|
|
|
|||||||
|
|||||||||
(Received for publication, October 12, 1994) From the
Aside from its
ability to bind to DNA in a site-specific manner, nothing is known
about the mechanisms by which
DNA fragments and oligonucleotides
used as hybridization probes for library screening were made
radioactive by random priming and end labeling according to standard
procedures(12) . Libraries were hybridized in 3 Genomic and cDNA clones were subcloned
into pBluescript/SK- (Stratagene) and sequenced using Sequenase
(version 2, U. S. Biochemical Corp.). DNA sequences were analyzed using
version 7 of the University of Wisconsin Genetics Computer Group
software package.
Figure 1:
Structure of
human PRDII-BF1, oligonucleotide probes derived from PRDII-BF1, and
partial sequence of the mouse cDNA clone CC-1. A, schematic
representation of PRDII-BF1. Zinc finger sequences (Z) and
acidic domains (A) are indicated by shaded and blackboxes, respectively. The locations of start and
stop codons are shown above the cDNA. Overlapping
oligonucleotides derived from the PRDII-BF1 cDNA sequence that were
used to screen a C
Figure 2:
Genomic Southern blot analysis of
Figure 3:
Restriction maps of
In
addition to the clones shown in Fig. 3A, we obtained
three other overlapping genomic clones that hybridized to the pYTN-8
insert (data not shown). The sequences of these clones are very similar
to the 3`-end of
Figure 4:
Alignment of the
Figure 5:
Northern blot analysis of
Figure 6:
All of the cDNA sequences except
for the testis clones MTC-1, MTC-6, and MTC-7 can be clearly aligned
with the ORF of PRDII-BF1 and show no evidence of alternative splicing
relative to the human cDNA. The 3`-end of MTC-1 and the 5`-ends of
MTC-6 and MTC-7 diverge from the PRDII-BF1 sequence and correspond
exactly to the intron between exons IX and X (Fig. 3) that has
been spliced out of the other overlapping cDNAs. This sequence is
flanked by consensus donor and acceptor splice sites (17) and
contains multiple stop codons that would result in the production of a
truncated protein if present in the mRNA (data not shown). In
addition to CC-1, we obtained two cDNAs (one of which was isolated
seven times) from the brain that encompass the 5`-end of the
A comparison of the cDNA and
genomic clones indicates that the
pCCRXN-1 was
transfected into To obviate the putative poisonous effect of the
Figure 7:
The genomic and cDNA clones that we have isolated indicate
that the
Figure 8:
Comparison of mouse and human cDNA clones
that have related zinc finger sequences. The amino acid (a.a.)
sequences of human clones MBP-2 and PRDII-BF1 and mouse clones
The intron/exon structure of the The large size of It is also interesting that the second (3`) set of zinc
fingers in None
of the cDNA clones that we isolated showed evidence of differential
splicing involving the zinc finger domains. There are multiple, cell
type-specific forms of The testis-specific
transcript described for the rat AT-BP2 gene (4) is also
produced by the The antisense experiments provide
direct functional in vivo evidence that The nucleotide sequence(s) reported in this paper has been submitted
to the GenBank(TM)/EMBL Data Bank with accession number(s)
L36825[GenBank]-L36829[GenBank].
Volume 270,
Number 3,
Issue of January 20, 1995 pp. 1221-1229
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
A-crystallin Binding Protein I
COMPLETE SEQUENCE, GENE STRUCTURE, EXPRESSION, AND FUNCTIONAL
INHIBITION VIA ANTISENSE RNA (*)
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
A-crystallin binding protein I (A-CRYBP1) is a
ubiquitously expressed DNA binding protein that was previously
identified by its ability to interact with a functionally important
sequence in the mouse A-crystallin gene promoter. Here, we have
cloned a single copy gene with 10 exons spanning greater than 70 kb of
genomic DNA that encodes A-CRYBP1. The mouse A-CRYBP1 gene
specifies a 2,688-amino acid protein with 72% amino acid identity to
its human homologue, PRDII-BF1. Both the human and the mouse proteins
contain two sets of consensus C
H zinc fingers
at each end as well a central nonconsensus zinc finger. The
A-CRYBP1 gene produces a 9.5-kb transcript in 11 different tissues
as well as a testis-specific, 7.7-kb transcript. A-CRYBP1 cDNA
clones were isolated from adult mouse brain and testis as well as from
cell lines derived from mouse lens (TN4-1) and muscle
(CC). A single clone isolated from the muscle
C
C library contains an additional exon near
the 5`-end that would prevent production of a functional protein if the
normal translation start site were utilized; however, there is another
potential initiation codon located downstream that is in frame with the
rest of the coding region. In addition, we identified multiple cDNAs
from the testis in which the final intron is still present. Finally, we
used an antisense expression construct derived from an
A-CRYBP1
cDNA clone to provide the first functional evidence that A-CRYBP1
regulates gene expression. When introduced into the TN4-1
mouse lens cell line, the antisense construct significantly inhibited
expression from a heterologous promoter that utilized the A-CRYBP1
binding site as an enhancer.
A-CRYBP1 (
)is a DNA binding protein that was
identified by its ability to recognize the sequence 5`-GGGAAATCCC-3` at
position -66/-57 of the gene encoding mouse
A-crystallin, a protein expressed almost exclusively in the ocular
lens(1, 2) . Mutations in the A-CRYBP1 binding
site caused significant reductions in A-crystallin promoter
function in transient transfection assays, suggesting that factors that
bind to this site are likely to play a critical role in regulating
A-crystallin gene expression(1, 3) . A partial
2.5-kbp A-CRYBP1 cDNA encoding the carboxyl-terminal portion of a
protein that contains two contiguous zinc fingers was isolated from a
mouse lens-derived cell line library(1) . DNA sequence analysis
revealed that A-CRYBP1 is homologous to a rat protein, AT-BP2,
that interacts with the -antitrypsin promoter (4) as well as to a human protein called PRDII-BF1, MBP-1, or
HIV-EP1, that binds to similar sequences in the interferon 
gene
promoter(5) , the MHC H2-K
gene
promoter(6) , and the HIV-1 viral enhancer(7) ,
respectively. The peptide sequence of PRDII-BF1 inferred from
overlapping cDNAs revealed a 2,717-amino acid protein with one pair of
consensus zinc fingers near the amino terminus and a similar pair near
the carboxyl terminus corresponding to those in
A-CRYBP1(5) . The amino- and carboxyl-terminal zinc
fingers of PRDII-BF1 can independently recognize the same DNA binding
sites(5) . The A-CRYBP1 transcript is approximately 9.5 kb
long in newborn mice, suggesting that the mouse gene encodes a protein
of comparable size with human PRDII-BF1(1) .A-CRYBP1 might regulate gene
expression in the mouse. In this report, we describe the cloning and
characterization of the mouse A-CRYBP1 gene. We chose to clone the
gene for several reasons. First, the A-CRYBP1/PRDII-BF1 binding
site is very similar to the recognition sequence of the transcription
factor NF-
B that has been implicated in the regulation of many
human and murine genes(8) , raising the possibility that some
of the promoters that are thought to interact with NF-B might also
interact with A-CRYBP1/PRDII-BF1. Studies of the A-CRYBP1
gene may help to elucidate additional pathways of regulation for an
array of genes. Second, Kantorow et al.(9) showed
that an anti-A-CRYBP1 antibody recognizes proteins of different
sizes on Western blots of mouse tissue culture cells, and Muchardt et al.(10) have identified alternatively spliced
transcripts of PRDII-BF1 by polymerase chain reaction in human cell
lines, suggesting that a thorough knowledge of the intron/exon
structure of the A-CRYBP1 gene is necessary to determine if
alternative RNA splicing is involved in the generation of variant
A-CRYBP1 proteins with potentially different capabilities for
transcriptional regulation. Finally, obtaining a clone of the
A-CRYBP1 locus will allow us to selectively mutate this gene
through the process of homologous recombination in embryonic stem cells
followed by the incorporation of those stem cells into mouse
blastocysts(11) .
Cloning, Screening, and Sequence
Analysis
Genomic clones were isolated from a library of
partially digested DBA/2J mouse genomic DNA in the EMBL-3 cloning
vector (Clontech) and from a library of 129SV mouse DNA in Fix II
(Stratagene). Brain and testis cDNA clones were isolated from libraries
of randomly primed poly(A)
RNA from adult BALB/c mice;
the brain clones were in the vector SWAJ-2, and the testis clones were
in
Max1, respectively (Clontech). A single clone in
Zap II
(Stratagene) was also obtained from a library of randomly primed
poly(A)
RNA isolated from the C
C mouse myoblast cell line. Finally, additional clones were
isolated from the
TN4-1 lens cell line cDNA library
previously described(1) .
SSC,
0.3 M Na
citrate
2HO, pH 7.0, 2%
Ficoll, 2% bovine serum albumin, 2% polyvinylpyrrolidone, 1% SDS, 0.05 M HEPES, and 0.018 mg/ml denatured salmon sperm DNA. After
hybridization, the filters were washed several times in 2 SSC.
Hybridization and washes were both performed at 65 °C when DNA
fragments were used as probes and at 55 °C when oligonucleotides
were used as probes. Oligonucleotides were synthesized on an Applied
Biosystems DNA synthesizer.Genomic Southern Blot Analysis
The insert
from the CC cell line cDNA clone, CC-1, was
hybridized in 0.5 M sodium phosphate, 7% SDS, and 1 mM EDTA at 65 °C to a ``zoo blot'' containing 8 µg
of EcoRI-digested genomic DNA from mice and humans and an
amount of DNA corresponding to the same number of genome equivalents of
other species (BIOS). The blot was initially washed with 1
SSC,
0.1% SDS at 52 °C and exposed to x-ray film. After an initial
exposure, the blot was washed again with 0.2 SSC, 0.1% SDS at
65 °C and re-exposed to film.Northern Blot Analysis
20 µg of total
RNA isolated from tissues of 3-week-old adult mice were subjected to
electrophoresis in 1% agarose gels containing 6.2% formaldehyde and
transferred to nylon membranes (BIOS). The blot was prehybridized in 5
SSC, 5 Denhardt's solution, 0.1% SDS, and 0.1
mg/ml salmon sperm DNA followed by hybridization in 50% formamide, 5
SSPE, 5 Denhardt's solution, 0.1% SDS, and 0.1
mg/ml salmon sperm DNA at 42 °C. The blot was washed three times in
0.5 SSC, 0.2% SDS at room temperature and then once in 0.25
SSC, 0.2% SDS at 50 °C. Hybridization probes were
restriction fragments excised from the CC cDNA
clone and a human glyceraldehyde-3-phosphate dehydrogenase cDNA clone
(Clontech).
Antisense Transfections
An A-CRYBP1
antisense construct, pCCRXV-1, was created by cloning the
CC cDNA insert in reverse orientation into the EcoRI site of pCDNA-1/Neo (Invitrogen).
TN4-1 cells
were transfected by the calcium phosphate precipitation
method(13) . Conditions for cell growth, transfection, and cell
harvesting were previously described(14) . To select for cells
that express the bacterial neomycin resistance gene, the medium was
replaced 48 h after transfection with fresh medium containing 500
µg/ml G418 (Life Technologies, Inc.), and the cells were refed
every other day. Assays for CAT activity and -galactosidase
activity were previously described(15, 16) .
Cloning the NH
Before attempting to clone the
-terminal Coding
Sequences of A-CRYBP1A-CRYBP1 gene, we wanted to obtain a cDNA corresponding to the
5`-end of the gene. Two overlapping oligonucleotides (Fig. 1A) representing the 5`-end of the open reading
frame (ORF) of the human PRDII-BF1 cDNA (5) were synthesized
and used to screen a cDNA library created from randomly primed
poly(A) RNA of C
C mouse
muscle cells. A single 1.2-kbp-long cDNA clone, CC-1, was isolated. DNA
sequence analysis of CC-1 (Fig. 1B) revealed an ORF
following a short 5`-untranslated region. The ORF shows 79% nucleotide
sequence identity and 76% deduced amino acid sequence identity with the
amino-terminal open reading frame of PRDII-BF1(5) . These data
are consistent with the 77% nucleotide and 75% deduced amino acid
sequence identity levels previously identified between the 3`-coding
sequences of human PRDII-BF1 and mouse
A-CRYBP1(1) . The
major difference between CC-1 and PRDII-BF1 is the presence of a 65-bp
insert containing a termination codon in CC-1 (Fig. 1B). There is, however, another ATG located
within the insert (Fig. 1B) that is in the same reading
frame as the remainder of the protein coding sequence in this cDNA (Fig. 1B).
C muscle cell cDNA library
are shown below the PRDII-BF1 cDNA; the initiation ATG is underlined. B, 5`-terminal sequence of the
C
C cell cDNA clone CC-1. Sequences in CC-1
that show similarity to the open reading frame of PRDII-BF1 are
conceptually translated below the nucleotide sequence. Amino
acids that are identical to PRDII-BF1 are underlined;
nucleotides that are inserted relative to the PRDII-BF1 cDNA sequence
are shaded. A conceptual translation of the inserted sequence
is shown in the two different reading frames (ORF1 and ORF2) that align
with the PRDII-BF1 open reading frame.
Genomic Southern Analysis of
To confirm that A-CRYBP1A-CRYBP1 is encoded by a
single copy gene and to test the possibility that A-CRYBP1
homologues might also be present in other species, a Southern blot of
genome equivalents of EcoRI-digested DNA from taxonomically
diverse species was hybridized with the CC-1 cDNA clone. CC-1 does not
contain any zinc finger sequences that might cross-hybridize to
unrelated zinc finger-containing genes. A low stringency wash (Fig. 2A) revealed hybridization bands of 3.7, 4.3,
9.1, and 19 kbp in mouse. Intense bands are also present in human and
frog, and fainter bands can be seen in mussel, fruit fly, and nematode
DNA. The apparent background hybridization in chicken DNA makes
individual bands difficult to discern. However, our previous isolation
of a cDNA from the chicken lens that is clearly homologous to
A-CRYBP1 indicates that at least one gene is related to
A-CRYBP1 in this species(3) . When the blot was washed
again under highly stringent conditions (Fig. 2B), only
bands in the mouse and human lanes remained visible, suggesting that
the A-CRYBP1 gene has sustained appreciable sequence changes
during the course of evolution.
A-CRYBP1-related sequences in various eukaryotic species. A blot
containing genome equivalents of EcoRI-digested DNA from the
indicated species was hybridized with the insert from the A-CRYBP1
cDNA clone CC-1. The blot was exposed after an initial low stringency
wash (A) and then re-exposed after a high stringency wash (B).
Isolation of Genomic Clones
Approximately
one million plaques from a DBA/2J mouse genomic DNA library were
screened with both CC-1, which encodes the amino terminus of
A-CRYBP1, and with the insert from pYTN-8, which encodes the
carboxyl terminus of A-CRYBP1(1) . Fig. 3A shows restriction maps of the clones that were isolated and the
regions of each clone that were sequenced. These sequences have been
placed in GenBank. We were unable to isolate a DBA/2J genomic clone
corresponding to nucleotides 6706-6808 in PRDII-BF1, the human
homologue of the mouse A-CRYBP1(5) . However, we were able
to obtain a clone (MG17-4) from a 129SV mouse genomic library
encompassing this sequence (Fig. 3A). The inserted
sequence that was observed in CC-1 relative to PRDII-BF1 (see Fig. 1B) is located in the genomic clone MG62-1 (Fig. 3A). The 65-bp insert is flanked by consensus
splice donor and acceptor sequences (17) (data not shown). The
intron/exon structure of the A-CRYBP1 based on the sequences of
genomic and cDNA (see below) clones is shown in Fig. 3B. Since the GC-rich nature of the 5`-end of the
gene prevented the establishment of the transcription start site, the
numbering of the exons must remain tentative. The sizes of the
restriction fragments in EcoRI-digested mouse genomic DNA that
are expected to hybridize to CC-1 based on restriction mapping of
genomic clones MG10-1, MG62-1, and MG4-2 (Fig. 3A) are indicated in Fig. 3B. The
4.3-, 3.7-, and >9.9-kbp fragments are consistent with the bands
seen in Fig. 2. A 2.8-kbp band is not seen in Fig. 2, but
there is a hybridizing fragment of 9.1 kbp. These data are consistent
with A-CRYBP1 being encoded by a single copy gene with a
polymorphism in one of its EcoRI restriction sites.
A-CRYBP1 genomic
clones and intron/exon structure of the A-CRYBP1 gene. Panel
A, restriction maps of lambda clones isolated from a DBA/2J mouse
genomic library (MG10-1, MG18-1, MG62-1, MG
4-2, and MG49-1) and from a 129SV mouse genomic library
(MG17-4). Shadedbars indicate the regions of
each clone that were sequenced. solidbars labeled
with Romannumeralsabove the maps represent
exons; the exon labeled with an asterisk in MG62-1 (exon
III) corresponds to the inserted sequence in the cDNA clone CC-1 (see Fig. 1B). Regions of overlap between clones are
indicated by verticaldottedlines. The
locations of the initiation and termination codons based on a
comparison with PRDII-BF1 are indicted in MG10-1/MG18-1 and
MG49-1, respectively. B indicates BamHI site. PanelB, intron/exon structure of the A-CRYBP1
gene. Filledboxes correspond to protein coding
exons; the openbox represents a 5`-untranslated
exon; stippledboxes with Z and Aunderneath denote the locations of consensus zinc fingers
and acidic domains, respectively. The cDNA clones CC-1 and
pYTN8-1 that were used to isolate the genomic clones are shown
aligned with the corresponding portions of the A-CRYBP1 gene. The
sizes of restriction fragments in EcoRI-digested mouse genomic
DNA that should hybridize to each exon in CC-1 are shown below.
A-CRYBP1 but contain a number of missense and
nonsense mutations that would occlude the production of a functional
protein. Further analysis of these clones revealed the presence of a
retroviral long terminal repeat 1,082 bp upstream of a termination
codon homologous to that in pYTN-8(1) . Therefore, it appears
that these three clones represent a partial or complete A-CRYBP1
pseudogene.Sequence Comparisons between
The deduced amino acid sequence of A-CRYBP1 and
PRDII-BF1A-CRYBP1
was compared with that of PRDII-BF1 (Fig. 4). The extra peptide
at the NH-terminal region of A-CRYBP1 was not included
in this comparison (see Fig. 1B). The overall identity
between the two protein sequences is 72%. Like PRDII-BF1(5) ,
A-CRYBP1 contains two sets of consensus zinc fingers at each end
of the protein as well as one central nonconsensus zinc finger. The
zinc finger sequences are very highly conserved between mouse and
human, the only difference being a single serine-threonine substitution
in one of the carboxyl-terminal fingers. The amino- and
carboxyl-terminal zinc fingers are both followed by a conserved stretch
of acidic amino acids that probably represents a domain of interaction
with other proteins (18) . A potential nuclear localization
signal in PRDII-BF1 that was identified by Van't Veer et al.(19) is also present in A-CRYBP1.
A-CRYBP1 and
PRDII-BF1 amino acid sequences. The protein sequences of A-CRYBP1
and PRDII-BF1, derived from conceptual translations of DNA, were
aligned using the algorithm of Needleman and Wunsch (45) as
adapted by version 7 of the University of Wisconsin Genetics Computer
Group sequence analysis software. Verticallines indicate identical amino acids between the mouse (M) and
human (H) sequences. Amino acids corresponding to the inserted
sequence in Fig. 1B (exon III) were deleted from the
A-CRYBP1 sequence prior to alignment. Boxes indicate the
zinc finger domains identified by Fan and Maniatis(5) . Arrows indicate the locations of introns in the mouse
A-CRYBP1 gene, and the exons from which the A-CRYBP1 peptide
sequence is derived are indicated by Romannumeralsabove the mouse sequence. Boldunderlines define stretches of acidic amino acids flanking the zinc finger
regions; a putative nuclear localization signal is also underscored with a thinnerline.
Northern Analysis of
Previously, we showed that the 3`-A-CRYBP1
TranscriptsA-CRYBP1
cDNA clone, pYTN-8, hybridized to a single 9.5-kb transcript in
RNA from newborn mouse lens, brain, spleen, thymus, and liver as well
as RNA from
TN4-1 lens-derived tissue culture
cells(1) . We have confirmed this result with the CC-1 5`-cDNA
probe and showed also the presence of the 9.5-kb A-CRYBP1
transcript in lung, kidney, skeletal muscle, and intestine (data not
shown). In addition, there is a prominent 7.7-kb A-CRYBP1
transcript that is found only in testis and not ovary (Fig. 5A). A control hybridization with a
glyceraldehyde-3-phosphate dehydrogenase cDNA probe indicated that
there was no generalized RNA degradation in the testis preparation (Fig. 5B). The rat homologue of A-CRYBP1, AT-BP2,
also produces this testis-specific transcript(4) , which can be
detected using a cDNA probe corresponding exactly to the homologous
sequence of pYTN-8(1, 4) . The fact that this RNA is
not observed in ovaries (Fig. 5) indicates that the message is
testis-specific rather than germ cell-specific. We also hybridized a
blot containing total RNA from four different tissues of 2-day-old,
2-week-old, and adult rats with CC-1. These experiments revealed a
single 9.5-kb transcript that is present throughout postnatal
development (data not shown).
A-CRYBP1
message in adult mouse ovary and testis. A, a blot containing
20 µg of total RNA isolated from adult mouse ovaries (O)
and testes (Te) was hybridized with the insert from the
CC muscle cell line cDNA clone, CC-1. B, the blot was reprobed with the human
glyceraldehyde-3-phosphate dehydrogenase
cDNA.
Analysis of
To
confirm that the coding sequences identified in the A-CRYBP1 cDNA ClonesA-CRYBP1
genomic clones are actually transcribed into mRNA, we isolated
additional A-CRYBP1 cDNA clones. Restriction fragments
encompassing the entire A-CRYBP1 putative ORF were used to screen
cDNA libraries derived from the brain and testes of adult mice and from
the TN4-1 mouse lens-derived cell line. The cDNA clones that
were isolated (Fig. 6) were sequenced either entirely or from
each end. With the exception of two silent site substitutions, the
protein coding sequences in the cDNAs are all identical to the genomic
sequences, consistent with A-CRYBP1 being encoded by a single copy
gene ( (1) and see below).
A-CRYBP1 cDNA clones. cDNA libraries
made from RNA isolated from TN4-1 mouse lens-derived cells,
adult mouse brain, and adult mouse testes were screened with
restriction fragments encompassing the protein coding regions of
genomic clones MG4-2 and MG49-1 (Fig. 3A).
The top of the figure depicts the locations of exons above a
consensus A-CRYBP1 cDNA. Note that exon III is present only in
CC-1. Z/A indicates the location of consensus zinc
finger/acidic domains; Z denotes the central nonconsensus zinc
finger. cDNA clones isolated from the brain (MBC1-13), testis
(MTC1-8), and TN4-1 (p2A-5 and p4B-1) libraries are
represented by diagonalstripes, verticalstripes, and solidblack, respectively,
and are aligned with the consensus cDNA. Clone CC-1, isolated from
CC mouse muscle cells, and the initial
A-CRYBP1 cDNA clone, pYTN-8, are also shown. Sequences in cDNA
clones that are not present in the consensus cDNA are indicated by openbars.
A-CRYBP1 gene. Both of the brain cDNAs lack exon III, which is
found only in CC-1 (Fig. 1B and Fig. 3). Of the
three 5`-cDNAs, CC-1 contains the shortest stretch of 5`-untranslated
sequence, and this sequence is identical to that immediately preceding
the initiation codon in the genomic clones MG10-1 and
MG18-1 (Fig. 3). Comparison of the 5`-untranslated region
of MBC-1 with the MG10-1 genomic sequence reveals the presence of
a 3.3-kb intron located 104 bp upstream of the initiation ATG (Fig. 3). The sequence of the 5`-end of MBC-3,5,6,7,9,10,11
diverges from the genomic sequence downstream from this intron. The
5`-terminal sequence of this clone does not appear to be present in our
genomic clones based on both DNA hybridization experiments and direct
sequence comparisons. Furthermore, the point at which the genomic
sequence diverges from that of MBC-3,5,6,7,9,10,11 does not correspond
to a splice acceptor site (data not shown)(17) . Therefore, we
believe that the unidentified 5`-terminal sequence of
MBC-3,5,6,7,9,10,11 is a cloning artifact. The 5`-end of MBC-1 lies
within a region of genomic DNA in MG10-1 that is extremely rich
in guanosines and cytosines (data not shown). Attempts to verify the
5`-end of the A-CRYBP1 message by primer extension have been
unsuccessful, probably as a result of RNA secondary structures that are
stabilized by these GC residues.A-CRYBP1 gene comprises 10 exons (Fig. 2B). The cDNA clones show evidence for
alternative RNA splicing involving the third exon, which is present
only in CC-1, as well as differential splicing of the final intron,
which is present in three of the testis cDNA clones but not in any of
the other cDNAs that span this region. The introns separating the 10
A-CRYBP1 exons are all bounded by consensus splice donor and
acceptor sites (17) except for the intron that immediately
precedes exon VII with the carboxyl-terminal zinc fingers (Fig. 3). In this intron, the splice donor sequence is
``GC'' rather than the consensus ``GT''.The Effect of an
With knowledge of the A-CRYBP1 Antisense Expression
Construct on a Heterologous Promoter Containing A-CRYBP1 Binding
SitesA-CRYBP1 gene structure in
hand, we are ready to conduct functional studies on the role of
A-CRYBP1 in regulating gene expression and embryonic development.
As an initial experiment, we attempted to interfere with A-CRYBP1
production in tissue culture cells by transfecting them with an
A-CRYBP1 antisense construct. It has been shown for a number of
eukaryotic genes that the expression of antisense RNA complementary to
a particular mRNA will reduce the amount of the protein encoded by that
message (see (20) for a review). To make an A-CRYBP1
antisense construct, we ligated the CC-1 cDNA clone in reverse
orientation downstream of the cytomegalovirus promoter in the vector
pCDNA-1/NEO (Invitrogen). The resulting construct, pCCRXN-1, also
contains the bacterial neomycin resistance gene linked to the Rous
sarcoma virus promoter and should confer neomycin (G418) resistance to
cells transfected with this construct(21) .TN4-1 mouse lens-derived cells, and stably
transformed cell lines were established by selection with G418. To
assay for reduced A-CRYBP1 activity, the cells were transiently
transfected with plasmids containing either one (PCS-15) or four
(pCS-31) copies of the A-CRYBP1 binding site placed upstream of
the herpes simplex virus thymidine kinase promoter which, in turn, was
fused to the bacterial CAT reporter gene. It was previously shown that
these reporter constructs produced significant levels of CAT activity
in transfected TN4-1 cells and that the level of CAT
activity was directly proportional to the number of A-CRYBP1
binding sites present in the construct(14) . In the present
experiments, we found that the pCCRXN-1 construct was not maintained
intact in stably transformed lines of TN4-1 cells. A likely
explanation for this result is that the antisense construct was
poisonous to the cells, and only cells with deletions in the
A-CRYBP1 portion of the plasmid could survive the G418 selection
process.A-CRYBP1 antisense construct, we measured CAT activity in
TN4-1 cells that were transiently cotransfected with
pCCRXN-1 and either of the two reporter constructs pCS-15 or pCS-31. Fig. 7shows CAT activity levels normalized to
-galactosidase activity from a cotransfected bacterial lacZ-expressing plasmid. In cells cotransfected with pTKCAT, a
reporter plasmid lacking the A-CRYBP1 binding site, there was
almost no CAT activity irrespective of the presence of the antisense
construct (pCCRXN-1). Cells transfected with the pCS-15, which has one
A-CRYBP1 binding site, had low levels of CAT activity when
cotransfected with either the antisense (pCCRXN-1) or control
(pCDNA-1/Neo) plasmid, although there is a slight but statistically
significant reduction in CAT levels with the antisense plasmid. Cells
cotransfected with pCS-31, which contains four A-CRYBP1 sites, and
the control plasmid (pCDNA-1/Neo) show almost 10 times as much CAT
activity as the cells transfected with pCS-15. When cells were
cotransfected with pCS-31 and the antisense-containing plasmid
(pCCRXN-1), CAT activity was reduced by approximately 50%. Therefore,
the A-CRYBP1 antisense construct appears to have a significant
inhibitory effect on the activity of a promoter in which the
A-CRYBP1 binding site is the only enhancer element.
A-CRYBP1 antisense experiments. CAT
activity levels normalized to -galactosidase activity in
cotransfected TN4-1 cells are shown for each set of
transfections. TN4-1 cells were transfected with 1 µg of
the lacZ-expressing plasmid pCMV (Clontech), 12 µg of
either the pCS-15 or pCS-31 reporter constructs, and 20 µg of
either pCDNA-1 or pCCRXN-1 (denoted as pCCR for simplicity).
CAT activity is expressed in counts/min/-galactosidase activity.
-galactosidase activity was measured in absorbance units at A. Each verticalbar represents an average of six separate transfections; the standard
errors are indicated on each bar.
A-CRYBP1 protein is encoded by a gene with 10 exons
spanning greater than 70 kbp of DNA. Several other cDNAs have been
isolated from rat, human, and mouse that contain zinc fingers very
similar to those of A-CRYBP1 and PRDII-BF1. The human cDNA, MBP-2,
encodes a 2,500-amino acid protein with two sets of zinc fingers spaced
comparably with those of PRDII-BF1(19) . However, MBP-2 lacks a
central, nonconsensus zinc finger, and although there are patches of
similarity between MBP-2 and PRDII-BF1 (Fig. 8), the overall
sequence identity between these two proteins is only 33%. A partial
cDNA clone, Rc-1, isolated from mouse cells encodes the
carboxyl-terminal portion of a protein with one set of
A-CRYBP1-like zinc fingers adjacent to a conserved acidic domain,
but the only similarity between the remainder of the encoded proteins
of the two mouse cDNAs is confined to a stretch of 37 amino acids (22) (Fig. 8). Partial cDNAs isolated from rat
(AT-BP1/AGIE-BP1) and human (KBP-1) appear to be homologous to MBP-2
and Rc-1, respectively(4, 23, 24) . Fig. 8also shows an alignment of the zinc finger sequences of
A-CRYBP1, PRDII-BF1, MBP-2, and Rc-1. The similarity between the
zinc finger sequences in Fig. 8is reflected in the ability of
the corresponding proteins to recognize the same DNA binding sites in
gel mobility shift
assays(1, 4, 5, 10, 21, 22, 23, 24) .
However, each of these proteins preferentially binds to specific DNA
sequences that differ for each protein. Possibly, several different
mammalian proteins have evolved from an ancestral A-CRYBP1-like
protein, and during the course of evolution, these factors acquired
subtle variations in DNA binding specificity and, perhaps, changes in
their ability to interact with other transcription factors. Our genomic
hybridization data suggest that homologues of these genes might be
present in non-mammalian species. Structural and functional studies of
A-CRYBP1-like proteins from distantly related species may
elucidate the utility of having multiple factors that recognize related
DNA binding sites. Such studies may also reveal evidence for
co-evolution between these factors and the gene promoters with which
they interact.
A-CRYBP1 and Rc-1 were aligned using the algorithm of Needleman
and Wunsch (45) as adapted by version 7 of the University of
Wisconsin Genetics Computer Group. The regions of highest similarity in
the resulting alignments are indicated by the lines that
connect the paired cDNAs. The numbers represent the percentage
of amino acid sequence identity in each region. ZF/A indicates
zinc finger/acidic regions; ZF indicates a nonconsensus zinc
finger that is present in A-CRYBP1 and PRDII-BF1 but not in the
other two cDNAs. The zinc finger sequences from each clone are aligned
at the bottom of the figure.
A-CRYBP1 gene
has several interesting features. The first is the presence of exon III
that is found only in the one cDNA clone, which was isolated from the
CC muscle cell line library. The inclusion of
exon III sequences in mRNAs would prevent the formation of a functional
protein if the normal translation start site were utilized. There is
another ATG, located further downstream, that is in the same ORF as the
rest of the
A-CRYBP1 protein coding sequence. However, other
studies have shown that reinitiation of translation at downstream ATGs
usually occurs with very low efficiency(25) . Perhaps the
inclusion of exon III sequences by alternative splicing serves to
modulate the amount of A-CRYBP1 protein without changing the level
of mRNA. There are precedents for employing alternative exons as a
means of down-regulating protein expression. For example, the Drosophila sex-lethal gene produces a male-specific transcript
containing an in-frame stop codon that results in a non-functional
protein only in male flies(26) , and a majority of the
transcripts produced by the cH-ras gene contain an extra exon
that leads to the production of a truncated, non-functional
protein(27) . The CC cell culture from
which the cDNA library was constructed consisted largely of
differentiating myoblasts. (
)It has been shown that
CC cells undergo extensive changes in gene
expression during differentiation into myotubes(28) , and it is
possible that a reduction in
A-CRYBP1 levels may be necessary for
affecting some of these changes.A-CRYBP1
exon V is another striking feature of the A-CRYBP1 gene. The
average vertebrate exon size is 137 nucleotides, and very few are
greater than 600 nucleotides in length(29) . However, it has
been recently shown that increasing the size of an exon in vivo does not necessarily interfere with RNA splicing(30) . In
light of the belief that many genes have evolved by a process of exon
duplication and shuffling(31) , exon V may represent the
structure of an ancestral A-CRYBP1 gene. The recruitment of the
other smaller exons during the course of evolution might have occurred
to alter or extend the functional role of the A-CRYBP1 protein in
regulating gene expression as postulated above for differentiating
myoblasts.A-CRYBP1 is encoded by a separate exon, and the
nucleotides corresponding to the acidic domain adjacent to these
fingers are located on yet another exon. This structure imbues the
A-CRYBP1 gene with the potential flexibility to produce
functionally distinct protein isoforms via alternative splicing of
these two domains. In addition, the intron that immediately precedes
the carboxyl-terminal zinc finger-containing exon (exon VII) has a
nonconsensus 5`-splice site in which ``GC'' replaces the
almost universal ``GT'' at this position(17) . The
mouse A-crystallin gene also has an intron with the same
noncanonical 5`-splice site(32) . In the case of
A-crystallin, the intron is spliced inefficiently, resulting in
two forms of the A-crystallin protein that differ by the presence
or absence of the exon that precedes the aberrant splice site.A-CRYBP1 as judged by Western blots where
an antibody specific for the carboxyl-terminal half of A-CRYBP1
recognizes 50-, 90-, and 200-kDa proteins, which may involve
alternative RNA splicing in those cell lines(9) . Moreover,
Muchardt et al.(10) detected in human cells, using a
sensitive polymerase chain reaction technique, rare PRDII-BF1
transcripts that contained only the amino- or the carboxyl-terminal
zinc fingers. One of these transcripts results from a deletion in which
a nucleotide corresponding to bp 13 of A-CRYBP1 exon VI is joined
to the sequence beginning with nucleotide 204 in A-CRYBP1 exon IX.
The other transcript contains a deletion corresponding exactly to exon
V of A-CRYBP1. These results also suggest that there is at least
some conservation of exon structure between the mouse gene and its
human homologue that has not yet been isolated.A-CRYBP1 gene in the mouse. A number of other
mammalian genes produce different sized mRNAs that are detected only in
the testis. These messages may arise from testis-specific sites of
transcription
initiation(33, 34, 35, 36) ,
polyadenylation(37, 38) , or alternative
splicing(39, 40, 41) . In several cases, it
has been demonstrated that the testis-specific transcript is produced
specifically by spermatogenic
cells(37, 40, 42) . We isolated a number of
A-CRYBP1 cDNA clones from a mouse testis library. Three of these
clones were derived from mRNAs in which the last intron had not been
removed. Since the testis-specific transcript observed on Northern
blots is smaller than the ubiquitous A-CRYBP1 transcript, there
must be another difference between the two mRNAs that is not reflected
in the cDNAs that we isolated.A-CRYBP1 is
involved with regulating gene expression via the promoter sequence
GGGAAATCCC. Although the antisense construct produces a significant
decrease in CAT activity, there is still a substantial amount of CAT
activity remaining in the cotransfected cells. This residual CAT
activity probably results, in part, from incomplete inhibition of
A-CRYBP1 by the antisense message, but it might also reflect the
ability of other proteins to regulate transcription through this
10-base pair sequence. Besides the related zinc finger proteins
mentioned above, NF-B and members of the c-rel family of
transcription factors bind to a consensus sequence (GGGRNNYYCC) that
corresponds to the site in our reporter constructs(8) .
Furthermore, the promoter sequence, PRDII, to which PRDII-BF1 binds in
the human interferon- gene, is recognized by at least five
different factors including NF-B, PRDII-BF1, and HMG I(Y) (43, 44) . However, our experiments provide the first
functional evidence that A-CRYBP1 is at least one of the proteins
that is involved in regulating gene expression using this sequence from
the promoter of the mouse A-crystallin gene.
)A-CRYBP1,
A-crystallin binding protein I; kb, kilobase(s); bp, base pair(s);
kbp, kilobase pair(s); ORF, open reading frame; CAT, chloramphenicol
acetyltransferase.)
We thank Dr. Rashmi Gopal-Srivastava for providing the
CC muscle cell cDNA library.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  X. Yang, J. Li, H. Qin, H. Yang, J. Li, P. Zhou, Y. Liang, and H. Han Mint Represses Transactivation of the Type II Collagen Gene Enhancer through Interaction with {alpha}A-crystallin-binding Protein 1 J. Biol. Chem., May 13, 2005; 280(19): 18710 - 18716. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Zhao, Y. Yang, C. M. Rizo, P. A. Overbeek, and M. L. Robinson Insertion of a Pax6 Consensus Binding Site into the {alpha}A-Crystallin Promoter Acts as a Lens Epithelial Cell Enhancer in Transgenic Mice Invest. Ophthalmol. Vis. Sci., June 1, 2004; 45(6): 1930 - 1939. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Tanaka, Y. Matsumoto, F. Nakatani, Y. Iwamoto, and Y. Yamada A Zinc Finger Transcription Factor, alpha A-Crystallin Binding Protein 1, Is a Negative Regulator of the Chondrocyte-Specific Enhancer of the alpha 1(II) Collagen Gene Mol. Cell. Biol., June 15, 2000; 20(12): 4428 - 4435. [Abstract] [Full Text]
|
|
|
|
|
|
I. Hjelmsoe, C. E. Allen, M. A. Cohn, E. M. Tulchinsky, and L.-C. Wu The kappa B and V(D)J Recombination Signal Sequence Binding Protein KRC Regulates Transcription of the Mouse Metastasis-associated Gene S100A4/mts1 J. Biol. Chem., January 14, 2000; 275(2): 913 - 920. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Ilagan, A. Cvekl, M. Kantorow, J. Piatigorsky, and C. M. Sax Regulation of alpha A-crystallin Gene Expression. LENS SPECIFICITY ACHIEVED THROUGH THE DIFFERENTIAL PLACEMENT OF SIMILAR TRANSCRIPTIONAL CONTROL ELEMENTS IN MOUSE AND CHICKEN J. Biol. Chem., July 9, 1999; 274(28): 19973 - 19978. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |