FOXD4a and FOXD4b, two new winged helix transcription factors, are expressed in human leukemia cell lines

Gene 294 (2002) 131–140 www.elsevier.com/locate/gene

FOXD4a and FOXD4b, two new winged helix transcription factors, are expressed in human leukemia cell lines Bettina S. Freyaldenhoven*, Cora Fried, Klaus Wielckens Institute of Clinical Chemistry, University of Cologne, Joseph-Stelzmann-Strasse 9, 50924 Cologne, Germany Received 12 February 2002; received in revised form 24 April 2002; accepted 14 May 2002 Received by R. Di Lauro

Abstract Winged helix factors are important regulators of embryonal development and tissue differentiation. They are also involved in translocations found in acute leukemias and solid tumors. We have detected transcripts from five known and four novel winged helix genes in leukemia cell lines and CD34 1 blood progenitor cells by reverse trancription–polymerase chain reaction with degenerate primers on the highly conserved DNA binding domain. The genomic clones coding for two new winged helix proteins, FOXD4a and FOXD4b were isolated by high-stringency hybridization of a human phage library. FOXD4a and FOXD4b are encoded by a 1319 and 1250 bp single exon coding for a winged helix DNA binding domain, an amino-terminal acidic region and a carboxy-terminal proline- and alanine-rich region which correspond to putative transcriptional regulatory motifs. TATA box, CCAAT box, and transcription factor binding motifs have been identified in the 5 0 region of the genes. In addition, foxD4a and foxD4b cDNA has been isolated from NB-4 mRNA. The fox genes are transcribed in a tissue-restricted pattern in adult and fetal human tissues. FoxD4a and foxD4b mRNA was expressed in the leukemia cell lines KG-1, Kasumi, NB-4, HL-60, U937, THP-1, HEL, U266, Jurkat, and Raji. It has already been shown that winged helix factors are also involved in carcinogenesis. Based upon these studies, our results suggest that FOXD4a and FOXD4b may play a role in leukemogenesis. q 2002 Elsevier Science B.V. All rights reserved. Keywords: FOX transcription factor; Gene structure; Leukemia

1. Introduction In hematopoiesis, mature blood cells of different lineages originate from pluripotent blood progenitor cells. Differentiation programs leading to cell maturation are commonly regulated by growth factors and transcription factors (Metcalf, 1989; Orkin, 1995). Transcription factors control activity of certain groups of lineage-specific genes which, in return, account for the phenotype of certain cell populations. In nearly all classes of transcription factors members were identified which are involved in normal hematopoietic differentiation or in leukemias (Nichols and Nimer, 1992). Winged helix factors constitute a family of transcription

Abbreviations: aa, amino acid(s); GAPDH, glyceraldehyde-3-phosphate dehydrogenase; FKHR, forkhead related; FREAC, forkhead-related activator; HNF, hepatocyte nuclear factor; HWH-1, human winged helix factor-1; .nt, nucleotide(s); PCR, polymerase chain reaction; poly(A) 1, polyadenylated RNA; RT, reverse trancription * Corresponding author. Tel.: 149-221-478-4454; fax: 149-221-4785273. E-mail address: [email protected] (B.S. Freyaldenhoven).

factors conserved from yeast to humans. In metazoa they are important regulators of embryonal development and tissue differentiation (Lai et al., 1993; Costa, 1994). Prototypical members are the Drosophila protein forkhead and the rat hepatocyte nuclear factor 3 (HNF-3a, b, and g) proteins (Weigel et al., 1989; Lai et al., 1990, 1991). All members share a highly conserved DNA binding domain of 101 amino acids referred to as winged helix based upon its crystallographic structure (Clark et al., 1993). Most of the targeted mutagenesis of winged helix genes performed in mice has demonstrated so far a crucial function for embryonal development in vertebrates (Ang and Rossant, 1994; Weinstein et al., 1994; Xuan et al., 1995; Hatini et al., 1996). In addition, it was shown that mutation in the whn winged helix gene of mice is responsible for interference with normal hair growth and thymus development in nude mice and rats (Nehls et al., 1994). Winged helix factors are also involved in aging processes. The winged helix factor Daf-16 can double the life span of Caenorhabditis elegans (Lin et al., 1997). There is accumulating evidence that winged helix factors are involved not only in the control of embryonal develop-

0378-1119/02/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(02)00702-3

132

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

ment and tissue-specific gene expression, but also contribute to oncogenesis (Vogt et al., 1997). The retroviral oncogene v-qin, a member of the winged helix family of transcription factors isolated from Avian Sarcoma Virus 31, induces fibrosarcoma in chickens and transforms chicken embryo fibroblasts in vitro (Li and Vogt, 1993). In alveolar rhabdomyosarcoma, a t(2;13)(q35q14) translocation fuses part of the PAX3 gene, and a t(1;13)(p36q14) translocation part of the PAX7 gene, with the winged helix factor FKHR. In the chimeric proteins the activation domains of PAX proteins are replaced by the carboxy-terminal part of FKHR (Galili et al., 1993; Shapiro et al., 1993). Furthermore, winged helix factors have also been found as fusion partners in chimeric proteins isolated from acute lymphocytic and myeloid leukemias. In a cell line established from a child with acute lymphocytic leukemia, a t(X;11)(q13q23) translocation creates a fusion of the MLL (ALL1) gene and the gene coding for the winged helix protein AFX-1 (Parry et al., 1994). Recently another winged helix gene, AF6q21, was identified as fusion partner of the MLL gene in a t(6;11)(q21q23) translocation from a patient with acute myeloblastic leukemia (Hillan et al., 1997). In addition we have shown in previous studies that overexpression of normal winged helix proteins can induce aberrant cell proliferation (Freyaldenhoven et al., 1997a). Based upon these data we were interested in learning whether winged helix factors are also involved in differentiation processes of normal human hematopoiesis and leukemia. 2. Materials and methods 2.1. Tissue culture and isolation of primitive hematopoietic precursors Cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and the Department of Human and Animal Cell Cultures (DSMZ, Braunschweig, Germany), except for FH109 cells, which were a gift from Dr. Michael Lu¨ bbert (University of Freiburg, Germany).The cell lines were cultured in RPMI 1640 medium supplemented with 10% FCS, 100 IU penicillin, and 100 mg/ml streptomycin (Gibco BRL, Karlsruhe, Germany). RNA from CD34 1/HLA-DR 2 cells was a gift from Dr. Reinhard Henschler (University of Freiburg, Germany). The cells were isolated from normal volunteer peripheral blood as previously described (Brandt et al., 1990) using two-color fluorescence-activated flow cytometric cell sorting. 2.2. Isolation of winged helix factor transcripts by polymerase chain reaction Partially degenerate oligonucleotides were synthesized to two highly conserved regions in the winged helix DNA binding domain, ALITMAIL for the 5 0 primer and MFDNGSFL for the 3 0 primer. The actual primer sequences which were

flanked by the restriction sites EcoRI (CGGAATTC) and XbaI (GCTCTAGA) were: 5 0 primer-GCTCTAGAC(G/A/ T)CTCAT(C/T)(A/G)(C/T)(C/G/T)ATGGC(C/G/T)ATCC, 216-fold degeneracy; 3 0 primer-CGGAATTCAG(G/A)AA (G/A)CT(G/C)CC(G/A)TT(G/C)TCGAAC, 32-fold degeneracy. These sequences were chosen based on the codon usage at these positions for HNF-3a and -b, FREAC 1-7, CWH 1-3, MFH-1, Genesis, HFH-2, and FKH 5-3. First, 4 mg of total RNA were treated with 3 ml RNasefree DNase for 1 h at 37 8C in a final volume of 10 ml. Then 1 ml N6 random primers (200 ng) were added, annealed for 10 min at 70 8C, and cooled down on ice. The cDNA synthesis was carried out in a final volume of 20 ml, containing 1£ reverse transcriptase buffer, 0.5 mM deoxynucleotides, 10 mM dithiothreitol, 1 ml RNasin (Promega Biotech, Madison, WI) and 200 units of Moloney sarcoma virus reverse transcriptase (Pharmacia, Piscataway, NJ). After 1 h at 42 8C, the reaction was stopped by heating for 5 min at 94 8C. Thereafter 2.5 units of Taq polymerase (Perkin-Elmer/ Cetus, Norwalk, CT), 1£ polymerase buffer, 0.1 mM deoxynucleotides and 30 pmol of each primer were added to 10 ml cDNA synthesis product in a final volume of 50 ml. The target was then amplified for 30 cycles at 94 8C for 1 min, 55 8C for 1 min, and 72 8C for 1 min. 2.3. Cloning The amplified reaction was electrophoresed on a 1.5% agarose gel and the 210 bp fragments were excised. The isolated fragments were ligated into the pBluescript SK 2 vector (Stratagene, LaJolla, CA) by the sites EcoRI and XbaI. The primer sequences were flanked by these restriction sites. For each cell line 20 clones were sequenced by automated fluorescence sequencing. Sequencing of these PCR fragments revealed four novel winged helix domains and five known winged helix domains. The novel winged helix factors were called Myeloid factor-a, -b, -g, -d. The GenBank accession numbers for the clones are AF343004, AF343005, AF343006, and AF343007. The sequences of Myeloid factor-a and Myeloid factor-g were submitted to Dr. D. Martinez at http://www.biology.pomona.edu/ fox.html so that they could be assigned a standard name according the unified nomenclature for winged helix proteins. The assigned names are FOXD4a for Myeloid factor-a and FOXD4b for Myeloid factor-g. The sequences of Myeloid factor-b and Myeloid factor-d were too short to be designated a fox name. The [a- 32P]dCTP random primer-labeled transcripts of foxD4a and foxD4b were used to screen for the full length genes in a human genomic EMBL3/SP6 phage library (Clontech, Palo Alto, CA). These probes contained the region encoding the conserved winged helix DNA binding domain. Hybridization was performed in 50% formamide, 5£ SSC, 20 mM Na2HPO4/NaH2PO4 (pH 7.6), 7% SDS, 1% polyethylene glycol (Mr 20,000), and 0.5% nonfat powdered milk at 42 8C. Final high-stringency washes were carried out

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

in 0:1£ SSC/0.1% SDS at 65 8C prior to autoradiography. To assign the phage clones to the original transcripts, we again performed PCR reactions with degenerate winged helix primers as stated above, cloned three clones for each reaction in the pBluescript vector SK 2 (Stratagene) and sequenced the clones on both strands by automated fluorescence sequencing. From screening approximately 6 £ 105 recombinants, 46 positive clones were obtained, of which eight proved to be from independent clones, as assessed by partial sequence analysis. Two of these eight clones were studied further because they were found predominantly in CD34 1 blood progenitor and CD34 1 leukemia cells. They were named Myeloid factor-a ( ¼ FOXD4a) and Myeloid factor-g ( ¼ FOXD4b). In addition to the four transcripts of Myeloid factors-a, -b, -g, and -d, described above, we identified one other new winged helix factor, not present in the cell lines examined which we named HWH-1. The GenBank accession number for this identified clone is AF343008. For both FOXD4a and FOXD4b we obtained a 12 kb phage insert which was cut into 3 fragments of 7, 3, and 2 kb by EcoRI and cloned into the PUC18 vector (Stratagene). To localize the winged helix domain on these fragments, we repeated the above described PCR with the partially degenerate primers. The winged helix domains of FOXD4a and FOXD4b were localized in the 7 kb fragments. These 7 kb fragments were subsequently sequenced by the company 4Base Lab (Reutlingen, Germany) using automatic fluorescent sequencing on both strands, with a redundancy of 3–8 reactions for each base pair. Sequence analysis was performed by using the Lasergene software from DNASTAR (Madison, WI). 2.4. Isolation of foxD4a and foxD4b cDNA First we isolated polyadenylated RNA from 5 £ 106 NB-4 cells with the Oligotex Direct mRNA Mini kit (Qiagen, Hilden, Germany). Then we synthesized cDNA from the poly(A) 1 RNA with the SUPERScript Preamplification System for first-strand cDNA synthesis (Gibco BRL). In summery, one mg of poly(A) 1 RNA was mixed with 1 ml of oligo(dT) primer (0.5 mg/ml). This was incubated in a

total of 12 ml DEPC-treated water for 10 min at 70 8C and cooled on ice for 1 min. Two microliters of 10£ PCR buffer, 2 ml 25 mM MgCl2, 1 ml 10 mM dNTP mix, 2 ml 0.1 M DTT, and 0.5 ml RNasin were added and incubated at 42 8C for 5 min. Then 1 ml SUPERScript II reverse transcriptase was added. The reaction was incubated another 50 min at 42 8C before being stopped by incubation at 70 8C for 15 min. Then 2.5 ml Rnase H was added and the reaction was finally incubated for 20 min at 37 8C. Ten microliters of the cDNA were used for each PCR amplification of the Myeloid factor fragments using the protocol described under Section 2.2. Each fragment was analysed on a 1% agarose gel and then directly sequenced on both strands by automated fluorescence sequencing. The pairs used for amplification of overlapping foxD4a and foxD4b cDNA fragments are given in Tables 1 and 2. 2.5. Northern blot analysis For Northern blot analysis, 10 mg of total RNA was electrophoresed on a 1% agarose/formaldehyde gel and transferred to a Hybond-N membrane (Amersham Corp., Arlington Heights, IL) by capillary action. Hybridization was performed as described for the screening procedure. As probes we used the same [a- 32P]dCTP random primerlabeled fragments which were used for the fox factor expression studies (see above) and which did not contain sequences of the highly conserved winged helix DNA binding domain. 2.6. Expression analysis by RT–PCR Total RNA was isolated using the Tripure isolation reagent (Roche Diagnostics, Indianapolis, IN). First, 3 mg of RNA was treated with 3 ml RNase-free DNase in a total volume of 10 ml for 1 h at 37 8C. Then the cDNA synthesis and PCR amplification were performed as described above. As primers for foxD4a we used: 5 0 primer GCTCTAGAGATGAAGACGAGGAGGAGGCGG; 3 0 primer CGGAATTCCGGCTGCCGGGCATCTTCAGAG. For foxD4b the following primers were used: 5 0 primer GCTCTAGACTTTGCTGCAAGTGTCCGCCGC; 3 0 primer CGGAATTCCACTGCCTGATACCTCAGCAGC.

Table 1 Primer pairs for amplification of overlapping foxD4a cDNA fragments 50 50 50 50 50 50 50 50 50 50 50 50

Primer 1 forward (nt 821) Primer 1 reverse (nt 1141) Primer 2 forward (nt 938) Primer 2 reverse (nt 1141) Primer 3 forward (nt 938) Primer 3 reverse (nt 1503) Primer 4 forward (nt 1483) Primer 4 reverse (nt 1859) Primer 5 forward (nt 1837) Primer 5 reverse (nt 2039) Primer 6 forward (nt 2018) Primer 6 reverse (nt 2203)

133

GCTCTAGAAGCACCTGCTCCGCCATGAACT CGGAATTCCGGCTGCCGGGCATCTTCAGAG GCTCTAGAGATGAAGACGAGGAGGAGGCGG CGGAATTCCGGCTGCCGGGCATCTTCAGAG GCTCTAGAGATGAAGACGAGGAGGAGGCGG CGGAATTCGCGGCGTGTGCAGCAGGTAGAG GCTCTAGACTCTACCTGCTGCACACGCCGC CGGAATTCGACAAACTCTGCGCAGCCCCTG GCTCTAGAACAGGGGCTGCGCAGAGTTTGTC CGGAATTCCCACCGCCTGATACCGCAGCAG GCTCTAGACTGCTGCGGTATCAGGCGGTGG CGGAATTCGCAAGGTGGAGTGAGCAGCTGC

134

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

Table 2 Primer pairs for amplification of overlapping foxD4b cDNA fragments 50 50 50 50 50 50 50 50 50 50 50 50 50 50

Primer 7 forward (nt 816) Primer 7 reverse (nt 955) Primer 8 forward (nt 934) Primer 8 reverse (nt 1148) Primer 9 forward (nt 934) Primer 9 reverse (nt 1510) Primer 10 forward (nt 1489) Primer 10 reverse (nt 1959) Primer 11 forward (nt 1489) Primer 11 reverse (nt 1866) Primer 12 forward (nt 1844) Primer 12 reverse (nt 1959) Primer 13 forward (nt 1938) Primer 13 reverse (nt 2094)

The PCR products were electrophoresed on a 1% agarose gel and blotted onto a Hybond-N membrane (Amersham) by capillary action. The blots were hybridized, as described for the screening procedure, with probes labeled with [a- 32P]dCTP by random priming. The probes were generated by PCR with the fox factor primers described above on foxD4a and foxD4b 7 kb DNA fragments which were cloned into the PUC18 vector (Stratagene). Hybridization probes did not include sequences that code for the conserved DNA binding domain. As a control for the expression studies, we amplified b2-microglobulin-mRNA from the analysed cell lines with the following primers: 5 0 primer ACCCCCACTGAAAAAGATGA; 3 0 primer ATCTTCAAACCTCCATGATG. For expression analysis of the tissue samples, we used adult and fetal multiple tissue cDNA panels (Clontech). For each reaction, 5 ml cDNA was used for PCR amplification as described above. As a control we amplified GAPDH-mRNA from the tissues with the following primers: 5 0 primer TGAAGGTCGGAGTCAACGCATTTGGT; 3 0 primer CATGTGGGCCATGAGGTCCACCAC. Again we electrophoresed the products on a 1% agarose gel and blotted them onto a Hybond-N membrane (Amersham) by capillary action. Hybridization was performed as described for the expression studies on the cell lines. The same [a- 32P]dCTP random primer-labeled probes were used.

3. Results 3.1. Isolation of transcripts from novel and known winged helix factors from leukemia cell lines and CD34 1 blood progenitor cells We isolated transcripts of five known and four new winged helix factor genes by PCR with degenerate primers of the highly conserved winged helix DNA binding domain from different leukemia cell lines and CD34 1 blood progenitor cells (Fig. 1). We used four AML cell lines of different maturity (KG-1, Kasumi, HL-60, NB-4), one CML cell line (K-

GCTCTAGAAGCACCTGCTCCGCCATGAACT CGGAATTCTCGTCTTCCACCTCGTCTTCAT GCTCTAGAATGAAGACGAGGTGGAAGACGA CGGAATTCCGGCTGCCGGGCATCTTCAGAG GCTCTAGAATGAAGACGAGGTGGAAGACGA CGGAATTCGCGGCGTGTGCAGCAGGTAGAG GCTCTAGACTCTACCTGCTGCACACGCCGC CGGAATTCGCGGCGGACACTTGCAGCAAAG GCTCTAGACTCTACCTGCTGCACACGCCGC CGGAATTCGACAAACTCTGCGCAGCCCCTG GCTCTAGAACAGGGGCTGCGCAGAGTTTGTC CGGAATTCGCGGCGGACACTTGCAGCAAAG GCTCTAGACTTTGCTGCAAGTGTCCGCCGC CGGAATTCCACTGCCTGATACCTCAGCAGC

562) and CD34 1 blood progenitor cells. In contrast to earlier studies (Hromas et al., 1993; Hromas et al., 1994) a set of primers was chosen which recognizes most of the, up-tonow, known sequences of the winged helix family. As controls for contamination with genomic DNA, reactions without addition of reverse transcriptase were performed. The newly identified fragments were named Myeloid factor-a ( ¼ FOXD4a), -b, -g ( ¼ FOXD4b), and -d. The known genes isolated are freac-3, freac-4, freac-5, freac-9, and FoxD3. Twenty clones for each set of cells were examined. PCR artifacts which might influence the result were regarded as unlikely since each transcript was identified by the same sequence several times and in different cell lines.

Fig. 1. (a) An ethidium bromide agarose gel electrophoresis of RT–PCR products from different cell lines and CD34 1 blood progenitor cells is shown. Plus indicates reactions with reverse transcriptase, minus shows mock control reactions without reverse transcriptase. The arrow denotes the approximate 210 bp product. This was isolated and cloned into the EcoRI and XbaI sites of pBluescript SK 2 . Clones were screened for winged helix domain homology by automated fluorescence sequencing. Markers are in the second lane from the left (M). In the far left lane a positive control (W) is shown with amplification on the known winged helix transcription factor CWH-2 cloned into pBluescript SK 2 . The intense band below the indicated winged helix domain band are the degenerate primers. (b) Control reactions with a b2-microglobulin probe are shown which demonstrate the integrity of the RNAs used. The following cell lines were used for RT–PCR reactions: 1, Kasumi; 2, KG-1; 3, NB-4; 4, HL-60; 5, K562; 6, CD34 1 human blood progenitor cells isolated from peripheral blood.

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

3.2. Isolation of fox genes from a phage library The transcripts of the DNA binding domains of FOXD4a and FOXD4b were used for screening the human genomic phage library EMBL3/SP6 (Clontech) because transcripts of FOXD4a and FOXD4b were predominantly found in CD34 1 blood progenitor and CD34 1 leukemia cells while Myeloid factor-b and -d transcripts showed a more ubiquitous expression in the cells analysed. We chose a genomic instead of a cDNA library, first because all genes could be isolated from the same source and second, cDNA libraries of the cell lines used were not commercially available. Third, it is known that it is difficult to isolate full-length winged helix genes from cDNA libraries because of regions which are highly GC-rich such as transcriptionally regulatory motifs. Fourth, we needed the genomic sequence in the long-term for promoter studies. It was an important requisite for this approach that some of the winged helix factors known to date are intronless. If introns are present one of them is often localized in a region within the winged helix domain and would be visible with our chosen primers. We screened 6 £ 105 phages with each probe under high stringency conditions. Despite this high stringency, we could not exclude cross-hybridization with closely related winged helix genes. Therefore, we isolated ten clones for each hybridization probe. To assign the 40–50 kb phage clones to the original transcripts, we again performed PCR on the winged helix domain with the degenerate primers. The 210 bp winged helix fragments were subcloned in pBluescriptVector SK 2 (Stratagene). Sequencing of three clones for each phage allowed matching the phage clones to the original winged helix transcripts. We chose two clones for further characterization: FOXD4a and FOXD4b. Both transcripts were found predominantly in early CD34 1 blood progenitor cells, including KG-1 and Kasumi cells. 3.3. Isolation of foxD4a and foxD4b cDNA We isolated the foxD4a and foxD4b cDNA by amplifying overlapping gene fragments by RT–PCR from NB-4 mRNA. The cDNAs include the complete sequence from the starting ATG until the stop codon of the genes (nt 821– 2203 of foxD4a and nt 816–2094 of foxD4b). The sequence of the fragments was determined by automated fluorescence sequencing on both strands. Sequences derived from the cDNAs were identical to the determined sequences of the genomic clones. Therefore we can conclude that both genes are intronless. In addition, this approach provides evidence that the genomic clones are not pseudogenes of splice variants. 3.4. Sequence and structure of FOXD4a and FOXD4b A 1319 bp sequence codes for the open reading frame of FOXD4a and a 1250 bp sequence codes for the open reading frame of FOXD4b (Fig. 2). Translation in both factors starts with a Kozak-consensus-sequence and follows on a stop

135

codon. The coding region of FOXD4a and FOXD4b contains a winged helix domain (aa 108–203). In the 5 0 region of the winged helix domain is an acidic region (aa 19–45) and in 3 0 of the winged helix domain is a prolineand alanine-rich region (aa 211–301), which could act as transcriptionally regulatory motifs. In the 3 0 region the stop codon is followed by a classical polyadenylation signal (nt 2441 foxD4a, nt 2779 foxD4b). Upstream of both factors there is a region with typical elements of a promoter region with CCAAT box (nt 207 foxD4a), TATA box (nt 646 foxD4a), Pu box (nt 566 foxD4a) and transcription factor binding sites such as three AP-1 sites (nt 102, nt 472, nt 567 foxD4a), two STAT-2 sites (nt 251, nt 515 foxD4a), one AP-2/SP-1 site (nt 91 foxD4a), one ets site (nt 191 foxD4a), one HNF-5 site (nt 223 foxD4a), one SP-1 site (nt 325 foxD4a), one E2F site (nt 392 foxD4a), and one HNF-3 site (nt 524 foxD4a). A computer-assisted search in the NCBI GenBank database revealed a homology of foxD4a and foxD4b to rp-2, a winged helix gene which was identified in a human DNA sequence from clone RP11-395L14 on chromosome 22q13.32–13.33 (accession number AL078621) by S. Blakey. Sequence homology of foxD4a to rp2 is 97% on the nucleotide level within the coding region, while homology in the 5 0 untranslated region and homology in the 3 0 untranslated region are 98%, respectively. FoxD4b is 94% homologous to rp2 within the coding region, 98% homologous in the 5 0 untranslated region and 99% homologous in the 3 0 untranslated region. FoxD4a is 93% homologous to foxD4b within the coding region, in the 5 0 untranslated region 96% homologous and in the 3 0 untranslated region identical. On the amino acid level, homology is much lower. FOXD4a is 64% homologous to RP2. FOXD4b is 69% homologous to RP2, and FOXD4a is 76% homologous to FOXD4b. On the nucleotide level there are two relevant differences between the fox factors and rp2. In position 950 of foxD4a there is a 12-nucleotide deletion which is not detectable in foxD4b or rp2. Therefore, this region was chosen as a forward primer for expression studies. FoxD4b, on the other hand, contains a 53-nucleotide insertion at position 1916 which is not present in the sequences of foxD4a or rp2. This region was also chosen as primer design for the expression studies. 3.5. Fox factor mRNA expression in leukemia cell lines Fox factor expression was examined by RT–PCR. Primers were chosen specifically for each fox factor sequence, within a 12-nucleotide deletion region of foxD4a in comparison to the sequence of foxD4b and a 53-nucleotide insertion region of foxD4b, in comparison to the sequence of foxD4a (see Section 2). This eliminates cross-hybridization between both factors, because of the otherwise close sequence similarity on the nucleotide level. The following cell lines were examined for the expression studies: KG-1 (early myeloblasts), Kasumi (myeloblasts), NB-4 (promyelocytes), HL-60 (early

136

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

promyelocytes), U937 (monoblasts), THP-1 (monoblasts), HEL (erythroblasts), U266 (plasma cells), Jurkat (T cells), Raji (B-lymphoblasts), HepG2 (hepatocellular carcinoma cells), and FH109 (embryonal lung fibroblasts). Northern blot analysis on HL-60 cells, U937 cells and NB-4 cells showed bands of approximately 1.5 kb length for both Myeloid factor-a and Myeloid factor-g (Fig. 3a,b).

FoxD4a was positive in RT–PCR in all cell lines except Jurkat, which became only weakly positive after long-term exposure (Fig. 4a). FoxD4b was positive in all cell lines except Kasumi, which became positive only after longterm exposure (Fig. 4b). RT–PCR analysis of b2-Mikroglobulin-mRNA, as an internal control for the integrity of the RNA, was positive in all cell lines (Fig. 4c). Mock reactions

Fig. 2. A comparison of the amino acid homology between FOXD4a, FOXD4b, and RP2. In the top line the consensus sequence is shown. From amino acids 19 to 45 of the consensus sequence, there is an acidic region rich in aspartate and glutamate. In positions 108 to 203, the highly conserved winged helix DNA binding domain is found. From amino acids 211 to 301, there is a proline- and alanine-rich region. The MegAlign program of the Lasergene Software from DNASTAR (Madison, WI) was used with the Clustal method for the comparison.

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

137

Fig. 3. Northern blot analysis of the expression patterns of foxD4a (a) and foxD4b (b) in three hematopoietic cell lines (1, HL-60; 2, U937; 3, NB-4). The size of the hybridization species is indicated on the right and the markers (M) are shown on the left. Both factors are expressed in all three cell lines.

Fig. 4. RT–PCR and Southern blot analysis of the expression patterns of foxD4a (a) and foxD4b (b) in several hematopoietic and non-hematopoietic cell lines (1, KG-1; 2, Kasumi; 3, NB-4; 4, HL-60; 5, U937; 6, THP-1; 7, HEL; 8, U266; 9, Jurkat; 10, Raji; 11, HepG2; 12, FH109). The size of the hybridization species is indicated on the right with the arrow. (c) Control reactions with a b2-microglobulin probe are shown which demonstrate the integrity of the RNAs.

138

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

muscle and in all fetal tissues except lung, heart, and spleen (Figs. 5b and 6b). RT–PCR analysis for GAPDH-mRNA, as an internal control for the integrity of the RNA was positive in all panels (Figs. 5c and 6c).

4. Discussion

Fig. 5. PCR and Southern blot analysis of the expression patterns of foxD4a (a) and foxD4b (b) in adult multiple tissue cDNA panels (1, heart; 2, brain; 3, kidney; 4, liver; 5, lung; 6, pancreas; 7, placenta; 8, skeletal muscle). The size of the hybridization species is indicated on the right with the arrow. (c) Control reactions with a GAPDH probe are shown which demonstrate the integrity of the cDNAs.

without reverse transcriptase were negative in all cases (data not shown). This indicates that we did not amplify from DNA. 3.6. Tissue-specific expression of fox factor mRNAs To examine tissue-specific expression of the fox factor mRNAs, we performed RT–PCR analysis on multiple tissue cDNAs (Clontech) with subsequent Southern blot analysis. The adult tissue panel contained cDNA of heart, brain, kidney, liver, lung, pancreas, placenta, and skeletal muscle. In the human fetal tissue panel, we examined cDNA of brain, kidney, liver, lung, heart, skeletal muscle, spleen, and thymus. FoxD4a mRNA was positive in all adult cDNA panels except pancreas, and with strong signals in placenta and skeletal muscle and all fetal cDNA panels except brain (Figs. 5a and 6a). FoxD4b mRNA was detectable in all adult tissues except liver, placenta, and skeletal

Fig. 6. PCR and Southern blot analysis of the expression patterns of foxD4a (a) and foxD4b (b) in fetal multiple tissue cDNA panels (1, brain; 2, kidney; 3, liver; 4, lung; 5, heart; 6, skeletal muscle; 7, spleen; 8, thymus). The size of the hybridization species is indicated on the right with the arrow. (c) Control reactions with a GAPDH probe are shown which demonstrate the integrity of the cDNAs.

We were interested in identifying further winged helix factor genes in leukemia cell lines and CD34 1 blood progenitor cells in order to study their expression patterns, especially in leukemia cell lines, and to examine in the future their possible involvement in leukemogenesis. In this paper we describe the isolation and expression pattern of two new winged helix factor genes. Altogether, we isolated transcripts of five known and four yet unknown winged helix factor genes. The transcripts of the known genes derive from the winged helix transcription factors FREAC-3, FREAC-4, FREAC-5, FREAC-9, and FOXD3. Two of the newly isolated factors which were named FOXD4a and FOXD4b were studied further. Both factors contain a highly conserved winged helix DNA binding domain. In the 5 0 untranslated regions of the fox genes, we found typical elements of a promoter region with a TATA box, CCAAT box and transcription factor binding sites. We identified one binding site for another winged helix transcription factor, HNF-3. It is possible that there is a hierarchy of regulation between winged helix factors. Further studies will show if HNF-3 regulates FOXD4a and FOXD4b.) Within the coding region, 5 0 of the winged helix domain there is an acidic region, and in 3 0 of the winged helix domain we found a proline- and alanine-rich region. Based upon their amino acid compositions, those regions might act as transcriptionally regulatory motifs, when compared to such motifs in other winged helix factors (Freyaldenhoven et al., 1997a,b). Surprisingly, foxD4a and foxD4b showed a high degree of sequence homology to each other on the nucleotide level as well as to a winged helix factor named rp2 for which no further information was available except the sequence that was deposited in the GenBank. Thus, we first thought that all three sequences derive from the same gene. When we compared the sequence in greater detail, however, it became obvious that two marked differences between the three factors exist. In position 950 of foxD4a there was a 12-nucleotide deletion which was not detectable in foxD4b and rp2. This deletion affects an acidic domain (aa 211–301) which could act as transcriptionally regulatory motif. The second difference was a 53-nucleotide insertion in foxD4b at position 1916 which was not present in the sequences of foxD4a and rp2. No domain identified so far is affected by this insertion. Therefore, homology between the three factors is much lower on the amino acid level. FOXD4a is 64% homologous to RP2, FOXD4b is 69% homologous to RP2, and FOXD4a is 76% homologous to FOXD4b. This low amino acid homology renders it more likely that

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

FOXD4a, FOXD4b, and Rp2 are encoded by three separate genes. We could also detect two different cDNAs for foxD4a and foxD4b which are identical in sequence to the genomic clones. This provides evidence that the genomic clones are not pseudogenes of splice variants. It is possible that originally duplication and subsequent mutations from one single gene led to different genes and transcripts. Gene duplication is a common phenomenon in the family of winged helix transcription factors (Lehmann et al., 2000). On the other hand it is also possible that mistakes in sequencing have taken place. We have sequenced our genes on the cDNA level and as genomic fragments several times on both strands with the same results. Therefore we think it is more likely that there are sequencing mistakes in the rp2 sequence deposited in the GenBank leading to frameshifts in the amino acid sequence. For example, the foxD4a gene deviates at amino acid position 294 from the previously published rp-2. The conceptual translation of the rp-2 sequence in three frames shows a frameshift in this region. This would explain why we can detect high sequence homology on the nucleotide level and much lower homology on the amino acid level. In a second set of experiments the expression of the fox factors was studied. We used the differences between the genes to choose primer pairs for RT–PCR analysis in order to eliminate cross reactions. The amplified fragments also did not contain sequences from the highly conserved DNA binding domain. Though we did not show a striking difference of expression pattern between the fox factors in the leukemia cell lines examined, we demonstrated a different expression pattern in the adult and fetal tissues. FoxD4a was found in adult brain but not in fetal brain. FoxD4b transcripts were not present in adult skeletal muscle but in fetal muscle tissue. In addition both fox factors are expressed in a variety of leukemia cell lines. Several lines of evidence point to the involvement of winged helix transcription factors in regulatory processes of hematopoiesis. Many members of the winged helix family are important for the development of mesodermal structures from which the hematopoietic system also originates (Lai et al., 1993; Costa, 1994). In Drosophila, blood cells are produced in the lymph glands where the prototypical winged helix protein Forkhead is expressed (Kuzin et al., 1994). Expression of the Drosophila forkhead gene is regulated by trithorax. MLL (ALL1), the human homolog of trithorax is recombined with multiple different gene loci in subsets of acute leukemia (Gu et al., 1992). The co-acting of transcription factors in hierarchies is often found in embryonal development and suggests that similar cascades of transcription factors are also involved in the regulation of human hematopoiesis, including winged helix transcription factors. A computer-assisted search of PCR-selected DNA binding sequences for winged helix factors in promoter data bases (Overdier et al., 1994) revealed as putative target genes, among other, products of genes which are involved in regulatory processes of hematopoiesis (Retinoid acid

139

receptor b, Oct3, HoxB2, and Cyclin D1). FAST-1, regulating the Mix.2 gene promoter, is the first winged helix factor gene that was found to be a mediator protein of TGF-b. TGF-b is an important signal for the control of cell proliferation and cell differentiation as well as mesodermal development (Chen et al., 1996). In nearly all classes of transcription factors which are subdivided by their DNA binding domain and/or dimerization domain, there are members for which an expression has been demonstrated during normal hematopoietic differentiation or a participation in leukemias has been found. Except for the translocated winged helix fusion genes, few winged helix factors have so far been found to be expressed in leukemia cells. In earlier studies, transcripts of winged helix genes have been found by RT–PCR with primers of the highly conserved DNA binding domain in the human erythroleukemia cell line HEL, and three new winged helix transcription factors were isolated (Hromas et al., 1993). In another study of the same group, winged helix transcripts were identified in cells with the enriched phenotype CD34 1/HLA-DR 2, and one new winged helix transcription factor was isolated (Hromas et al., 1994). In both studies, primers used for RT–PCR were chosen based upon the winged helix sequences known by the time of the first study (1993). This choice of primers, however, would not recognize a large part of the winged helix genes described afterwards. Furthermore, neither study contained detailed expression research on the identified genes. A correlation between the grade of cellular differentiation and winged helix factor expression was not examined. It has already been shown for the murine winged helix factor Genesis that overexpression in the myeloid mouse cell line 32Dcl3, with addition of granulocyte colony-stimulating factor and interleukin-3 withdrawal, caused continuous cell proliferation without differentiation, instead of the typical apoptosis. This study suggests a possible role of Genesis in preserving the undifferentiated phenotype of myeloid cells (Sutton et al., 1996; Xu et al., 1998). Future direction will show whether the fox factors are also differentially expressed during myeloid and monocyte differentiation. This can be demonstrated in cell lines such as HL60 cells, NB-4 cells, or U937 cells where differentiation into the granulocyte or monocyte lineage is possible (Lill et al., 1995; Liu and Wu, 1992). It will also be an important task to examine overexpression of the fox factors in leukemia cell lines with subsequent induction of differentiation, in analogy to the overexpression studies with the murine winged helix factor Genesis. These studies will further elucidate the role of winged helix factors in leukemias.

Acknowledgements The authors thank Christel Weiß-Fuchs for technical assistance. This work was supported by a grant of B.S.F. from the Bundesministerium fu¨ r Bildung, Wissenschaft, Forschung und Technologie, Federal Republic of Germany.

140

B.S. Freyaldenhoven et al. / Gene 294 (2002) 131–140

References Ang, S.L., Rossant, J., 1994. HNF3b is essential for node and notochord formation in mouse development. Cell 78, 561–574. Brandt, J., Srour, E., Van Besien, K., Briddell, R., Hoffman, R., 1990. Cytokine-dependent long term marrow culture of highly enriched precursors of hematopoietic progenitor cells from human bone marrow. J. Clin. Invest. 86, 932–941. Chen, X., Rubock, M.J., Whitman, M., 1996. A transcriptional partner for MAD proteins in TGF-b signalling. Nature 383, 691–696. Clark, K.L., Halay, E.D., Lai, E., Burley, S.K., 1993. Co-crystal structure of the HNF-3/fork head DNA-recognition motif resembles histone H5. Nature 364, 412–420. Costa, R.H., 1994. In: Tronche, F., Yaniv, M. (Eds.). Liver-Specific Gene Transcription, CRC Press, Boca Raton, FL, pp. 183–205. Freyaldenhoven, B.S., Freyaldenhoven, M.P., Iacovoni, J.S., Vogt, P.K., 1997a. Aberrant cell growth induced by avian winged helix proteins. Cancer Res. 57, 123–129. Freyaldenhoven, B.S., Freyaldenhoven, M.P., Iacovoni, J.S., Vogt, P.K., 1997b. Avian winged helix proteins CWH-1, CWH-2, and CWH-3 act as transcriptional repressors. Oncogene 15, 483–488. Galili, N., Davis, R.J., Fredericks, W.J., Mukhopadhyay, S., Rauscher III, F.J., Emanuel, B.S., Rovera, G., Barr, F.G., 1993. Fusion of a fork head domain gene to PAX3 in solid tumour alveolar rhabdomyosarcoma. Nat. Genet. 5, 230–235. Gu, Y., Nakamura, T., Alder, H., Prasad, R., Canaani, O., Cimino, G., Croce, C.M., Canaani, E., 1992. The t(4;11) chromosome translocation of human acute leukemias fuses the ALL-1 gene, related to Drosophila trithorax, to the AF-4 gene. Cell 71, 701–708. Hatini, V., Huh, S.O., Herzlinger, D., Soares, V.C., Lai, E., 1996. Essential role of stromal mesenchyme in kidney morphogenesis revealed by targeted disruption of winged helix transcription factor BF-2. Genes Dev. 10, 1467–1478. Hillan, J., Le Coniat, M., Jonveaux, P., Berger, R., Bernard, O.A., 1997. AF6q21, a novel partner of the MLL gene in t(6;11)(q21;q23), defines a forkhead transcriptional factor subfamily. Blood 90, 3714–3719. Hromas, R., Moore, J., Johnston, T., Socha, C., Klemsz, M., 1993. Drosophila forkhead homologues are expressed in lineage-restricted manner in human hematopoietic cells. Blood 81, 2854–2859. Hromas, R., Klemsz, M., Amaravadi, L., Hufford, T., Huang, I., Desai, A., Srour, E., Bruno, E., Hoffman, R., 1994. Drosophila forkhead homologues are expressed in CD34 1/HLA-DR 2- primitive human hematopoietic progenitors. Leukemia Lymphoma 15, 439–444. Kuzin, B., Tillib, S., Sedkov, Y., Mizrokhi, L., Mazo, A., 1994. The Drosophila trithorax gene encodes a chromosomal protein and directory regulates the region-specific homeotic gene fork head. Genes Dev. 8, 2478– 2490. Lai, E., Prezioso, V.R., Smith, E., Litvin, O., Costa, R.H., Darnell Jr, E., 1990. HNF-3a, a hepatocyte-enriched transcription factor of novel structure is regulated transcriptionally. Genes Dev. 4, 1427– 1436. Lai, E., Prezioso, V.R., Tao, W., Chen, W.S., J Jr, E., Darnell Jr, J.E., 1991. Hepatocyte nuclear factor 3a belongs to a gene family in mammals that is homologous to the Drosophila homeotic gene fork head. Genes Dev. 5, 416–427. Lai, E., Clark, K.L., Burley, S.K., Darnell Jr, J.E., 1993. Hepatocyte nuclear factor 3/fork head or ‘winged helix’ proteins: A family of diverse biologic function. Proc. Natl. Acad. Sci. USA 90, 10421–10423.

Lehmann, O.J., Ebenezer, N.D., Jordan, T., Fox, M., Ocaka, L., Payne, A., Leroy, B.P., Clark, B.J., Hitchings, R.A., Povey, S., Khaw, P.T., Bhattacharya, S.S., 2000. Chromosomal duplication involving forkhead transcription factor gene FOXC1 causes iris hypoplasia and glaucoma. Am. J. Hum. Genet. 67, 1129–1135. Li, J., Vogt, P.K., 1993. The retroviral oncogene qin belongs to the transcription factor family that includes the homeotic gene fork head. Proc. Natl. Acad. Sci. USA 90, 4490–4494. Lill, M.C., Fuller, J.F., Herzig, R., Crooks, G.M., Gasson, J.C., 1995. The role of the homeobox gene HOXB7 in human myelomonocytic differentiation. Blood 85, 692–697. Lin, K., Dorman, J.B., Rodan, A., Kenyon, C., 1997. daf16: An HNF-3/ forkhead family member that can function to double the life-span of Caenorhabditis elegans. Science 278, 1319–1322. Liu, M.Y., Wu, M.C., 1992. Induction of human monocyte cell line U937 and CSF-1 production by phorbol ester. Exp. Hematol. 20, 974–979. Metcalf, D., 1989. The molecular control of cell division, differentiation, commitment, and maturation in hematopoietic cells. Nature 339, 27–30. Nehls, M., Pfeifer, D., Schorpp, M., Hedrich, H., Boehm, T., 1994. New member of the winged-helix protein family disrupted in mouse and rat nude mutations. Nature 372, 103–106. Nichols, J., Nimer, S.D., 1992. Transcription factors, translocations, and leukemia. Blood 80, 2953–2963. Orkin, S.H., 1995. Transcription factors and hematopoietic development. J. Biol. Chem. 270, 4955–4958. Overdier, D.G., Porcella, A., Costa, R.H., 1994. The DNA-binding specificity of the hepatocyte nuclear factor 3/forkhead domain is influenced by amino acid residues adjacent to the recognition helix. Mol. Cell. Biol. 14, 2755–2766. Parry, P., Wei, Y., Evans, G., 1994. Cloning and characterization of the t(X;11) breakpoint from a leukemic cell line identify a new member of the forkhead gene family. Genes Chromosomes Cancer 11, 79–84. Shapiro, D.N., Sublett, J.E., Li, B., Downing, J.R., Naeve, C.W., 1993. Fusion of PAX3 to a member of the forkhead family of transcription factors in human alveolar rhabdomyosarcoma. Cancer Res. 53, 5108– 5112. Sutton, J., Costa, R., Klug, M., Field, L., Xu, D., Largaespada, D.A., Fletcher, C.F., Jenkins, N.A., Copeland, N.G., Klemsz, M., Hromas, R., 1996. Genesis, a winged helix transcriptional repressor with expression restricted to embryonic stem cells. J. Biol. Chem. 271, 23126– 23133. Vogt, P.K., Li, J., Freyaldenhoven, B.S., 1997. Relevations of a captive: retroviral Qin and the oncogenicity of winged helix proteins. Virology 238, 1–7. Weigel, D., Ju¨ rgens, G., Ku¨ ttner, F., Seifert, E., Ja¨ ckle, H., 1989. The homeotic gene fork head encodes a nuclear protein and is expressed in the terminal regions of the Drosophila embryo. Cell 57, 645–658. Weinstein, D.C., Ruiz i Altaba, A., Chen, W.S., Hoodless, P., Prezioso, V.R., Jessell, T.M., Darnell Jr, J.E., 1994. The winged-helix transcription factor HNF-3b is required for notochord development in the mouse embryo. Cell 78, 575–588. Xu, D., Yoder, M., Sutton, J., Hromas, R., 1998. Forced expression of Genesis, a winged helix transcriptional repressor isolated from embryonic stem cells, blocks granulocytic differentiation of 32D myeloid cells. Leukemia 12, 207–212. Xuan, S., Baptista, C.A.:., Balas, G., Tao, W., Soares, V.C., Lai, E., 1995. Winged helix transcription factor BF-1 is essential for the development of the cerebral hemispheres. Neuron 14, 1141–1152.