. 3b). Collectively, these findings
demonstrate that SNPs in CCR5 may have arisen by a nested
mutational process and that this locus represents a complex multi-allelic system. A clear understanding of the extent and organization of genetic variation in human CCR5, including
the classification system we developed herein, permitted us to
investigate the role of CCR5 haplotypes in
vertical/horizontal transmission and disease progression in infected
HIV-1 adults and children (16).2
Functional Effects of Variation in a cis-Regulatory Region of Human
CCR5: Haplotype-specific Differences in Transcriptional
Activity--
Because of the influential effect that CCR5
haplotypes had on HIV-1 transmission/disease progression (14-17), we
next initiated studies to determine the potential mechanisms underlying
their effects. To test the hypothesis that polymorphisms in the
cis-regions of human CCR5 haplotypes confer
haplotype-specific differences in transcriptional activity, we cloned
CCR5 2761 to 1814 derived from haplotypes representing
several of the CCR5 haplogroups upstream of the luciferase
ORF. These constructs represent haplotype-specific CCR5
promoter constructs because they included the cis-region of
CCR5 that encompassed the polymorphic sites that demarcated the major haplogroups. We found significant differences among the
luciferase activity of the five haplotype-specific promoter constructs tested, with the HHA-specific promoter construct
demonstrating the least promoter activity (Fig.
4a).
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 4.
Haplotype-specific promoter efficiency and
differential nuclear factor binding associated with CCR5
SNPs. a, mean (±S.E.) luciferase activity of
haplotype-specific promoter constructs made from the region spanning
CCR5 2761 to 1814 is shown. The number of transfections
in K562 cells performed with each haplotype-specific construct is
included within each bar graph. *, p < 0.01 for the
difference between HHA versus HHE, HHF, or HHG;  , for the
difference between HHC versus HHF. Similar results were
obtained in the Jurkat T-cell line (data not shown). b,
EMSAs were performed with K562 (K), THP-1 (T), Jurkat (J), or no (N)
nuclear extracts and radiolabeled oligonucleotides designated as
2733A and 2733G or 1835C and 1835T (Table I). These
oligonucleotides span the region encompassing the SNPs at
CCR5 2733 and 1835, respectively, and differ by the
indicated single nucleotide (A/G or C/T; Table I). Lanes 1-4,
5-8, 17-20, and 21-24 illustrate the interaction of
the radiolabeled oligonucleotides 2733A, 2733G, 1835C, and
1835T with nuclear proteins, respectively. The competition with the
homologous (lanes 9-11 and 25-27), heterologous
(lanes 12-14 and 28-30), or nonspecific
(lanes 15 and 16, and 31 and
32) oligonucleotides is also shown. The triangles
above lanes 9-11, 12-14, 25-27, and
28-30 denote increasing concentrations (4, 40, and 400 ×)
of the unlabeled oligonucleotide indicated. The competition-binding
assays for 2733A/G and 1835C/T were with nuclear extracts from
THP-1 and K562 cells, respectively. For 1835C/T oligonucleotides,
results similar to K562 were obtained using Jurkat nuclear extracts.
NF, nuclear factor; NS, nonspecific; oligo,
oligonucleotide.
|
|
Functional Effects of Variation in a cis-Regulatory Region of Human
CCR5: Altered DNA/Nuclear Factor Binding--
Another potential
mechanism by which polymorphisms in the cis-region of
CCR5 haplotypes could influence CCR5 expression is by
binding to distinct nuclear factors or altering the avidity of binding
to a transcription factor. This might also provide a reason for the
haplotype-specific differences in transcriptional activity. Computer
algorithms suggested that the cis-regions surrounding the
SNPs that delimit the major human haplogroups could potentially bind to
distinct transcription factors (Table
II). We used EMSA and supershift assays
to determine whether polymorphisms at 2733, 2554, 2459, 2135,
2132, 2086, or 1835 bound nuclear proteins with differing avidity
or whether a particular polymorphism bound a novel transcription
factor. Table III summarizes the findings of these studies.
View this table:
[in this window]
[in a new window]
|
Table II
Transcription factors predicted to bind cis-regions that encompass
human CCR5 SNPs
Two computer algorithms (TESS and MatInspector program) were
used to predict the transcription factors that could potentially bind
to the SNPs in a cis-regulatory region (2761 to 1835) of
human CCR5.
|
|
View this table:
[in this window]
[in a new window]
|
Table III
Summary of EMSA findings for cis-regions spanning SNPs at CCR5 2733,
2554, 2459, 2135, 2132, 2086, and 1835
Oligonucleotides that encompassed different SNPs (Table I) in a
cis-regulatory region of CCR5 (2761 to 1835)
were incubated with nuclear extracts derived from different cell lines
or PHA-stimulated human PBMCs (PHA-blasts). Whether a SNP resulted in
the differential (diff.) nuclear factor binding and/or altered affinity
of nuclear factor binding is indicated. Results of supershift assays
and the competition experiments with oligonucleotides containing the
consensus binding sequences for the indicated nuclear factors are also
shown. The single letters within the parentheses indicate the cells
from which the nuclear extracts were prepared and used in EMSAs (J,
Jurkat; K, K562; T, THP-1; and P, PHA blast). + indicates that
competition or supershift was present; indicates that competition or
a supershift was absent. NT indicates not tested. "Identical
binding" indicates that the nuclear factor binding pattern of the two
polymorphic oligomers to nuclear factors in a given cell line or in
PHA-blasts was similar. Note, the binding pattern among the cell lines
or between the cell lines and the PHA-blasts was not necessarily
identical.
|
|
We found that SNPs in human CCR5 cis-regulatory sequences
were associated with three different kinds of alterations in
DNA/nuclear factor binding. First, gain-of-nuclear factor binding:
radiolabeled 2733G but not 2733A oligomers (Table I) bound to a
novel nuclear factor (NF1) in three cell human cell lines (Fig.
4b and Table III). When using 2733G, two additional bands
were found below NF1 in Jurkat cells, and it was difficult to exclude
the possibility that one of these bands corresponded to that observed
in EMSAs with 2733A (Fig. 4b). These two bands were also
found in EMSAs with the 2733G oligomer in K562 or THP1 cells;
however, the intensity of these bands was significantly lower than that
found in Jurkat cells (Fig. 4b, compare lane 8 with lanes 6 and 7). Only NF1 is highlighted in
Fig. 4b because this gain-of-nuclear factor binding was
consistently present in the three cell lines tested.
Second, loss-of-nuclear factor binding: radiolabeled 1835C oligomer
bound specifically to two novel nuclear factors (NF2 and NF3), but in
contrast the radiolabeled 1835T oligomer bound to only one of these
factors (NF2; Fig. 4b and Table III). The radiolabeled
1835C and 1835T oligomers also bound an additional nuclear factor
that appears as a band below NF3, and this band is competed weakly by
unlabeled homologous or heterologous oligomers but not by a nonspecific
oligomer (Fig. 4b). Notably, the altered DNA/NF1-NF3 binding
was observed in human cell lines but not in PHA-blast cells, suggesting
that this effect may be cell type- and/or cell activation-specific.
Finally, a SNP in human CCR5 was associated with altered
avidity of nuclear factor binding. This observation was illustrated by
the altered binding pattern of several members of the NF-B family of
transcription factors to the cis-region that encompasses the
human CCR5 G2554T polymorphism (Fig.
5a). Radiolabeled 2554G and
2554T oligomers each bound to four nuclear factor complexes present
in nuclear proteins derived from human PHA blasts (NF4-NF7; Fig.
5b). In homologous competition assays, unlabeled 2554G and 2554T oligomers competed for the NF4-NF7 bound to the radiolabeled 2554G and 2554T oligomers, respectively, and the efficiency for
competition was NF6 = NF7 > NF4 = NF5 (Fig.
5b, compare intensity of bands in lanes 2 and
3; and data not shown for 2554G competition). By
densitometric analysis, the intensities of the NF5 and NF4 complexes
that bound to the 2554T probe were 198 ± 28.13 and 371 ± 42% greater than the intensities of these complexes that bound to the
2554G probe, respectively (n = 5 separate experiments using nuclear extracts from the same donor). Notably, this altered DNA/NF-B binding was observed in nuclear extracts from human PHA
blasts but not from unstimulated Jurkat T-cells, again highlighting that these altered DNA/nuclear protein interactions to human CCR5 SNPs are likely to be cell type- and/or cell activation
state-specific.
View larger version (41K):
[in this window]
[in a new window]
|
Fig. 5.
Differential binding of
NF-B/Rel family members and two as yet
unidentified nuclear factors to polymorphisms at human CCR5
2554. a, the cis-region spanning
human CCR5 2554 and the homologous region in non-human
primates share sequence similarities with the consensus binding site
for members of the NF-B/Rel family of transcription factors.
CCR5 2554G is found in HH-A, -E, -F, and -G and non-human
primates. 2554T is found in HH-B, -C, and -D. Boxes
highlight differences among the different primates. b, EMSA
and competition experiments were performed with nuclear extracts from
PHA-stimulated human PBMCs with radiolabeled oligomers that contain
either 2554G or 2554T (Table I). Both the radiolabeled 2554G
(lane 1) and 2554T (lane 2) oligomers bound to
four distinct nuclear factor complexes (NF4-NF7); however, the 2554T
oligomer bound to these factors with greater avidity than the 2554G
oligomer (compare intensity of bands in lanes 1 and
2; see "Results" for densitometric analysis). The
unlabeled 2554T oligomer but not an unrelated oligomer competed
efficiently and specifically for the binding of NF4-NF7 bound to the
radiolabeled 2554T oligomer (lane 3, and data not shown).
The unlabeled 2554G oligomer also competed efficiently and
specifically for the binding of radiolabeled 2554G oligomer (data not
shown). Lanes 4 and 5 show the competition with
excess of the unlabeled consensus NF-B-binding oligomer.
c, supershift experiments to determine the identity of the
nuclear proteins bound by the 2554T oligomer in PHA-stimulated human
PBMCs. Radiolabeled 2554T or consensus NF-B oligomers (Table I)
were incubated with extracts from human PHA-blasts. The complexes were
then electrophoretically resolved after addition of the antibodies to
the proteins indicated. No interaction between the nuclear protein
complexes and an isotype control antibody was observed (data not
shown). Note, complexes equivalent to NF6 and NF7 did not bind to the
radiolabeled NF-B oligomer, and the intensity of the NF4 and NF5
bands is greater in lane 10 versus lane 1. d, efficiency of competition of unlabeled 2554T and
2554G oligomers for NF-B family of transcription factors. The
nuclear proteins in PHA blasts that bound a radiolabeled consensus
NF-B-binding oligomer was competed with excess unlabeled NF-B
(lanes 2-5), 2554G (lanes 6-9), or 2554T
(lanes 10-13) oligomers. An equimolar amount of unlabeled
2554T oligomer was much more efficient in competing NF-B/Rel
family of proteins than the 2554G oligomer (compare lanes
6-9 versus 10-13). However, compared with
the unlabeled consensus NF-B oligomer (lanes 2-5), both
the 2554G (lanes 6-9) and 2554T (lanes
10-13) oligomers were significantly less efficient in competing
for NF-B proteins.
|
|
Because the region encompassing human CCR5 G2554T was
predicted to bind to members of the NF-B family of transcription
factors (Table II and Fig. 5a), we examined the ability of
an unlabeled oligomer that contains a canonical NF-B-binding site to
compete for the NF4-NF7 bound to the labeled 2554G or 2554T
oligomers. The unlabeled consensus NF-B-binding oligomer competed
for the NF4-NF7 bound to the labeled 2554G or 2554T oligomers, but
the efficiency of competition was different than that for unlabeled 2554G/2554T and was NF5 > NF4 = NF7 > NF6 (Fig.
5b; compare intensity of bands in lanes 4 and
5 to lanes 2 and 3, and data not shown
for competition to 2554G).
It was not clear from these EMSA competition assays (Fig.
5b) whether each or only a subset of the nuclear complexes
that bound to the labeled 2554G and 2554T oligomers,
i.e. NF4-NF7, contained components of the NF-B/c-Rel
family of nuclear factors. Thus, it was important to compare directly
the nuclear protein binding pattern of radiolabeled 2554T or 2554G
to that of the radiolabeled consensus NF-B-binding oligomer (Table
I; Fig. 5c). The radiolabeled consensus NF-B-binding
oligomer bound to a complex of factors that appeared to include all or
portions of NF4 and NF5 but not NF6 or NF7 (Fig. 5c, compare
lanes 1 and 10). In contrast, both radiolabeled
2554T oligomer (Fig. 5c) and 2554G (data similar to Fig.
5b, lane 1) bound NF4-NF7. Notably, the
consensus NF-B-binding oligomer bound to NF-B family members with
greater avidity than the 2554T oligomer (Fig. 5c, compare intensity of bands in lanes 1 and 10). These
findings suggested that the cis-regions surrounding human
CCR5 G2554T bound to various members of the NF-B
proteins as well as to two nuclear proteins (NF6-NF7) that are not
bound by an oligomer that contains a consensus NF-B-binding site.
To confirm that NF4 and NF5 but not NF6 or NF7 contain members of the
NF-B family of transcription factors, we conducted supershift assays
with antibodies to p65, p50, c-Rel, RelB, and p52. By supershift
assays, only p50 was detected in the NF4 complex bound to radiolabeled
2554T oligomer, whereas parts of NF5 were supershifted by antibodies
to p65, p50, and c-Rel (Fig. 5c, lanes 1-9). Note that the
complex supershifted by c-Rel antibody comigrates with NF7. Overall,
the pattern of the supershifted bands indicated that NF4 and NF5 bound
to the radiolabeled 2554T oligomer contained p65 homodimers, p50/p65
heterodimers, p50 homodimers, and c-Rel/p50 heterodimers (Fig.
5c, lanes 1-9). No supershift was observed with
antibodies to p52 or RelB (Fig. 5c, and data not shown). Notably, antibodies to the NF-B family members did not induce a
supershift of NF6 or NF7 bound to the labeled 2554G (data not shown)
or the 2554T oligomers (Fig. 5c, lanes 1-9). In control experiments, the identical antibodies were directed against the nuclear
proteins that bound to the radiolabeled consensus NF-B-binding oligomer, and a similar supershift pattern was observed (Fig. 5c, lanes 10-18).
Direct evidence that 2554T and 2554G bound to the NF-B family of
transcription factor
TOP
ABSTRACT
INTRODUCTION
