HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 290/3/G466    most recent 00431.2005v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Rowley, C. W. Articles by Simon, T. C. Search for Related Content PubMed PubMed Citation Articles by Rowley, C. W. Articles by Simon, T. C.

LIVER AND BILIARY TRACT

Mechanisms of mutual functional interactions between HNF-4 and HNF-1 revealed by mutations that cause maturity onset diabetes of the young

²Division of Biology and Biomedical Sciences, and ¹Department of Pediatrics, Washington University School of Medicine, St. Louis, Missouri

Submitted 13 September 2005 ; accepted in final form 6 October 2005

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Hepatic nuclear factor (HNF)-4 and HNF-1 are key endodermal transcriptional regulators that physically and functionally interact. HNF-4 and HNF-1 cooperatively activate genes with binding sites for both factors, whereas suppressive interactions occur at regulatory sequences with a binding site for only one factor. The liver fatty acid binding protein gene (Fabp1) has binding sites for both factors, and chromatin precipitation assays were utilized to demonstrate that HNF-4 increased HNF-1 Fabp1 promoter occupancy during cooperative transcriptional activation. The HNF4 P2 promoter contains a HNF-1 but not HNF-4 binding site, and HNF-4 suppressed HNF-1 HNF4 P2 activation and decreased promoter HNF-1 occupancy. The apolipoprotein C III (APOC3) promoter contains a HNF-4 but not HNF-1 binding site, and HNF-1 suppressed HNF-4 APOC3 activation and decreased HNF-4 promoter occupancy. Maturity onset diabetes of the young (MODY) as well as defects in hepatic lipid metabolism result from mutations in either HNF-4 or HNF-1. We found that MODY missense mutant R127W HNF-4 retained wild-type individual Fabp1 activation and bound to HNF-1 better than wild-type HNF-4, yet did not cooperate with HNF-1 or increase HNF-1 Fabp1 promoter occupancy. The R127W mutant was also defective in both suppressing HNF-1 activation of HNF4 P2 and decreasing HNF-1 promoter occupancy. The HNF-1 R131Q MODY mutant also retained wild-type Fabp1 activation and bound to HNF-4 as well as the wild type but was defective in both suppressing HNF-4 APOC3 activation and decreasing HNF-4 promoter occupancy. These results suggest HNF-1-HNF-4 functional interactions are accomplished by regulating factor promoter occupancy and that defective factor-factor interactions may contribute to the MODY phenotype.

transcription factor; hepatic nuclear factor-4; hepatic nuclear factor-1

HNF-1 and HNF-4 bind directly to one another, and two types of functional interactions have been identified between these two factors (17, 22). A cooperative interaction can occur for target genes with binding sites for both factors in their regulatory sequences (9, 17). Cooperative activation is characterized by significantly greater target gene activation by the two factors together compared with the sum of the individual activations of each factor separately. A second type of factor-factor functional interaction is a single-site tethering interaction, where one factor binds to the DNA and the second factor binds to the first factor to effect regulation. HNF-4 can act as a coactivator for HNF-1 at target genes that contain an HNF-1 but not HNF-4 binding site, where HNF-4 binds to HNF-1 rather than the DNA and contributes to target gene activation (11). HNF-1 can inhibit HNF-4 activation of genes with only an HNF-4 binding site (22). These functional interactions are dependent on direct binding of the two factors to one another (17, 22). HNF-1 binds to HNF-4 through the HNF-4 AF domain at residues 337–368 (23). HNF-4 binds to HNF-1 through interaction with HNF-1 residues 280–440 (23).

In addition to these functional and physical interactions, HNF-4 and HNF-1 are linked through the disease maturity onset diabetes of the young (MODY). MODY results from haploinsufficiency of the HNF-4 gene (HNF4, MODY1) or HNF-1 gene (TCF1, MODY3) as well as four other pancreatic transcription factors (40). MODY is characterized by a defect in insulin secretion arising after adolescence (31), but extrapancreatic phenotypes that are independent of diabetes are also observed in MODY patients. MODY1 patients display disorders in hepatic lipid metabolism, with lowered levels of HNF-4 target genes apolipoprotein AI, apolipoprotein AII, apolipoprotein CIII (APOC3), and lipoprotein a (36, 43). MODY3 patients exhibit decreased serum levels of the HNF-1 target gene apolipoprotein M (38). MODY is an autosomal dominant disease caused by complete inactivation of one TCF1 or one HNF4 allele. However, most MODY1 and MODY3 mutations so far identified are missense mutations with a single amino acid substitution, and these patients have a phenotype indistinguishable from those with allelic inactivation (13, 40). Although many missense mutations result in loss of DNA binding, nuclear localization, or transactivation function, other mutations have no obvious effect on individual target gene activation (4, 12, 25, 29, 32, 48, 52, 55, 56). These findings suggest that the mutations are defective in some function besides direct target gene activation.

We obtained evidence that MODY mutants are specifically defective in cooperatively activating the liver fatty acid binding protein gene (Fabp1) (9). Fabp1 provides a model for endodermal gene regulation because it is highly expressed in hepatocytes and enterocytes, and expression is primarily regulated at the transcriptional level (3). A transgene constructed from rat Fabp1 nucleotides –596 to +21 relative to the start site of transcription is active in murine hepatocytes and intestinal epithelial cells (49). HNF-1 and HNF-4 activate this transgene through interactions with cognate sites in the proximal promoter (9), and HNF-4 (18) and HNF-1 (1) are required for in vivo hepatic Fabp1 expression. HNF-4 and HNF-1 cooperatively activate the Fabp1 transgene in cultured cells, and we found that the MODY3 mutant R131Q retained wild-type activation of the Fabp1 transgene but had decreased cooperative activation with HNF-4.

We have now utilized MODY mutations as a tool to investigate the functional interactions between HNF-4 and HNF-1. MODY missense mutants for both HNF-1 and HNF-4 were identified that are defective in cooperative interactions and also in single-site suppressive interactions. We found that cooperative interactions are accompanied by increased factor binding to the gene regulatory sequence and suppressive interactions by decreased factor binding. The MODY mutants retain wild-type factor-factor binding but are deficient in altering DNA binding of the opposite factor.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Plasmids. Expression plasmids for all transcription factors were in the pSG5 plasmid (Stratagene) as previously reported (9), utilizing a commercial kit to introduce mutations (QuikChange, Stratagene). The expression plasmid for HNF-4 is pTS276, for HNF-1 is pTS158, for HNF-4 R127W is pTS294, and for HNF-1 R131Q is pTS269. Promoter transgenes expressed human growth hormone (hGH) or firefly luciferase reporters, and similar results were obtained with either reporter. pTS245 contains Fabp1 nucleotides –596 to +21 with a hGH reporter with the first intron steroid hormone binding destroyed by mutagenesis (9). pTS247 contains the same transgene except that the Fabp1 HNF-4 binding site is mutagenized (9). A plasmid with the Fabp1 promoter driving firefly luciferase (pTS388) was created by releasing the Fabp1 promoter from pTS10 (44) with BamHI/KpnI and ligating this fragment into pGL3-basic (Promega) cut with BglII/KpnI. An HNF4 P2 transgene plasmid (pTS462) was constructed by amplifying the promoter sequence –285 to +2 (50) from human genomic DNA utilizing PCR and inserting this sequence into pGL3-basic (Promega) with BglII (5') and HindIII (3') sites introduced in the PCR primers. An identical transgene with a hGH reporter was created by amplifying the same HNF4 P2 sequence and exchanging it for the Fabp1 sequence in pTS245 with KpnI/BamHI. The APOC3 promoter plasmid was a kind gift from Bernard Laine (Institut National de la Santé et de la Recherche Médicale, Lille, France) and includes promoter nucleotides –1390 to +21 (33). Glutathione-S-transferase (GST) fusion protein expression plasmids for HNF-4 (pTS377) and HNF-1 (pTS359) were constructed by PCR amplification of the transcription factor coding sequence from pTS276 or pTS158 and insertion into pGEX-2T (Amersham). Pull-down assay prey synthesis templates were constructed by amplifying the desired sequence from pTS276 or pTS158 and inserting the fragment into pSG5 to create pTS414 (HNF-4 residues 51–174), pTS437 (R127W mutant HNF-4 residues 51–174), pTS442 (HNF-1 residues 1–282), and pTS443 (HNF-1 residues 1–282 with the R131Q mutation). Plasmids with HNF-1 residues 1–282 had a coding sequence for an additional 10 methionines added to the carboxy terminus for increased radiolabeling specific activity. All sequences amplified by PCR or subjected to site-directed mutagenesis were verified by sequencing.

Cell culture and transfections. Caco-2, HeLa, and HepG2 cells were obtained from the American Type Culture Collection and maintained as recommended. Transient transfections in Caco-2 or HepG2 cells were performed by calcium phosphate precipitation (9), and those in HeLa cells were performed with Superfect reagent (Qiagen) according to the manufacturer's protocol. All assays were performed on cells in six-well plates harvested 48 h after transfection. Transfections with hGH reporter transgenes included expression plasmids for firefly luciferase to control for transfection efficiency (pGL3, Promega). Transgene activity was measured with hGH radioimmunoassay (Nichols or Diagnostics Systems Laboratories) of the cell media changed 24 h before the assay or with a commercial kit (Promega) for firefly luciferase in cell lysates. Transfections with firefly luciferase reporter transgenes included expression plasmids for Renilla luciferase to control for transfection efficiency (pRL-TK, Promega). The total amount of DNA in each well was kept constant in each experiment with addition of pSG5. All transfections were performed with triplicate wells of cells for each condition, and transgene activity in each well was calculated as the ratio of transgene reporter to pGL3 or pRL-TK luciferase activity. Transgene activity is reported as the average of the triplicate value with error bars indicating standard deviation. Statistical significance was calculated with a two-tailed t-test, and error propagated for calculated values. All transfections were performed at least twice with similar results.

Binding assays. EMSAs were performed as previously described (9) except that nuclear extracts were prepared from Caco-2 cells transfected with the HNF-4 expression plasmid or control pSG5. Antibodies purchased from Santa Cruz Biotechnology for HNF-4 (sc-8987) and c-myc (sc-764) were both raised in rabbits.

Solution binding assays with GST fusion proteins were performed essentially as described (8). Radiolabled prey was synthesized from pSG5 plasmid templates, and aliquots were subjected to SDS-PAGE to verify equivalent synthesis between wild-type and mutant factors. Equivalent amounts of prey were utilized in each solution binding assay.

Chromatin immunoprecipitation. Chromatin immunoprecipitation (ChIP) assays were performed on HepG2 cells transiently transfected using the Superfect reagent (Qiagen). Cells were transfected in 10-cm dishes with a plasmid mixture consisting of 4 µg of each transgene and transcription factor expression plasmid. Cell harvest was performed 48 h after transfection.

Formaldehyde was added to cell culture medium to a final concentration of 1%, plates were rocked at room temperature for 10 min, and glycine was then added to a final concentration of 130 mM for an additional 5 min. All subsequent steps were performed at 4°C. Cells were washed twice with PBS and suspended in 1.4 ml PBS by scraping. Cells were centrifuged at 14,000 g for 1 min, resuspended in 1 ml PBS containing 1 mM PMSF, centrifuged as before, resuspended in 1 ml PBS with PMSF, centrifuged as before, and resuspended in 1 ml cell lysis buffer [5 mM PIPES (pH 8.0), 85 mM KCl, and 0.5% Nonidet P-40 (NP-40)] containing a protease inhibitor cocktail (Roche no. 1836853) and 1 mM PMSF. Cells were incubated for 10 min and centrifuged at 2,000 g for 5 min, and the pellet was resuspended in 1 ml nuclei lysis buffer [50 mM Tris (pH 8.1), 10 mM EDTA, and 1% SDS] with 0.5 mM PMSF and protease inhibitor cocktail. After a 10-min incubation, samples were sonicated five times for 10 s (Fisher Sonic Dismembrator 100 with microprobe) and then centrifuged at 14,000 g for 15 min. Supernatants were diluted with 1 ml dilution buffer [16.7 mM Tris (pH 8.1), 167 mM NaCl, 1.2 mM EDTA, and 1.1% Triton X-100], and 100 µl of a 50% suspension of blocked protein-Sepharose A beads were added (Amersham no. 17-09741). Beads were blocked by incubation as a 10% suspension in dilution buffer containing 2.0 µg/ml sheared herring sperm DNA for 4–24 h. Samples were centrifuged at 14,000 g for 5 min; 100 µl of the supernatant were reserved as the "input" sample, and the remainder was incubated with the appropriate antibody overnight on a turning wheel. Antibodies were from Santa Cruz Biotechnology (sc-8986 HNF-1 and sc-6556 HNF-4).

Blocked protein-Sepharose A (100 µl) was added to each sample, and incubation was continued for 1 h. Samples were centrifuged for 5 min at 14,000 g, and the Sepharose washed as follows: once with wash buffer 1 [20 mM Tris (pH 8.0), 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, and 01.% SDS], four times with wash buffer 2 [20 mM Tris (pH 8.0), 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, and 01.% SDS], once with wash buffer 3 [10 mM Tris (pH 8.0), 250 mM LiCl₂, 1 mM EDTA, 1% deoxycholate, and 1% NP-40], and three times with Tris-EDTA [10 mM Tris (pH 8) and 0.1 mM EDTA]. Each wash was for 10 min, recovering beads by 3-min centrifugation at 14,000 g. Pellets were resuspended in elution buffer (50 mM NaHCO₃ and 1% SDS), agitated for 15 min at room temperature, and then centrifuged for 3 min at 14,000 g. Elution was repeated, and the two supernatants were combined. All samples, including the input samples, had their total volume adjusted to 300 µl with water, and NaCl was adjusted to 0.3 M. Samples were heated at 67°C for 5 h, and then 70 µl of PK buffer [1.25% SDS, 0.5 M Tris (pH 7.5), and 25 mM EDTA] were added, followed by 125 µl proteinase K solution (20 µg/µl water, Amersham). Samples were incubated at 50°C for 1 h, and DNA was then purified using a commercial kit (Qiagen PCR purification kit). Purified DNA was heated at 95°C for 10 min.

The presence of transgene sequences in the immunoprecipitated DNA was assessed with PCR as previously described (47) using 3 µl of each immunoprecipitated sample or 1 µl of each input sample as the template. Amplification was performed for 25–28 cycles to ensure that results were obtained in the linear range. PCR primers were 5'-TTAGAACAAACTTCTGCCTTGCCCATTCTG-3' and 5'-TACCAACAGTACCGGAATGCCAAGCTTACT-3' for Fabp1-luciferase, 5'-TTGCTGCATCTGGACACCCTGCCTCAGGCC-3' and 5'-TACCAACAGTACCGGAATGCCAAGCTTACT-3' for APOC3-luciferase, and 5'-TTAGAACAAACTTCTGCCTTGCCCATTCTG-3' and 5'-TTGCCGCTAGTGAGCTGTCCACAGGACCCT-3' for HNF4 P2-hGH. All ChIP assays were performed at least twice with similar results.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
HNF-4 activates the Fabp1 transgene through direct interaction with a cognate binding site in the proximal promoter. A transgene constructed from nucleotides –596 to +21 of Fabp1 is active in murine hepatocytes (49) and HepG2 cells in transient transfection assays (9). Consensus binding sites for HNF-4 and HNF-1 are in close proximity in the proximal Fabp1 promoter (Fig. 1A), and transgene activation by HNF-4 or HNF-1 in HepG2 cells is dependent on these cognate sites (9). HNF-1 and HNF-4 cooperatively activate the Fabp1 transgene, and the MODY3 R131Q HNF-1 mutant is specifically defective in this cooperation (9). We (9) have previously established that HNF-1 and HNF-1 bind to the HNF-1 site and that mutagenizing the HNF-1 site abolished transgene activation by these factors. The putative HNF-4 binding site was more carefully characterized to explore HNF-1-HNF-4 cooperative interactions. Transient transfection assays in HepG2 cells demonstrated that mutagenizing the putative HNF-4 binding site reduced but did not abolish HNF-4 transgene activation (9) (Fig. 1B). However, HNF-4 did not activate the mutagenized transgene in HeLa cells (Fig. 1B). The residual HNF-4 activation of the mutagenized transgene in HepG2 cells may be due to interaction of endogenous HNF-4 with HNF-1 or other factors present in HepG2 but not HeLa cells. EMSAs were employed to determine whether HNF-4 bound directly to the Fabp1 site (Fig. 1, C and D). Incubation of the radiolabled Fabp1 HNF-4 binding site sequence with nuclear extracts prepared from Caco-2 cells transfected with HNF-4 resulted in the formation of numerous complexes (Fig. 1D, lane 2). Inclusion of native and mutagenized Fabp1 sequences as competitors revealed that several low-mobility complexes were specific for the Fabp1 sequence (Fig. 1D, compare lanes 3 and 4 with lanes 7 and 8). However, the authentic APOC3 HNF-4 site sequence competed only for one of these bands (Fig. 1D, arrow, lanes 5 and 6), which was also the only complex to bind to the HNF-4 antibody (Fig. 1D, lane 9). The other specific complexes did not form with the APOC3 probe (Fig. 1D, lane 12) and may include peroxisome proliferator-activated receptors (PPAR) and related binding partners that also bind to the Fabp1 HNF-4 site (data not shown and Refs. 37 and 44). The Fabp1 HNF-4 binding site contains one nucleotide that does not match the consensus sequence (41) (Fig. 1C). A comparison of competition EMSAs reveals that the Fabp1 site binds HNF-4 less avidly than the APOC3 site, which conforms exactly to the consensus. Both the Fabp1 and APOC3 site oligonucleotides abolished HNF-4 complex formation with the Fabp1 probe at a 64- and 128-fold molar excess (Fig. 1D, lanes 3–6). However, the Fabp1 competitor at 64-fold molar excess did not completely abolish the HNF-4 complex with the APOC3 probe, whereas no complex was detected in the presence of a 64-fold molar excess of APOC3 probe (Fig. 1D, compare lane 13 with lane 15). In addition, considerably more HNF-4 complex formed with the APOC3 probe than with the Fabp1 probe, and the EMSAs were conducted at the same time with probes of similar specific activity. These results demonstrate that HNF-4 binds directly to the Fabp1 HNF-4 site, although with less avidity than to a perfect consensus site.

View larger version (57K): [in this window] [in a new window]   Fig. 1. Hepatic nuclear factor (HNF)-4 directly activates the fatty acid binding protein 1 gene (Fabp1) through a cognate site in the proximal promoter. A: the rat Fabp1 promoter transgene sequence, numbered relative to the start site of transcription (+1) with factor binding sites numbered by the most proximal consensus site nucleotide. B: transient transfections comparing HNF-4 activation of the native Fabp1 promoter transgene and that of a transgene with the HNF-4 site mutagenized as shown in C. Triplicate wells of cells were transfected with 2 µg transgene, 2 µg HNF-4 expression plasmid, and 50 ng pGL3 control luciferase expression plasmid each. Transgene reporter levels were normalized to luciferase activity in each well, and values are expressed as fold activation over the native Fabp1 transgene activity in the absence of HNF-4. Bar height is the average value and error bars are SD. Mutagenesis of the putative HNF-4 site decreased HNF-4 activation in HepG2 cells (left; P = 0.002) and eliminated activation in HeLa cells (right; P > 0.5). C: nucleotide sequence of the HNF-4 consensus binding site (53), the Fabp1 -55 HNF-4 binding site, the mutagenized Fabp1 site with loss of HNF-4 activation, and a characterized apolipoprotein CIII (APOC3) HNF-4 binding site (24). The Fabp1 nucleotide deviant from the consensus is in bold, mutagenized bases are underlined, and numbers indicate the position of the 3' nucleotide relative to the transcriptional start site. D, left: EMSA with radiolabeled double-stranded Fabp1 probe as in C incubated with (lane 2) or without (lane 1) nuclear extracts of cells transfected with HNF-4 expression plasmid. Reactions in lanes 3 and 4 contained a 64- and 128-fold molar excess of unlabeled probe competitor, lanes 5 and 6 contain competitor APOC3 oligonucleotide as in C, and lanes 7 and 8 contain competitor mutant Fabp1 oligonucleotide. One shifted complex (arrow) is competed by the authentic HNF-4 sequence and the Fabp1 HNF-4 site but not the mutagenized site. This same band is lost when binding reactions include an antibody to HNF-4 (lane 9) but not an antibody to c-myc (lane 10). Right, an identical EMSA to that in the left except that the probe is radiolabled APOC3 oligonucleotide.

HNF-4 increases wild-type but not R131Q mutant HNF-1 promoter occupancy during cooperative interactions.

View larger version (26K): [in this window] [in a new window]   Fig. 2. Mature onset diabetes of the young (MODY) mutants defective in HNF-1-HNF-4 transcriptional cooperativity and partner recruitment. A: a transient transfection in HeLa cells compares Fabp1 transgene activation by HNF-1 and the HNF-1 MODY3 R131Q mutant. Triplicate wells of cells were transfected with 2 µg transgene, varying amounts of wild-type (WT) or mutant HNF-1 expression plasmid, and 1 µg pGL3 control luciferase expression plasmid each. The transgene reporter level was normalized to luciferase activity in each well, and values are fold activation over Fabp1 transgene activity without added HNF-1. Symbols are average values and error bars are SD. The R131Q mutant retains near WT activity at all doses tested (P = 0.779 for 0.625 µg expression plasmid, P = 0.067 for 0.125 µg expression plasmid, P = 0.373 for 0.250 µg expression plasmid, and P = 0.070 for 0.5 µg expression plasmid). B: transient transfections in HepG2 cells compare cooperative Fabp1 transgene activation by WT or R131Q HNF-1 with HNF-4 as in A except that each well contains 2 µg transgene, 2 µg each transcription factor expression plasmid, and 50 ng pGL3. Solid bars show activation in the presence of both factors, and open bars show the calculated sum of the separate activations. R131Q mutant cooperative activity is 58% of that of WT HNF-1 (P = 0.0003). C: chromatin immunoprecipitation (ChIP) comparison of Fabp1 occupancy by WT or R131Q HNF-1 during transient transfections in HepG2 cells under the same conditions as in B except that a luciferase reporter Fabp1 transgene was utilized. Transcription factors included in each transfection are indicated above each lane. Control transfections were performed with added HNF-4 and HNF-1, but either antibody (lane 7) or transgene (lane 8) was omitted. The bottom row (input) is amplification performed on aliquots from each sample before immunoprecipitation. The addition of HNF-4 increases promoter occupancy by HNF-1 but not R131Q. D: a transient transfection assay in HeLa cells compares Fabp1 transgene activation by WT and R127W MODY1 mutant HNF-4 as in A. Activity of WT and mutant factors is indistinguishable at all doses tested (P > 0.5). E: transient transfection assays in HepG2 cells compare WT and R127W HNF-4 Fabp1 cooperativity with HNF-1 as in B. Activity of HNF-1 with R127W is not significantly different than the sum of the individual activations (P = 0.205).

The R127W HNF-4 MODY mutant is defective in cooperative gene activation with HNF-1.

View larger version (26K): [in this window] [in a new window]   Fig. 5. The HNF-1 R131Q mutant is defective in a single-site suppressive functional interaction with HNF-4. A: transient transfection assays in HepG2 cells utilizing an APOC3 transgene, which contains a HNF-4 but not HNF-1 binding site. Triplicate wells of cells were transfected with 1 µg transgene, 2 µg HNF-4 expression plasmid, varying amounts of WT or mutant HNF-1 expression plasmid, and 1 µg pRL-TK each. The transgene reporter level was normalized to control luciferase activity in each well, and values are fold transgene activation over that with no added factors (bottom dotted line). Symbols are averages and error bars are SD. HNF-4 activated the transgene 6.23-fold (top dotted line), and WT HNF-1 suppressed this activation in a dose-dependent manner (; P = 0.0472 at 0.25 µg, P = 0.0022 at 0.5 µg, P = 0.0034 at 1.0 µg, and P = 0.0001 at 2.0 µg). R131Q also inhibited HNF-4 activation (; P = 0.0630 at 0.25 µg, P = 0.0554 at 0.5 µg, P = 0.0257 at 1.0 µg, and P = 0.0004 at 2.0 µg) but was less effective than the WT (P = 0.4490 at 0.25 µg, P = 0.0026 at 0.5 µg, P = 0.0006 at 1.0 µg, and P = 0.00003 at 2.0 µg). In addition, WT HNF-1 alone suppressed APOC3 transgene activity (; P = 0.1463 at 0.25 µg, P = 0.0194 at 0.5 µg, P = 0.0161 at 1.0 µg, and P = 0.0058 at 2.0 µg). The R131Q mutant alone also inhibited at higher doses (; P = 0.8559 at 0.25 µg, P = 0.4687 at 0.5 µg, P = 0.0813 at 1.0 µg, and P = 0.0007 at 2.0 µg) but was less effective at lower doses (P = 0.1203 at 0.25 µg, P = 0.0133 at 0.5 µg, P = 0.0165 at 1.0 µg, and P = 0.1280 at 2.0 µg). B: ChIP comparison of HNF4 P2 promoter occupancy by HNF-4 during a transfection identical to that in A at the 2 µg factor added level. Transcription factors included in each transfection are indicated above each lane. Control transfections were performed with both HNF-4 and HNF-1 added but omitting antibody (lane 5) or transgene (lane 6). The bottom row (input) is amplification performed on aliquots from each sample in the top row before immunoprecipitation. The addition of HNF-1 but not R131Q reduced the amount of HNF-4 associated with the APOC3 promoter.

View larger version (42K): [in this window] [in a new window]   Fig. 3. HNF-4 R127W acquires increased binding to HNF-1, and HNF-1 R131Q binds to HNF-4 as well as the WT. A: equivalent amounts of 35S-radiolabled WT or R127W HNF-4 prey were incubated with glutathione-S-transferase (GST)-HNF-1 or GST alone as bait. WT HNF-4 bound to GST-HNF-1 (lane 2) with greater avidity than binding to GST alone (lane 1). More R127W mutant prey bound to HNF-1 than WT HNF-4 (compare lane 3 with lane 2), whereas binding to GST was minimal (lane 4). B: identical binding assay to that shown in A except that prey was HNF-4 amino acid (AA) residues 51–174. WT residues 51–174 do not bind to HNF-1, but the mutant residues acquire binding capability. C: solution binding similar to that in A except that bait is GST-HNF-4 and prey is WT or R131Q HNF-1. The R131Q mutant binds more avidly to both GST and HNF-4 than WT HNF-1, which does not bind appreciably to GST alone. D: solution binding assay similar to that in B except that HNF-1 residues 1–280 were utilized as prey. Both WT and R131Q mutant HNF-1 residues 1–280 bind specifically to HNF-4.

Wild-type but not R127W mutant HNF-4 suppresses HNF-1 activation of genes with a HNF-1 but not HNF-4 binding site by decreasing HNF-1 binding to the promoter.

View larger version (27K): [in this window] [in a new window]   Fig. 4. The HNF-4 R127W mutant is defective in a suppressive single-site interaction with HNF-1. A: transient transfection assays in HepG2 cells with the HNF4 P2 transgene, which contains a HNF-1 but not HNF-4 binding site. Triplicate wells of cells were transfected with 1 µg transgene, varying amounts of WT or mutant HNF-4 or HNF-1 expression plasmids, and 300 ng pRL-TK each. Transgene reporter values were normalized to pRL-TK activity in each well, and values are expressed as fold activation over the HNF4 P2 transgene without added factors (bottom dotted line). Symbols are averages and error bars are SD. HNF-1 activated the transgene 2.51-fold (top dotted line), and WT HNF-4 suppressed this activation in a dose-dependent manner (; P = 0.0015 at 0.25 µg, P = 0.0014 at 0.5 µg, P = 0.0005 at 1.0 µg, and P = 0.0004 at 2.0 µg). R127W also inhibited HNF-1 activation at higher doses (; P = 0.8467 at 0.25 µg, P = 0.4475 at 0.5 µg, P = 0.005a at 1.0 µg, and P = 0.0036 at 2.0 µg) but was less effective than the WT at all doses (P = 0.0025 at 0.25 µg, P = 0.0048 at 0.5 µg, P = 0.0002 at 1.0 µg, and P = 0.0005 at 2.0 µg). WT HNF-4 alone suppressed HNF4 P2 transgene activity (; P = 0.0029 at 0.25 µg, P = 0.0028 at 0.5 µg, P = 0.0036 at 1.0 µg, and P = 0.0282 at 2.0 µg). The R127W mutant also inhibited transgene activity (; P = 0.0981 at 0.25 µg, P = 0.0038 at 0.5 µg, P = 0.0081 at 1.0 µg, and P = 0.0068 at 2.0 µg) but was less effective at the lowest dose (P = 0.0272 at 0.25 µg, P = 0.8358 at 0.5 µg, P = 0.0298 at 1.0 µg, and P = 0.5251 at 2.0 µg). B: ChIP comparison of HNF4 P2 promoter occupancy by HNF-1 during a transfection identical that in A at the 2 µg dose. Transcription factors included in each transfection are indicated above each lane. Control transfections were performed with both HNF-4 and HNF-1 added but with either antibody (lane 5) or transgene (lane 6) omitted. The bottom row (input) is amplification performed on aliquots from each sample in the top row before immunoprecipitation. The addition of HNF-4 but not R127W reduced the amount of HNF-1 associated with the HNF4 P2 promoter.

Wild-type but not R131Q mutant HNF-1 suppresses HNF-4 activation of genes with a HNF-4 but not HNF-1 binding site by decreasing HNF-4 bound to the promoter. HNF-1 can suppress target genes that lack an HNF-1 binding site through binding to HNF-4 and interfering with HNF-4 activation (23). The HNF-1 R131Q mutant might be predicted to retain this single-site regulatory ability because it binds to HNF-4 as well as wild-type HNF-1 (Fig. 3) and exhibits wild-type target gene activation (Fig. 2). The suppressive activity of wild-type and R131Q mutant HNF-1 was compared with transgenes constructed from APOC3 proximal promoter sequences, which contain a HNF-4 but not HNF-1 binding site (23, 26). Transient transfection assays in HepG2 cells revealed that HNF-1 suppressed HNF-4 APOC3 transgene activation in a dose-dependent fashion and that the R131Q HNF-1 mutant was defective in this suppression (Fig. 5A). Wild-type and mutant HNF-1 factors also inhibited basal transgene activity, with the mutant less effective than the wild type. This inhibition presumably reflects the contribution of endogenous HNF-4 in HepG2 cells to basal transgene activity.

ChIP assays were utilized to compare HNF-4 APOC3 promoter occupancy in the presence and absence of wild-type and R131Q mutant HNF-1 during a transfection similar to that shown in Fig. 5A (Fig. 5B). Expression of HNF-4 resulted in increased HNF-4 APOC3 promoter occupancy (Fig. 5B, lanes 1 and 2). Expression of wild-type but not R131Q mutant HNF-1 in addition to HNF-4 decreased HNF-4 promoter occupancy (Fig. 5B, compare lanes 3 and 4 with lane 2). The HNF-4 R127W mutant activated the APOC3 transgene as well as wild-type HNF-4, as previously reported (29), but no defect was observed in the inhibition of R127W by HNF-1 compared with wild-type HNF-4 (data not shown).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
HNF-4 and HNF-1 exhibit different functional interactions in different contexts, although the molecular mechanisms of cooperative and suppressive interactions are unknown. Cooperative interactions have been reported for genes with binding sites for both factors in their regulatory sequences (15, 19, 34, 35, 42). We found that HNF-4 increased Fabp1 promoter occupancy by HNF-1, suggesting that two-site cooperativity is achieved at least in part by increased factor binding. One possible mechanism for increased binding in the presence of both factors is formation of a ternary complex among both factors and DNA that is more stable than separate binary complexes between each factor and DNA. The MODY mutants provide evidence that a mechanism in addition to simple binding stabilization is involved. The R131Q MODY3 HNF-1 mutant binds to both DNA and HNF-4 as well as the wild type, yet the presence of HNF-4 does not increase R131Q Fabp1 promoter occupancy. This observation indicates that some mechanism beyond stabilized binding is responsible for increased promoter occupancy and cooperative transcriptional activation. One candidate for this mechanism is a cofactor or other transcription factor that is required for recruitment of one factor by another, and the mutants are defective in interacting with this third protein. Support for this hypothesis comes from the finding that HNF-4 MODY mutants R154X and E276Q are specifically deficient in recruiting the transcriptional coactivator p300 (12). Cooperativity between HNF-1 and HNF-4 binding to closely spaced sites has also been demonstrated at enhancer sequences (15, 27), and the increased regulatory sequence occupancy may also be a contributing mechanism for enhancer function.

Physical binding between HNF-1 and HNF-4 allows single-site functional interactions. The HNF4 P2 promoter has a HNF-1 binding site that is critical for expression but no HNF-4 binding site (16, 50). However, HNF-4 suppresses HNF4 P2 activity in vivo (5). We found that HNF-4 suppressed HNF-1 HNF4 P2 promoter activation by decreasing HNF-1 bound to the promoter. This finding suggests that single-site suppression is accomplished by sequestration of the DNA binding factor away from the regulatory sequence by the second factor. As for the two-site cooperative interaction, this sequestration is not simply due to binding between the two factors, because the R127W MODY1 mutant binds to HNF-1 better than wild-type HNF-4 yet is defective in sequestration. An activational single-site functional interaction has also been reported for HNF-4 to HNF-1 at genes with a HNF-1 but not HNF-4 binding site (11). This coactivational function of HNF-4 to HNF-1 was observed utilizing transgenes with abbreviated promoter sequences containing a HNF-1 site. The presence of other factor binding sites such as HNF-6 in the HNF4 P2 promoter may influence factor interactions (5). It is also possible that HNF-6 or other factors or cofactors may directly mediate the interaction of HNF-4 and HNF-1 during single-site interactions at the HNF4 P2 and APOC3 promoters.

A reciprocal single-site functional interaction has also been reported for genes with a HNF-4 but not HNF-1 binding site in their regulatory sequences. In this case, HNF-1 suppresses HNF-4 target gene activation (23). We found that the R131Q MODY3 HNF-1 mutant was defective in suppressing HNF-4 activation of the APOC3 promoter, which contains a HNF-1 but not HNF-4 binding site. Wild-type HNF-1 decreased HNF-4 promoter occupancy, demonstrating a sequestration mechanism. The R131Q HNF-1 mutant was also defective in sequestration despite the wild-type ability to bind to HNF-4, suggesting a mechanism besides simple physical binding. An alternative explanation for changes in transcription factor promoter occupancy is that expression of one factor in the transient transfection assays affects the amount of the second factor through alterations in protein production or stability. However, identical transfections containing both HNF-1 and HNF-4 were conducted with three different target genes, and because in each case a different effect on factor promoter occupancy was observed, this suggests that one factor does not significantly alter the level of the other.

The MODY mutants reveal details of the physical interactions between the HNF-1 and HNF-4. The binding domains for these two factors have been mapped to the AF2 domain of HNF-4 (residues 337–368) and residues 280–444 of HNF-1 (23). In addition, there is evidence that other domains affect binding, where MODY missense mutations in the ligand binding domain of HNF-4 (A223F and E276Q) decrease binding of HNF-4 to HNF-1 (11). In contrast, the R127W mutation confers a new DNA binding capability to the HNF-4 DNA binding domain. It is interesting to note that HNF-1 MODY3 missense mutants P447L and P519L bind to cofactors CREB binding protein (CBP) and p300/CBP associated factor better than wild-type HNF-1, although the cofactors lacked histone acetyltransferase activity when bound to the mutant but not wild-type protein (46).

Previous studies have demonstrated that HNF-4 residues 227–445 could be localized to the nucleus by HNF-1 residues 1–440 but not residues 1–280, implying that the interacting domain was within HNF-1 residues 280–440 (23). We found that HNF-1 residues 1–282 were also capable of binding to full-length HNF-4 in solution. The full-length R131Q mutant exhibited somewhat increased binding to HNF-4. However, mutant but not wild-type HNF-1 also bound to GST, indicating altered factor binding characteristics. HNF-1 R131Q residues 1–280 displayed binding to HNF-4 similar to wild-type HNF-1.

We and others have found that the MODY R127W HNF-4 (29) and R131Q HNF-1 (52) mutants retain wild-type individual activation of target genes, suggesting that defective factor-factor interactions may contribute to the MODY phenotype. An alternative explanation for normal target gene activation is that the mutations are nonfunctional polymorphisms rather than MODY mutations, as has been suggested for R127W (29). This possibility is unlikely because the R127W mutant segregates with the MODY phenotype in two kindreds (6, 14, 54), as does the R131Q HNF-1 mutation (7, 20). Our finding of functional defects in both mutants also lends credence to their genetic link with the diabetic phenotype. In addition to the R131Q MODY3 and R127W MODY1 mutants, many other MODY missense mutants have wild-type target gene activation, including the HNF-4 MODY1 mutant V255M (25, 29) and HNF-1 MODY3 mutants G191D (55), T301I (56), G47E (4), S256T (4), A276D (4), and M522V (4). Because a similar phenotype develops in patients with either allelic deletion or missense mutations, factor interactions may contribute to the disease phenotype.

The data obtained with ChIP demonstrates that cooperativity may be achieved through increased promoter binding of one factor by another and suppression through decreased promoter binding of one factor by another. The increase or decrease in promoter occupancy requires more than simple binding to the opposite factor because R127W and R131Q mutants do not have decreased binding. These findings suggest that any mutation that affects the factor conformation might result in defective recruitment/sequestration of the other factor and thus a loss of factor-factor binding, either directly or through mediation of a cofactor or another transcription factor. MODY missense mutations do not cluster in specific protein domains but are distributed throughout the entire HNF-1 (13) and HNF-4 (25, 36) sequences. The results obtained in these studies through overexpression of transcription factors and target transgenes demonstrates mechanisms by which transcription factor interactions may regulate target genes. However, in vivo interactions in the context of development and in the presence of chromatin can differ from those observed with in vitro systems (39), and the in vivo relevance of these mechanisms and their potential importance for regulating native target genes will require further investigation.

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grant DK-56361 (to T. C. Simon), NIDDK Washington University Diabetes Research Training Center Pilot and Feasibility Award DK-20579 (to T. C. Simon), and NIDDK Washington University Digestive Diseases Research Core Center Grant DK-52574.

	ACKNOWLEDGMENTS

 
We thank Shiming Chen for assistance with the chromatin immunoprecipitation assays.

	FOOTNOTES

 

Address for reprint requests and other correspondence: T. C. Simon, Dept. of Pediatrics, Washington Univ. School of Medicine, Campus Box 8208, St. Louis, MO 63110(e-mail: simon_t{at}kids.wustl.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 

1. Akiyama TE, Ward JM, and Gonzalez FJ. Regulation of the liver fatty acid-binding protein gene by hepatocyte nuclear factor 1 (HNF-1). Alterations in fatty acid homeostasis in HNF-1-deficient mice. J Biol Chem 275: 27117–27122, 2000.[Abstract/Free Full Text]
2. Bach I, Mattei MG, Cereghini S, and Yaniv M. Two members of an HNF-1 homeoprotein family are expressed in human liver. Nucleic Acids Res 19: 3553–3559, 1991.[Abstract/Free Full Text]
3. Bass NM, Manning JA, Ockner RK, Gordon JI, Seetharam S, and Alpers DH. Regulation of the biosynthesis of two distinct fatty acid-binding proteins in rat liver and intestine. Influences of sex difference and of clofibrate. J Biol Chem 260: 1432–1436, 1985.[Abstract/Free Full Text]
4. Bjorkhaug L, Sagen JV, Thorsby P, Sovik O, Molven A, and Njolstad PR. Hepatocyte nuclear factor-1 gene mutations and diabetes in Norway. J Clin Endocrinol Metab 88: 920–931, 2003.[Abstract/Free Full Text]
5. Briancon N, Bailly A, Clotman F, Jacquemin P, Lemaigre FP, and Weiss MC. Expression of the 7 isoform of HNF4 is activated by HNF6/OC-2 and HNF1, and repressed by HNF41 in the liver. J Biol Chem 24: 33398–33408, 2004.
6. Bulman MP, Ellard S, and Hattersley AT. R127W in HNF4 is a loss-of-function mutation causing maturity-onset diabetes of the young (MODY) in a UK Caucasian family. Diabetologia 43: 1203, 2000.[CrossRef][ISI][Medline]
7. Cao H, Shorey S, Robinson J, Metzger DL, Stewart L, Cummings E, and Hegele RA. GCK and HNF1A mutations in Canadian families with maturity onset diabetes of the young (MODY). Hum Mutat 20: 478–479, 2002.[Medline]
8. Divine J, Staloch L, Haveri H, Jacobsen C, Wilson D, Heikinheimo M, and Simon T. GATA-4, GATA-5, and GATA-6 activate the rat liver fatty acid binding protein gene in concert with HNF 1. Am J Physiol Gastrointest Liver Physiol 287: G1086–G1099, 2004.[Abstract/Free Full Text]
9. Divine JK, McCaul SP, and Simon TC. HNF-1 and endodermal transcription factors cooperatively activate Fabpl: MODY3 mutations abrogate cooperativity. Am J Physiol Gastrointest Liver Physiol 285: G62–G72, 2003.[Abstract/Free Full Text]
10. Dogra SC and May BK. Liver-enriched transcription factors, HNF-1, HNF-3, and C/EBP are major contributors to the strong activity of the chicken CYP2H1 promoter in chick embryo hepatocytes. DNA Cell Biol 16: 1407–1418, 1997.[ISI][Medline]
11. Eeckhoute J, Formstecher P, and Laine B. Hepatocyte nuclear factor 4 enhances the hepatocyte nuclear factor 1-mediated activation of transcription. Nucleic Acids Res 32: 2586–2593, 2004.[Abstract/Free Full Text]
12. Eeckhoute J, Formstecher P, and Laine B. Maturity-onset diabetes of the young Type 1 (MODY1)-associated mutations R154X and E276Q in hepatocyte nuclear factor 4 (HNF4) gene impair recruitment of p300, a key transcriptional co-activator. Mol Endocrinol 15: 1200–1210, 2001.[Abstract/Free Full Text]
13. Ellard S. Hepatocyte nuclear factor 1 (HNF-1) mutations in maturity-onset diabetes of the young. Hum Mutat 16: 377–385, 2000.[CrossRef][ISI][Medline]
14. Furuta H, Iwasaki N, Oda N, Hinokio Y, Horikawa Y, Yamagata K, Yano N, Sugahiro J, Ogata M, Ohgawara H, Omori Y, Iwamoto Y, and Bell GI. Organization and partial sequence of the hepatocyte nuclear factor-4/MODY1 gene and identification of a missense mutation, R127W, in a Japanese family with MODY. Diabetes 46: 1652–1657, 1997.[Abstract]
15. Gregori C, Porteu A, Lopez S, Kahn A, and Pichard AL. Characterization of the aldolase B intronic enhancer. J Biol Chem 273: 25237–25243, 1998.[Abstract/Free Full Text]
16. Hansen SK, Parrizas M, Jensen ML, Pruhova S, Ek J, Boj SF, Johansen A, Maestro MA, Rivera F, Eiberg H, Andel M, Lebl J, Pedersen O, Ferrer J, and Hansen T. Genetic evidence that HNF-1-dependent transcriptional control of HNF-4 is essential for human pancreatic -cell function. J Clin Invest 110: 827–833, 2002.[CrossRef][ISI][Medline]
17. Hatzis P and Talianidis I. Regulatory mechanisms controlling human hepatocyte nuclear factor 4 gene expression. Mol Cell Biol 21: 7320–7330, 2001.[Abstract/Free Full Text]
18. Hayhurst GP, Lee YH, Lambert G, Ward JM, and Gonzalez FJ. Hepatocyte nuclear factor 4 (nuclear receptor 2A1) is essential for maintenance of hepatic gene expression and lipid homeostasis. Mol Cell Biol 21: 1393–1403, 2001.[Abstract/Free Full Text]
19. Hu C and Perlmutter DH. Regulation of ₁-antitrypsin gene expression in human intestinal epithelial cell line Caco-2 by HNF-1 and HNF-4. Am J Physiol Gastrointest Liver Physiol 276: G1181–G1194, 1999.[Abstract/Free Full Text]
20. Iwasaki N, Oda N, Ogata M, Hara M, Hinokio Y, Oda Y, Yamagata K, Kanematsu S, Ohgawara H, Omori Y, and Bell GI. Mutations in the hepatocyte nuclear factor-1/MODY3 gene in Japanese subjects with early- and late-onset NIDDM. Diabetes 46: 1504–1508, 1997.[Abstract]
21. Jiang G, Lee U, and Sladek FM. Proposed mechanism for the stabilization of nuclear receptor DNA binding via protein dimerization. Mol Cell Biol 17: 6546–6554, 1997.[Abstract]
22. Kritis AA, Ktistaki E, Barda D, Zannis VI, and Talianidis I. An indirect negative autoregulatory mechanism involved in hepatocyte nuclear factor-1 gene expression. Nucleic Acids Res 21: 5882–5889, 1993.[Abstract/Free Full Text]
23. Ktistaki E and Talianidis I. Modulation of hepatic gene expression by hepatocyte nuclear factor 1. Science 277: 109–112, 1997.[Abstract/Free Full Text]
24. Ladias JA, Hadzopoulou-Cladaras M, Kardassis D, Cardot P, Cheng J, Zannis V, and Cladaras C. Transcriptional regulation of human apolipoprotein genes ApoB, ApoCIII, and ApoAII by members of the steroid hormone receptor superfamily HNF-4, ARP-1, EAR-2, and EAR-3. J Biol Chem 267: 15849–15860, 1992.[Abstract/Free Full Text]
25. Lausen J, Thomas H, Lemm I, Bulman M, Borgschulze M, Lingott A, Hattersley AT, and Ryffel GU. Naturally occurring mutations in the human HNF-4 gene impair the function of the transcription factor to a varying degree. Nucleic Acids Res 28: 430–437, 2000.[Abstract/Free Full Text]
26. Lavrentiadou SN, Hadzopoulou-Cladaras M, Kardassis D, and Zannis VI. Binding specificity and modulation of the human ApoCIII promoter activity by heterodimers of ligand-dependent nuclear receptors. Biochemistry 38: 964–975, 1999.[CrossRef][Medline]
27. Matsumura K, Saito T, Takahashi Y, Ozeki T, Kiyotani K, Fujieda M, Yamazaki H, Kunitoh H, and Kamataki T. Identification of a novel polymorphic enhancer of the human CYP3A4 gene. Mol Pharmacol 65: 326–334, 2004.[Abstract/Free Full Text]
28. Mendel DB, Hansen LP, Graves MK, Conley PB, and Crabtree GR. HNF-1 and HNF-1 (vHNF-1) share dimerization and homeo domains, but not activation domains, and form heterodimers in vitro. Genes Dev 5: 1042–1056, 1991.[Abstract/Free Full Text]
29. Navas MA, Munoz-Elias EJ, Kim J, Shih D, and Stoffel M. Functional characterization of the MODY1 gene mutations HNF4 (R127W), HNF4 (V255M), and HNF4 (E276Q). Diabetes 48: 1459–1465, 1999.[Abstract]
30. Odom DT, Zizlsperger N, Gordon DB, Bell GW, Rinaldi NJ, Murray HL, Volkert TL, Schreiber J, Rolfe PA, Gifford DK, Fraenkel E, Bell GI, and Young RA. Control of pancreas and liver gene expression by HNF transcription factors. Science 303: 1378–1381, 2004.[Abstract/Free Full Text]
31. Owen K and Hattersley AT. Maturity-onset diabetes of the young: from clinical description to molecular genetic characterization. Best Pract Res Clin Endocrinol Metab 15: 309–323, 2001.[CrossRef][Medline]
32. Oxombre B, Kouach M, Moerman E, Formstecher P, and Laine B. The G115S mutation associated with maturity-onset diabetes of the young impairs hepatocyte nuclear factor 4 activities and introduces a PKA phosphorylation site in its DNA-binding domain. Biochem J 383: 573–580, 2004.[CrossRef][ISI][Medline]
33. Oxombre B, Moerman E, Eeckhoute J, Formstecher P, and Laine B. Mutations in hepatocyte nuclear factor 4 (HNF4) gene associated with diabetes result in greater loss of HNF4 function in pancreatic -cells than in nonpancreatic -cells and in reduced activation of the apolipoprotein CIII promoter in hepatic cells. J Mol Med 80: 423–430, 2002.