colony-stimulating factor - NCBI

Proc. Natl. Acad. Sci. USA Vol. 83, pp. 7633-7637, October 1986 Biochemistry

Isolation and characterization of the cDNA for murine granulocyte colony-stimulating factor (murine fibrosarcoma NFSA cells/cDNA sequence/protein homology/growth factor/differentiation)

MASAYUKI TSUCHIYA, SHIGETAKA ASANO, YOSHITo KAZIRO, AND SHIGEKAZU NAGATA Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108, Japan

Communicated by Charles Yanofsky, June 26, 1986

In this report, we describe the isolation of the murine G-CSF cDNA from a recombinant X phage library prepared from the murine fibrosarcoma NFSA cell line, which produces G-CSF constitutively. The murine G-CSF cDNA was identified by using cross-hybridization with human G-CSF cDNA under a low-stringency condition. The cDNA was expressed in monkey COS cells under the simian virus 40 (SV40) early promoter. The protein produced by COS cells had an ability to stimulate the granulocyte colony formation in bone marrow cells and to support the proliferation of murine NFS-60 myeloid leukemic cells.

A cDNA sequence coding for murine granuABSTRACT locyte colony-stimulating factor (G-CSF) has been isolated from a cDNA library prepared with mRNA derived from murine fibrosarcoma NFSA cells, which produce G-CSF constitutively. Identification of murine G-CSF cDNA was based on the cross-hybridization with human G-CSF cDNA under a low-stringency condition. The cDNA can encode a polypeptide consisting of a 30-amino acid signal sequence, followed by a mature G-CSF sequence of 178 amino acids with a calculated Mr of 19,061. The nucleotide sequence and the deduced amino acid sequence of murine G-CSF cDNA were 69.3% and 72.6% homologous, respectively, to the corresponding sequences of human G-CSF cDNA. The murine G-CSF cDNA, when introduced into monkey COS cells under the simian virus 40 promoter, could direct the synthesis of a protein that can stimulate the granulocyte colony formation from mouse bone marrow cells and support the proliferation of murine NFS-60 myeloid leukemia cells.

MATERIALS AND METHODS Cell Lines and Isolation of mRNA. The murine fibrosarcoma cell line NFSA was kindly provided by Mikio Shikita (National Institute of Radiological Sciences). The cells were grown in Dulbecco's minimal essential medium (Nissui Seiyaku, Tokyo). Total cellular RNA was extracted from about 4 x 107 cells by the guanidine thiocyanate method (11), and poly(A)+ RNA was selected by oligo(dT)-cellulose column chromatography (12). Construction of the cDNA Library. Double-stranded DNA complementary to NFSA mRNA was synthesized as described (13), methylated with EcoRI methylase (New England Biolabs), and ligated with the EcoRI linker. The double-stranded DNA was then digested with EcoRI and size-fractionated on 1.2% agarose gel (Low Gel Tempurature, Bio-Rad). DNA ranging from 1,200 to 1,800 bp was recovered, and ligated with an EcoRI-digested Xgtl0 vector (14). The hybrid DNA was packaged in vitro, and a cDNA library consisting of 1.0 x 106 plaque-forming units (pfu) was constructed on Escherichia coli C600 high-frequency lysogeny cells.

Colony stimulating factors (CSFs) have been identified as factors that can allow proliferation and differentiation of hematopoietic progenitor cells from bone marrow on semisolid culture systems (1, 2). In the murine system, four different CSFs-i.e., granulocyte-macrophage CSF (GMCSF), granulocyte CSF (G-CSF), macrophage CSF (MCSF), and interleukin 3 (IL-3)-have been highly purified and characterized (2), and the gene structures for two of them (GM-CSF, IL-3) have been determined (3-5). Among these CSFs, murine G-CSF has been purified initially from the medium conditioned by lung cells from mice injected with bacterial endotoxin (6). The purified murine G-CSF has a Mr of 24,000-25,000 (6) and is distinguished from other CSFs by its ability to stimulate exclusively neutrophilic granulocyte colony formation from bone marrow cells and to induce the terminal differentiation of myeloid leukemia cells such as WEHI-3B D+ in vitro (6). Recently, we (7, 8) and others (9) have reported the isolation and expression of the cDNA for human G-CSF from cDNA libraries constructed with mRNA prepared from human carcinoma cells that produce G-CSF constitutively. It was found that there are two different G-CSF mRNAs (G-CSFa and G-CSFb mRNAs) for human G-CSF (8). GCSFa and G-CSFb mRNAs can code for polypeptides consisting of 207 and 204 amino acids, respectively, both of which are functionally active. Although G-CSF has no apparent species specificity between the murine and human system (10), the availability of the purified murine G-CSF by recombinant DNA technology might prove valuable in the study of the in vivo and in vitro functions of G-CSF in the murine model system.

Hybridization and DNA Sequence Analysis. Blot (15) and plaque hybridization (16) were carried out as described (17) except that the hybridization temperature was lowered to 37°C and the filter was washed at 42°C in 15 mM NaCl/1.5 mM sodium citrate, pH 7.0/0.1% NaDodSO4. The nucleotide sequence of cDNA was determined by the chain-termination method after subcloning into M13 phage derivatives (18). Transfection of COS Cells and in Vitro Colony-Formation Assay. COS cells (2 x 106) in 10 ml of Dulbecco's minimal essential medium were transfected with 20 ,Ag of plasmid DNA. At 72 hr after transfection, the medium was collected and assayed for G-CSF activity as described (19). In brief, 5 x 104 bone marrow cells from a C3H/He mouse were cultured with or without 0.1 ml of test sample in 1 ml of McCoy's 5A medium containing 40% horse serum and 0.3% agar. After incubation at 37°C in humidified 5% CO2 in air for 5 days, colonies consisting of >50 cells were counted. Abbreviations: CSF, colony-stimulating factor; G-CSF, M-CSF, and GM-CSF, CSFs that can stimulate colony formation of granulocytes, macrophages, and granulocytes/macrophages, respectively; IL-3, interleukin 3; SV40, simian virus 40; bp, base pairs.

The publication costs of this article were defrayed in part by page charge

payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

7633

7634

Biochemistry: Tsuchiya et al.

Cell-Proliferation Assay. The NFS-60 cell line (20) obtained from James N. Ihie (National Cancer Institute-Frederick Cancer Research Facility, Frederick, MD) was routinely maintained in RPMI 1640 medium (GIBCO) containing 10% fetal calf serum (Microbiological Associates), 50 puM 2mercaptoethanol, and 4% of the COS cells' supernatant (a gift from Ken-ichi Arai, DNAX Institute) transfected with mouse IL-3 expression plasmid (3) or 10 ng of the purified human G-CSF per ml (19). The proliferation assay was carried out in 96-well microtiter plates as described (21) with a minor modification. Samples to be examined were serially diluted 1:2 in 50 tkl of RPMI medium and mixed with 50 ,ul of NFS-60 cells (1 x 106 cells per ml), which had been washed extensively with the medium without IL-3 or G-CSF. The cultures were incubated for 24 hr at 370C, after which 0.25 ACi (1 Ci = 37 GBq) of [3H]thymidine (specific activity, 2 Ci/mmol; Amersham) was added per well and further incubated for 6 hr at 37TC. The cells were subsequently harvested with an automated cell harvester unit (Titertek, Flow Laboratories) onto filter paper and were assayed for [3H]thymidine incorporation.

RESULTS Presence of mRNA Homologous to Human G-CSF mRNA in Murine Fibrosarcoma NFSA Cells. Murine fibrosarcoma NFSA cells constitutively produce proteins having CSF activity (22). CSF proteins produced by NFSA cells were separated into two fractions. One stimulates colony formation of mainly granulocytes and is active on human as well as mouse bone marrow cells, while the other stimulates colony formation of macrophages from only mouse bone marrow cells (23). Since murine G-CSF is known to work on human cells (10), the former fraction was thought to be the murine equivalent of human G-CSF. To examine this possibility, mRNA from NFSA cells was analyzed by blot hybridization using human G-CSF cDNA as probe. Under a low-stringency hybridization condition, a single band of about 1.5 kb could be detected in mRNA from NFSA cells with human G-CSF cDNA (pBRG-4 cDNA in ref. 7) as a probe (Fig. 1). The size of the hybridizing mRNA was slightly smaller than the human G-CSF mRNA in human CHU-2 squamous carcinoma cells (7), which produce G-CSF constitutively. The result indicates that NFSA cells synthesize an mRNA highly homologous to human G-CSF mRNA. Identification of Murine G-CSF cDNA Clone. To isolate a cDNA clone containing the sequence coding for murine G-CSF, a cDNA library was constructed with mRNA from murine NFSA cells. The double-stranded cDNA, ranging from 1.2 to 1.8 kb, was size-fractionated on an agarose gel electrophoresis, and a cDNA library was prepared with the XgtlO vector system (14). By screening 1 x 106 clones with human G-CSF cDNA, a total of 24 hybridizing clones was obtained. Ten of the 24 clones were plaque-purified, and DNA was prepared from each clone. EcoRI digestion of the recombinant DNAs released fragments of 1.3-1.8 kb, and DNA from one clone having the =1.4-kb cDNA insert was subcloned to the EcoRI site of pBR327 (denoted as pMG2) and further characterized by restriction enzyme analysis and nucleotide sequence determination. Nucleotide Sequence Analysis of Murine G-CSF cDNA. Fig. 2 shows the restriction map of pMG2 and the sequence analysis strategy. The sequence was determined on both strands of the cDNA, crossing the restriction fragment junction. The nucleotide sequence (1,363 bp) is shown in Fig. 3, together with the deduced amino acid sequence. The

pMG2 cDNA contains a single open reading frame containing three potential initiation codons (nucleotide numbers 68-70, 95-97, and 104-106) like human G-CSF cDNA (7). Since the nucleotide sequences surrounding the first methionine codon are most optimal to the consensus sequence of higher-

Proc. Natl. Acad. Sci. USA 83 \1 1

(1986)

2 3

( WI' -

4.9 3

4.3. 92' - 0 -

FIG. 1. Blot-hybridization analysis of mRNA derived from the human CHU-2 and murine NFSA cell lines. Poly(A)+ RNAs were electrophoresed through 1.2% agarose gel containing formaldehyde and blotted onto a nitrocellulose filter. The EcoRI fragment of human G-CSF cDNA (pBRG-4 in ref. 7) was labeled by nick-translation (15) with [32P]dATP (specific activity, 3,000 Ci/mmol) and was hybridized as described. Lanes: 1, 2 ug of poly(A)+ RNA from human CHU-2 cells; 2 and 3, 5 and 10 ,ug of poly(A)+ RNA from murine NFSA cells, respectively; M, size markers (32P-labeled DNA fragments) run in parallel. Sizes are given in kb. Ori, origin of electrophoresis.

eukaryote mRNA (24), the first ATG codon was tentatively assigned as the initiation codon. The initiation codon is followed by 207 codons before a TAG termination codon at nucleotide positions 692-694 is encountered. The 3' noncoding region of 669 nucleotides contains the AATAAA polyadenylylation signal (25) at positions 1,346-1,351, although the poly(A) tract is not included in this cDNA. On the basis of the homology with the amino acid sequence of human G-CSF (7), Val-1 was assigned to the NH2-terminal amino acid of the mature, secreted murine G-CSF. The peptide containing amino acid residues -30 to -1 probably serves as a signal peptide for the secretion. The mature murine G-CSF consists of 178 amino acids with a calculated Mr of 19,061 and no potential N-glycosylation site (Asn-XaaThr) (26) is found. However, it is possible that murine G-CSF is O-glycosylated because the natural G-CSF purified from the mouse lung-conditioned medium has a Mr of 24,00025,000 and has been thought to be glycosylated (6). Actually, the common sequences for galactosamine attachment (27) -I

0

ZZ

Qq

L

I

:1

CL.

z :z

C. r:z

;_:

.

ll!

_

Q, Z,

-o

_

Z,

-,!

_

:

.L 200 bp

FIG. 2. Restriction map of murine G-CSF cDNA (pMG2) and the strategy for nucleotide sequence determination. The untranslated sequences are represented by a line, while the coding sequences are boxed. The dark portion indicates the sequence coding for mature G-CSF protein, while the white region represents the sequence coding for the putative 30-residue signal sequence. Arrows show the direction and the length of sequence determined by each independent

experiment.


Proc. Natl. Acad. Sci. USA 83 (1986)

10

30

50

70

7635 90

GTATAAAGGCCCCCTGGAGCTGGGCCCTGGCAGAGCCCAGAGCTGCAGCCCAGATCACCCAGAATCC ATG GCT CAA CTT TCT GCC CAG AGG Met Ala Gin Leu Ser Ala Gln Arg -30 110 130 150 170 CGC ATG AAG CTA ATG GCC CTG CAG CTG CTG CTG TGG CAA AGT GCA CTA TGG TCA GGA CGA GAG GCC GTT CCC CTG GTC ACT GTC AGC GCT Arg Met Lys Leu Met Ala Leu Gln Leu Leu Leu Trp Gin Ser Ala Leu Trp Ser Gly Arg Glu Ala Val Pro Leu Val Thr Val Ser Ala -20 -10 -1 1

190 210 230 250 270 CTG CCA CCA TCC CTG CCT CTG CCC CGA AGC TTC CTG CTT AAG TCC CTG GAG CAA GTG AGG AAG ATC CAG GCC AGC GGC TCG GTG CTG CTG Leu Pro Pro Ser Leu Pro Leu Pro Arg Ser Phe Leu Leu Lys Ser Leu Glu Gln Val Arg Lys lie Gln Ala Ser Gly Ser Val Leu Leu 10 20 30 290 310 330 350 GAG CAG TTG TGT GCC ACC TAC AAG CTG TGT CAC CCC GAG GAG CTG GTG TTG CTG GGC CAC TCT CTG GGG ATC CCG AAG GCT TCC CTG AGT Glu Gln Leu Cys Ala Thr Tyr Lys Leu Cys His Pro Glu Glu Leu Val Leu Leu Gly His Ser Leu Gly Ile Pro Lys Ala Ser Leu Ser 40 50 60 370 390 410 430 450 GGC TGC TCT AGC CAG GCC CTG CAG CAG ACA CAG TGC CTA AGC CAG CTC CAC AGT GGG CTC TGC CTC TAC CAA GGT CTC CTG CAG GCT CTA Gly Cys Ser Ser Gln Ala Leu Gln Gin Thr Gln Cys Leu Ser Gin Leu His Ser Gly Leu Cys Leu Tyr Gln Gly Leu Leu Gin Ala Leu 70 80 90 470 490 510 530 TCG GGT ATT TCC CCT GCC CTG GCC CCC ACC TTG GAC TTG CTT CAG CTG GAT GTT GCC AAC TTT GCC ACC ACC ATC TGG CAG CAG ATG GAA Ser Gly Ile Ser Pro Ala Leu Ala Pro Thr Leu Asp Leu Leu Gln Leu Asp Val Ala Asn Phe Ala Thr Thr Ile Trp Gln Gln Met Glu 100 110 120 550 570 590 610 630 AAC CTA GGG GTG GCC CCT ACT GTG CAG CCC ACA CAG AGC GCC ATG CCA GCC TTC ACT TCT GCC TTC CAG CGC CGG GCA GGA GGT GTC CTG Asn Leu Gly Val Ala Pro Thr Val Gln Pro Thr Gln Ser Ala Met Pro Ala Phe Thr Ser Ala Phe Gln Arg Arg Ala Gly Gly Val Leu 130 140 150 650 670 690 710 GCC ATT TCG TAC CTG CAG GGC TTC CTG GAG ACG GCT CGC CTT GCT CTG CAC CAC TTG GCC TAG ACCTGAGCAGAAAGCCCTTTCCAGATAGTTTATTT Ala Ile Ser Tyr Leu Gin Gly Phe Leu Glu Thr Ala Arg Leu Ala Leu His His Leu Ala End 160 170 750

770

790

810

830

ATCTCTATTTAATATTTATGCATATTTAAGCCTACTATTTAAAGACAAAGACGAGAAAATGGAGCTCTAAGCTTCTAGATCATTCTCTCCACTTCCGAGTTTTGTTCTCCTGCTTAGAG 870

890

910

930

950

1050

1070

CAGAGAGAGAAGGCTCTTGTGTCCTCCTGTGGAGGCCAGGGAAGGAGATGGGTAAATACCAAGTATTGATTCCTGCTGCTGCTCCAGGCACCCAGTTCTGTGGCAGTACCCCCAAAAAA 970

990

1010

1030

TCAGTGAGCCCTGCCGTGCTGAGGCACCATCTCAGGGGGGCCCAGGCAGCATCTGGTCTCCCTTCCGGGGGACAAGACATCCCTGTTTAATATTTAAACAGCAGTGTTCCCAAACTGGG 1090

1110

1130

1150

1170

1190

TTCTTATATCCCTTGCTCTGGTCAACCAGGTTGCAGGGTTTCCTGTCCTCACAGGAACGAAGTCCCTAAAGAAACAGTGGCAGCCAGGTTTAGCCCCGGAATTGACTGGATTCCTTTTT 1210

1230

1250

1270

1290

1310

TAGGGCCCTGCTGGCCTGGAAGTTGGAGTGGGGGGCAGAGGAGGCAGGAGGAAGCCTGGGGGGGGGGTTGGCATGGAGGGAGGCCTTCCCATCCACCCTCACCCTCCACCCCACCTGTC 1330

1350

ACTATAGCCAAGCTTGCGGATAATAAAGTGTGGTGTTCC

FIG. 3. Nucleotide sequence and the deduced amino acid sequence of pMG2 encoding murine G-CSF. Numbers above and below each line refer, respectively, to nucleotide position and amino acid position. Amino acids are numbered starting at Val-1 of the mature G-CSF sequence, which was postulated on the basis of homology with human G-CSF sequence (7, 8). The mature G-CSF sequence is preceded by a 30-residue putative signal sequence (-1 to -30).

several points (for example, amino acid positions 7, 12, and 102) of the murine G-CSF sequence. Expression of Murine G-CSF cDNA in Monkey COS Cells. To prove that the cDNA insert of pMG2 can code for murine G-CSF, we expressed the cDNA in the mammalian cell system. The 745-bp Apa IlAha III fragment containing the coding sequence for G-CSF (see Fig. 2) was recloned into the expression vector pdKCR (28, 29), and the resultant plasmid (pMGD2) was used to transfect monkey COS cells. After 72 hr, the protein secreted into the medium was assayed for G-CSF activity by using mouse bone marrow cells. The supernatant of COS cells transfected with either the murine or the human G-CSF expression plasmid (pMGD2 or pHGV21 in ref. 8) could stimulate colony formation, while little CSF activity was detected in the supernatant of COS cells transfected with pdKCR vector plasmid DNA (Table 1). The colonies stimulated by the COS cell supernatant were mainly neutrophilic granulocyte colonies, and little colonies consisting of either macrophages or eosinophilic granulocytes were observed. Recently, Holmes et al. have established several IL-3dependent myeloid leukemia cell lines from mouse spleen cells transfected with Cas-Br-M murine leukemia virus (20). Although most of these cell lines require murine IL-3 for their proliferation, NFS-60 cells are able to grow in response to the purified murine G-CSF (30). We tested the ability of murine and human COS G-CSF to support the proliferation of occur at

NFS-60 cells. Rapidly growing NFS-60 cells were extensively washed free of endogenous factors, and the dependency of [3H]thymidine incorporation on growth factors was evaluated. NFS-60 cells were factor-dependent for proliferation, and this requirement could be met by the human G-CSF purified from the conditioned medium of CHU-2 cells (19), human COS G-CSF, or murine COS G-CSF, but not by the supernatant of COS cells transfected with pdKCR vector plasmid (Fig. 4). The half-maximal incorporation was obtained with Table 1. Colony-stimulating activity in the supernatant of COS cells Colonies formed, no.

COS-cell Transfection vector Control supernatant,,l pdKCR pMGD2 pHGV21 G-CSF Saline 20 15, 15 74, 77 66, 90 85, 75 18, 10 4 4, 20 72, 81 75, 76 0.8 12,17 54, 48 48, 49 The supernatants of COS cells transfected with the indicated plasmid DNA were collected at 72 hr after transfection and aliquots (20, 4, and 0.8 ,ul) were assayed for CSF activity as described in 1 ml of the bone marrow culture in duplicate. As controls, the purified human G-CSF (10 ng) or saline was added to the assay mixture. Numbers of colonies consisting of more than 50 cells were counted on day 5.

7636


Proc. Natl. Acad Sci. USA 83

cells. By screening the library with human G-CSF cDNA as a hybridization probe, a full-length murine G-CSF cDNA clone was isolated (Fig. 3). The nucleotide sequence of murine G-CSF cDNA exhibits a strong homology (69.3%) in both coding and noncoding regions with that of human G-CSF cDNA (7-9). The high degree of homology is strong evidence that the cloned cDNA is the murine equivalent of human G-CSF cDNA. The murine cDNA could direct the synthesis of a protein in COS cells, which has an authentic G-CSF activity (Table 1). A comparison of the deduced amino acid sequences of human G-CSFa (7) and murine G-CSF is presented in Fig. 5. The overall homology of the mature human and murine G-CSF polypeptides is =72.8%, which is much higher than the homology (50%) observed between human and murine GM-CSF (31, 32). This result agrees with the observation that G-CSF has no apparent species specificity between human and murine systems (10), while GM-CSF shows species specificity (31, 33). Murine and human G-CSF proteins contain five cysteine residues of which four residues are found in the homologous positions (Fig. 5). There is evidence that the murine G-CSF molecule contains internal disulfide bonds (6), suggesting this may also be the case with human G-CSF. The alignment of the amino acid sequence of murine G-CSF with that of human G-CSF has indicated that the murine cDNA isolated in this report is the counterpart of human G-CSFb cDNA, which has a nine-nucleotide (three amino acids) deletion at the amino acid residue 35 of the mature G-CSFa polypeptide (Fig. 5; refs. 7 and 8). In the human system (8), a single chromosomal gene for G-CSF is interrupted by four introns. At the 5' end of the intron 2, two splice donor sequences are arranged in tandem, 9 bp apart, and two G-CSF mRNAs are generated by alternative use of the two splice donor sequences. Southern hybridization analysis of DNA from mouse spleen cells with mouse G-CSF cDNA as probe has suggested the existence of a single gene for G-CSF in mouse haploid genome (M.T. and S.N., unpublished results). Whether two different mRNAs exist also for murine G-CSF as in the human system remains to be

100 C :8

80 )F

A

0 C.

A

60)t ._c . 40

ou

'E

20

m,

1

2

4

8 16 Dilution

32 64

FIG. 4. Response of NFS-60 cells to G-CSF. NFS-60 cells were incubated for 24 hr with various doses of the purified human G-CSF (o), the supernatants of COS cells transfected with human G-CSF expression vector (pHGV21) (e), murine G-CSF expression vector (pMGD2) (A), or pdKCR vector (x). Cells were then pulsed with 0.25 GCi of [3H]thymidine for 6 hr, and the incorporated radioactivity was determined. [3H]Thymidine incorporation is expressed as a percentage of the maximal incorporation (110,000 cpm). The incorporation without growth factors was 11,500 cpm. The purified human G-CSF was diluted starting from 2.5 ng/ml, while the supernatants of COS cells were diluted starting from a 1:300 diluted sample (for human COS-G-CSF) or a 1:200 diluted sample (for mouse COS-G-CSF).

0.2 ng of the purified human G-CSF per ml (10 pM) which is almost the same concentration obtained with the purified murine IL-3 (20).

DISCUSSION Previously, the human CHU-2 carcinoma cell line from an oral cavity tumor and the 5637 bladder carcinoma cell line were successfully used to isolate human G-CSF cDNA (7-9). In this report, NFSA fibrosarcoma cells were found to synthesize murine G-CSF mRNA constitutively (Fig. 1), and a cDNA library was constructed with mRNA from the NFSA Mouse: Human:

Mouse: Human:

Mouse: Human: Mouse: Human:

Mouse: Human: Mouse: Human:

Among many IL-3-dependent cell lines established from mouse spleen cells transformed with Cas-Br-M murine leu-

-20

-10

1 -1 Arg Glu Ala Val Pro Leu Val Thr Val Ser Ala Leu Gin Glu Ala Thr Pro Leu --- --- --- --- --- Gly -1 I 30 Glu Gin Val Arg Lys lie Gin Ala Ser Gly Ser Val Glu Gin Val Arg Lys lie Gin Gly Asp Gly Ala Ala

50 His Pro Glu Glu Leu His Pro Glu Glu Leu 50 80 Gin Gin Thr Gin Cys Gin Leu Ala Gly Cys Pro Pro 100 130 Leu Leu

10 20 Pro Pro Ser Leu Pro Leu Pro Arg Ser Phe Leu Leu Lys Ser Leu Pro Ala Ser --- Ser Leu Pro Gin Ser Phe Leu Leu Lys Cys Leu 10

40 Leu Leu Glu Gin Leu Cys Ala Thr Tyr Lys Leu tys Leu Gin Glu Lys Leu|Val Ser GlulCys Ala Thr Tyr Lys Leu Cys 40

30

20

Mouse: Human:

seen.

-10 -20 -30 Met Ala Gin Leu Ser Ala Gin Arg Arg Met Lys Leu Met Ala Leu Gin Leu Leu Leu Trp Gin Ser Aia Leu Trp Ser Gly Met Ala Gly Pro Ala Thr Gin Ser Pro Met Lys Leu Met Ala Leu Gin Leu Leu Leu Trp His Ser Ala Leu Trp lhr Val -30

Mouse: Human:

(1986)

60 70 Val Leu Leu Gly His Ser Leu Gly Ile Pro Lys Ala Ser Leu Ser Gly Cys Ser Ser Gin Ala Leu Val Leu Leu Gly His Ser Leu Gly Ile Pro Trp Ala Pro Leu Ser Ser Cys Pro Ser Gin Ala Leu 60 70 90

100

Leu Ser Gin Leu His Ser Gly Leu Cys Leu Tyr Gin Gly Leu Leu Gin Ala Leu Ser Gly lie Ser Leu Ser Gin Leu His Ser Gly Leu Phe Leu Tyr Gin Gly Leu Leu Gin Ala Leu Glu Gly lie Ser 80

90

110

120

Ala Leu Ala Pro Thr Leu Asp Leu Leu Gin Leu Asp Val Ala Asn Phe Ala Thr Thr Ile Trp Gin Gin Met Glu Asn Glu Leu Gly Pro Thr Leu Asp Thr Leu Gin Leu Asp Val Ala Asp Phe Ala Thr Thr Ile Trp Gin Gin Met Glu Glu 110 140

120 150

Gly Val Ala Pro Thr Val Gin Pro Thr Gin Ser Ala Met Pro Ala Phe Thr Ser Ala Phe Gln Ars Arg

Gly Met Ala Pro Ala Leu Gin Pro Thr Gin Gly Ala 130 160 Val Leu Ala lie Ser Tyr Leu Gin GlY Phe Leu Glu Thr Val Leu Val Ala Ser His Leu Gin Ser Phe Leu Glu Val 160

Ala Gly Met Pro Ala Phe Ala Ser Ala Phe Gin Arg Arg Ala Glo 150 140 170 Ala Arg Leu Ala Leu His His Leu Ala Ser Tyr Arg Val Leu Arg His Leu Ala Gin Pro 170

Glb GCl

FIG. 5. Comparison of murine and human G-CSF amino acid sequences. The numbers above and below each row refer to amino acid positions of murine G-CSF and human G-CSF, respectively. The two sequences are aligned to get the maximum homology between them, and the identical amino acids are underlined. Human G-CSF amino acid sequence is the G-CSFa amino acid sequence deduced from pBRG-4 cDNA (7). The boxed three amino acids, Val-Ser-Glu at the residue numbers 36-38 cannot be found on the human G-CSFb sequence (8).

Biochemistry: Tsuchiya et al. kemia virus (20), only NFS-60 cells could respond to G-CSF for its proliferation (30), and the dependency of NFS-60 cells on G-CSF was confirmed by the recombinant human and murine G-CSF (Fig. 4). On the other hand, G-CSF can induce the terminal differentiation of the murine myelomonocytic leukemia cell line WEHI-3B D+ (6) but not of NFS-60 cells (ref. 30; S.N., unpublished data). By using these two cell lines as a model system, it will be possible to study the proliferation and differentiation of the progenitor cells of granulocytes. In particular, it will be interesting to examine the receptor for G-CSF on both cell lines and the intracellular mechanisms for proliferation and differentiation of cells generated by the G-CSF-receptor complex. We thank Dr. H. Nomura for assaying the G-CSF activity in COS cell supernatant with mouse bone marrow cells. We are grateful to Dr. J. N. Ihle for his kind supply of murine NFS-60 cells and valuable information prior to publication. We also thank Dr. M. Shikita for NFSA cells and Dr. K. Arai for murine IL-3. This work was supported in part by Grants-in-Aid from the Ministry of Education, Science and Culture of Japan. 1. Burgess, A. W. & Metcalf, D. (1980) Blood 56, 947-958. 2. Metcalf, D. (1985) Science 229, 16-22. 3. Yokota, T., Lee, F., Rennick, D., Hall, C., Arai, N., Mosmann, T., Nabel, G., Cantor, H. & Arai, K. (1984) Proc. Natl. Acad. Sci. USA 81, 1070-1074. 4. Fung, M. C., Hapel, A. J., Ymer, S., Cohen, D. R., Johnson, R. M., Campbell, H. D. & Young, I. G. (1984) Nature (London) 307, 233-237. 5. Gough, N. M., Gough, J., Metcalf, D., Kelso, A., Grail, D., Nicola, N. A., Burgess, A. W. & Dunn, A. R. (1984) Nature (London) 309, 763-767. 6. Nicola, N. A., Metcalf, D., Matsumoto, M. & Johnson, G. R. (1983) J. Biol. Chem. 258, 9017-9023. 7. Nagata, S., Tsuchiya, M., Asano, S., Kaziro, Y., Yamazaki, T., Yamamoto, O., Hirata, Y., Kubota, N., Oheda, M., Nomura, H. & Ono, M. (1986) Nature (London) 319, 415-418. 8. Nagata, S., Tsuchiya, M., Asano, S., Yamamoto, O., Hirata, Y., Kubota, N., Oheda, M., Nomura, H. & Yamazaki, T. (1986) EMBO J. 5, 575-581. 9. Souza, L. M., Boone, T. C., Gabrilove, J., Lai, P. H., Zsebo, K. M., Murdock, D. C., Chazin, V. R., Bruszewski, J., Lu, H., Chen, K. K., Barendt, J., Platzer, E., Moore, M. A. S., Mertelsmann, R. & Welte, K. (1986) Science 232, 61-65. 10. Nicola, N. A., Begley, C. G. & Metcalf, D. (1985) Nature (London) 314, 625-628. 11. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J. & Rutter,

Proc. Natl. Acad. Sci. USA 83 (1986)

7637

W. J. (1979) Biochemistry 18, 5294-5299. 12. Aviv, H. & Leder, P. (1972) Proc. Natl. Acad. Sci. USA 69, 1408-1412. 13. Hall, A. & Brown, R. (1985) Nucleic Acids Res. 13, 5255-5268. 14. Huynh, T., Young, R. A. & Davis, R. W. (1985) in DNA Cloning Techniques, A Practical Approach, ed. Glover, D. M. (IRL, Oxford). 15. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY). 16. Benton, W. D. & Davis, R. W. (1977) Science 196, 180-182. 17. Wahl, G. M., Stem, M. & Stark, G. R. (1979) Proc. Natl. Acad. Sci. USA 76, 3683-3687. 18. Messing, J. (1983) Methods Enzymol. 101, 20-78. 19. Nomura, H., Imazeki, I., Oheda, M., Kubota, N., Tamura, M., Ono, M., Ueyama, Y. & Asano, S. (1986) EMBO J. 5, 871-876. 20. Holmes, K. L., Palaszynski, E., Fredrickson, T. N., Morse, H. C., III, & Ihle, J. N. (1985) Proc. Natl. Acad. Sci. USA 82, 6687-6691. 21. Ihle, J. N., Keller, J., Henderson, L., Klein, F. & Palaszynski, E. (1982) J. Immunol. 129, 2431-2436. 22. Sakai, N., Shikita, M., Tsuneoka, K., Bessho, M. & Hirashima, K. (1984) Gann 75, 355-361. 23. Bessho, M., Shikita, M., Tsuneoka, K., Sakai, N. & Hirashima, K. (1984) Gann 75, 993-1001. 24. Kozak, M. (1986) Cell 44, 283-292. 25. Proudfoot, N. J. & Brownlee, G. G. (1976) Nature (London) 263, 211-214. 26. Winzler, R. J. (1973) in Hormonal Proteins and Peptides, ed. Ii, C. H. (Academic, New York). 27. Takahashi, N., Takahashi, Y. & Putnam, F. W. (1984) Proc. Natl. Acad. Sci. USA 81, 2021-2025. 28. O'Hare, K., Benoist, C. & Breathnach, R. (1981) Proc. Natl. Acad. Sci. USA 78, 1527-1531. 29. Fukunaga, R., Sokawa, Y. & Nagata, S. (1984) Proc. Natl. Acad. Sci. USA 81, 5086-5090. 30. Weinstein, Y., Ihle, J. N., Lavu, S. & Reddy, E. P. (1986) Proc. Natl. Acad. Sci. USA 83, 5010-5014. 31. Lee, F., Yokota, T., Otsuka, T., Gemmel, L., Larson, N., Luh, J., Arai, K. & Rennick, D. (1985) Proc. Natl. Acad. Sci. USA 82, 4360-4364. 32. Wong, G. G., Witek, J. S., Temple, P. A., Wilkens, K. M., Leary, A. C., Luxenberg, D. P., Jones, S. S., Brown, E. L., Kay, R. M., Orr, E. C., Shoemaker, C., Golde, D. W., Kaufman, R. J., Hewick, R. M., Wang, E. A. & Clark, S. C. (1985) Science 228, 810-815. 33. Lusis, A. J., Quan, D. H. & Golde, D. W. (1981) Blood 57, 13-21.