| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Department of Cell Signaling, DNAX Research Institute of Molecular and Cellular Biology, Palo Alto, CA 94304;
Department of Microbiology, Keio University School of Medicine, Shinjuku-ku, Tokyo, Japan; and
Department of Molecular and Developmental Biology, Institute of Medical Science, University of Tokyo, Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, CD40 ligand, and Fas ligand (2, 3, 4). NFAT consists of two
components: a pre-existing cytoplasmic component, which translocates
into the nucleus upon calcium mobilization, and an inducible nuclear
component composed of members of the activating protein-1
(AP-1)4 family of
transcription factors (5, 6). Translocation of the cytoplasmic
component is dependent on the activity of the
calcium/calmodulin-dependent phosphatase calcineurin and is sensitive
to the actions of the immunosuppressants cyclosporin A and FK506 (2, 7, 8). The search for genes encoding cytoplasmic components of NFAT, collectively termed NFAT family proteins, led to the isolation of four distinct cDNAs that shared sequence similarities. They encoded proteins that were named NFAT1(NFATp), NFATc, NFATx/NFAT4/NFATc3, and NFAT3 (9, 10, 11, 12, 13). All four NFAT proteins can activate transcription driven by the NFAT site from the IL-2 promoter (9, 10, 11, 12). However, the individual roles of NFAT family proteins in regulating gene expression are still unclear. NFAT family proteins are characterized by a highly homologous domain, which is weakly related to the DNA binding domain of Rel family proteins (9, 10, 14, 15). The existence of this characteristic Rel similarity domain allowed us to divide NFAT family proteins into three distinct regions: an amino-terminal region, a carboxyl-terminal region, and a middle region containing the Rel similarity domain. The main function of the Rel similarity domain is DNA binding and protein-protein interactions with the AP-1 family of transcription factors. Indeed, the Rel similarity domains from all four NFAT family proteins are capable of binding cooperatively with AP-1 family proteins at the IL-2 promoter NFAT site (9, 12, 14, 16). On the other hand, the amino-terminal region of NFAT1(NFATp) and NFATx/NFAT4 is important and sufficient for binding to calcineurin and seems to regulate nuclear translocation (16, 17, 18). The amino-terminal region shows significant, albeit weak, sequence similarity among NFAT family proteins and is characterized by the presence of three perfect repeats for NFAT1(NFATp), NFATc, and NFATx and one for NFAT3, containing the sequence SPXXSPXXSPXXXXX(D,E)(D,E), referred to as SP boxes (10). Finally, in contrast to the middle and the amino-terminal regions, little is known about the carboxyl-terminal region, which shows no significant sequence similarities. Moreover, the carboxyl-terminal region is absent in the NFATc protein (12).
Recently, we cloned a cDNA encoding a member of the NFAT family proteins, referred to as NFATx, and showed that NFATx specifically bound the NFAT binding sequence, that it was capable of activating the IL-2 promoter, and that this molecule was expressed predominantly in the thymus (10). In the present work we identified a 15-amino acid transactivation sequence in the carboxyl-terminal end of NFATx. In addition, we isolated cDNAs encoding three isoforms of NFATx, i.e., NFATx2, NFATx3, and NFATx4. NFATx was renamed NFATx1. Interestingly, these isoforms have altered carboxyl-terminal ends and lost NFATx1 transactivation sequence, which explains the strong transactivation activity of NFATx1 compared with those of the other isoforms.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cDNAs encoding NFATx2 and NFATx3 were isolated by screening
Jurkat cDNA libraries in 
Ziplox (Life Technologies, Gaithersburg,
MD) as previously described (10). NFATx4 was isolated by screening a
cDNA library obtained from T Ag-transformed human diploid fibroblasts
(19), using as a probe a DNA fragment spanning the whole Rel similarity
domain of human NFAT1(NFATp). Five positive cDNA clones, which had a
fragment of approximately 3 kb, were identified as encoding the new
member, designated NFATx4.
Plasmid constructions
NFATx2 and NFATx3 isoforms and NFATx1 deletion expression
plasmids were constructed using pME-NFATx as a starting plasmid.
pME-NFATx is expression plasmid for NFATx1 under the control of the
SR promoter in vector pME18S and has been described previously (10);
it was renamed pME-NFATx1 in this study. The
PvuII-AflIII fragment (positions 2044–3152
according to the sequence of Masuda et al. (10)) of pME-
NFATx1 was replaced by the corresponding fragments containing
NFATx2- or NFATx3-specific sequences for pME-NFATx2 or pME-NFATx3. For
the construction of pME-NFATx1
CAA, pME-NFATx1CPS, and
pME-NFATx1C, a part of the carboxyl-terminal coding region of
pME-NFATx1 was deleted by AflIII (position 3152),
PstI (position 3110), and BstEII (position 2298)
digestion, respectively. The plasmid pME-NFATx1CBA was made by
deleting most of the carboxyl-terminal coding region of NFATx1
(positions 2298–3152) using BstEII and AflIII.
For construction of pME-NFATx1CBP, the 156-bp PstI
fragment (positions 3110–3266) from the very end of the
carboxyl-terminal coding region of the pME-NFATx1 was ligated into the
BstEII site on the pME-NFATx1C. pME-NFATx4 was formed by
ligation of the EcoRI fragment, which contained full-length
NFATx4, into the EcoRI site of pME18S. Plasmid pNFAT72Luc
was constructed by inserting three copies of the murine IL-2 promoter
NFAT site (-290 to -261) into the BglII site of the
pmoIL-2-72Luc (20). All plasmids for fusions between the GAL4 DNA
binding domain and NFATx1 were constructed using pM in the TransAct
Assay Kit (Clontech, Palo Alto, CA) as a starting plasmid. For
construction of pGAL4-X1BB, pGAL4-X2BB, and pGAL4-X1PP, the
BstEII fragment from pME-NFATx1 or -NFATx2 and the
PstI fragment (positions 3110–3226) from pME-NFATx1 were
ligated into the SmaI site of pM, respectively. pGAL4-X1PA
was made by AflIII partial digestion of pGAL4-X1PP. For the
construction of pG5Luc, which is a reporter plasmid with five GAL4 DNA
binding sites, the PvuII fragment (positions 308–736) from
pG5CAT in the TransAct assay kit was ligated into the SmaI
site of pUC00Luc (21). To form pGAL4-X1WT, synthesized primer (21-mer
5'-CCGGAATTCGACATCACTTTA-3') and template (54-mer
5'-TTATCTAGACATGTCTCTCCCAATTATCTCGTTCACATCATCTAAAGTGATGTC-3')
were annealed and filled in with Klenow fragment. After digestion with
EcoRI and XbaI, this fragment was ligated into
the EcoRI-XbaI site of pM plasmid. For alanine
substitution of 15-amino acid trans-activation sequence
(pGAL4-X1 M1
pGAL4-X1 M10), mutant 54-mer templates were used, which
contain 5'-GCT-3' nucleotides instead of wild-type nucleotides encoding
one of 15 amino acid residues.
DNA transfections
Transfection of Jurkat cells or 293 cells was performed by the
DEAE-dextran method (22) or the calcium phosphate precipitation method
(23), respectively. Transfection efficiency was monitored by
cotransfection of pCMV/SEAP (Tropix, Bedford, MA) for Jurkat cells and
pBC12/RSV/SEAP (Tropix) for 293 cells. The plasmids pCMV/SEAP
and pBC12/RSV/SEAP contain the CMV promoter/enhancer and the RSV long
terminal repeat driving secreted alkaline phosphatase expression,
respectively. For transfection into Jurkat cells, 1 x
106 cells were transfected with 3 µg of reporter plasmid
pNFAT72Luc, 3 µg of internal control plasmid pCMV/SEAP, and 6 µg of
NFATx1 deletion expression plasmid (pME- NFATx1CAA,
pME-NFATx1CPS, pME-NFATx1C, pME- NFATx1CBA, or
pME-NFATx1CBP) or NFATx isoform expression plasmid (pME-NFATx1,
pME-NFATx2, pME-NFATx3, or NFATx4). Semiconfluent 293 cells (35-mm
dish) were transfected with 1 µg of reporter plasmid NFATLuc (20) or
pG5Luc, 0.25 µg of internal control plasmid pBC12/RSV/SEAP (Tropix),
and 2 µg of GAL4 fusion expression plasmid (pGAL4-X1BB, pGAL4-X1PP,
pGAL4-X1PA, pGAL4-X1WT, or pGAL4-X1 M1pGAL4-X1 M10) or NFATx isoform
expression plasmid. The transfection assay was repeated at least three
times, and luciferase activity was normalized to protein mass and
alkaline phosphatase activity.
RT-PCR analysis
Total RNA (2 µg) from Jurkat cells was isolated using an RNeasy kit (Qiagen, Chatsworth, CA). Poly(A)+ RNAs (0.5 µg) from thymus, skeletal muscle, and kidney were purchased from Clontech (Palo Alto, CA). The RNA was reverse transcribed at 42°C, using the Superscript preamplification system (Life Technologies) in a 20-µl reaction mixture as described by the manufacturer. For PCR amplification, 1 µl of the cDNA synthesis mixture described above was used. Primer pairs used for RT-PCR of NFATx mRNA isoforms were sense 20-mer P1 (5'-CTGGACAGCACTCAACTCAA-3'), which is common for all NFATx mRNA isoforms, and antisense 20-mer P2 (5'-CTGGACATGAGAGCACACTG-3'), P3 (5'-GCCAAAACGTAGGTCTCAAC-3'), or P4 (5'-TGAACTCCTGACCTTGTGAT-3'), which are specific for NFATx2, NFATx3, and NFATx4, respectively. An antisense 18-mer P5 (5'-CCTTGGGAAACAGAAATC3') was also used with P1 to check relative expression of NFATx2 and NFATx3 to NFATx1. The cDNA was amplified for 30 cycles at an annealing temperature of 55°C, and PCR products were then analyzed by electrophoresis on 3% NuSieve agarose gels (FMC Bioproducts, Rockland, ME).
Northern blot analysis
Northern (RNA) blots of poly(A)+ RNA from different human tissues were purchased from Clontech (no. 7759-1, lot 58630, and no. 7760-1, lot 58500). The blots contained 2 µg of poly(A)+ RNA/lane, and the quality of RNA was controlled by the manufacturer by hybridization of a representative blot for each lot of RNA with a human ß-actin control probe. To detect all mRNA isoforms of NFATx, we used a cDNA probe spanning the NFATx sequence (positions 1156–2614) (10). For the NFATx2 isoform, a 243-bp DNA fragment was generated from the plasmid pME-NFATx2 by PCR using oligonucleotides 5'-ACCAATTTATATCTGACTTG-3' and 5'-GGCTGCAGTAAGCACTGTTA-3' for the sense and antisense primers, respectively. The 243-bp fragment was purified and cloned into the pCRII vector (Invitrogen, San Diego, CA), and the 106-bp EcoRI-XbaI fragment was used for the NFATx2-specific hybridization probe. DNA probes were labeled by random priming, and hybridization was conducted in Quick Hyb Hybridization Solution (Stratagene, La Jolla, CA) with 200 µg/ml salmon sperm DNA at 68°C for 1.5 h. The filter was washed in 2x SSC/0.1% SDS at room temperature and subsequently in 0.1x SSC/0.1% SDS at 60°C, then exposed to x-ray film.
Electrophoretic mobility shift assays (EMSA)
EMSAs were conducted with 10-µg protein extracts as described previously (24). The oligonucleotides used in EMSAs contained the following sequence (only one strand is shown; sequence overhangs are in lower case): NFAT site from the human IL-2 (25), 5'-gatcGGAGGAAAACTGTTCATACAGAAGGCGT-3'. Extracts from transfected COS-7 cells were prepared, and fractions containing purified AP-1 were obtained as described previously (24).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We recently reported domain analysis of NFATx1 and found that
transcriptional activation mediated by NFATx1 was significantly reduced
by deletion of the carboxyl-terminal domain (16). This observation
focused our attention on the carboxyl-terminal region of NFATx1 in
terms of transactivation function. We then made a series of constructs
deleting part of this region from NFATx1 and checked transactivation in
Jurkat cells by measuring the luciferase activity of the extract of
Jurkat cells that was cotransfected with the reporter plasmid
pNFAT72Luc and with either an expression plasmid of the deleted
NFATx1 carboxyl-terminal region or the empty vector (pME18S). The
plasmid pCMV/SEAP was used as an internal control to monitor
transfection efficiency. The transfected Jurkat cells were incubated at
37°C for 40 h and stimulated with a combination of PMA and
A23187 for 8 h before harvest (Fig. 1
A). The expression plasmid
containing a deletion of 31 amino acids from the end of NFATx1
(pME-NFATx1CAA) still conferred the same transactivation activity as
the full-length construct of NFATx1 (pME-NFATx1). On the other hand,
the expression plasmid containing a 47-amino acid deletion from the end
of NFATx1 (pME-NFATx1CPS) showed a severely decreased
transactivation activity compared with pME-NFATx1. These results
clearly indicated that the 16-amino acid sequence of NFATx1
(QDITLDDVNEIIGRDM) were important for transactivation activity mediated
by NFATx1.
|
B,
the expression construct pME-NFATx1C, which had most of the
carboxyl-terminal domain of NFATx1 deleted, conferred no increase in
transactivation activity. Interestingly, transfection of the expression
plasmid pME-NFATx1CBP, containing this deletion fused with the end
of the carboxyl-terminal region, including the 15-amino acid sequence
(DITLDDVNEIIGRDM) that we had identified, resulted in the same
level of increased activity as pME-NFATx1. However, consistent with the
previous results, the expression plasmid pME-NFATx1CBA containing
this large deletion fused with the end of that carboxyl terminus that
lacked the 15-amino acid sequence conferred no increase in activity.
From these results, we conclude that the carboxyl-terminal region of
NFATx1 has a role in its transactivation activity and that a sequence,
consisting of the 15-amino acid sequence (DITLDDVNEIIGRDM), is
essential and apparently sufficient for the transactivation activity
mediated by NFATx1 in Jurkat cells. The results also indicate that
other sequences in the carboxyl terminus of NFATx1 can be deleted
without detrimental effects on its transactivation activity as measured
in our assays. Transactivation sequence of NFATx1
To assess whether the 15 amino acid residues of NFATx1 we
identified had any intrinsic transactivation potential, chimeric
molecules were generated by fusing portions of the NFATx1
carboxyl-terminal sequence to the DNA binding domain of GAL4 (Fig. 2A). These constructs were
assayed for their ability to transactivate reporters driven by GAL4 DNA
binding sites in the human embryonic kidney HEK 293 cells. We included
a pM vector that encodes only the GAL4 DNA binding domain as a negative
control. The results showed that the entire carboxyl-terminal region of
NFATx1 (amino acids 761-1075) was sufficient for transcriptional
activation (X1BB). Moreover, 46-amino acid residues from the very end
of the carboxyl terminus of NFATx1 (amino acids 1030–1075) and
15-amino acid residues (amino acids 1030–1044) were also found to
activate transcription just as efficiently as the whole
carboxyl-terminal region of NFATx1 did (X1PP and X1PA). These results
indicate that the sequence, consisting of 15 amino acids from the
carboxyl terminus of NFATx1, was an intrinsic transactivation sequence.
|
B, most of the amino
acid residues we substituted, except asparagine (M5), were critical for
the potential as a transactivation sequence. Notably, the mutation on
the hydrophobic residue (leucine, valine, or isoleucine) completely
abrogated the intrinsic transactivation potential (M1, M4, M7, and M8). Isolation of mRNA isoforms for NFATx
We previously reported the cloning of NFATx1, a member of the NFAT
family (10). During further screening for NFAT cDNA clones from the
Jurkat cDNA library, we isolated two additional isoforms of NFATx,
designated NFATx2 and NFATx3. These two clones had distinct 104- and
89-bp insertion sequences at the same position, position 3131 according
to the sequence of Masuda et al. (10), in the 3'-coding region of the
NFATx1 cDNA sequence (Fig. 3A). The insertions,
displaying no homology to sequence in NFATx1 or to any other sequence
in the available databases, resulted in a frame shift and replacement
of the most carboxyl-terminal 40 residues of the NFATx1 by 33 and 30
unrelated amino acids for NFATx2 and NFATx3, respectively (Fig. 3, B and C).
|
, A–C). Nucleotide sequences were deposited in the
GenBank sequence databank and are not shown here. In short, we found
four distinct cDNA clones of NFATx, designated NFATx1, NFATx2, NFATx3,
and NFATx4, which are most likely derived by alternative splicing at a
common site in the coding sequence. The absence of a consensus splice
donor and acceptor sequences at the junctions (position 3131) makes it
unlikely that these isoforms represent incompletely spliced transcripts
and likely that alternative exons were used for these isoforms.
Verification of this hypothesis, however, awaits deciphering of the
precise exonic structure of the NFATx gene. Most interestingly,
junctions for these isoforms are located in the 15-amino acid
transactivation sequence that we identified in the NFATx1.
Consequently, these three isoforms have different corresponding
sequences to the transactivation sequence of NFATx1. Starting from the
first ATG codon, NFATx1 has 1,075 amino acid residues, NFATx2 has 1,068
residues, NFATx3 has 1,065 residues, and NFATx4 has 1045 residues, with
predicted molecular masses of 115,592, 115,310, 115,124, and 112,625
Da, respectively, and they differ after amino acid residue 1035. Tissue distribution of NFATx mRNA isoforms
Our previous report showed preferential expression of NFATx in
thymus and leukocytes (10). However, we noticed that when we used a
newly designed cDNA probe (positions 1156–2614 in the NFATx1 sequence)
that should hybridize to all four isoforms of NFATx, we could also find
a reasonable hybridizing band with an apparent size of 7.0 kb in RNA
from testis, ovary, skeletal muscle, and kidney (Fig. 4A). This result suggested
that our previously reported expression profile was specific to the
NFATx1 isoform and that the tissue distributions of the other NFATx
isoforms were distinct.
|
A). We did amplify a
specific segment of NFATx2 or NFATx3, but not of NFATx4, using total
RNA from Jurkat cells (Fig. 4B, lane 6). This
result indicated that both NFATx2 and NFATx3, but not NFATx4 mRNA, were
expressed in Jurkat T cells. Since we are interested in the expression
of NFATx isoforms, especially in skeletal muscle and kidney where
NFATx1 was not strongly expressed, we next checked for the expression
of these NFATx isoforms in thymus, skeletal muscle, and kidney using
same RT-PCR analysis. As shown in Figure 4B, lanes 7to 9, the NFATx2 segment was detected in all these
tissues. The NFATx3 segment was detected in thymus and kidney, but not
in skeletal muscle. The NFATx4 segment was detected in thymus and
skeletal muscle, but not in kidney. These results suggested that the
expression patterns of NFATx isoforms are distinct in those tissues.
To compare the relative expression levels of NFATx2 and
NFATx3 mRNA to that of NFATx1, we designed an oligonucleotide
primer pair (P1 and P5) amplifying a region that contained the common
insertion site in the 3' coding region of NFATx (Fig. 3A).
Contrary to the RT-PCR analysis described above (Fig. 4B),
this primer pair should amplify NFATx1-, NFATx2-, and NFATx3-specific
segments simultaneously. The results using total RNA from Jurkat cells
revealed two amplification products of 244 and 348 bp, which
corresponded to NFATx1 and NFATx2, respectively. Interestingly, the
lower band corresponding to NFATx1 was dominant compared with the upper
band, which corresponded to NFATx2 (Fig. 4C, lane
2). In addition, this result did not change even if we used
poly(A)+ or total RNA from Jurkat cells that were treated
with PMA and A23187 for 3 h (data not shown). Although we did
amplify an NFATx3-specific segment using total RNA from Jurkat cells by
RT-PCR with primer pair P1 and P3 (Fig. 4B, lane
6), we did not detect a band corresponding to NFATx3 (333 bp) in
this RT-PCR analysis using primer pair P1 and P5, perhaps because the
assay was not sensitive enough to detect the low amounts of specific
NFATx3 mRNAs. These results indicated that NFATx1, NFATx2, and NFATx3
mRNAs were expressed in Jurkat T cells and that NFATx1
appeared to be the predominant species.
We further investigated the relative expression of NFATx isoforms in
thymus, skeletal muscle, and kidney using same primers (P1 and P5)as
those used for RT-PCR analysis. We detected the same two amplification
products as in the Jurkat cells (Fig. 4C, lanes
6, 9, and 12). Interestingly, the thymus and
kidney exhibited basically the NFATx1 mRNA isoform (lower
band, lanes 6 and 12), while skeletal muscle
exhibited primarily the NFATx2 mRNA isoform (upper
band, lane 9). These results indicated that NFATx1 and
NFATx2 mRNA isoforms were expressed in those tissues, however with
different ratios. Again, we did not detect a band corresponding to
NFATx3 in these tissues by this analysis. Judging from these results,
although we did detect the specific NFATx3 mRNA (Fig. 4B, lanes 6, 7, and 9), the
amount is very low compared with those of NFATx1 and NFATx2 mRNA in
Jurkat cells, thymus, and kidney.
To compare the relative expression levels of the NFATx2 in various
tissues, we then hybridized a human multiple tissue blot with an
NFATx2-specific probe consisting of the insertion sequence of NFATx2
cDNA. NFATx2 mRNA, which had the same apparent size (7.0 kb) as NFATx
mRNA in Figure 4A, was detected very strongly in RNA from
skeletal muscle and weakly in RNA from testis, spleen, thymus,
prostate, ovary, small intestine, heart, placenta, and pancreas (Fig. 4D). We did not detect NFATx2 expression in
kidney. We also hybridized the same tissue blot with an NFATx3- or
NFATx4-specific probe. However, neither NFATx3 nor NFATx4 mRNA was
detected even in very low stringent washing conditions (data not
shown). Taken together, the NFATx isoforms with altered carboxyl
terminus are expressed differentially in different tissues, possibly
generated by tissue-specific alternative splicing.
Transcriptional activity of NFATx isoforms
It is interesting to determine whether the transcription activity
of the NFATx isoforms is modulated by alteration of last amino acid
residues in Jurkat T cells as well as HEK 293 cells, since all three
isoforms isolated in this study lack the transactivation sequence of
NFATx1. The activity of the NFATx isoforms was checked by the transient
transfection assay described above. The transfected Jurkat cells were
then divided into three sets and incubated at 37°C for 40 h. The
cells were left unstimulated or were stimulated either with PMA or with
a combination of PMA and A23187 for 8 h before harvest. The
activity attributed to endogenous NFAT was detected as the increase in
luciferase activity upon PMA/A23187 stimulation of cells transfected
with only empty vector, pME18S (Fig. 5A). Cotransfection of NFATx1
with reporter plasmid caused increase in NFAT-dependent transactivation
by eightfold over the endogenous NFAT activity. Notably,
transactivation activity mediated by NFATx2, NFATx3, or NFATx4 was
markedly lower than that mediated by NFATx1. Stimulation by both PMA
and A23187 was strictly required for transactivation in all cases (data
not shown).
|
A). The expression plasmid of each
NFATx isoform was cotransfected with the reporter plasmid NF-ATLuc,
which contains the luciferase gene under the control of seven copies of
the murine IL-2 distal NFAT site. The plasmid pBC12/RSV/SEAP was also
included to monitor for transfection efficiency. Two sets of
transfection were performed for each NFATx isoform and were incubated
at 37°C for 40 h. One set was left without stimulation, while
the other set was stimulated with PMA and A23187 for 8 h. Similar
to results with Jurkat cells, the extract from cells transfected with
pME-NFATx1 showed 10 times greater activity when cells were stimulated
by PMA/A23187 compared with mock transfectant (pME18S), whereas
transfection with pME-NFATx2, pME-NFATx3, or pME-NFATx4 gave much lower
activity compared with NFATx1 cDNA (Fig. 5B). It is unlikely
that lower induction of luciferase activity by transfection of the
NFATx isoform expression plasmid (pME-NFATx2, -NFATx3, or -NFATx4) is
due to a failure in expression, because comparable NFATx isoform
expression was confirmed by immunoblotting using an antiserum against a
common peptide sequence of NFATx isoforms (data not shown).
To test whether the NFATx2-, NFATx3-, or NFATx4-specific sequence
affects NFATx DNA binding, we transfected an expression plasmid
containing the full-length cDNA of NFATx isoforms into COS-7 cells and
prepared nuclear and cytosolic extracts from the transiently
transfected cells. Western blot analysis using an antiserum against the
peptide in the amino terminus of NFATx (16) revealed very similar
levels of the NFATx isoform proteins in extracts of the transfected
cells (data not shown). We also transfected the empty expression vector
(pME18S) as a mock control. EMSAs were then conducted with a
double-stranded oligonucleotide containing the NFAT-binding site (Fig. 5C). The results demonstrated the presence of DNA binding
activities in extracts from cells transfected with all NFATx isoforms.
This activity was absent from cells transfected with the expression
vector alone.
In agreement with previous characterization of NFAT in T cells (5, 6),
exogenous AP-1 polypeptides were needed to reconstitute NFAT binding by
the cytoplasmic extract from NFAT-transfected cells. Nuclear extracts,
on the other hand, did not require exogenous AP-1 (Fig. 5C).
These results collectively indicate that no difference was found in the
characteristics of DNA binding to the NFAT site among NFATx isoforms
and further support the idea that a 15-amino acid sequence in the
carboxyl terminus of NFATx1 is important for transactivation activity
in at least T cells (Jurkat) and kidney cells (293).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, A and
B) and confirmed that these amino acids comprise a
transcriptional activation sequence (Fig. 2A). Furthermore,
amino acid substitution proves that this sequence constitutes a motif
required for activation (Fig. 2B). Interestingly, activation
mediated by this 15-amino acid sequence is lost when its hydrophobic
residues (leucine, valine, and isoleucine) are nonconservatively
replaced by alanine. Consistently, those hydrophobic residues are
perfectly conserved among NFAT family proteins (see below). The
mechanism of the transactivation remains unknown, but this 15-amino
acid transactivation sequence increases the transactivation function of
NFATx1 by enhancing its ability to interact with coactivator
or other components of the general transcription apparatus, and alanine
substitution of this sequence may cause disruption of this interaction.
The transcription-activating domains of other transcription factors
have often been classified according to the predominant amino acids
contained therein. For example, acidic domain, glutamine-rich, and
proline-rich (26) as well as glycine-rich (27) or isoleucine-rich (28)
activating segments have been identified. Notably, the 15-amino acid
transactivation sequence in NFATx1 does not contain typical amino acid
residues for transactivation domains, except for the five acidic amino
acids (glutamic and aspartic). Although some transcription-activating
domains can be induced to fold into -helixes (26), it appears the
transactivation sequence in NFATx1 does not allow a helix/coil
structure as predicted by the algorithm AGADIR (29).
We also have isolated the cDNAs encoding three isoforms of NFATx that
have different carboxyl-terminal ends in this study (Fig. 3, A and B). As a result, the transactivation
sequence of NFATx1 is replaced for these isoforms. Consistently,
transactivation activity mediated by NFATx isoforms NFATx2, NFATx3, and
NFATx4 was lower than that for NFATx1 in Jurkat as well as 293 cells
(Fig. 5, A and B). We also showed that the
different carboxyl-terminal ends did not affect DNA binding (Fig. 5C). In fact, the NFATx isoforms were shown to have
relatively similar binding affinities to the NFAT site using a
competition experiment with increasing amounts of unlabeled probe in
the EMSA (data not shown). These data indicated the importance of the
transactivation sequence of NFATx1 in Jurkat and 293 cells and also
suggested that NFATx has a transactivation sequence as a module, and
this sequence is replaced by a different sequence for a different
isoform. It is known that truncation of the activation domain of
several transcription factors, such as STAT, retinoic acid receptor,
and hepatocyte nuclear factor, turn them into dominant negative
repressors of the normal forms (30, 31, 32). However, a carboxyl-terminal
deletion mutant of NFATx1 (NFATx1C) and newly isolated NFATx
isoforms (NFATx2, NFATx3, and NFATx4) that showed lower transactivation
activity than that of NFATx1 (Figs. 1B, 5A, and
5B) did not affect NFATx1-mediated transactivation as
measured in a cotransfection assay (data not shown).
Independently, Hoey et al. previously reported the sequences of three
isoforms of NFATx/NFAT4, designated NFAT4a, NFAT4b, and
NFAT4c (9). Sequence comparison between isoforms of NFAT4 and NFATx
showed that NFATx2, which we isolated in this study, was identical with
NFAT4c. Taken together, NFATx/NFAT4 has at least six isoforms, named
NFATx1, NFATx2(NFAT4c), NFATx3, NFATx4, NFAT4a, and NFAT4b.
Unlike NFATx isoforms, the divergence of the sequences of
NFAT4a and NFAT4b isoforms begins just downstream of the Rel
similarity region, where the carboxyl-terminal regions are much shorter
(9 and 40 amino acid residues for Figure 4, A and
B, respectively). Recently, we also cloned mouse
counterparts of NFATx1 and NFATx2, designated mNFATx1 and
mNFATx2, from a murine EL-4 cDNA library (33). Moreover, a truncated
version of the mouse counterpart of human NFATx1, designated NFATc3,
was also isolated from a cDNA library derived from the thymus of a TCR
transgenic mouse by Ho et al. (13).
The existence of isoforms has been reported for NFAT1(NFATp). Three isoforms of NFAT1(NFATp) (NFAT1-A, NFAT1-B, and NFAT1-C) were isolated from a Jurkat cDNA library (34, 35). As with NFATx isoforms, the very ends of the carboxyl-terminal region of NFAT1(NFATp) isoforms are replaced by unrelated amino acid residues. Similarly, an NFATc isoform with a distinctive sequence of the carboxyl-terminal region, designated NFATc.ß, was recently reported (36). It appears, then, that the existence of multiple isoforms is a characteristic of NFAT proteins (Table I).
We previously reported Northern blot analysis data showing that NFATx
was expressed predominantly in thymus and at lower levels in peripheral
blood leukocytes. In that analysis, we used as a probe BglII
cDNA fragment (positions 2698–3235) that contained a common sequence
of all NFATx isoforms and an NFATx1-specific sequence (10). In the
present study we used a different cDNA probe (positions 1156–2614)
whose sequence was common in all four NFATx isoforms (Fig. 4A). Using this probe, we detected significant NFATx
expression in various tissues, such as testis, ovary, skeletal muscle,
and kidney as well as in thymus and leukocytes. These observations
prompted us to examine the possibility of differential expression of
the NFATx mRNA isoforms. Interestingly, RT-PCR and Northern blot
analyses with probes specific for either NFATx1 or NFATx2 showed a
significant difference in the ratio of the two NFATx mRNA isoforms in
different tissues. Although our RT-PCR analysis showed the existence of
NFATx3 and NFATx4 mRNA species in several tissues with distinct
expression patterns (Fig. 4B), we could not detect those
mRNA species by Northern blot analysis using multiple tissue blot,
suggesting that NFATx3 and NFATx4 isoforms are relatively minor species
compared with NFATx1 and NFATx2 isoforms in those tissues. On
the other hand, NFATx1 and NFATx2 are major mRNA species, at least in
thymus, leukocytes, and skeletal muscle tissues. The thymus and
leukocytes expressed primarily NFATx1 mRNA, while skeletal muscle
expressed predominantly NFATx2 mRNA (Fig. 4, C and
D). Accordingly, the cDNA of NFAT4c, which is identical with
that of NFATx2, was cloned from a skeletal muscle tissue cDNA library
(9). Although our RT-PCR analysis showed that NFATx1 mRNA was present
(Fig. 4C, lane 12), it is notable that the
BglII cDNA fragment from NFATx1 did not strongly hybridize
with mRNA from kidney. It is thus possible that another NFATx isoform
is expressed in kidney.
Since the tissues in which we detected expression of NFATx mRNA isoforms, such as thymus, skeletal muscle, and kidney, consist of various types of cells, it is important to identify what types of cells in tissues express those isoforms to investigate their function. In addition, we cannot completely rule out the possibilities that expression of NFAT genes are under translational control and that NFAT proteins are not necessarily found in all tissues in which the genes are transcribed. The use of isoform-specific Abs should help us resolve these issues.
The role of NFAT proteins in skeletal muscle and kidney tissues is unknown. One potential target for NFAT proteins in these tissues could be the IL-15 gene, whose mRNA expression was observed outside the lymphoid system; in particular, skeletal muscle, kidney, placenta, and lung expressed IL-15 mRNA strongly, whereas lymphoid organs, such as spleen and thymus, expressed lesser quantities of the transcript (37, 38). Alternatively, it is also possible that target genes of NFAT proteins are other than cytokine genes. Interestingly, recent gene targeting and a transgene approach suggest that NFAT family proteins have unique target genes other than cytokine genes in heart (39, 40, 41). No obvious abnormality was reported in the skeletal muscle and kidney of mice deficient in NFAT1(NFATp) or NFATc protein (39, 40, 42). The development of mice deficient in other NFAT proteins (NFATx and NFAT3) will help us understand the roles of the NFAT family proteins in these tissues.
Significantly, the transactivation sequence of NFATx1 that we
identified in this study was also found in NFAT1-C, one of the
NFAT1(NFATp) isoforms, and in NFAT3 (Fig. 6). Furthermore, a GAL4 DNA binding
domain fused to the very end of carboxyl-terminal region of NFAT1-C or
NFAT3, which contains a sequence corresponding with the NFATx1
transactivation sequence also showed transactivation potential (R.
Imamura, E. S. Masuda, and N. Arai, unpublished observation).
Consistently, Luo et al. showed that NFAT1-C was the most potent
transcriptional activator among NFAT1(NFATp) (35). In addition, they
showed that a GAL4 DNA binding domain fused to the carboxyl-terminal
domain of NFAT1-C contained intrinsic transactivation potential.
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 The nucleic acid sequences reported in this paper have been submitted to the GenBank with accession numbers U85428, U85429, and U85430.
3 Address correspondence and reprint requests to Dr. Naoko Arai, Department of Cell Signaling, DNAX Research Institute of Molecular and Cellular Biology, 901 California Ave., Palo Alto, CA 94304-1104. E-mail address:
4 Abbreviations used in this paper: AP-1, activating protein-1; RSV, Rous sarcoma virus; SEAP, secreted placental alkaline phosphatase; EMSA, electrophoretic mobility shift assay.
Received for publication December 2, 1997. Accepted for publication May 29, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Bourajjaj, A.-S. Armand, P. A. da Costa Martins, B. Weijts, R. van der Nagel, S. Heeneman, X. H. Wehrens, and L. J. De Windt NFATc2 Is a Necessary Mediator of Calcineurin-dependent Cardiac Hypertrophy and Heart Failure J. Biol. Chem., August 8, 2008; 283(32): 22295 - 22303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Kaminuma, F. Kitamura, N. Kitamura, T. Hiroi, H. Miyoshi, A. Miyawaki, and S. Miyatake Differential Contribution of NFATc2 and NFATc1 to TNF-{alpha} Gene Expression in T Cells J. Immunol., January 1, 2008; 180(1): 319 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Dienz, S. M. Eaton, T. J. Krahl, S. Diehl, C. Charland, J. Dodge, S. L. Swain, R. C. Budd, L. Haynes, and M. Rincon Accumulation of NFAT mediates IL-2 expression in memory, but not naive, CD4+ T cells PNAS, April 24, 2007; 104(17): 7175 - 7180. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Dong, S. Kim, S. Hong, J. Das Gupta, K. Malathi, E. A. Klein, D. Ganem, J. L. DeRisi, S. A. Chow, and R. H. Silverman From the Cover: An infectious retrovirus susceptible to an IFN antiviral pathway from human prostate tumors PNAS, January 30, 2007; 104(5): 1655 - 1660. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. B. Hock and M. A. Brown Nuclear Factor of Activated T Cells 2 Transactivation in Mast Cells: A NOVEL ISOFORM-SPECIFIC TRANSACTIVATION DOMAIN CONFERS UNIQUE Fc{epsilon}RI RESPONSIVENESS J. Biol. Chem., July 11, 2003; 278(29): 26695 - 26703. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. van Rooij, P. A. Doevendans, C. C. de Theije, F. A. Babiker, J. D. Molkentin, and L. J. De Windt Requirement of Nuclear Factor of Activated T-cells in Calcineurin-mediated Cardiomyocyte Hypertrophy J. Biol. Chem., December 6, 2002; 277(50): 48617 - 48626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Amasaki, S. Adachi, Y. Ishida, M. Iwata, N. Arai, K.-i. Arai, and S. Miyatake A Constitutively Nuclear Form of NFATx Shows Efficient Transactivation Activity and Induces Differentiation of CD4+CD8+ T Cells J. Biol. Chem., July 5, 2002; 277(28): 25640 - 25648. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Yoshida, I. Ishikawa, Y. Ono, T. Imai, R. Suzuki, and O. Yoshie An Activation-Responsive Element in Single C Motif-1/Lymphotactin Promoter Is a Site of Constitutive and Inducible DNA-Protein Interactions Involving Nuclear Factor of Activated T Cell J. Immunol., September 15, 1999; 163(6): 3295 - 3303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Chuvpilo, A. Avots, F. Berberich-Siebelt, J. Glockner, C. Fischer, A. Kerstan, C. Escher, I. Inashkina, F. Hlubek, E. Jankevics, et al. Multiple NF-ATc Isoforms with Individual Transcriptional Properties Are Synthesized in T Lymphocytes J. Immunol., June 15, 1999; 162(12): 7294 - 7301. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Liu, E. S. Masuda, L. Tsuruta, N. Arai, and K.-i. Arai Two Independent Calcineurin-Binding Regions in the N-Terminal Domain of Murine NF-ATx1 Recruit Calcineurin to Murine NF-ATx1 J. Immunol., April 15, 1999; 162(8): 4755 - 4761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. S. Stevenson, M. F. Gomez, D. C. Hill-Eubanks, and M. T. Nelson NFAT4 Movement in Native Smooth Muscle. A ROLE FOR DIFFERENTIAL Ca2+ SIGNALING J. Biol. Chem., April 27, 2001; 276(18): 15018 - 15024. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
X174 DNA with
HaeIII as a size marker. For a control, expression
plasmids of pME-NFATx1, pME-NFATx2, pME-NFATx3, and pME-NFATx4 were
also amplified with each primer set (lane 2–5).
C, Comparison of relative expression levels of NFATx
mRNA isoforms in human tissues using a primer pair (P1 and P5)
amplifying a region that contained the common insertion site in the
3'-coding region of NFATx. This primer pair amplifies
NFATx1-, NFATx2-, and NFATx3-specific segments
simultaneously. Lanes 1 and 5 represent a
digest of 
