| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Induction of MHC Class II Genes in G1B Cells: Identification of a Novel and Functionally Critical Leucine-Rich Motif (62-LYLYLQL-68) in the Regulatory Factor X 5 Transcription Factor1
*
UNC Lineberger Comprehensive Cancer Center and Department of Immunology and Microbiology and
Curriculum in Oral Biology, School of Dentistry, University of North Carolina, Chapel Hill, NC 27599
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Here, we show that the defect in G1B, one such
cell line, is attributed to the lack of functional RFX5, the largest
subunit of RFX. The RFX5 gene isolated from G1B cells
contains two separate single-base pair mutations. One alteration does
not exhibit a phenotype, whereas a leucine-to-histidine mutation
eliminates DNA-binding and transactivating functions. This mutation
lies outside of previously defined functional domains of RFX5 but
within an unusual, leucine-rich region (62-LYLYLQL-68). To further
investigate the significance of the leucine-rich region, we targeted
all neighboring leucine residues for mutagenesis. These mutants were
also unable to transactivate a MHC class II reporter gene, confirming
that these leucine residues play an essential role in RFX activity and
characterize a novel leucine-rich motif. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
response of most genes requires a common cascade of activating
proteins that regulate the induction of a variety of genes. Many of
these molecules have been well defined by biochemical and genetic
approaches. However, the IFN- induction of MHC class II molecules
requires additional specific transcriptional activators. The strongest
evidence for this requirement comes from the analyses of somatic mutant
cell lines that are normal in their IFN- responses except for the
lack of induction of genes encoding MHC class II and associated
molecules (1, 2). This report focuses on the understanding
of the molecular and genetic basis for one such mutant cell
line. The MHC class II molecules, HLA-D, play key regulatory roles in T cell-mediated immune responses by presenting antigenic peptides on cell surfaces for recognition by class II-restricted T cells. Molecules other than the conventional MHC class II molecules are required for optimal processing and presentation of a spectrum of Ags. These include the invariant chain (Ii)4 (3) and the nonconventional DM molecules. The expression of HLA-D, Ii, and DM are in general synchronously regulated, and their levels fluctuate coordinately upon modulation (3, 4, 5). Their presence is restricted to APCs, such as B cells, macrophages, thymic epithelia, and activated T cells. Their level of expression is primarily regulated by the activity of transcription factors (6, 7, 8, 9, 10). Therefore, understanding the players and mechanisms of transcriptional regulation of these genes will have significant biologic ramifications.
The MHC class II promoters as well as the Ii and DM promoters are unique for the presence and stereochemical arrangement of three regulatory elements, the S/W, X, and Y (also known as a CCAAT motif) elements (10, 11, 12, 13). The factors that bind to X and Y boxes have been identified as the multimeric regulatory factor X (RFX; Refs. 14 and 15) and trimeric nuclear factor Y (NF-Y) DNA binding proteins (16), respectively. An additional factor, known as X2 binding protein (X2BP), binds to the X2 element present in some promoters (17). Binding of factors to the X and Y elements serves novel roles, such as opening up previously closed promoters in vivo and promoting accessibility to regulatory elements (18, 19).
Mutant cell lines that are defective in the expression of HLA-D
molecules provide an important means for understanding the
transcriptional regulation of MHC class II and related genes. These
systems include cells derived from patients afflicted with bare
lymphocyte syndrome (BLS), as well as cells generated in vitro by
negative selection for the loss of HLA-D expression. The latter
includes both B cell lines that have lost constitutive MHC class II
expression and IFN--defective cell lines that have selectively lost
the IFN- induction of HLA-D expression (1, 20, 21, 22, 23). The
mutant EBV-transformed B cell lines fall into four complementation
groups, and the genetic defects in all of these cell lines have been
cataloged as defects either in the class II transactivator (CIITA)
(24) or in the individual components of RFX (25, 26). To date, mutant cells defective in NF-Y have not been
found, and this is presumably due to the critical role NF-Y plays in
the regulation of a large repertoire of CCAAT-containing promoters
(19, 27, 28).
The focus of this study, RFX5, is defective in the complementation group C of HLA-D-negative cell lines. The gene for RFX5 was identified by complementation cloning using the BLS cell line SJO (25). RFX5 belongs to a family of novel DNA binding proteins, including among its members RFX1 to RFX4 (29). This family shares an unusual and highly conserved DNA binding domain (DBD) (30, 31). It has been shown that RFX5 is the largest component, 75 kDa (3), of the multimeric nuclear complex RFX and that this subunit directs the binding of this complex exclusively to the X element in MHC class II and related genes (8, 10, 32).
In parallel to the study of mutant B cell lines, our earlier
examination of IFN--defective mutant cell lines revealed that one of
the mutant cell lines, namely G3A, is defective in the expression of
CIITA, whereas G1B is defective in X box binding activity, presumably
RFX (4). This present report constitutes the first
detailed analysis of the defect in G1B, which lies in two point
mutations of the RFX5 subunit of RFX. One of these two mutations
(Leu66His) is responsible for IFN-
nonresponsiveness, and it lies outside of the previously defined
DNA-binding domain. Although this mutation does not affect the level of
RFX5 expression, it eliminates the DNA binding activity and
transactivation potential of RFX5. To ascertain the significance of
this leucine-rich region, mutations in neighboring leucines were
generated in vitro. Analysis of these novel RFX5 mutants revealed that
alteration of any of the leucines eliminated transactivation function
of RFX5. These results describe an important characterization of the
G1B mutant RFX5 (G1B-RFX5) and have important implications for
suppressing the function of RFX5 and therefore modulating MHC class II
gene expression. Equally significant, this study identified an unusual,
but pivotal leucine-rich motif that is important for the function
of RFX5.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human fibrosarcoma cell lines that are defective in IFN-
induction of MHC class II gene expression were generated earlier
(1). G1B, G2A, G3A, and G4A HLA-D-negative cell lines were
generously provided by George R. Stark and Catherine Mao (The Cleveland
Research Foundation, Cleveland, OH). The 2fTGH cell line (wild-type
(wt) control) derived from the human sarcoma line HT1080 (American Type
Culture Collection, Manassas, VA) only expresses high levels of MHC
class II with IFN- induction. For the experiments here, these lines
were cultured in DMEM supplemented with 10% FBS and 2 mM
L-glutamine.
Several human B lymphoid cell lines were utilized in this study. The SJO human B cell line is defective in RFX binding and belongs to the complementation group C of MHC class II-deficient patients with BLS (33, 34, 35). The RJ2.2.5 cell line lacking CIITA activity is a MHC class II-negative mutant B cell line (36). Raji is a human EBV-positive Burkitt’s lymphoma cell line that expresses high levels of MHC class II molecules. Namalwa cells, a gift from R. Roeder (Rockefeller University, New York, NY), were used as a positive human B cell control. All the B cell suspensions were maintained in RPMI 1640 supplemented with 7.5–10% FBS and 2 mM L-glutamine.
Ab generation and purification
Synthetic peptides of human RFX5 were designed and prepared by David Klapper (University of North Carolina). Polyclonal rabbit antisera were generated against these peptides by Rockland Immunochemicals (Gilbertsville, PA). The peptide used as the immunogen for anti-RFX5 amino terminus was NH2-CAEDEPDAKSPKTGGR-COOH and that for the anti-RFX5 carboxyl terminus was NH2-CNKDLKEHVLQSSLSQEHKD-COOH. Each was conjugated to keyhole limpet hemocyanin before immunization. High titer antisera was subsequently purified by passage over protein A/G Sepharose columns as described by the manufacturer (Pierce, Rockford, IL).
EMSA, oligonucleotide competitions, and Ab supershift
Crude nuclear extracts were prepared as described (37). Gel-shift reactions were performed essentially as previously described (38), with slight modifications. Binding reactions (20 µl volume) were conducted in 12% glycerol, 60 mM KCl, 12 mM HEPES (pH 7.9), 5 mM MgCl2, 0.06 mM EDTA, 0.5 mM DTT, and 0.5 mM PMSF with the nonspecific competitor mix of 1 µg poly(dI-dC) (Pharmacia, Piscataway, NJ) and 0.5 µg sheared salmon sperm DNA (Sigma, St. Louis, MO). To eliminate RFX5 family members that bind preferentially to methylated DNA, 50 ng of a high affinity methylated pBR322 oligonucleotide, 5'-GATMGCMGTGAMGATC-3' (where M = 5-methyl cytosine), was included in each binding reaction. Approximately 0.2–0.5 ng of the S/WXY DNA probe and 4–6 µg of nuclear extract were used. Binding reactions were incubated for 40 min on ice and then for 20 min at 30°C to promote inclusion of the X2BP. The samples were resolved by electrophoresis through 4% (40:1 acrylamide:bis-acrylamide) native polyacrylamide gels with a TGE running buffer (25 mM Tris-HCl, pH 7.9, 190 mM glycine, and 1 mM EDTA) and electrophoresed for 2.6 h at 200 V in a 4°C room. The S/WXY probe was generated by PCR using the p152DRA-chloramphenicol acetyltransferase (CAT) (3) plasmid (18) and the following radioactive primers: DRA152s1, 5'-GAACGGAGTATCTTGTGTC-3', and DRA48s2, 5'-CAAATCAAT TACTCTTTGG-3'.
The sequences of the oligonucleotides used for competition are: Y box, 5'-AAATATTTTTCTGATTGGCCAAAGAGTAAT-3', and S/W box, 5'-CTTGTGTCCTGGACCCTTTGCAAGAACCCTTC-3'. The X1, X2, X1/Y, X2/Y, and mut oligonucleotide competitors are all based on a 60-bp HLA-DRA fragment spanning the X/Y motif (–118 to –59) with appropriate mutations underlined in each of the elements. The X1 competitor sequence is 5'-GAACCCTTCCCCTAGCAACAGATTGTGAGTCTCAAAATATTTTTCGGAGGTTCCAAAGAG-3'. The X2 competitor is 5'- GAACCCTTAGAACTAGTCCAGATGCGTCATCTCAAAATATTTTTCGGAGGTTCCAAAGG-3'. The X1/Y and X2/Y competitors are the same as the X1 and X2 competitors, respectively, except that the Y box is not mutated. The mut competitor contains mutations in all three sites.
Ab supershift reactions were performed as stated above, except 2 µl of protein A/G affinity-purified preimmune or immune Ab were added to each reaction for 2 h at 4°C before the addition of the oligonucleotide probe. The probe was then added, and the reactions were incubated as described above. The protein-DNA complexes were resolved by electrophoresis as described.
Western blot analysis
About 25 million cells were lysed in 50 mM HEPES (pH 7.9), 500 mM KCl, 10% glycerol, 0.1 mM EDTA, 0.5% Nonidet P-40, DTT, and PMSF at 4°C for 1 h. Samples were prepared and analyzed by SDS-PAGE electrophoresis and electroblotting to nitrocellulose as described (39). Protein A/G affinity-purified rabbit anti-RFX5 Ab and anti-NF-YA (18) polyclonal sera were used at a final concentration of 2 µg/ml to probe the blotted proteins, as described previously (40). To detect the primary Ab-protein complexes, the blots were incubated with secondary goat anti-rabbit Abs conjugated with horseradish peroxidase (Bio-Rad Laboratories, Gaithersburg, MD) at a dilution of 1:4000. Signals were visualized using the enhanced chemiluminescence detection system (Amersham Life Science, Arlington Heights, IL) and autoradiography.
DNA constructs
Complementary DNAs were prepared from total RNA isolated from
the wt control 2fTGH and the IFN--defective G1B cell lines. The
cDNAs were generated using mouse mammary tumor virus reverse
transcriptase (RT; Bio-Rad Laboratories) and random hexamers. These
cDNA samples were subjected to PCR amplification using the Extend Long
Template PCR kit (Boehringer Mannheim, Indianapolis, IN) and the
following primers: 5'-CAAGTGCCCTCATGCCGGGATGG-3' and
5'-CATGGGGGTGTTGCTTTTGGGTCTTTATGC-3' (13). The PCR
product of the appropriate size was then subcloned into the TA cloning
vector (Invitrogen, Carlsbad, CA). Subsequently, a DNA fragment
containing RFX5 was inserted directionally into the
HindIII and EcoRV sites of the pcDNA3 expression
vector for functional testing in transient transfection studies. The
point mutants Leu66His and
Pro409Arg were created by replacing either the
upstream HindIII (from pcDNA3 polycloning site) to
AflII (897) or downstream AflII (897) to
EcoRV (from pcDNA3 polycloning site) wt RFX5 DNA
restriction fragments with the corresponding G1B mutant DNA fragments,
respectively. The pDRCAT300 reporter plasmid containing 
300 bases of
the HLA-DRA promoter just upstream of the transcription
start site has been described (41). All plasmid DNAs were
purified using Qiagen columns (Qiagen, Santa Clarita, CA) before
transfection analyses. Plasmid DNAs were subjected to dideoxy
sequencing using Sequenase, version 2.0, according to the
manufacturer’s protocol (Amersham Life Science).
Transient transfection and CAT activity assay
Transient transfections were performed by electroporation using 200 mV, 960 µF, and 3–4 x 106 SJO cells with a gene pulser (Bio-Rad Laboratories) as described (42). After 48 h, cells were harvested and CAT enzyme activity in the transfected cell extracts was measured as described (43). The radioactivity on the TLC plates was quantitated by PhosphorImager scanning (Molecular Dynamics, Sunnyvale, CA). Then, percentage acetylation and fold induction, defined as the ratio of percentage acetylation of transfectant with RFX5 expression construct to the percentage acetylation of sample transfected with pcDNA3 vector alone, were determined.
Stable transfection and FACS analysis
Each construct (pREP4 or pcDNA3 empty vectors and wt or G1B RFX5 vectors) carrying the G418 drug resistance marker was electroporated into SJO cells as described above. After 48 h, the cultures were subjected to selection with increasing amounts of G418 (Bio-Rad Laboratories) at doses ranging from 400 to 1200 µg/µl over a 4-wk period. To prepare stably transfected G1B lines, pREP8 vector alone or carrying wt RFX5, along with the hygromycin drug resistance marker, was introduced into cells by the standard calcium phosphate transfection method. After 48 h of incubation, transfected G1B cells were selected by treatment with increasing doses of hygromycin (75 to 150 µg/µl) (Boehringer Mannheim). For analysis of surface MHC class II expression, the cells were stained first with the mouse anti-human HLA-DR-specific Ab L243 at a dilution of 1:25 and then goat anti-mouse Ig fluorescein-conjugated secondary Ab (PharMingen, San Diego, CA) at a dilution of 1:300. The stained cells were examined by flow cytometry (Becton Dickinson, Franklin Lakes, NJ).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
induction of MHC class
II gene expression. The absence of X box promoter binding activity and
the correlative lack of transcriptional activation capability in G1B
cells have been demonstrated previously (4). This report
describes the localization of this defect to the largest subunit of
RFX, specifically to RFX5. In vitro multiprotein/DNA complexes are formed with the S/WXY domain of HLA-DRA and contain the known MHC class II binding proteins
Due to the complexity of gel-shift complexes formed on the S/WXY
DNA, it is important to establish the pattern before defining the
specific defect in the G1B cell line. Fig. 1
displays the protein/DNA complexes
formed with a B cell nuclear extract (lane 1)
followed by specific oligonucleotide competition to identify the
individual components of each band (lanes 2–8). The
predominant complexes formed in the absence of competition are three
closely spaced slow migrating bands labeled NF-Y/RFX/X2BP, RFX/X2BP,
and NF-Y/RFX. In addition, two weaker and faster migrating complexes
are detected: RFX and NF-Y. The bands are designated by the proteins
present in the complex as defined by competition (Fig. 1A)
and Ab reactivity (i.e., Fig. 2), and
they agree with an earlier report (38). Competition with
the Y box element abolishes two of the prominent bands, each containing
NF-Y, leaves the RFX/X2BP complex intact, and increases the amount
migrating as RFX alone (Fig. 1A, lane 2).
Competition with the X1 box element, the site of RFX binding,
alone or in combination with the Y box oligo, abolishes nearly all of
the upper prominent bands as well as the RFX band (Fig. 1A,
lanes 3 and 6). The X2 element is a weak
competitor, but it eliminates the uppermost NF-Y/RFX/X2BP band and
partially diminishes the RFX/X2BP band (Fig. 1A, lane
4). This is consistent with the X2 box protein present in these
two uppermost complexes. The S motif does not compete for any of the
bands (Fig. 1A, lane 5). Competition with the X2
and Y oligos excludes all protein-shifted DR complexes, except for the
RFX-DNA complex (Fig. 1A, lane 7). An oligo
mutated at the S/W, X1, X2, and Y sites does not block the formation of
specific complexes (Fig. 1A, compare lane 8 with
lane 1).
|
|
We initially observed that extracts from G1B cells do not support
X box binding as demonstrated by the lack of formation of protein-DNA
complexes by gel-shift analysis (4). In this study, we
compared the X box binding activity of four cell lines defective in
IFN- induction of MHC class II genes, including the G1B, G2A, G3A,
and G4A lines (1), as well as for other class II-deficient
cell lines (i.e., SJO and RJ2.2.5) (see Fig. 1
B). To
specifically visualize the RFX complex, a series of competitors
including S/W, Y, X2, and methylated DNA were used in this gel-shift
experiment to eliminate the binding of proteins to the S/W, Y/CCAAT,
and X2 boxes and the binding of RFX1, respectively. Based on
comparisons to the band migration patterns in Fig. 1A, an
RFX specific protein-DNA complex is missing in the G1B extract, while
it is present in the other IFN- mutant cell lines, G2A, G3A, and G4A
(Fig. 1B, compare lane 2 with lanes 3,
4 and 5). The absence of RFX binding activity in the
G1B mutant is consistent with previous reports that the HLA-DRA
promoter is bare in this line as shown by in vivo footprinting analyses
(4, 44). Likewise, extracts prepared from SJO cells that
lack functional RFX protein do not bind to DNA containing the X1 box,
whereas the binding in other B cells (i.e., Namalwa and Raji) is normal
(Fig. 1B, compare lane 7 with lanes
6 and 8). On the other hand, RJ2.2.5 cells that
lack a functional CIITA protein, and are thereby deficient in class II
expression, exhibit normal RFX binding activity (Fig. 1B,
lane 9).
The RFX5 protein from the G1B cells is expressed at a normal level
We wanted to investigate the properties of the RFX5 protein and to
confirm the composition of the DNA-protein gel-shift complexes. So, two
new Abs specific for RFX5 were generated. As shown in Fig. 2
(left), the Ab to the carboxyl-terminal domain of
RFX5 recognizes a single 75-kDa band in Western blot analysis of a G3A
nuclear extract as predicted from the amino acid sequence and
biochemical purification (25). Both amino- and
carboxyl-terminal-specific RFX5 Abs were used in an in vitro binding
assay, and both clearly shifted the RFX, the NF-Y/RFX, and RFX/X2BP
complexes (Fig. 2, right). The shifted complexes migrate
very slowly and overlap the position of the NF-Y/RFX/X2BP band, thereby
obscuring the effect of Abs on this band. Neither Ab affected the
migration of the NF-Y band. In addition, Abs to NF-
Bp50 had no
affect (data not shown).
To directly examine the expression of the RFX5 protein in G1B cells, we
utilized the new Ab generated against the carboxyl-terminal RFX5
peptide (see Fig. 2, left, and Materials and
Methods). This Ab was used for Western analysis of whole-cell
extracts prepared from 2fTGH, G1B, and SJO cells. The Western blot
shown in Fig. 3 indicates that the
anti-RFX5 Ab detects a 75-kDa band in both the mutant G1B and its
wt parental line, 2fTGH. Furthermore, G1B cells contain similar amounts
of RFX5-specific protein as found in 2fTGH (compare lanes
4–6 with 1–3). This result suggests that there is no
gross alteration or loss of RFX5 protein in the G1B mutant. In
contrast, the 75-kDa protein in SJO detected by the RFX5 Abs was
greatly diminished (see Fig. 3, lanes 7–9), as expected on
the basis of earlier work (25). Although we routinely see
a doublet at 75-kDa for 2fTGH and G1B cells and a faint band for SJO
cells, the specificity of the second band is unknown.
|
RFX5 cDNA clones from G1B and wt cells were prepared to
investigate molecular aberrations in the gene derived from G1B cells.
RFX5-specific primers for PCR amplification of
reverse-transcribed RNA were designed according to the published
RFX5 sequence (25). To eliminate PCR artifacts,
several precautions were taken. First, at least two different RT-PCR
preparations were made using a mixture of Taq and Pwo DNA
polymerases for higher fidelity. Subsequently, multiple clones were
studied. As predicted by the initial report characterizing RFX5
(25), RT-PCR amplification gave rise to DNA fragments that
were 1900 bp in length for both wt and G1B. Finally, the entire
length of multiple clones derived from different RT-PCR preparations
for both wt and G1B-RFX5 cDNAs was sequenced.
Comparison of mutant to wt RFX5 sequences by
dideoxynucleotide sequence analyses revealed the presence of
single-base pair point mutations in G1B. Unexpectedly, we discovered
that RFX5 from G1B is altered from wt at two positions,
specifically at nucleotides 358 (T to A) and 1387 (C to G) (see Fig. 4A). This is true for multiple
clones synthesized from different RT-PCR preparations. Both of these
transition mutations lead to the replacement of nonpolar amino acids
with basic positively charged amino acids: leucine to histidine
(Leu66His) and proline to arginine
(Pro409 Arg) (Fig. 4A). This mutant
RFX5 is hereafter referred to as G1B-RFX5.
|
To understand the consequences of these mutations, we examined the
function of RFX5 isolated from G1B cells in promoter transactivation
and in promoting class II expression on the surfaces of cells. First,
expression vectors encoding wt or G1B-RFX5 clones were
transiently transfected along with the class II DR promoter-CAT
reporter construct, pDRCAT300, in SJO cells, which is deficient of RFX5
protein (see Fig. 3). In the absence of expressed RFX5, the
HLA-DR promoter is not active in SJO cells (25, 35). However, DR promoter induction is restored
(nearly 4-fold over pcDNA3 vector alone) when wt RFX5 is
introduced into SJO cells (Fig. 4B). In contrast,
G1B-RFX5 expression clones do not reconstitute promoter
activity when transfected into SJO cells (Fig. 4B),
indicating noncomplementarity of RFX mutations in SJO and G1B
cells.
Second, class II molecules do not appear on the surface of SJO cells
that stably harbor G1B-RFX5 plasmid. In contrast, SJO cells
stably transfected with wt RFX5 show a positive shift in
HLA-DR surface expression as determined by FACScan analyses (see Fig. 4C). Taken together, these results confirm that the G1B-RFX5
that we isolated is unable to activate class II promoter function.
Complementation of G1B with cloned RFX5 restores class II surface expression
To underscore that the defect in G1B cells resides in RFX5 alone,
we stably transfected G1B cells with the cloned wt RFX5 (or
vector alone as a negative control). Class II expression was induced in
these stable lines by treating with IFN-, and then surface
expression was measured by flow cytometry. The results presented in
Fig. 5 show a positive shift for G1B
cells carrying wt RFX5, indicating that the lack of surface class II
molecules in G1B cells is remedied by complementation with wt
RFX5 (Fig. 5, compare upper right and
left). Conversely, untransfected G1B or cells transfected
with empty vector do not show a shift in class II expression with
IFN- induction (Fig. 5, lower left and right,
respectively).
|
To define the effect of each individual point mutation found in
G1B-RFX5, additional clones carrying either one or the other
single-base pair mutation were generated by exchanging restriction
fragments between the wt and G1B mutant clones. The promoter-activating
function of each of these mutants was tested in the SJO transfection
assay described above. Our results indicate that the
Leu66His clone, having a mutation just upstream
of the DBD homologous region, is unable to support the induction of
the HLA-DR promoter (Fig. 6B). On the other hand, the
Pro409Arg mutation present in the proline-rich
region of RFX5 appears to have little to no adverse effect on
activating function, as the promoter induction for this construct was
approximately equivalent to that seen with wt RFX5 (Fig. 6B).
|
Discovering that the nonfunctioning Leu66His
mutation in G1B cells was positioned in an unusual sequence of RFX5
(62-LYLYLQL-68) rich in leucine residues (see Table I), we wondered whether the leucines in
this region were significant for the proper regulatory function of RFX.
To test this possibility, each of four leucines was mutagenized to
histidines (similar to the original G1B mutation). Then, SJO cells were
transfected with the pDRCAT300 reporter plus expression constructs
carrying each of these mutations. Subsequent assays measuring CAT
activity shown in Fig. 7B
revealed that alteration of any of the leucine residues eliminated
transactivation, indicating that each leucine in this region is
necessary for RFX function. Interestingly, this leucine-rich motif is
only found in the RFX5 protein but not in the other RFX family
members.
|
|
Finally, upon examination of the DBD sequence of RFX5
(30), we noted the presence of four cysteine residues that
are uniquely found in RFX5 but not in the other RFX family members.
This parallels the unique presence of the leucine-rich sequence in
RFX5. We questioned whether these unique cysteine residues were
required either for supporting the intramolecular structure of RFX5 by
disulfide bonds or for regulation of RFX5 activity by cellular redox
cycles (45). To determine the significance of these amino
acids for the activity of RFX5, we altered each of the cysteines to
serine residues by site-directed mutagenesis and then introduced the
expression constructs along with the pDRCAT300 reporter into SJO cells.
HLA-DR promoter activation as assessed by CAT assays using
extracts of transfected cells was not altered from wt levels upon loss
of any of the cysteine residues (see Fig. 7B). This
indicates that the cysteines, unlike the leucine residues, are
dispensable for RFX5 activity in promoter transactivation.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Along the same vein, four mutant cell lines, G1 to G4, were genetically
chosen on the basis of their selective loss of IFN--induced MHC
class II expression while retaining expression of other IFN--induced
genes (1). Analyses of these and other mutant lines
clearly indicate that the IFN- induction of MHC class II genes
requires additional molecules defined by the G1 to G4 mutant cell lines
which are not required for the IFN- induction of other genes such as
IRF-1, TAP-1, and LMP-7
(1). Previously, we performed the initial characterization
of the G3A cell line, which appears to lack the proper induction of the
CIITA transcript (4). That report showed that CIITA is
necessary for the IFN- induction of MHC class II promoters as well
as the Ii promoter.
In this report, we correlate the loss of IFN- induction of MHC class
II genes with the identification of a molecular defect in an essential
regulator, namely RFX5. Here, we present evidence that a nonconserved
amino acid change in the RFX5 chain in the G1B mutant cell line is
sufficient to abolish the binding capacity of RFX as well as the
transactivating function of this essential regulator. The mutated amino
acid is located in a region that has not been previously recognized as
important for the function of RFX5 and may delineate a new leucine-rich
functional domain. We have further confirmed this finding by showing
that new RFX5 leucine mutants created in vitro are incapable of
transactivating a class II promoter, suggesting the identification of
residues essential for RFX activity.
The RFX5 protein from G1B contains two mutant residues; however, only
one of the two appears to contribute to the mutant phenotype. The
critical leucine-to-histidine mutation in G1B-RFX5 resides in
amino acid residue 66, which is located upstream of the region of RFX5
defined by sequence homology as the region responsible for DNA binding
activity (i.e., amino acids 93–167) (30, 31). An analysis
of the region surrounding residue 66 shows that it shares no homology
with any of the previously defined RFX family members (see Table I).
Interestingly, this residue lies within a short fragment that is rich
in leucines (62-LYLYLQL-68). Curiously, we discovered that mutation of
Leu62, Leu64, or
Leu68 to histidine also abrogates the
transactivation potential of RFX5. Based on multiple algorithms for
predicting protein secondary structure, such as nnPredict and Secondary
Structure Prediction (49, 50), the RFX5 leucine-rich
region will very likely form a ß sheet. This potential diminishes
when the histidine-altered sequences are analyzed by these programs,
suggesting that a change in protein conformation occurs.
There is a well-established precedence for leucine motifs playing
critical roles in regulatory proteins. Foremost, the leucine zipper
motif present in transactivators such as myc, GCN4, and C/EBP is
required to direct protein dimerization through hydrophobic
interactions between opposing helical faces (51, 52). A
growing family of proteins that participate in protein-protein
interactions, mediate hormone signals, regulate gene expression, and
support processes requiring cell-to-cell adhesion contain leucine-rich
repeats (LRR) (53). The LRR from these proteins share
common hydrophobic and hydrophilic features with each repeat, generally
consisting of the following basic sequence:
XLa/bXLa/bLS/TN/QXN/QalS/TXG/PG/PG/PXXalXXLX, where a/b is
acidic/basic, al is aliphatic, and X is any amino acid. These proteins
tend to form ß sheets and amphipathic structures. A few notable
examples include CIITA with three LRR repeats (53); adenyl
cyclase that interacts with Ras to transmit hormone signals
(53); leucine-rich proteoglycans that regulate the
assembly of matrix proteins that affect cellular growth, repair, and
differentiation (54); and the GPIb-V-IX protein, in which
a point mutation that changes leucine (at 129) to proline in a LRR
reduces binding of von Willebrand factor and results in a bleeding
disorder (55). A recent report characterizing the
structure and function of the novel transcriptional coactivator p/CIP
(56) describes a novel leucine motif. Through its
interactions with CBP, the p/CIP transactivator mediates
hormone-receptor signals to activate gene expression. This protein
contains six leucine-charged domains with a consensus sequence of
LX1X2LL, where
X2 is Q or Y in four of six repeats, in addition
to three other large protein-interacting domains. Finally, Rowland and
Segall (57) recently described yet another unique
leucine-rich sequence found in TFIIIA, specifically
343-LEKRLESGENGLNLLL-358. They showed that changing two or four of
the leucines to alanine in almost any combination resulted in a
defective protein that was no longer able to transactivate a 5S RNA
template, nor was it able to support cell viability (57).
In light of these examples, the presence of the LYLYLQL leucine-rich
motif in RFX5, which interacts with other proteins to regulate the
constitutive and IFN--inducible MHC class II gene expression
(18, 32, 58, 59), is significant.
Although the precise mechanism by which a leucine-to-histidine change at residues 62, 64, 66, or 68 affects DNA binding remains to be elucidated, there are several possibilities. First, mutations in the 62-LYLYLQL-68 region may affect the three-dimensional structure of either RFX5 or, specifically, the DBD defined by sequence homology, thereby interfering with the binding of RFX to the X1 element. As a result, the DBD might not be accessible or able to bind to the DNA. Second, it is possible that this region is itself a DBD, required for optimal binding of RFX to its cognate class II promoter sequences. Finally, we favor the hypothesis that any change in this region may affect the local and/or overall structure of RFX5, therefore preventing the proper assembly of the subunits of RFX and/or the interactions with other proteins. To test this idea, immunoprecipitation experiments were conducted to investigate the proteins [RFX5 subunits, 36 kDa (26) or 41 kDa (15), or other transactivating factors, i.e., CIITA, X2BP, and NF-Y] that are associated with wt and mutant RFX5 proteins in vivo (data not shown). Unfortunately, in our initial experiments analyzing wt and G1B RFX5, we did not see additional proteins associated convincingly with the immunoprecipitated RFX5 (W. J. Brickey and J. P.-Y. Ting, unpublished observation, and U. Nagarajan and J. M. Boss, unpublished observation). We interpret these negative results as due either to the unstable, low abundant, or transient nature of protein-protein interactions or to a diminished affinity or specificity of the RFX5 Abs for native RFX5 protein, because the Abs were generated against a peptide synthesized in vitro. It is notable that the analyses of protein-protein interactions with RFX5 up to this point have been primarily performed with in vitro generated recombinant proteins (59) or highly enriched protein fractions (60).
Understanding the critical role of RFX in regulating promoter accessibility will be directly related to defining its interaction with other regulatory factors, as well as its DNA binding activity. There is strong evidence showing that the MHC class II promoters require and share a conserved, highly organized stereochemical arrangement of regulatory elements (12, 61) and multiple interacting proteins that bind to those elements (10, 62). Previous studies have shown that RFX or an X1 box binding protein can interact with the NF-Y transcription factor (18, 38), as well as the X2BP (58, 63, 64). This was revealed using assays that measure direct biochemical interaction or the association-dissociation rates of protein-DNA interactions. In addition, an in vivo footprinting approach of transfected promoter constructs bearing mutations in the X1, X2, or Y elements used in our laboratory showed interdependency of protein-DNA interactions in intact cells (18, 19). Also, recent data using a yeast two-hybrid system implicate an interaction of RFX with CIITA (59).
In contrast to the critical role of the leucine-rich region, altering any of the RFX5-unique cysteines in the DBD has no adverse effects on class II promoter transactivation in our system. Our results indicate that the possibility that RFX might be regulated by cellular reduction/oxidation systems at the cysteines in the DBD is diminished. Also, the Pro409Arg mutation found in G1B-RFX5 at the border of the proline-rich region in RFX5 does not diminish transactivation function. These results indicate that these individual amino acids do not play specific roles and/or do not alter protein structure or conformation to affect interactions of RFX with DNA or with other proteins.
In summary, this report identifies the molecular defect in the G1B cell
line that is selectively defective in the IFN- induction of MHC
class II genes and their associated genes. Even though two different
mutations in the G1B-RFX5 gene were found, only one of the
two contributes to the mutant phenotype. This single-residue mutant
(Leu66His) resulted in a drastic change in
phenotype and the total loss of MHC class II expression, primarily
through the absence of promoter transactivation function. Consistent
with this, we found that mutants with neighboring leucines altered to
histidine lacked transactivation function as well. This
characterization should pave the way for further examination of this
and other regulators of MHC class II gene induction.
|
Acknowledgments |
|---|
-defective HLA-negative cell lines and David Klapper for the
design and synthesis of RFX5 peptides. W. J. B. thanks
Guoxuan Li for his technical assistance in the affinity purification of
Abs used in this work, Keh-Chuang Chin for providing a sample of total
RNA isolated from G1B cells, and Allison Kron for her assistance with
the preparation of this manuscript. |
Footnotes |
|---|
2 Present address: H. Lee Moffitt Cancer Center, Department of Biochemistry and Molecular Biology, University of South Florida, 12901 Magnolia Drive, Tampa, FL 33612.
3 Address correspondence and reprint requests to Dr. Jenny P.-Y. Ting, UNC Lineberger Comprehensive Cancer Center, Campus Box 7295, Room 209, University of North Carolina, Chapel Hill, NC 27599-7295. E-mail address:
4 Abbreviations used in this paper: Ii, invariant chain; RFX, regulatory factor X; NF-Y, nuclear factor Y; X2BP, X2 binding protein; BLS, bare lymphocyte syndrome; CIITA, class II transactivator; DBD, DNA binding domain; MMTV, mouse mammary tumor virus; CAT, chloramphenicol acetyltransferase; RT, reverse transcriptase; wt, wild type; LRR, leucine-rich repeats.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Proc. Natl. Acad. Sci. USA 90:2880. mutants reveals that defects in CIITA or RFX result in defective class II MHC and Ii gene induction. Immunity 1:687.[Medline]
-induced expression of human class II-associated invariant chain gene. J. Immunol. 146:3183.[Abstract]
inducible expression of the HLA-DRA gene. J. Biol. Chem. 267:23728.-MHC class II signal transduction pathway reveals that type I and type II interferon systems share common signalling component(s). EMBO J. 11:1351.[Medline]
/ß signaling pathway. Cell 70:313.[Medline]
-helix and ß-strand segments of globular proteins. Comput. Appl. Biosci. 10:661.. Blood 86:3805.This article has been cited by other articles:
|
|
 |
|
  E. Klapisz, I. Sorokina, S. Lemeer, M. Pijnenburg, A. J. Verkleij, and P. M. P. van Bergen en Henegouwen A Ubiquitin-interacting Motif (UIM) Is Essential for Eps15 and Eps15R Ubiquitination J. Biol. Chem., August 16, 2002; 277(34): 30746 - 30753. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Jabrane-Ferrat, N. Nekrep, G. Tosi, L. J. Esserman, and B. M. Peterlin Major Histocompatibility Complex Class II Transcriptional Platform: Assembly of Nuclear Factor Y and Regulatory Factor X (RFX) on DNA Requires RFX5 Dimers Mol. Cell. Biol., August 1, 2002; 22(15): 5616 - 5625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. K. Sengupta, J. Fargo, and B. D. Smith The RFX Family Interacts at the Collagen (COL1A2) Start Site and Represses Transcription J. Biol. Chem., July 5, 2002; 277(28): 24926 - 24937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Roy, J. Hu, Q. Meng, Y. Xia, P. S. Shapiro, S. P. M. Reddy, L. C. Platanias, D. J. Lindner, P. F. Johnson, C. Pritchard, et al. MEKK1 plays a critical role in activating the transcription factor C/EBP-beta -dependent gene expression in response to IFN-gamma PNAS, June 11, 2002; 99(12): 7945 - 7950. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Li, J. A. Harton, X. Zhu, and J. P.-Y. Ting Downregulation of CIITA Function by Protein Kinase A (PKA)-Mediated Phosphorylation: Mechanism of Prostaglandin E, Cyclic AMP, and PKA Inhibition of Class II Major Histocompatibility Complex Expression in Monocytic Lines Mol. Cell. Biol., July 15, 2001; 21(14): 4626 - 4635. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Dovhey, N. S. Ghosh, and K. L. Wright Loss of Interferon-{{gamma}} Inducibility of TAP1 and LMP2 in a Renal Cell Carcinoma Cell Line Cancer Res., October 1, 2000; 60(20): 5789 - 5796. [Abstract] [Full Text]
|
|
|
|
|
|
X.-S. Zhu, M. W. Linhoff, G. Li, K.-C. Chin, S. N. Maity, and J. P.-Y. Ting Transcriptional Scaffold: CIITA Interacts with NF-Y, RFX, and CREB To Cause Stereospecific Regulation of the Class II Major Histocompatibility Complex Promoter Mol. Cell. Biol., August 15, 2000; 20(16): 6051 - 6061. [Abstract] [Full Text]
|
|
|
|
|
|
J. Villard, M. Peretti, K. Masternak, E. Barras, G. Caretti, R. Mantovani, and W. Reith A Functionally Essential Domain of RFX5 Mediates Activation of Major Histocompatibility Complex Class II Promoters by Promoting Cooperative Binding between RFX and NF-Y Mol. Cell. Biol., May 15, 2000; 20(10): 3364 - 3376. [Abstract] [Full Text]
|
|
|
|
|
|
J. Hu, S. K. Roy, P. S. Shapiro, S. R. Rodig, S. P. M. Reddy, L. C. Platanias, R. D. Schreiber, and D. V. Kalvakolanu ERK1 and ERK2 Activate CCAAAT/Enhancer-binding Protein-beta -dependent Gene Transcription in Response to Interferon-gamma J. Biol. Chem., January 5, 2001; 276(1): 287 - 297. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
