0161-5890/92 $5.00 + 0.00 Pergamon Press Ltd

Molecular Immunology, Vol. 29, No. 10, pp. 1165-I 174, 1992 Printed in Great Britain.

CHARACTERIZATION OF THE S-FLANKING TRANSCRIPTIONAL REGULATORY REGION OF THE HUMAN Fey RECEPTOR GENE, Fey RIIA E. M&ENzIE,*$** MARGARET A. KELLER,*? DIANA L. CASSEL,§T/ ALAN D. SCHREIBER,§~ ELIAS SCHWARTZ,*$ 11SAUL SURREY*11and ERIC F. F&PPAPORT*$

STEVEN

*Division of Hematology, Graduate Studies Group, University of Pennsylvania Oncology, Hospital of

Children’s Hospital of Philadelphia, Philadelphia, U.S.A.; tBiomedica1 and fDepartments of Pediatrics, $Medicine and IlHuman Genetics, School of Medicine, Philadelphia, U.S.A.; and IDivision of Hematology/ the University of Pennsylvania, Philadelphia, Pennsylvania, U.S.A.

(First received 11 December

1991; accepted in revised form 20 ~April 1992)

Abstract-The human Fey receptor gene Fey RIIA is expressed in platelets, neutrophils, monocytes and macrophages. Understanding the regulation of expression of Fey RIIA will enhance our knowledge of regulated gene expression and immune function in these cells. We cloned a 3.65 kb region of the 5’ end of the Fey RIIA gene and characterized 3.4 kb of previously unreported sequence of the S-flanking region. Primer extension studies and RNase protection analyses of mRNA from HEL, K562 and U937 cells revealed multiple transcription start sites. One transcription start site mapped to a S-untranslated (SUT) exon approximately 1 kb 5’ to the ATG translation initiation codon, while a second start site mapped near the ATG codon. Reverse transcription combined with PCR (RT-PCR) employing an oligonucleotide in the putative 5’UT exon and an antisense oligonucleotide in the translated region yielded products which confirm that transcription starts in this 5’UT exon 881 bp upstream of the ATG codon. Sequence analysis of the RT-PCR products showed two related RNA splice products which use alternative 3’-consensus AG splice acceptor sites. Fey RIIA mRNA thus has three distinct potential 5’UT regions, two alternatively spliced forms from the start site in the 5’UT exon and the third from the start site near the ATG codon. Comparisons of the human Fey RIIA 5’-flanking region whith human Fey RI and mouse Fey RIIP genes as well as with other genes expressed in megakaryocytes, neutrophils and monocytes reveal structural similarities and shared promoter elements.

INTRODUCTION

Fey receptors (Fey R) are membrane glycoproteins which bind the Fc domain of IgG and, thus, served to link humoral and cellular components of the immune system. Three distinct classes of Fey R have been described at the protein and cDNA levels: Fey RI (CD64), Fey RI1 (CD32) and Fey RI11 (CD 16). The genomic 5’-regulatory region has been identified only for Fey RI (van de Winkel et al., 1991; Pearse et al., 1991). Fey RI1 is the most widely distributed class, present on platelets, neutrophils, monocytes, macrophages, and lymphocytes. Three different human genes coding for Fey RI1 have been identified and isolated: Fey RIIA, Fey RIIB and Fey RIIC (Qui et al., 1990). Fey RIIA, which is restricted to cells of the megakaryocyte and myeloid lineages, mediates several functions. Stimulation of monocyte Fey RI1 leads to antibodydependent cellular cytotoxicity, superoxide production and phagocytosis (Schreiber er al., 1989; Unkeless, 1989; van de Winkel and Anderson, 1991). Crosslinking of platelet Fey RIIA following the binding of oligomeric **To whom correspondence should be addressed, at Division of Hematology, Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104, U.S.A.

IgG results in platelet activation (King et al., 1990; Anderson and Anderson, 1990). The Fey RI1 cDNA structure consists of a 5’-untranslated region, a signal sequence region, an extracellular domain with homology to members of the immunoglobulin gene superfamily, a transmembrane region, a cytoplasmic domain and a 3’-untranslated region. Human Fey RI1 cDNA clones isolated from several cell lines exhibit structural heterogeneity (Brooks et al., 1989; Stuart et al., 1989; Seki, 1989). The extracellular domains of the three isotypes are highly homologous, but the cytoplasmic domains diverge considerably. This cytoplasmic diversity is believed to underlie the ability of this class of receptors to mediate different functions for different cell types. In addition to the membrane-spanning forms of Fey RI1 we have demonstrated a form of Fey RIIA in which the transmembrane exon is spliced out, generating a soluble form of this receptor (Walterhouse et al., 1988; Rappaport et al., 1992). Fey RIIA expression is increased by interferon gamma (IFN-y) and granulocyte-macrophage colony-stimulating factor (GM-CSF) and decreased by glucocorticoids (Comber et al., 1989; Rossman et al., 1991). The effects of IFN-y and the glucocorticoid dexamethasone in monocytic cells are due in part to effects on relative transcriptional activity. We cloned and determined the DNA sequence of the 5’ end of the Fey RIIA gene in

1165

1166

STEVEN E. MCKENZIEet al.

order to understand the regulation of Fey RIIA transcription in the megakaryocytic, granulocytic and monocytic lineages and the mechanism of action of transcriptional modulators. In this study we report the sequence of the 5’-flanking region of human Fey RIIA, identification and mapping of multiple transcription start sites, analysis of FcyRIIA transcripts which differ in their 5’-untranslated sequence and delineation of genomic elements which may mediate the regulation of transcription.

MATERIALS AND METHODS Genomic clones. A 15.7 kb genomic clone for the Fey RIIA gene was isolated from an EMBL 3A library constructed by partial Suu 3A digestion of high molecular weight genomic DNA extracted from human white blood cells (Sambrook et al., 1989). The library was screened with a 0.8 kb HEL cell Fey RIIA cDNA clone isolated in our laboratory (Rappaport et al., 1991). The genomic clone, 1Fc8-1, extended from upstream of the first protein-coding exon through the first exon in the cytoplasmic domain of the protein. A 3.65 kb Sal I/&o RI fragment mapping upstream of the ATG translation initiation codon was subcloned into pGEM3Z (Promega Scientific, Madison, U.S.A.) and its DNA sequence was determined. DNA sequence analysis. DNA sequence analysis was accomplished either by dideoxy-terminator sequencing with T7 DNA polymerase and 32P-dCTP (Sequenase, United States Biochemicals Corporation, Cleveland, U.S.A.) or automated fluoresence-based sequencing with dye-labeled primers or terminators and Tug polymerase (Sequencer 373A, Applied Biosystems, Foster City, U.S.A.). Cells. The human megakaryocytic cell line, HEL (ATCC TIB 180), the human monocytic cell line, U937 (ATCC CRL 1593) and the human multipotential hematopoietic cell line, K562 (ATCC CCL 243) were obtained from the American Type Culture Collection, Rockville, U.S.A. The human megakaryocytic cell line CHRF-28811 was obtained from David P. Witte, University of Cincinnatti College of Medicine, Cincinnatti, U.S.A. (Fugman, 1990). The human megakaryocytic cell line Dami was obtained from Dr Sheryl Greenberg, Harvard Medical School, Boston, U.S.A. (Greenberg, 1988). All cell lines were maintained in RPM1 1640 supplemented with 10% (v/v) Serum Plus (Hazleton Biologics, Lenexa, USA; Dami cells were grown in 10% (v/v) horse serum) and with 2 mM L-glutamine, 1X BME vitamin solution and 100 U penicillin/l00 pg streptomycin per ml (Gibco). RNA isolation. RNA was prepared using guanidinium thiocyanate and extraction with cold acid phenol (Chomczynski and Sacchi, 1987) and was resuspended in diethyl-pyrocarbonate (DEPC)-treated water and stored at - 70°C. RNA concentrations were determined spectrophotometrically and quality was assessed by electrophoresis on non-denaturing agarose gels. Poly A + RNA was prepared by the PolyATract method

(Promega) according to the manufacturer’s specifications. Preparation of oligonucleotides. Oligonucleotides were synthesized on a DNA Synthesizer 380B (Applied Biosystems). Oligonucleotides used as hybridization probes or in primer-extension studies were labeled with T4 polynucleotide kinase according to standard procedures (Sambrook et al., 1989). The following primers were used for primer extension and RT-PCR: 16M 10M - 7cs - 12KS

S’CAATGGTTGAAGCAGCCACAGGTT3’ S’TCTGAGACATTTGGGTCTCCATAG3’ S’CATAGAGGGAAAACCTACTTGCA3’ S’CTTTACTAGCTTACTCCCAA3’.

Primer extension. Primer extension followed the method of Sambrook et al. (1989). Briefly, 100 ng of a labeled antisense oligonucleotide was purified using a G50 spin column and resuspended in DEPC-treated water. Labeled primers (10-25 ng) were hybridized at 37°C or 42°C to 10 pg of poly A + RNA or 50 pg of total RNA in 80% (v/v) formamide/ mM PIPES after heating to 80°C for 3 minutes. Both poly A- RNA and yeast tRNA served as controls. Following overnight hybridization, the hybrids were precipitated with ethanol and then were resuspended in RT buffer (50 mM TrisHCl, pH 8.3; 75 mM KCl, 3 mM MgCl,, 0.01 M DTT, 1 mM dNTP mix (Pharmacia), 50 ng/pl Actinomycin D, 1 U/cl1 RNasin). Moloney-Murine Leukemia Virus Reverse Transcriptase (M-MLV RT, 200 U, Gibco BRL) was added and extension products were synthesized at 42°C for 2 hours. The products were precipitated using ethanol, resuspended in 4 ,ul 1 x TE and 4 ~1 formamide loading buffer and then electrophoresed on a 5% (w/v) denaturing polyacrylamide gel and analyzed by autoradiography overnight at -70°C on Kodak XAR film (Eastman Kodak, Rochester, U.S.A.). A sequence ladder generated by 32P dCTP-based sequencing of the genomic clone with the same primer (16M) used in primer extension served as a molecular size marker, along with radiolabeld 0X174-Hae III fragments (Gibco BRL). RNase protection. RNase protection assays were performed using the pGEM3Z riboprobe system (Promega). Radiolabeled single-stranded RNA probes were transcribed in vitro using SP6 RNA polymerase. The riboprobes extended from the SP6 promoter site in p4.1F to either the HindIII, Ava I or Dpn I sites. Each probe contained the first signal peptide exon Sl (85 bp) and 131 bp of the intron 3’ to this exon. The riboprobes (1 x lo5 dpm) were annealed to RNA at 42°C overnight followed by digestion with RNases A (10 pg/ml) and Tl (200 U/ml) at 37°C for 1 h using the Ambion RPA II kit (Ambion, Austin, U.S.A). Protected fragments were separated on a 5% (w/v) denaturing polyacrylamide gel and analyzed by autoradiography overnight at -70°C on Kodak XAR film (Eastman Kodak). Reverse transcription-PCR (RT-PCR). First-strand cDNA was synthesized from 10 pg total RNA using M-MLV RT (Gibco BRL) in a total volume of 20 p 1 by a modification of the procedure described by Witsel and Schook (1990). The first-strand cDNA reaction mix

Human Fey RIIA gene S-flanking region contained 1.3 PM Random Primer (Promega), 500 PM each dATP, dCTP, dGTP and TTP (Pharmacia); 1X RT Buffer (50 mM Tris-HCl, pH 8.3; 75 mM KC1 and 3 mM MgCl,, BRL/Life Technologies), 10 mM dithiothreitol and 500 U M-MLV RT. The reaction mixture was held at room temperature for lOmin, incubated at 42°C for 1 h and then heated at 90°C for 10 min. The reaction was then diluted 1:2 by addition of 20 ~1 of 1X RT buffer. 12 ~1 of the diluted cDNA reaction mix were amplified in a total volume of 100 ~1 containing 50mM KCl, 10mM Tris-HCl, pH 8.3, 1.5mM MgCl,, 0.001% (w/v) gelatin, 200 PM each dATP, dCTP, dGTP and TTP, 1 PM 5’ primer (sense); 1 PM 3’ primer (antisense and 1.5 units AmpliTaq DNA polymerase (Perkin-Elmer/Cetus, Norwlak, U.S.A.). 50 ~1 of paraffin oil was then placed on top to prevent evaporation and the mixture was amplified for 40 cycles using a Perkin-Elmer/Cetus Thermal Cycler. Each cycle consisted of a denaturation step (94°C for 1.Omin), an annealing step (60°C for 1.5 min) and an elongation step (72°C for 1.0 min), in that order. Analysis of PCR products. Aliquots (lo-20 pL) of the PCR product were analysed by agarose gel electrophoresis for size determination. Southern blotting to Zetabind membranes (Cuno, Meriden, U.S.A.) was performed according to the manufacturer’s specifications. The membranes were pre-hybridized in 6 x SSC, 50 mM sodium phosphate, pH 6.8, 5 x Denhardt’s solution and 100 mg/ml salmon sperm DNA for at least 1 h, and hybridized for 16 h at 37°C in the same solution containing 1 x lo6 dpm/ml of labeled olignoncleotide probe. Blots were washed once at room temperature for 15 min and once at 50°C for 15 min with 6 x SCC, 0.05% sodium pyrophosphate and were then exposed to Cronex film at room temperature for 1-5 h (DuPont), PCR products were run on 1.5% (w/v) agarose gels in TAE buffer, bands were cut from the gel, purified with GeneClean (BiolOl, La Jolla, U.S.A.), and the DNA sequence determined using a cycle sequencing strategy employing a fluorescence-based automated sequencer and dye-labeled terminators. Mapping

of upstream start sites in genomic

clone.

Radiolabeled primer-extension products were hybridized to Southern blots containing fragments of the genomic clone. The fragments were generated either by restriction enzyme digestion or by PCR amplification. First, the genomic clone was digested to completion in three separate reactions with HindIII, Ava I or Dpn I. Second, pairs of primers were used for PCR amplification of the genomic clone to yield fragments of varying lengths including some whose 3’-end was 400 bp upstream of the ATG codon. The fragments generated by each of these reactions were separated on agarose gels and transferred to a nylon membrane. Hybridization with the radiolabeled primer-extension products was performed essentially as described above, except that the probe consisted of half of a primer-extension product mix, with the other half loaded on a polyacrylamide gel for analysis.

1167 RESULTS

Characterization of genomic clones and sequence

We isolated

a 15.7 kb genomic clone for human

Fey RIIA by screening a human genomic library with an 800 bp Fey RIIA cDNA isolated in our laboratory.

Restriction enzyme digestion and targeted sequence analysis revealed that the genomic clone 1Fc8-1 extended from 5’ to the ATG translation initiation codon through 3’ to the first cytoplasmic exon (Fig., 1). We subcloned a 3.65 kb Sal I-EcoRI fragment of the 5’ end of 1 Fc8-1 and determined its restriction digestion map and DNA sequence (Fig. 1, GenBank accession number pending). The clone, called p4. lF, includes on its 5’-end 3.4 kb of previously unreported sequence while the 3’ end includes 85 bp of the first protein-coding exon (Sl) followed by 131 bp of intron. The sequence of the p4.1F insert was determined in both directions in is entirety. Figure 1 depicts the sequence and major features of 2000 bp 5’ to the ATG translation initiation codon. Determination of the transcription start sites Primer extension. Primer extension of RNA from the HEL cell line revealed multiple transcription start sites giving rise to three specific bands (280, 240 and 80 bp) when oligonucleotide primer 16M was used, which extends from +42 to + 66 bp beyond the ATG initiation codon (Fig. 2). The same results were obtained with RNA from U937 and K562 cells (data not shown). By subtraction of the 66 bp within the translated region from the size of the bands, the sizes of the 5’ UT regions corresponding to the primer extension products are calculated to be approximately 214, 174, and 14 bp. RNase protection studies were done to independently confirm the primer extension results. RNase protection. RNase protection was performed with riboprobes synthesized from the SP6 promoter 3’ to the p4.1F genomic insert in pGEM. A riboprobe containing 336 bp 5’ to ATG as well as 85 bp of translated region was created by linearizing p4.1F at a Dpn I site (Fig. 3A). If the multiple transcription start sites suggested from the primer-extension studies were contained within this 336 bp region and not generated from upstream splicing events, then protected mRNA species of approximately 100, 260 and 300 nt would be seen following RNase digestion. Only two specifically protected clusters of fragments sized at 100 and 129 nt are seen in Fig. 3B in lane 2 with HEL cell polyA+ RNA; they are absent in the polyA_ (lanes 1 and 7), and tRNA controls, (lane 8). As the amount of input RNA from U937 cells was doubled, these specifically protected complexes doubled in intensity, as expected for specific protection (20 pg in lanes 3 and 4 versus 40 pg in lanes 5 and 6). Identical results were seen with additional riboprobes, one extending 201 nt (Aua I site) 5’ to the ATG codon and the other 43 nt (Hind111 site, Fig. 3C, lanes l-10). Hence, the specifically protected 100 and 129 nt products with these probes indicate protection of 15 and 44 nt of 5’UT region.

1168

STEVEN E. MCKENZIE

S

EEB

et al.

EE

hFc 8-l 5UT Sl S2

Ll EC1 EC2 TM

1 kb I

c1-

\

ADDAH II

E

llllI

I

p4.1F

I

8

u

5UT

Sl

1 kb

Fig.l(A) GTTGGACTGA TGGCCTGAGA AGGATAACAT AGTGTATGGG ATGTCTGAAT ACATTCTCTG ACCTCCCTGG GATACGTCGC CTCAGGAGGT CAAGATTCCC TCTTCCTTAG ACTAAATACA TAGCATAAAG ATTATTCTAA CAAAACAACT MTAATAATT TTAATTTTAT CTGTCTCTAT CTCCAATGAA ATTGGATGTA TTCAAATGTT UTTATAACC CCAAACCACC =a AACATAGAGG TTTCAAGAAA AATTTTCTAC TTATCTTCTT TGGTCAATTT AGAATTTAAC AGACGGAGTC GGCTCACTGC TCCCGAGTAG TATTATTAGT CTCCTGACCT CAGGCTTGAA AAGAATACTC CGGGGTTTTC GCAGAGTGAA AATTGAGGAC

GGTGGGGTAT ACAATTTGTT TCTTTTGGCT TATTCCCCTA CTTCCATAAA TAAGATACTC GGTCTTTACT TATACTTCCC CCAGACTATG ACTGCACCAA GCACTGGCAA GGGGTGTCAC ATAGTAATTA GTACTTTATA CTATAAGGGA TTTTTATTTA AGAGATGAGA TATCTKTAA

TATTCTTGTT TTTTTTGAAA TAATAGATAG

CCATTATTTG 0 TGCTCCGCAT GAAAACCTAC AGAAPATTTC TTTAAGTTTC CCCAGGCTCT TGGCTCCTGT CAGCGATTAT TCGCTCTTTC AAGCTCTGCC CTGGGACTAC AGAGACGGGG CGTGATCCAC

CCACCGCGCC TAAGGAGGGG ACAAGGACAC CTCCTAACTG TGACGACAGC

TTTTCTAf&C

TTGTCTCTTA

ATGACTATGG

AGACCCAAAT (r CCATTGACAG

GCTGCTTCAA

AGTATATCTC GCTCCTCTCC GGAATCTCTC CAGGCCCCCC TCCACAGAAT AGCTTACTCC ACCCTGGACT GGCATGATAA CAGTGGTTCT TGAGCACCCT TATCTTCTCA CATTATTTM CAATTAATTT GGTACTATTC ATAAAATCCC TTTCACTCTG TCTTGACGGT GATTTTTTTT TGGAATTTTT TGCCTAAATC TGTTCCATTT

TTGGGTAAGC TTCTGCCACC GAAGTGTGCA TAGTCCTAGC TCAGACCCAT ATGTCTCTAG CAACTTGGCA GCATCCTGTA CTAGCAACCA CCACCTGCTC TTTCAGGTAA CAGGCCCATC CATTTTCTGA ATCTGACATT TATTTCATTA CCATTTCATT TTGCCCAGGA TTTCCTAAAA TTTCACACCT AGGTGATAAG CCCTCCAA?A TGAAAGAW

CGCAACAATA MGTGTTTCT GCTTACAAAT ATGTTTGm P&.TATGTGT

-1851 -1801 -1751 -1701 -1651 -1601 -1551 -1501 -1451 -1401 -1351 -1301 -1251 -1201 -1151 -1101 -1051 -1001 -951 -901

GGGTAGTTCC

TTACAATTTC

TGTAATGGAA

-851

TTGCAGTTTC AGTAGGAGAA TGTCCTATGG TCTCCCCTCT TGGGAACCAA

AGTGGGAGAA AATATTGTGA GTATAAAAAA TCTCCGTCAG CTTCTCTAAT TTTTTTTTTT GAGTACAGTG CGCCATTCTC GCCCGGCTAA TTAGCCAGGA CTCCCAAAGT

AAGGTTCCCG AX+ACAAGa TAGCTGGACA GCTTCATTTC TCCTTGGCTA

TGCCTCAGCC ATTTTTTTTG TGGTCTCGAT GCTGAGATTA TAAACAGAAA TAACCCTCCC TATCTATAAA ECAGAAACAG TTCCTTCCTC

-801 -751 -701 -651 -601 -551 -501 -451 -401 -351 -301 -251 -201 -151 -101 -51

CAGTGCTGGG

-1

GTATGTCCCA

GAAACCTGTG

+50

GCTGGGTGAG

TGAGGGTCAT

+lOO

GTATAGAAAT

CCTTTTTTTT GCCCAAGCTG

TCCGGGTTCA CTGCCACCGC TTTCACCGTG CCGCCTTGGC CGGCCGACCA TATACCGGCC AGCTTGGTGT GAATTGTTCC TGCACAAGAA AAACCCACTG F GTCTCAGAAT TTTTGCTGCT

GGAATTATCT

AGGAGAACAG GTAATAGACT TGATTGGTTB GATGAACCAT 0 GACGTTGGCA

AGAGTATAGA TGCATATGTG TTTAATTTTC CATGATCACA ATACCCTGTC CTCTGCTGGC CTCCATGGAG TGCCTTGACT CTATGGCGAG TGGGAAACAT AATACCACCC CATAAAAGAG TATTGGTGGT ATTTAACCTT AATTTACAAA TACTTTCTAA TGGTCTCGAA

TTTTTTTTTG GCGCTATCTG

-1951 -1901

Fig.l(B) Fig. 1. Physical map of the genomic clones and sequence of 2000 bp of the 5’ flanking region of the human Fey RIIA gene. (A) Physical map: 1Fc8-1 is the genomic clone from which p4.1F was subcloned. S = Sal I, :E = EcoRI, H = HindIII, A = Ava I, D = Dpn I, B = BarnHI. Exons are boxed; the 5’UT exon described in this report is hatched. S = signal peptide exon, EC = extracellular, TM = transmembrane, C = cytoplasmic, 3’UT = 3’untranslated (B) Sequence: + 1 is the ATG translation initiation codon, shown in bold. Transcription start sites found at positions -881 and - 15 are indicated by 0. The 5’ intron (GT) splice site at position -752 and alternative 3’ intron (AG) splice sites at positions - 109 and -42 are in bold and underlined. * denotes the 5’ end of 10M (+ 29) and oligonucleotide - 7CS at - 848; -= denotes the 3’ end of antisense oligonucleotides 16M (+ 66M). CCAAT and TATA box elements are underlined. Additional elements which may be involved in the regulation of transcription as defined in Table 1 are in italics.

Human Fey RIIA gene S-flanking region

PA(+)

PA(-)

-310

280 240

/z 271,281 i-234

1169

products were hybridized to Southern blots containing size-fractionated fragments of the genomic clone. Specific genomic fragments which were positive for annealing of this probe were identified. These included fragments whose 3’-end mapped more than 400 bp 5’ to the ATG. These results supported our hypothesis that transcription began in a discrete 5’ untranslated exon (data not shown). The sequence and restriction map of our genomic clone (Fig. 1) permitted identification of an area of common overlap among the positive fragments approximately 1 kb 5’ to the ATG, which suggested the location of the upstream start sites. The sequence in this region was examined for putative start sites by identifying appropriately positioned CCAAT and TATA elATG 1

I3361

_ I2OU

1

A

l

I431 H--l

Fig. 3(A)

80 -72 Fig. 2. Mapping of transcription start sites by primer extension. Using HEL cell RNA, specific bands of 280, 240 and 80 bp were seen with polyA + RNA, (left lane), but not polyA RNA (right lane). This primer (16M) was associated with moderate background of non-specific bands with polyA_ RNA and tRNA. Position of molecular weight markers is indicated on the right. Protection of 15 nt SUT region confirms the transcription start site at approximately - 14 predicted in the primer-extension studies. Since no specifically protected bands in the size range of 140 to 500 nt could be seen in the RNase protection experiments, transcription start sites predicted by the 240 and 280 bp primer-extension products did not map within the 336 bp 5’ to ATG. Furthermore, the primer-extension results did not predict a start site 44 nt upstream from the ATG. Sequence analysis of this region of the genomic clone showed however that near the site 44 nt upstream from the ATG was a potential intron 3’-splice acceptor site. These observations led to our hypothesis that other start sites mapped further upstream in a 5’-untranslated exon. Therefore, they were not contained within the region corresponding to the probes in these RNase protection studies. Mapping of upstream start sites in the genomic clone. In an effort to locate the upstream start sites within the 5’-flanking sequence, radiolabeled primer-extension

-310 -271,281

-234

-194

-118

12345678 Fig. 3(B)

1170

STEVEN E.

MCKENZIE et (11

- 234

-194

RT-PCR with primers jiom the putative 5’-UT tj.\Ton. RT-PCR with each sense oligonucleotide in the putative 5’UT exons (- 12KS or -7CS) and the antisense oligonucleotide 10M in the protein-coding region was performed (Fig. 4A). The - 12KS and 10M pair yielded no products, while the - 7CS and 10M pair yielded three products with RNA from HEL, K562 and U937 cells, as well as megakaryocytic cell lines Dami and CHRF-288 (Fig. 4B). One (RI) is the same size as independently amplified genomic DNA; the other two products (R2, R3) were smaller, consistent with their origin from processed RNA. These results confirm that transcription is initiated far upstream of the ATG translation initiation codon, specifically in the exon containing the -7CS sequence. For direct characterization of the amplified Fey RIIA mRNA, the RT-PCR products were sequenced. Characterization

118

12345678910

Fig. 3(C) Fig. 3. Mapping of transcription start sites by RNase protection. (A) A schematic of the RNase protection experiments indicates the size and origin of the riboprobes with respect to the p4.lF genomic subclone and D, A and H restriction sites. (B) The autoradiograph is shown for the experiment using the riboprobe created with Dpn I digestion (total riboprobe length 553 nt, including 336nt 5’ of ATG). Specifically protected clusters of fragments are seen at 129 nt and 100 nt in HEL polyA + RNA (lane 2) but not polyA- RNA from HEL, U937 cells (lanes 1, 7) or in tRNA (lane 8). When the input amount of RNA from U937 cells was doubled, the intensity of the bands at N 100 and 129 nt doubled as expected for specific protection (20 pg in lanes 3 and 4 versus 40 pg in lanes 5 and 6). The position of molecular weight markers is indicated on the right. Note that the resolution of the polyacrylamide gel in the 100 to 130 nt region is such that bands within clusters are within 2 nt in size. Note also that no specifically protected complexes are seen in the region between 130 and 500 nt. (C) The autoradiograph is shown for the experiment using the riboprobe created with Hind111 digestion (total riboprobe length 260 nt, including 43 nt 5’ to ATG). Identical specific protection of 100 and 129 nt complexes is seen in total (lanes 1 and 2) and polyA+ (lane 4) but not polyA_ RNA (lanes 3 and 9) or tRNA (lane 10). The specific bands double in intensity when input total RNA is doubled from 20 pg (lanes 5 and 6) to 4Opg (lanes 7 and 8). ements and a GT 5’-consensus splice donor site within 200 bp downstream. Two such areas were identified, and one sense oligonucleotide from each area, designated - 7CS (position - 848, Fig. 1) and - 12KS (position - 1687, Fig. 1) were made. We hypothesized that these primers were located in 5’UT exon(s) which began with the upstream start sites.

of the 5’-end of FcyRIIA

mRNA

The size and sequence of the three RT-PCR products, shown in Fig. 4C, shows that two bands (R2-234 bp, R3-167 bp) were colinear with genomic sequence until a 5’-GT consensus splice donor signal (position -753 in Fig. 1). Their sequences then continued through one of two alternative 3’-AG consensus splice acceptor sites, 109 or 42 bp 5’ to the ATG codon (positions - 109, -42 Fig. 1). They each then extend into the next exon which includes both 5’-untranslated and protein-coding portions (Sl). The occurrence of a splice site at -42 is consistent with the RNase protection results, while no protected species corresponding to the - 109 splice site was detected. One RT-PCR band (Rl) corresponded to the same size (875 bp) and sequence of the independently -7CS

1QM

--_)

t

ATG

Rl GT-

a75 bp

-AG

GT-

167 bp Fig. 4(A)

G,Rl

12345676

Fig. 4(B)

Human Fey RIIA gene S-flanking a

1171

region

1875 bp)

CATAGAGGGA

AAACCTACTT

GCAGTTTCAG

TGGGAGAAAA

GGTTCCCGTT

AAACCTACTT

GChGTTTChG

TGGGAGAAAA

GGTTCCCGTT

CATAGAGGGA

AAACCTACTT

GCAGTTTCAG

TGGGAGAAAA

GGTTCCCGTT

AACAAGGTAA

R2

(234 bp)

CATAGAGGGA S

(167 bp)

TCAAGAAAAG

AAAATTTCAG

TAGGAGAAAA

TATTGTGAAA

TCAAGAAAAG

AAAATTTCAG

TAGGAGW

TATTGTGAAA

AACAAG----

TCAAGAAAAG

AAAhTTTCAG

TAGGAGMAA

TATTGTGAAA

AACFLAG----

GATTGGTT~ _ ______-__

CAGAAACAGA CAG-CAGA

TTTTCTACTT

(same as genomic) __

__

____---_--

ATTGAGGACT

GACGACAGCT

GCACAAGAAG

ATGAKCATT

TCCTTCCTCT

ATTGAGGACT ___---__-_

GACGACAGCT __--___--_

GCACAAGAAG ---__-----

ATGAACCATT ----------

TCCTTCCTCT ---__-----

TTTCTAA_GCT

TGTCTCTTAA

AACCCACTGG

ACGTTGGCAC

AGTGCTGGGA

TTTCTAAGCT

TGTCTCTTAA

AACCCACTGG

ACGTTGGCAC

AGTGCTGGGA

--------CT

TGTCTCTT~

AACCCACTGG

ACGTTGGCAC

AGTGCTGGGA

TGACTATGGA TGACTATGGA TGACTATGGA

GACCCAAATG GACCCAAATG GACCCAAATG

TCTCAGA TCTCAGA TCTCAGA

Fig. 4(C)

Fig. 4. Localization of the 5’UT exon by RT-PCR. (A) Schematic of the RT-PCR experiment and the origin of three species generated by the primers -7CS and 10M with respect to the genomic sequence. (B) Ethidium bromide-stained agarose gel depicting the three species amplified by primers -7CS and IOM. The species are G and RI (875 bp), R2 (234 bp) and R3 (167 bp). Lanes 1 and 8 are 0X174-Hae III molecular weight markers, while lane 2 is from the genomic clone p4.1F. The RNA sources include Dami (lane 3), CHRF 288 (lane 4), HEL (lane 5), U937 (lane 6) and K562 (lane 7). The 167 bp product is present in all RNA lanes but more readily visualized in lanes 3-5. (C) Sequence of RT-PCR products, with the sequences vertically aligned and gaps indicated by dashes (---). The intron splice sites GT and AG are underlined, as are the primer sequences. The ATG translation initiation codon is in bold. The 875 bp product has the identical sequence to the genomic. This product could arise either from an unprocessed RNA transcript or some genomic DNA contaminating the RNA preparation, or both.

amplified colinearity clone.

genomic between

PCR product this transcript

(G), indicating and the genomic

DISCUSSION

Our

first

step

in understanding the regulation of has been to clone a 3.65 kb region of the 5’ end of the Fey RIIA gene. We determined that its DNA sequence includes 3.4 kb of previously unpublished sequence of the S-flanking region. We identified two transcription start sites (the T at position -881 and the T at position - 15 in Fig. 1) associated with three possible 5’UT regions. Multiple complementary methods of RNA and DNA analysis were necessary to map the start sites and identify the heterogeneity of 5’ ends in Fey RIIA mRNA. Figure 5 depicts the organization of the 5’ end of the

Fey RIIA transcription

human Fey RIIA gene. The start site in the discrete 5’UT exon beginning 881 nt 5’ to the ATG is flanked by TATA and CCAAT box elements mapping 32 and 72 bp 5’,

respectively. There are no CCAAT or TATA consensus sequences for the initiation of transcription 15 bp 5’ to ATG, but there are GATA and SPl sites in the flanking region as well as homology at this putative start site to the initiator elements previously described for genes lacking a TATA box (Smale and Baltimore, 1989; Smale et al., 1990). Fey RIIA transcripts have been identified with 5’UT ends of at least three different sizes and sequences, as depicted in Fig. 5. In the mRNA with the longest 5’ UT region (237 nt) there is a sequence present that could form a stem-loop structure. This sequence is lacking in the alte~ativeIy sphced form (R3, 167 bp). This potential stem-loop s~onda~ structure in the region between 109 and 42 nt proximal to the ATG may interfere with hybridization of a labelled probe and may

1172

STEVEN

E. MCKENZIE et al.

C’UT-Sl

S2

EC1 EC2

TM

Cl

C2

C3-3’UT

Fig. 5. The organization of the 5’-end of the human Fey RIIA gene. Exons are depicted There are two potential transcription start sites indicated by the large arrows above the of the more 5’-start site is associated with alternative splicing of the 3’ end of the first indicated by the lines below the gene. The position of the ATG translation initiation indicated by the small vertical arrow.

have precluded detection of such a species in our RNase protection

assays.

We can compare the 5’-end of human Fey RIIA to murine Fey RIIB and human Fey RI. The 5’-end of the murine FcyRII/? gene is encoded by two discrete 5’ UT exons. In addition there are two transcription start sites, the major one 585 bp 5’ to the ATG translation initiation codon and a minor one 632 bp 5’ to ATG (Hogarth, 1991). When the major murine Fey RIIfl start site is used, an mRNA with a 5’UT region of 367 bp is produced, even longer than the 237 bp 5’UT region for the longest form of human Fey RIIA mRNA. The human Fey RI gene has two different transcription start sites, one of which is used following induction by IFN-y, but its 5’UT region is encoded by a single exon continuous with the ATG translation initiation codon (van de Winkel, 1991; Pearse, 1991). Whether lineage-specific factors or extracellular stimuli influence the choice of start site or 5’UT splice site is currently unknown. Of special interest to us is any possible linkage between the choice of start site or 5’UT splice site and the alternative splicing of the transmembrane exon which generates a soluble receptor in addition to the integral membrane form (Rappaport et al., 1991). Since increasing numbers of hematopoietic cell receptors important to the immune and inflammatory systems appear to occur in both soluble and integral membrane forms, it will be interesting to compare the organization of their 5’-flanking transcriptional regulatory regions and 5’UT regions (Pleiman et al., 1991). The

by boxes. lines. Use intron, as codon is

studies described here will provide the foundation for functional expression studies aimed at promoter and enhancer regions within and flanking the human Fey RIIA gene. Based on the location of the transcription start sites, we have identified several elements which may mediate the regulation of transcription within the 5’ flanking region, including potential binding sites for GATA-1 and elements that mediate response to glucocorticoids (GRE), interferon (IRE), cytokines (CK-l), Vitamin D and phorbol esters (AP-I), (Table 1). Of special interest are the elements which may mediate response to glucocorticoids and interferon, since the effects of glucocorticoids and IFN-?/ on Fey RIIA transcription have been described (Comber, 1989). The presence of elements for response to hematopoietic cytokines (e.g. GM-CSF), phorbol ester, and Vitamin D also is notable; we have identified effects of these modulators on FqRIIA protein expression (Rossman et al., 1991 and unpublished observations). Because human Fey RIIA is expressed in megakaryocytic and myeloid cells, the 5’-flanking regions from several genes expressed in these lineages were compared to this region of Fey RIIA. Fey RIIA has several potential binding sites for GATA-1 and GATA-2, hematopoietic transcription factors expressed in erythrocytes, megakaryocytes and mast cells (Martin and Orkin, 1987). There is also an element previously identified only in the platelet genes GPIIb and PF4, namely TTTTAT. In studies of the human GPIIb gene, this element bound a megakaryocyte-specific transcription factor (Uzan et al., 1991).

Table 1. Sequence elements in the 5’-flanking region of hFcy RIIA (1) Elements flanking the first transcription Binding site Sequence

start site (- 881, Fig. 1) Position(s)

GATAl, AGATAG GATA2 GGTCTCGAACTGTCTCT GRE CK-1 GAGATTCCAC AP-4 CAACTTGG AGAAATGAAGT IRE TTTTAT Vitamin D GGTGA (2) Elements flanking the second transcription Sequence Binding site SP-1 GATA GRE(-) -

CCCGCC AGATAG CTGACCTCGTGATCCA (T)n

-906, - 1141, - 1392, - 1431 -1159 -1184 - 1670 - 1875 - 1196, - 1239 - 1019 start site (- 15, Fig. 1) Position(s) -331 -161, -216 - 347 - 578

Shared with GPIIb, PF4 GPIbcr, globin HFcy RI, mFcy RII/3 G-CSF, GM-CSF, IL-3 mFcy RIIP hFcy RI, mIL-7R GPIIb, PF4 osteopontin, osteocalcin Shared with ubiquitous (see above) prolactin PF4

Human Fey RIIA gene S-flanking region The human platelet factor 4 gene has 54 consecutive Ts in a row, an area conserved in the rat PF4 gene and demonstrated to regulate transcription (Doi et al., 1987; Eisman et al., 1990). Fey RIIA has 27 consecutive Ts proximal to one start site, and short spans of 11 to 14 Ts in the region 5’ of the other start site. The significance of these areas remains to be studied. The occurrence of three different possible 5’-untranslated regions in Fey RIIA mRNAs may result in different relative translational efficiencies for the different transcripts. The effects of different 5’UT regions on the translational efficiency of mRNAs has been reviewed by Kozak (1991), and suggest significant effects of length and secondary structure on translation efficiency. For FqRIIA, the presence of a stem-loop in the mRNA with the longest 5’ UT is notable. Choice of the alternative splice acceptor site removes the stem-loop and makes a shorter mRNA transcript, perhaps rendering this mRNA more translatable. Further studies of the regulation of transcription of Fey RIIA and other Fey R-family genes may offer insights into the regulation of transcription for many genes in diverse lympho-hematopoietic lineages. Acknowledgements-This work was supported in part by research grants from the National Institutes of Health (DK16691 and HL 40387) and a training grant (ST32 HL07150). The Nucleic Acid/Protein Core facility of the Children’s Hospital of Philadelphia was essential for DNA sequence analysis and oligonucleotide synthesis. We thank Dr Mortimer Poncz and Dr Randi Isaacs for helpful discussions, and Dr Hugh Nicholas of the Pittsburgh Supercomputer Center for help with RNA folding studies.

REFERENCES Anderson G. P. and Anderson C. L. (1990) Signal transduction by the platelet Fc receptor. Blood 76, 11651172. Beato M. (1989) Gene regulation by steroid hormones. Cell 56, 335-344. Brooks D. G., Qiu W. Q., Luster A. D. and Ravetch J. V. (1989) Structure and expression of human IgG FcRII(CD32). 1. exp. Med. 170, 1369-1385. Chomczynski P. and Sacchi N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction. Anal. Biochem. 162, 156-159. Comber P. G., Rossman M. D., Rappaport E. F., Chien P., Hogarth P. M. and Schrieber A. D. (1989) Modulation of

human mononculear phagocyte Fey RI1 mRNA and protein. Cell Immun. 124, 292-307. Doi T., Greenberg S. M. and Rosenberg R. D. (1987) Structure of the rat Platelet Factor 4 gene: a marker for megakaryocyte differentiation. Mol. Cell Biol. 7, 898-904. Eisman R., Surrey S., Ramachandran B., Schwartz E. and Poncz M. (1990) Structural and functional comparison of the genes for human Platelet Factor 4 and PF4dt. Blood 76, 336344. Fugman D. A., Witte D. P., Jones C. L. A., Aronow B. J. and Liebetman M. A. (1990) In vitro establishment and characterization of a human megakaryoblastic cell line. Blood 75, 1252-1261. Greenberg S. M., Rosenthal D. S., Greeley T. A., Tantravahi R. and Handin R. I. (1988) Characterization of a

1173

new megakaryocytic cell line: the Dami cell. Blood 72, 1968-1977. Hogarth P. M., Witort W., Hullet M. D., Bonnerot C., Even J., Fridman W. H. and McKenzie I. F. C. (1991) Structure of the mouse /?Fcy receptor II gene. J. Immun. 146, 369-376. King M., McDermott P. and Schreiber A. D. (1990) Characterization of the Fey receptor on human platelets. Cell. Zmmun. 128,462476. Kozak M. (1991) Structural features in eukaryotic mRNAs that modulate the initiation of translation. J. biol. Chem. 266, 19867-19870. Liao J., Ozono K., Sone T., McDonnell D. P. and Pike J. W. (1990) Vitamin D receptor interaction with specific DNA requires a nuclear protein and 1,25-dihydroxyvitamin D,. Proc. natn Acad. Sci. USA 87, 9751-9755. Martin D. I. and Orkin S. H. (1990) Transcriptional activation and DNA binding by the erythroid factor GF-l/NFEl/Eryfl. Genes Dev. 4, 18861898. Majumdar S., Gonder D., Koutsis B. and Poncz M. (1991) Characterization of the human b-thromboglobulin gene. J. biol. Chem. 266, 5785-5789. Mitchell P. J. and Tijan R. (1989) Transcriptional regulation in mammalian cells by sequence-specific DNA-binding proteins. Science 245, 371-378. Nishizawa M. and Nagata S. (1990) Regulatory elements responsible for inducible expression of the Granulocyte Colony-Stimulating Factor gene in macrophages. Mol. cell. Biol. 10, 2002-2011. Noda M., Vogel R. L., Craig A. M., Prahl J., DeLuca H. F. and Denhardt D. T. (1990) Identification of a DNA sequence responsible for binding of the 1,25_dihydroxyvitamin D, receptor and 1,25-dihydroxyvitamin D, enhancement of mouse secreted phosphoprotein 1 (Spp-1 or osteopontin) gene expression. Proc. natn Acad. Sci. USA. 87, 9995-9999. Pearse R. P., Feinman R. and Ravetch J. V. (1991) Characterization of the promoter of the human gene encoding the high affinity IgG receptor: transcriptional induction by y-interferon is mediated through common DNA response elements. Proc. natn Acad. Sci. USA. 88, 11305-l 1309. Pleiman C. M., Gimpel S. D., Park L. S., Harada H., Taniguchi T. and Ziegler S. F. (1991) Organization of the murine and human Interleukin-7 receptor genes: Two mRNAs generated by differential splicing and presence of a Type I-Interferon-inducible promoter. Mol. cell. Biol. 11, 3052-3059.

Poncz M., Surrey S., LaRocco P., Weiss M. J., Rappaport E. F., Conway T. M. and Schwartz E. (1987) Cloning and characterization of a platelet factor 4 cDNA derived from a human erythroleukemic cell line. Blood 69, 219-223. Qiu W. Q., de Bruin D., Brownstein B. H., Pearse R. and Ravetch J. V. (1990) Organization of the human and mouse low-affinity Fey R genes: duplication and recombination. Science 248, 732-735. Rappaport E. F., Cassel D. L., Walterhouse D. O., McKenzie S. E., Surrey S., Keller M. A., Poncz M., Chien P., Schreiber A. D. and Schwartz, E. (1991) Characterization of a cDNA clone and transcripts coding for a soluble form of the human Fey receptor, Fey RIIA, submitted. Rossman M. D., Ruiz P., Gomez F., Rottem M., Comber, P. and Schreiber A. D. (1991) Modulation of macrophage Fey receptors by ffiM-CSF. Clin. Res. 39, 239A. Sambrook J., Fritsch E. F. and Maniatis T. (1989) Molecular cloning: A laboratory manual (second edition). Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

1174

STEVENE. MCKENZIE et al.

Schreiber A. D., Gomez F., Levinson A. I. and Rossman M. D. (1989) The Fey receptors on human macrophages. Transfusion Med. Rev. 3, 281-293. Seki T. (1989) Identification of multiple isoforms of the low-affinity human IgG Fc receptor. Immunogenetics 30, 5-12. Smale S. T., Schmidt M. C., Berk A. J. and Baltimore D. (1990) Transcriptional activation by Spl as directed through TATA or initiator: specific requirement for mammalian transcription factor IID. Proc. natn Acad. Sci. USA 87, 45094513. Smale S. T. and Baltimore D. (1989) The “initiator” as a transcription control element. Cell 57, 1033113. Stuart S. G., Simister N. E., Clarkson S. B., Kacinski B. M., Shapiro M. and Mellman I. (1989) Human IgG Fc receptor (hFcRI1; CD32) exists as multiple isoforms in macrophages, lymphocytes and IgG-transporting placental epithelium. EMBO J. 8, 3657-3666. Tsai S. F., Strauss E. and Orkin S. H. (1991) Functional analysis and in vivo footprinting implicate the erythroid transcription factor GATA-1 as a positive regulator of its own promoter. Genes Dev. 5, 919-931.

Unkeless J. C. (1989) Function and heterogeneity of human Fc receptors for immunoglobulin G. J. clin. Invest. 83, 355-361. Uzan G., Prenant M., Prandini M.-H., Martin F. and Marguerie G. (1991) Tissue-specific expression of the platelet GPIIb gene. J. biol. Chem. 266, 8932-8939. van de Winkel J. G., Ernst L. K., Anderson C. L. and Chiu I.-M. (1991) Gene organization of the human high affinity receptor for IgG, Fey RI (CD64). J. biol. Chem. 266, 13449-13455. van de Winkel J. G. and Anderson C. L. (1991) The biology of leukocyte IgG -Fc receptors. J. leukocyte Biol. 49, 51 l-524. Walterhouse D. O., Cassel D. L., Schreiber A. D., Meister R. P., Schwartz E. and Rappaport E. F. (1988) Characterization of HEL cell Fey RI1 cDNA clone lacking the sequence coding for the transmembrane region. Blood 72, 344 (abstr). Wenger R. H., Wicki A. N., Kieffer N., Adolph S., Hameister H. and Clemetson K. J. (1989) The 5’ flanking region and chromosomal localization of the gene encoding human platelet membrane glycoprotein Ibcc. Gene 85, 517-524. Witsel A. L. and Schook L. B. (1990) Clonal analysis of gene expression by PCR. Biotechniques 9, 318-322.