Search tips
Search criteria 


Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. 2007 October; 27(20): 7345–7353.
Published online 2007 August 6. doi:  10.1128/MCB.02201-06
PMCID: PMC2168911

Protein Kinase C-Mediated Modulation of FIH-1 Expression by the Homeodomain Protein CDP/Cut/Cux[down-pointing small open triangle]


Under normoxia, FIH-1 (factor inhibiting HIF-1) inhibits the transcriptional activity of hypoxia-inducible factor (HIF); however, under such conditions, we observed a significant level of HIF activity in renal cell carcinoma (RCC). This phenomenon could be attributed to a decrease in the level of functional FIH that has been identified in our previous work. Nonetheless, the molecular mechanism of FIH regulation in cancer, in particular RCC, was unclear until now. In this communication, we have demonstrated that in RCC, the Cut-like homeodomain protein (CDP/Cut) is involved in FIH transcriptional regulation and is controlled by a specific signaling event involving protein kinase C (PKC) ζ. Furthermore, we have defined a unique CDP/Cut binding site on the FIH promoter. With chromatin immunoprecipitation assays, we show that CDP binds to the FIH-1 promoter in vivo and that this binding is PKC ζ dependent. Moreover, we have also defined a potential phosphorylation site in CDP (serine 987) that modulates FIH expression. CDP/Cut is a transcriptional repressor that decreases FIH-1 expression and subsequently leads to a decrease in the repressor activity of FIH-1. Without this repression, HIF activity increases, allowing for the increased transcription of the genes it regulates, such as the vascular endothelial growth factor and GLUT-1 genes. Both CDP and HIF levels are increased in several cancers and are responsible for the metastatic progression of the tumors. Taken together, our results suggest for the first time a potential connection between CDP and FIH that could lead to the development of future therapeutic interventions.

Hypoxia-inducible factor (HIF) is a heterodimeric transcription factor that regulates a broad range of hypoxic responses in the cells and tissues of an organism (40, 41). These adaptive responses include the up-regulation of target genes involved in angiogenesis, erythropoiesis, matrix metabolism, glycolysis, vasomotor tone control, cell survival, and cell death (9, 27). To date, about 110 hypoxia-inducible genes have been found to be directly regulated by HIF-1 (14, 43).

HIF-1 is composed of two subunits that are both basic helix-loop-helix PAS proteins (40). The α subunit is rarely detectable in normoxia, but strikingly, it is induced under hypoxia in all cell types examined so far (12, 34-36). On the other hand, the β subunit, also known as the aryl hydrocarbon receptor nuclear translocator (ARNT), is not influenced by oxygen concentrations (15). It has been found that the hydroxylation of HIF-α subunits is important for their function as a transcription factor. In normoxia, HIF-αs are hydroxylated by prolyl hydroxylases (PHDs) and the asparagine hydroxylase FIH-1 (factor inhibiting HIF), members of the 2-oxoglutarate (2OG) superfamily, and Fe(II)-dependent dioxygenases (1, 11, 12, 14, 19, 21, 28-30, 31, 37). The PHDs, also known as EGLNs, hydroxylate at a specific proline residue (14). This hydroxylation recruits pVHL (von Hippel-Lindau gene product), where pVHL subsequently ubiquitinates HIF-α, designating it for degradation (13, 16, 18, 25, 26, 34, 44), while FIH-1 hydroxylates a specific asparagine residue within the HIF-α activation domain (21, 24). In normoxia, the FIH-1-mediated hydroxylation of the conserved asparagine residue at the COOH-terminal end of the HIF-α activation domain leads to the inhibition of p300/CBP recruitment (11, 17, 21, 22, 28). p300/CBP is a coactivator of HIF-1, and blocking its recruitment represses the transcriptional activity of HIF-1-targeted genes (2, 8, 15). One such target gene is the vascular endothelial growth factor gene, which has been shown to be inhibited by FIH-1 in cancer cells. Again, this modification is inhibited under hypoxia and leads to the transactivation of HIF-1.

Crystallographic studies suggest a unique active-site pocket and interactive sites for HIF-1 on FIH-1 (5, 22). Furthermore, it is very important that the length of the HIF substrate and the sequence surrounding the hydroxylation site play a crucial role in determining the catalytic efficiency of FIH-1 (7, 17, 23). The formation of an FIH-1 homodimer is required for substrate recognition and hydroxylase activity, and a single-point mutation (L340R) was found to be sufficient to disrupt the dimerization and ablate catalysis (20).

The FIH-1 gene is located on chromosome 10q24 (locus ID is 55662). FIH-1 contains the JmjC domain, which may be involved in chromatin structure and dynamics and signal transduction mechanisms (7). Metzen et al. have shown that the intracellular localization of FIH-1 was mainly in the cytoplasm and was not influenced by hypoxia (29). FIH-1 protein is widely expressed in human tissues and is thus potentially available for the regulation of HIF activity across a broad range of cells and culture conditions (38). Recent reports also indicate that FIH-1 may contribute to the regulation of PHD-2 and -3 expression in hypoxia, thus promoting HIF stability. It is also suggested that FIH-1 has important nonredundant effects in vivo on the expression of a range of HIF transcriptional targets in normoxia and even in severe hypoxia (38).

In summary, FIH-1 is a corepressor that interacts with HIF-α and the von Hippel-Lindau tumor suppressor protein to mediate the repression of HIF-1 transcriptional activity (24). Understanding FIH-1 regulation is essential to elucidate the HIF-1-mediated hypoxic response due to physiological stresses. Recent genetic studies did not detect any mutation in the FIH-1 gene that can be correlated with disease progression (31). This finding suggests a possible epigenetic control of FIH-1 expression responsible for FIH-1-associated pathogenesis. Our previous paper showed an increase in FIH-1 mRNA in renal cancer cells in the presence of the kinase-inactive form of protein kinase C (PKC) ζ. These data suggest that PKC ζ may play an important role in FIH-1 transcriptional regulation. In accordance with our previous results (6), we found that the FIH-1 protein level significantly increased when PKC ζ was blocked in renal cell carcinoma (RCC) cells using PKC ζ short interfering RNA (siRNA) or by overexpressing dominant-negative PKC ζ. To further study the mechanism of this event, we have cloned the FIH-1 promoter region and identified a cis element important for PKC ζ-mediated inhibition of FIH-1 transcription. Furthermore, we detected, via in vitro and in vivo assays, PKC ζ-mediated binding of the homeobox protein-CDP (CCAAT displacement protein), an important regulator of FIH-1 transcription, to the FIH-1 promoter.


Cell culture and reagents.

The human renal cancer cell line 786-O (ATCC no. CRL-1932) and the human embryonic kidney cell line HEK293 were maintained in Dulbecco modified Eagle medium with 10% fetal bovine serum (HyClone Laboratories, Logan, UT). 786-O is a human primary clear-cell carcinoma line with an epithelial morphology. Expression wild-type vectors as well as mutant plasmids of CDP/Cut were kind gifts from Alain Nepveu, McGill University, Quebec, Canada. The pxm/Cux/Mut CDP 6-mutant plasmid contains Ser402, Thr415, Ser789, Thr804, Ser972, and Ser987 and also Ser987 alone (pxm/Cux/CR#-HD/S987A), residues which were replaced with codons for Ala (4). Note that the Ser987 mutant should not be recognized by anti-CDP antibody (Santa Cruz Biotechnology, Inc.) as this is a much smaller version of the wild-type CDP as described earlier (4).

FIH-1 reporter constructs and deletion of the FIH-1 promoter.

A 6,950-bp FIH-1 promoter construct was cloned into pGL3-basic vector to create an approximately full-length FIH-1 promoter. Details were described in our previous paper (6). Based on this construct, a series of deletion mutants (see Fig. Fig.1)1) of the FIH-1 promoter were generated by direct PCR amplification from genomic DNA prepared from HEK293 cells by using the DNeasy kit (QIAGEN Inc., Valencia, CA). The length was counted with the transcription start site as 1, and upstream of the ATG codon was a minus count. PCR primers were designed containing an MluI site in the 5′ forward primers and an XhoI site in the 3′ reverse primers. The 5′ forward primer used for the −2.7 kb construct was 5′-CG AGC CGT CTA CAT GGT GTG TTA TAT TGA TC-3′, and the reverse primer is 5′-CCG CTC GAG CAT CTC TTT CAC CAT TC-3′. The 480-bp forward primer is 5′-CG ACG CGT CAT TTC TTC AGC AGT-3′, and its reverse primer is identical to the 2.7-kb reverse primer. The −220-bp forward primer is identical to that of the 480-bp construct, and its reverse primer is 5′-CCG CTC GAG CTC CGC CGC CAG GGA C-3′. The −180-bp forward primer is 5′-CAG CGC GTC TAC CAT TAG GAG GCG CAC-3′, the −114-bp forward primer is 5′-CGA CGC GTC GCA GCT CAT CTC TAT G-3′, and the −92-bp forward primer is 5′-CGA CGC GTT GGG TGT GTG GTT T-3′. The reverse primer for the 180-bp, 114-bp, and 92-bp constructs is identical to the 220-bp construct. PCR products were generated using DNA polymerase (Eppendorf-Brinkmann Instruments, Inc., Westburg, NY). PCR conditions were as follows: 98°C for 30 s and 94°C for 5 min, with 40 cycles of 94°C for 30 s, 65°C for 45 s, and 72 °C for 4 min, followed by 72°C for 10 min. PCR products were digested with MluI and XhoI restriction enzymes and ligated into the pGL3-basic vector. All the constructs generated with PCR were sequenced to monitor for any PCR-associated nucleotide incorporation errors. The automated sequencing was performed by the Mayo Clinic Molecular Biology Core facility.

FIG. 1.
FIH protein level was increased in the presence of PKC ζ DN. The 786-O RCC cells were transfected with 2 μg of expression vector plasmids carrying a dominant-negative mutant (DN) of PKC ζ or HA-FIH or empty vector in 60-mm tissue ...

Determination of FIH-1 potential transcriptional start site(s).

Briefly, the total RNA abstracted from 786-O cells was treated with calf intestinal phosphatase to remove the 5′ phosphates. This eliminates truncated mRNA and non-mRNA, leaving only full-length, capped mRNA. Intact full-length mRNA was then dephosphorylated with tobacco acid pyrophosphatase to remove the 5′ cap and leave a 5′ phosphate at the 5′ end of the mRNA. T4 RNA ligase was used to ligate the GeneRacer RNA oligonucleotide (sequence, 5′-CGA CUG GAG CAC GAG GAC ACU GAC AUG GAC UGA AGG AGU AGA AA-3′) to the 5′ end of the mRNA. This was followed with reverse transcription of the ligated mRNA by using SuperScript III reverse transcriptase and the GeneRacer random primers to create RACE (rapid amplification of cDNA ends)-ready first-strand cDNA with known priming sites at the 5′ end. We amplified the first-strand cDNA using a reverse FIH-1 gene-specific primer (5′-CGGT CAGC ACCA CAGG CTCC TCAT TC-3′) and the GeneRacer 5′ primer (5′-CGA CTG GAG CAC GAG GAC ACT GA-3′) at 94°C for 15 min, with 40 cycles of 94°C for 30 seconds, 60°C for 45 seconds, and 72°C for 45 seconds, followed by 72°C for 10 min. HotStar Taq DNA polymerase (QIAGEN Inc., Valencia, CA) was used for PCR. The PCR product was analyzed on a 2% agarose gel, and three bands were found, ranging from 150 bp to 500 bp. Three microliters of these PCR products was ligated into the pCR4-TOPO vector, and 36 clones were selected and sent for sequencing. Through BLAST, sequencing results were found to match the PubMed sequence.

Transfection and whole-cell extract preparation.

786-O cells and 293 cells (5 × 105) were seeded in a 60-mm dish 1 day before transfection. Transfections were carried out with the Effectene transfection kit (QIAGEN Inc., Valencia, CA). Briefly, 2 μg of the PKC ζ dominant-negative plasmid or the CDP dominant-negative plasmid was resuspended in DNA condensation buffer (150 μl) and 16 μl of Enhancer was added prior to incubation of the samples at room temperature for 5 min. Effectene (40 μl) was then added, and the whole mixture was incubated for another 10 min. Dulbecco modified Eagle medium with 10% fetal bovine serum (350 μl) was then added to the DNA mixtures. Forty-eight hours after transfection, cells were washed three times with phosphate-buffered saline and whole-cell extracts were prepared in accordance with the protocol listed below.

RNA interference for PKC ζ and CDP.

PKC ζ (catalog no. M-003526-02; Dharmacon, Chicago, IL), nontarget control (sc-36869), and the CDP siRNA pool (catalog no. sc-35051) were purchased from Santa Cruz Biotechnology, Inc., Santa Cruz, CA. Transfections were performed according to the instructions of the Transfect I kit (Dharmacon, Chicago, IL). Briefly, for mixture A, 95 μl of serum-free medium was mixed with 5 μl of transfect reagent I for 5 min at room temperature (RT); for mixture B, 10 μl of siRNA pool (10 μM) was mixed with 90 μl of serum-free medium for 5 min at RT. Mixture A was then added to mixture B and incubated for another 20 min at RT. The medium was then replaced with 800 μl of complete medium, added dropwise to the dish. For whole-cell lysate preparation and Western blot analysis, the cells were lysed 72 h after siRNA transfection. For the luciferase assay, 24 h after the siRNA transfection, samples were then transfected with 0.5 μg of FIH-1 promoter constructs using the Effectene transfection kit (described above) and incubated for an additional 48 h. The cells were then lysed with 1× lysis buffer for the luciferase assay.

Preparation of whole-cell extracts.

786-O cells and 293 cells were washed twice with 10 ml of cold phosphate-buffered saline, lysed with ice-cold RIPA lysis buffer (50 mM Tris [pH 7.5], 1% NP-40, 150 mM NaCl, 1 mM Na3VO4, 2 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, leupeptin [10 μg/ml], 0.5% aprotinin, 2 mM pepstatin A), incubated on ice for 30 min, and centrifuged at 14,000 rpm at 4°C for 10 min.

Western blot analysis.

The whole-cell extracts were separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis, and the immunodetection antibodies used were FIH-1 (catalog no. BC-100-428A; Novus Biologicals, Littleton, CO), CDP (catalog no. sc-13024; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), phospho-mitogen-activated protein (phospho-MAP) kinase (catalog no. 07-467; Millipore, Billerica, MA), phosphoserine (catalog no. AB1603; Chemicon, Temecula, CA), and PKC ζ (Santa Cruz Biotechnology). Primary incubations were followed with an appropriate secondary antibody and detected with the ECL enhancer reagent from Amersham Biosciences Corp. (Piscataway, NJ). For immunoprecipitation assays with the phosphoserine antibody, anti-PKC ζ antibody was used for Western blotting.

Luciferase reporter assays.

786-O cells and 293 cells (5 × 104) were seeded in a 12-well plate 1 day before transfection. All transfections were carried out using the Effectene reagent as described above. A total of 1 μg of FIH-1 promoter construct and 0.1 μg of pRL-TK Renilla luciferase vector (Promega, Pittsburgh, PA) were used for each transfection. The pRL-TK Renilla luciferase activity was used as the control for transfection efficiency. In some of the experiments, we equilibrated the luciferase activities by protein concentrations of the cell extracts with respect to the control vectors. Each transfection experiment was performed in triplicate and repeated for a minimum of three times. For cotransfection experiments, cells received 0.3 μg of different deletion constructs of pGL3-FIH-1, 0.1 μg of pRL-TK Renilla luciferase vector, 1 μg of the indicated expression plasmids (PKC ζ DN) or siRNA of CDP/Cut, and carrier DNA. Firefly luciferase and Renilla luciferase assays were performed using the Dual-Luciferase Reporter Assay System (Promega). Approximately 48 h after transfection, cells were washed twice with 1× phosphate-buffered saline and harvested with 200 μl of passive lysis buffer. Cell lysates were cleared by centrifugation, and 5 μl of lysate was added to 100 μl of firefly luciferase substrate. Light units were measured in a LB960 microplate luminometer. Renilla luciferase activities were measured in the same well after addition of 100 μl of the Stop and Glo reagent. The average was calculated among three wells; the ratio between the firefly luciferase and Renilla luciferase was calculated, and the relative change (n-fold) was obtained after comparing the results to the control sample. Data were expressed as the means ± standard deviations of triplicate values.


Double-stranded oligonucleotides generated from the single-stranded oligonucleotide list in Table Table11 were used as electrophoretic mobility shift assay (EMSA) probes. Sequence W is the wild-type FIH-1 promoter region between 114 bp and 92 bp; sequence M is the mutant sequence. The upper-strand (sense) oligonucleotide (30 ng) was 5′ end labeled using polynucleotide kinase with [γ-32P]ATP (Amersham Pharmacia Biotech). After the labeling reaction, a twofold excess of the lower-strand (antisense) oligonucleotide was annealed to the upper strand. Double-stranded DNA probes were then purified from the reaction mixture using a Bio-Gel P-100 column (Bio-Rad). Whole-cell extracts were isolated from 786-O cells and 293T cells. DNA-protein binding was performed in 0.5× Dignam buffer D (20 mM HEPES, pH 7.9, 100 mM KCl, 20% glycerol, 0.2 mM EDTA) supplemented with 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol, 10 mM MgCl2, and 100 μg/ml poly(dI-dC). Binding reactions were initiated with a 30,000-cpm addition of DNA probe in TE buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA) to 5 to 10 μl of whole-cell extracts. Electrophoresis was performed in 4% acrylamide gels, and the gels were vacuum dried and exposed to film for 16 to 48 h. Competition experiments were carried out in the same way as described above except that increasing amounts of double-stranded wild-type oligonucleotide (200-fold) were mixed with 30,000 cpm of M-1 probe (see Table Table1)1) and then added to the binding reaction mixture.

Oligonucleotides used for gel shift assay

ChIP assay.

The chromatin immunoprecipitation (ChIP) assay was performed using the ChIP assay kit from Upstate Biotechnology, Inc. (Lake Placid, NY). Briefly, 7 × 106 to 10 × 106 cells were used for each assay. Protein-DNA cross-linking was carried out by the addition of 1% formaldehyde directly to the cell cultures, followed by incubation at 37°C for 10 min. After the cells were thoroughly washed with ice-cold phosphate-buffered saline, the cells were scraped off and harvested. Cells were lysed with 200 μl of SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris, pH 8.1) supplemented with protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, and 1 μg/ml pepstatin A). Sonication was then performed on ice using a sonicator (Lab-Line Ultra Tip; Lab-Line Instrument. Inc.) preset for 10-s pulses with 10-s intervals. Ten repeated sonication cycles (as previously standardized by us) were applied to achieve chromatin fragmentation in the 200- to 1,000-bp range. Fragmented chromatin was diluted 10-fold in ChIP dilution buffer (0.01% SDS, 1% Triton X-100, 2 mM EDTA, 16.7 mM Tris, pH 8.1, and 150 mM NaCl). Diluted chromatin fragments were precleared by incubation with protein A agarose beads under constant rotation for 2 h at 4°C. For immunoprecipitations, an antibody specific for CDP-1 or the respective immunoglobulin G (IgG) control was used for an overnight incubation with constant rotation at 4°C. The protein-DNA-antibody complex was pulled down by protein A agarose-salmon sperm DNA beads. After thorough and sequential washings with low-salt, high-salt, and LiCl-containing buffers, the resulting immune complex was eluted with 1% SDS and 0.1 M NaHCO3. Formaldehyde cross-links were reversed by adding 5 M NaCl and heating at 65°C for 4 h. DNA fragments were then recovered by ethanol precipitation following proteinase K digestion and phenol-chloroform extraction.

PCR was performed with FIH promoter-specific primers amplifying the CDP binding region (forward, 5′-GTT CTT TAT GTC TCT GGA CCG-3′, and reverse, 5′-CCA CAC ACC CAA CGG AA AC-3′). The 200-bp amplicon was analyzed with 2% agarose gel electrophoresis.


FIH-1 protein was increased in the presence of overexpressing dominant-negative PKC ζ in 786-O cells.

In our previous paper, we have shown that the siRNA or kinase-inactive form of PKC ζ increased the total mRNA level of FIH-1 in renal cancer cells and other cell lines, including 293T cells. Because there was no FIH-1-specific antibody commercially available at that time, we were unable to determine whether the level of FIH-1 protein also increased similarly to its mRNA. Here, we confirmed the increase in FIH-1 protein by Western blotting in 786-O cells after they were transiently transfected with the kinase-inactive form of PKC ζ (Fig. (Fig.1A).1A). As a positive control to illustrate the inhibitory effect of PKC ζ, we observed a 50% decrease of phospho-MAP kinase in this condition (Fig. (Fig.1B).1B). This result not only confirms our previous findings of PKC ζ as an important transcriptional mediator of FIH-1 expression in cancer cells but also paves the way towards understanding how PKC ζ exerts its control. Previously, we observed a dose-dependent increase in FIH-1 full-length promoter activity with a dominant-negative form of PKC ζ, suggesting PKC ζ-mediated transcriptional control of the FIH-1 gene. As we did not observe any change in FIH-1 mRNA stability in the presence of the dominant-negative PKC ζ under our experimental setup (data not shown), we have focused our subsequent studies on PKC ζ-mediated transcriptional regulation of FIH-1, beginning with the identification of the FIH-1 transcriptional start site.

Determining the transcriptional start site(s) of the FIH-1 gene.

The NCBI website indicated a potential FIH-1 transcriptional start site located 20 bp upstream of the first FIH-1 translational start codon (ATG) (gene ID, 55662). In order to experimentally confirm the transcriptional start site, we examined the heterogeneous FIH-1 transcriptional start sites by using RNA ligase-mediated oligonucleotide-capping rapid amplification of cDNA ends (RML-RACE from Invitrogen Life Technologies). Our results indicated a potential FIH-1 transcription start site located 35 bp, not 20 bp, upstream of the ATG start codon (data not shown). According to our results, this is the first report of an FIH-1 transcription start site.

Potential transcriptional factor(s) regulated by PKC ζ for FIH-1 transcription.

Previously, we showed that the PKC ζ isoform inhibits the activity of the full-length FIH-1 promoter (6.9 kb in pGL3-basic vector). In this study, we tried to elucidate the molecular mechanism of PKC ζ-mediated FIH-1 transcriptional regulation. Different deletion fragments of the FIH-1 proximal promoter region were cloned into the pGL3-basic vector containing the luciferase reporter gene (e.g., 2.7 kb, 480 bp, 220 bp, 180 bp, 114 bp, and 92 bp) (Fig. (Fig.2A).2A). The basal activities of these deletion constructs are shown in Fig. Fig.2B2B.

FIG. 2.
Analysis of FIH-1 promoter activity. (A) Schematic diagram of a 6.95-kb fragment of genomic DNA containing the FIH-1 promoter cloned into the pGL3-basic reporter construct (pGL3Prom). A series of deletion constructs are also shown. The position of the ...

When cotransfected with the 2.7-kb or 480-bp construct of FIH-1, an increased FIH-1 transcriptional activity was observed when the PKC ζ dominant-negative plasmid (kinase-inactive PKC ζ) was overexpressed, compared to the empty control vector (Fig. (Fig.2C2C).

We further generated a 220-bp fragment of the FIH-1 promoter. The 5′ end of the 220-bp fragment was identical to the 480-bp fragment; however, the 3′ end stops immediately after the transcription start site. Interestingly, the 220-bp fragment showed a similar increase in promoter activity in the presence of dominant-negative PKC ζ. These results suggest that a cis element responsible for the PKC ζ-mediated transcriptional regulation of FIH-1 exists within the 220-bp region, upstream of the transcription start site (Fig. (Fig.2D2D).

We engineered further deletions within the 220-bp region of the FIH-1, yielding 180-bp, 114-bp, and 92-bp fragments. FIH-1 promoter activities with both the 180- and 114-bp fragments were increased when cells were cotransfected with the PKC ζ dominant-negative plasmid. Importantly, we did not observe any PKC ζ regulation with the 92-bp FIH-1 promoter construct (Fig. (Fig.2D).2D). Therefore, we concluded that the region between 114 bp and 92 bp may be important for regulation of FIH-1 transcription by PKC ζ. According to the Genomatix website (, this region contains a consensus binding site for Cut-like homeodomain proteins (CDP/Cut). Our next series of experiments were to determine the involvement of Cut in PKC ζ-mediated FIH-1 transcription.

CDP/Cut is involved in the transcriptional regulation of FIH-1.

CDP (CCAAT-displacement protein) is the human homologue of the Drosophila Cut protein (32). CDP/Cux/Cut proteins are an evolutionarily conserved family of proteins containing several DNA binding domains. CDP was found to be the DNA binding protein for several genes including histone nuclear factor D and CXCL1 (10, 33, 39). These proteins generally act as transcriptional repressors, although some data suggest their involvement in transcriptional activation. The human CUTL1 gene was mapped to 7q22, a chromosomal region that is frequently rearranged in various cancers. The function of CDP/Cut is regulated by several posttranslational modification events including phosphorylation, dephosphorylation, and acetylation (reviewed by Nepveu [32]). Previous studies showed that the transcriptional activity of Cut proteins was modulated by PKC (4).

Further indications that CDP/Cut plays an inhibitory role in FIH-1 transcription came from the comparison study of the basal FIH-1 promoter activity of the 114-bp and 92-bp deletion constructs. The promoter activity of the 114-bp and 92-bp constructs was measured in both 293T cells and 786-O cells. The transcriptional activity was significantly higher with the 92-bp construct than with the 114-bp construct in both cell lines, suggesting a potential transcriptional inhibitory function within the 92- and 114-bp region (Fig. (Fig.3A).3A). As mentioned above, this 22-bp region contains binding domains for CDP/Cut proteins. Therefore, it was reasonable to examine the involvement of the Cut protein in FIH-1 transcription.

FIG. 3.
Deletion mutations of FIH-1 promoter and the role of CDP/Cut in FIH-1 gene expression. (A) The basal FIH-1 promoter activity of 114-bp and 92-bp deletion constructs in 293T and 786-O cells. The luciferase activity (units/μg protein) of FIH-1 promoter ...

Inhibition of Cut leads to the increase in FIH-1 promoter activity.

We used Cut siRNA to deplete endogenous Cut protein from both 293T and 786-O cells. For the 114-bp FIH-1 promoter construct, an increase in promoter activity was observed in both 293T and 786-O cells when cotransfected with Cut siRNA compared to the nontargeted control. Interestingly, no significant difference in promoter activity was observed when the 92-bp FIH-1 promoter construct was transfected (Fig. (Fig.3B).3B). This result suggests that the Cut protein could repress FIH-1 transcription.

The FIH-1 protein level is increased by inhibition of CDP/Cut.

An increase in mRNA does not always lead to an increase in protein levels. In order to determine whether FIH-1 protein also increased due to the inhibition of CDP/Cut, we transfected 786-O and 293T cells with CDP/Cut siRNA. For the control experiments, cells were transfected with nontarget siRNA oligonucleotide. FIH-1 protein levels were determined by Western blot assays from the cell lysates (Fig. (Fig.3C).3C). An increase in FIH-1 protein was observed when the CDP/Cut expression was inhibited by siRNA. These data led to the question of whether CDP/Cut was a transcriptional regulator of the FIH-1 promoter.

CDP/Cut regulates FIH-1 transcription.

In order to verify the binding of CDP/Cut to the FIH-1 promoter, EMSAs were performed. We constructed a probe of the CDP binding region within the FIH-1 promoter (−114 to −92 bp, wild-type probe) as well as another probe of the same length and similar sequence except with six out of the nine core nucleotides from the CDP binding sequence within the FIH-1 promoter being mutated (mutated probe) (Table (Table1).1). A specific DNA-protein complex was detected with 786-O total cell extracts using the wild-type probe (Fig. (Fig.4,4, lane 2) but not with the mutant probe (Fig. (Fig.4,4, lane 9). The protein-DNA complex identified with the radiolabeled probe was specifically competed out with increasing concentrations of excess unlabeled wild-type probe (Fig. (Fig.4,4, lanes 3 to 5) but not with excess unlabeled mutant probe (Fig. (Fig.4,4, lanes 6 to 8), suggesting the specificity of CDP binding to the FIH-1 promoter region.

FIG. 4.
The CDP/Cut binding protein present in renal cancer cells recognizes specific sequences located between −114 and −92 of the FIH-1 promoter. The binding of 786-O cell nuclear extracts to 32P-labeled wild-type (W) and 32P-labeled mutant ...

PKC ζ regulates the binding of CDP/Cut to FIH-1 promoter DNA.

To address any role of PKC ζ in the binding of CDP to the FIH-1 promoter, we initially examined whether CDP phosphorylation is PKC ζ dependent in RCC cells, and later we also evaluated the consequences of this phosphorylation. We immunoprecipitated the cell extracts that overexpressed PKC ζ DN with anti-CDP antibodies and immunoblotted with antiserine antibodies. Figure Figure5A5A showed reduced phosphorylation of CDP in the presence of PKC ζ DN without changing any endogenous CDP expression. Then, we performed EMSAs with the wild-type probe and whole-cell extracts obtained from 293T cells transfected with PKC ζ siRNA. A decrease in DNA-protein binding was observed when PKC ζ was depleted with siRNA compared to the nontargeted control siRNA (Fig. (Fig.5B).5B). The efficiency of the siRNA was shown in Fig. Fig.5C.5C. These data suggest that PKC ζ increases CDP phosphorylation and binding to the FIH-1 promoter and most likely leads to inhibition of FIH-1 promoter activity.

FIG. 5.
PKC ζ influences CDP/Cut activity on the FIH-1 promoter. (A) Phosphorylation of CDP by PKC ζ in 786-O cells. Whole-cell lysates obtained from 786-O cells transfected with the PKC ζ DN and control plasmids were subjected to immunoprecipitation ...

ChIP assays to detect the binding of CDP/Cut to the FIH promoter.

To detect the in vivo binding of CDP to the FIH promoter region in 786-O cells, the ChIP assay was performed. Protein-cross-linked chromatin fragments were isolated from 786-O cells transfected with either the PKC ζ DN-overexpressing plasmid or the control plasmid. Immunoprecipitation of the chromatin fragments was then carried out with the CDP antibody. In control experiments, mouse IgG was used to immunoprecipitate 786-O cell lysates transfected with PKC ζ DN or control plasmids. After reverse cross-linking, DNA fragments were isolated and the CDP binding region on the FIH promoter was amplified by PCR with specific FIH promoter primers as described in detail in Materials and Methods. A prominent amplified product was detected when immunoprecipitation was carried out with the CDP antibody on cell lysates transfected with the control vector, suggesting specific binding of CDP at the FIH-1 promoter (Fig. (Fig.5D).5D). This result also confirms our previously described EMSA result in vivo. Interestingly, overexpression of the dominant-negative mutant of PKC ζ inhibited CDP binding to the FIH-1 promoter, therefore underscoring the importance of PKC ζ function in regulating FIH-1 promoter activity via CDP.

Defining the site(s) of CDP/Cut that modulates FIH-1 expression.

In order to define the domain or site(s) of CDP that regulates FIH-1 expression in RCC, we overexpressed wild-type and mutant CDP in RCC cells. We then evaluated the phosphorylation status of CDP and its mutant by immunoprecipitating CDP from RCC cell lysates overexpressing the wild-type CDP (CDP-WT) or mutant CDP (CDP-6 mutants; pxm/Cux/6-mutant plasmid contains Ser402, Thr415, Ser789, Thr804, Ser972, and Ser987 and also Ser987 alone [pxm/Cux/CR#-HD/S987A]; these residues were replaced with codons for Ala [4]) and immunoblotting with antiphosphoserine antibodies. Figure Figure66 showed that the phosphorylation level of CDP was dramatically decreased in the presence of the CDP-6 mutant compared to wild-type CDP (Fig. (Fig.6,6, top panel).

FIG. 6.
Phosphorylation site(s) of CDP that modulates FIH-1 expression. (Upper panel) Whole-cell lysates obtained from 786-O cells transfected with the CDP-6 mutants as well as CDP S987A mutant expression vectors and control plasmids were subjected to immunoprecipitation ...

We then questioned whether the CDP-6 mutation modulated FIH-1 expression. The level of FIH protein was measured in whole-cell lysates transfected with either wild-type CDP or the CDP-6 mutant. As shown in Fig. Fig.6,6, an increased FIH-1 protein level was observed in cells transfected with the CDP-6 mutant construct compared to that of the control vector and the level of FIH-1 protein was also decreased in CDP wild-type transfectants (lower panel). These data are in accordance with our previous finding that FIH-1 protein increased in the presence of CDP siRNA. Furthermore, we examined whether a CDP mutant containing S987A yielded an effect similar to that of the CDP-6 mutant. Interestingly, a significant decrease in CDP phosphorylation was observed in transfectants containing a CDP S987A mutant compared to those transfected with a control vector (Fig. (Fig.6,6, top panel). This indicated that this serine is one of the modification sites. Similarly, we observed increased FIH-1 expression in cell extracts from CDP S987A transfectants compared to wild-type CDP transfectants. Overall, our results suggest potential CDP sites that modify FIH-1 expression in RCC cells. As we reported earlier, CDP serine phosphorylation depends upon PKC ζ and this controls its interaction with the FIH-1 promoter and its subsequent expression. Taken together, these data suggest a unique control of gene expression by a PKC isoform on the transcription factor CDP, which leads to the modulation of FIH-1 and the potential control of HIF-dependent pathways and, ultimately, tumor angiogenesis.


In this communication, we addressed the mechanism of PKC ζ-mediated FIH-1 transcriptional control in RCC. With various deletion constructs of the FIH-1 promoter, we were able to identify a specific region of the FIH-1 promoter (−114 to −92 bp) as a regulatory cis element. This region contains the consensus sequence for CDP/Cut/Cux. CDP/Cut is a homeobox protein that usually represses gene transcription. In our FIH-1 gene transcriptional studies, we observed the involvement of PKC ζ, which might facilitate the binding of CDP to the FIH-1 promoter and thereby repress its transcription. We monitored FIH-1 promoter activity and its endogenous protein level in RCC after eliminating CDP function either by siRNA or by a dominant-negative approach. We observed an up-regulation of FIH-1 in RCC by inhibiting the function of CDP/Cut, suggesting its involvement in this pathway. The confirmation of the connection of CDP/Cut in FIH-1 transcription with its regulation by PKC ζ came from our EMSA and ChIP studies. These results indicated that active PKC ζ can increase CDP binding to the FIH-1 promoter region and lead to an increased inhibition of CDP on FIH-1 transcription. Moreover, we also observed specific serine residues, most likely CDP S987A, that are important for FIH-1 regulation through PKC ζ.

Even under normoxia, the up-regulation of HIF target genes is a common characteristic in cancer cells (3, 9, 42, 43). Clear-cell renal carcinoma is a good example where the vascular endothelial growth factor A gene, a target gene of HIF, is expressed at high levels in the presence of normal oxygen concentrations (6). Activation of oncogenes along with the inactivation of tumor suppressor genes might explain this increased HIF activity under normoxia. Furthermore, in our present study, we have shown an epigenetic regulation of FIH-1 in RCC that may account for this increased HIF activity. Therefore, it highlights an important regulatory mechanism of FIH-1 that leads to high HIF activity in cancer cells and could signify an important target for potential therapies.


We sincerely thank A. Nepveu for providing us with the reagents as well as helpful discussions during this study.

This work is partly supported by NIH grants CA78383 and HL70567 and also a grant from the American Cancer Society to D.M.


[down-pointing small open triangle]Published ahead of print on 6 August 2007.


1. Acker, T., and H. Acker. 2004. Cellular oxygen sensing need in CNS function: physiological and pathological implications. J. Exp. Biol. 207:3171-3188. [PubMed]
2. Arany, Z., L. E. Huang, R. Eckner, S. Bhattacharya, C. Jiang, M. A. Goldberg, H. F. Bunn, and D. M. Livingston. 1996. An essential role for p300/CBP in the cellular response to hypoxia. Proc. Natl. Acad. Sci. USA 93:12969-12973. [PubMed]
3. Axelson, H., E. Fredlund, M. Ovenberger, G. Landberg, and S. Pahlman. 2005. Hypoxia-induced dedifferentiation of tumor cells—a mechanism behind heterogeneity and aggressiveness of solid tumors. Semin. Cell Dev. Biol. 16:554-563. [PubMed]
4. Coqueret, O., G. Berube, and A. Nepveu. 1996. DNA binding by cut homeodomain proteins is down-modulated by protein kinase C. J. Biol. Chem. 271:24862-24868. [PubMed]
5. Dann, C. E., III, R. K. Bruick, and J. Deisenhofer. 2002. Structure of factor-inhibiting hypoxia-inducible factor 1: an asparaginyl hydroxylase involved in the hypoxic response pathway. Proc. Natl. Acad. Sci. USA 99:15351-15356. [PubMed]
6. Datta, K., J. Li, R. Bhattacharya, L. Gasparian, E. Wang, and D. Mukhopadhyay. 2004. Protein kinase C zeta transactivates hypoxia-inducible factor alpha by promoting its association with p300 in renal cancer. Cancer Res. 64:456-462. [PubMed]
7. Elkins, J. M., K. S. Hewitson, L. A. McNeill, J. F. Seibel, I. Schlemminger, C. W. Pugh, P. J. Ratcliffe, and C. J. Schofield. 2003. Structure of factor-inhibiting hypoxia-inducible factor (HIF) reveals mechanism of oxidative modification of HIF-1 alpha. J. Biol. Chem. 278:1802-1806. [PubMed]
8. Freedman, S. J., Z. Y. Sun, F. Poy, A. L. Kung, D. M. Livingston, G. Wagner, and M. J. Eck. 2002. Structural basis for recruitment of CBP/p300 by hypoxia-inducible factor-1 alpha. Proc. Natl. Acad. Sci. USA 99:5367-5372. [PubMed]
9. Greijer, A. E., P. van der Groep, D. Kemming, A. Shvarts, G. L. Semenza, G. A. Meijer, M. A. van de Wiel, J. A. Belien, P. J. van Diest, and E. van der Wall. 2005. Up-regulation of gene expression by hypoxia is mediated predominantly by hypoxia-inducible factor 1 (HIF-1). J. Pathol. 206:291-304. [PubMed]
10. Gupta, S., M. X. Luong, S. A. Bleuming, A. Miele, M. Luong, D. Young, E. S. Knudsen, A. J. Van Wijnen, J. L. Stein, and G. S. Stein. 2003. Tumor suppressor pRB functions as a co-repressor of the CCAAT displacement protein (CDP/cut) to regulate cell cycle controlled histone H4 transcription. J. Cell. Physiol. 196:541-556. [PubMed]
11. Hewitson, K. S., L. A. McNeill, M. V. Riordan, Y. M. Tian, A. N. Bullock, R. W. Welford, J. M. Elkins, N. J. Oldham, S. Bhattacharya, J. M. Gleadle, P. J. Ratcliffe, C. W. Pugh, and C. J. Schofield. 2002. Hypoxia-inducible factor (HIF) asparagine hydroxylase is identical to factor inhibiting HIF (FIH) and is related to the cupin structural family. J. Biol. Chem. 277:26351-26355. [PubMed]
12. Hirota, K., and G. L. Semenza. 2005. Regulation of hypoxia-inducible factor 1 by prolyl and asparaginyl hydroxylases. Biochem. Biophys. Res. Commun. 338:610-616. [PubMed]
13. Ivan, M., K. Kondo, H. Yang, W. Kim, J. Valiando, M. Ohh, A. Salic, J. M. Asara, W. S. Lane, and W. G. Kaelin, Jr. 2001. HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292:464-468. [PubMed]
14. Kaelin, W. G. 2005. Proline hydroxylation and gene expression. Annu. Rev. Biochem. 74:115-128. [PubMed]
15. Kallio, P. J., K. Okamoto, S. O'Brien, P. Carrero, Y. Makino, H. Tanaka, and L. Poellinger. 1998. Signal transduction in hypoxic cells: inducible nuclear translocation and recruitment of the CBP/p300 coactivator by the hypoxia-inducible factor-1alpha. EMBO J. 17:6573-6586. [PubMed]
16. Kim, W., and W. G. Kaelin, Jr. 2003. The von Hippel-Lindau tumor suppressor protein: new insights into oxygen sensing and cancer. Curr. Opin. Genet. Dev. 13:55-60. [PubMed]
17. Koivunen, P., M. Hirsila, V. Gunzler, K. I. Kivirikko, and J. Myllyharju. 2004. Catalytic properties of the asparaginyl hydroxylase (FIH) in the oxygen sensing pathway are distinct from those of its prolyl 4-hydroxylases. J. Biol. Chem. 279:9899-9904. [PubMed]
18. Kondo, K., J. Klco, E. Nakamura, M. Lechpammer, and W. G. Kaelin, Jr. 2002. Inhibition of HIF is necessary for tumor suppression by the von Hippel-Lindau protein. Cancer Cell 1:237-246. [PubMed]
19. Lahiri, S., A. Roy, S. M. Baby, T. Hoshi, G. L. Semenza, and N. R. Prabhakar. 2006. Oxygen sensing in the body. Prog. Biophys. Mol. Biol. 91:249-286. [PubMed]
20. Lancaster, D. E., L. A. McNeill, M. A. McDonough, R. T. Aplin, K. S. Hewitson, C. W. Pugh, P. J. Ratcliffe, and C. J. Schofield. 2004. Disruption of dimerization and substrate phosphorylation inhibit factor inhibiting hypoxia-inducible factor (FIH) activity. Biochem. J. 383:429-437. [PubMed]
21. Lando, D., D. J. Peet, J. J. Gorman, D. A. Whelan, M. L. Whitelaw, and R. K. Bruick. 2002. FIH-1 is an asparaginyl hydroxylase enzyme that regulates the transcriptional activity of hypoxia-inducible factor. Genes Dev. 16:1466-1471. [PubMed]
22. Lee, C., S. J. Kim, D. G. Jeong, S. M. Lee, and S. E. Ryu. 2003. Structure of human FIH-1 reveals a unique active site pocket and interaction sites for HIF-1 and von Hippel-Lindau. J. Biol. Chem. 278:7558-7563. [PubMed]
23. Linke, S., C. Stojkoski, R. J. Kewley, G. W. Booker, M. L. Whitelaw, and D. J. Peet. 2004. Substrate requirements of the oxygen-sensing asparaginyl hydroxylase factor-inhibiting hypoxia-inducible factor. J. Biol. Chem. 279:14391-14397. [PubMed]
24. Mahon, P. C., K. Hirota, and G. L. Semenza. 2001. FIH-1: a novel protein that interacts with HIF-1alpha and VHL to mediate repression of HIF-1 transcriptional activity. Genes Dev. 15:2675-2686. [PubMed]
25. Maxwell, P. H., M. S. Wiesener, G. W. Chang, S. C. Clifford, E. C. Vaux, M. E. Cockman, C. C. Wykoff, C. W. Pugh, E. R. Maher, and P. J. Ratcliffe. 1999. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399:271-275. [PubMed]
26. Maynard, M. A., and M. Ohh. 2004. Von Hippel-Lindau tumor suppressor protein and hypoxia-inducible factor in kidney cancer. Am. J. Nephrol. 24:1-13. [PubMed]
27. Mazure, N. M., M. C. Brahimi-Horn, M. A. Berta, E. Benizri, R. L. Bilton, F. Dayan, A. Ginouves, E. Berra, and J. Pouyssegur. 2004. HIF-1: master and commander of the hypoxic world. A pharmacological approach to its regulation by siRNAs. Biochem. Pharmacol. 68:971-980. [PubMed]
28. McNeill, L. A., K. S. Hewitson, T. D. Claridge, J. F. Seibel, L. E. Horsfall, and C. J. Schofield. 2002. Hypoxia-inducible factor asparaginyl hydroxylase (FIH-1) catalyses hydroxylation at the beta-carbon of asparagine-803. Biochem. J. 367:571-575. [PubMed]
29. Metzen, E., U. Berchner-Pfannschmidt, P. Stengel, J. H. Marxsen, I. Stolze, M. Klinger, W. Q. Huang, C. Wotzlaw, T. Hellwig-Burgel, W. Jelkmann, H. Acker, and J. Fandrey. 2003. Intracellular localisation of human HIF-1 alpha hydroxylases: implications for oxygen sensing. J. Cell Sci. 116:1319-1326. [PubMed]
30. Metzen, E., and P. J. Ratcliffe. 2004. HIF hydroxylation and cellular oxygen sensing. Biol. Chem. 385:223-230. [PubMed]
31. Morris, M. R., E. Maina, N. V. Morgan, D. Gentle, D. Astuti, H. Moch, T. Kishida, M. Yao, P. Schraml, F. M. Richards, F. Latif, and E. R. Maher. 2004. Molecular genetic analysis of FIH-1, FH, and SDHB candidate tumour suppressor genes in renal cell carcinoma. J. Clin. Pathol. 57:706-711. [PMC free article] [PubMed]
32. Nepveu, A. 2001. Role of the multifunctional CDP/Cut/Cux homeodomain transcription factor in regulating differentiation, cell growth and development. Gene 270:1-15. [PubMed]
33. Nirodi, C., J. Hart, P. Dhawan, N. S. Moon, A. Nepveu, and A. Richmond. 2001. The role of CDP in the negative regulation of CXCL1 gene expression. J. Biol. Chem. 276:26122-26131. [PMC free article] [PubMed]
34. Ohh, M., C. W. Park, M. Ivan, M. A. Hoffman, T. Y. Kim, L. E. Huang, N. Pavletich, V. Chau, and W. G. Kaelin. 2000. Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domain of the von Hippel-Lindau protein. Nat. Cell Biol. 2:423-427. [PubMed]
35. Safran, M., and W. G. Kaelin, Jr. 2003. HIF hydroxylation and the mammalian oxygen-sensing pathway. J. Clin. Investig. 111:779-783. [PMC free article] [PubMed]
36. Semenza, G. L. 2004. Hydroxylation of HIF-1: oxygen sensing at the molecular level. Physiology (Bethesda) 19:176-182. [PubMed]
37. Soilleux, E. J., H. Turley, Y. M. Tian, C. W. Pugh, K. C. Gatter, and A. L. Harris. 2005. Use of novel monoclonal antibodies to determine the expression and distribution of the hypoxia regulatory factors PHD-1, PHD-2, PHD-3 and FIH in normal and neoplastic human tissues. Histopathology 47:602-610. [PubMed]
38. Stolze, I. P., Y. M. Tian, R. J. Appelhoff, H. Turley, C. C. Wykoff, J. M. Gleadle, and P. J. Ratcliffe. 2004. Genetic analysis of the role of the asparaginyl hydroxylase factor inhibiting hypoxia-inducible factor (HIF) in regulating HIF transcriptional target genes. J. Biol. Chem. 279:42719-42725. [PubMed]
39. van Wijnen, A. J., M. F. van Gurp, M. C. de Ridder, C. Tufarelli, T. J. Last, M. Birnbaum, P. S. Vaughan, A. Giordano, W. Krek, E. J. Neufeld, J. L. Stein, and G. S. Stein. 1996. CDP/cut is the DNA-binding subunit of histone gene transcription factor HiNF-D: a mechanism for gene regulation at the G1/S phase cell cycle transition point independent of transcription factor E2F. Proc. Natl. Acad. Sci. USA 93:11516-11521. [PubMed]
40. Wang, G. L., B. H. Jiang, E. A. Rue, and G. L. Semenza. 1995. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2 tension. Proc. Natl. Acad. Sci. USA 92:5510-5514. [PubMed]
41. Wang, G. L., and G. L. Semenza. 1995. Purification and characterization of hypoxia-inducible factor 1. J. Biol. Chem. 270:1230-1237. [PubMed]
42. Wiesener, M. S., P. M. Munchenhagen, I. Berger, N. V. Morgan, J. Roigas, A. Schwiertz, J. S. Jurgensen, G. Gruber, P. H. Maxwell, S. A. Loning, U. Frei, E. R. Maher, H. J. Grone, and K. U. Eckardt. 2001. Constitutive activation of hypoxia-inducible genes related to overexpression of hypoxia-inducible factor-1alpha in clear cell renal carcinomas. Cancer Res. 61:5215-5222. [PubMed]
43. Yeo, E. J., Y. S. Chun, and J. W. Park. 2004. New anticancer strategies targeting HIF-1. Biochem. Pharmacol. 68:1061-1069. [PubMed]
44. Yu, F., S. B. White, Q. Zhao, and F. S. Lee. 2001. HIF-1alpha binding to VHL is regulated by stimulus-sensitive proline hydroxylation. Proc. Natl. Acad. Sci. USA 98:9630-9635. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)