Eur. J. Immunol. 1992. 22: 1157-1163

David G. Jackson., Derek N. J. Hart*, Gary Starling. and John I. Bell. Molecular Immunology Group., Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Headington and Department of Haematology., Christchurch Hospital, Christchurch

1157

CMFW35, a new Ig superfamily member related to the poly Ig receptor

Molecular cloning of a novel member of the immunoglobulin gene superfamily homologous to the polymeric immunoglobulin receptor The CMRF35 monoclonal antibody recognizes a cell membrane antigen present on the surface of monocytes, neutrophils, a proportion of peripheral blood Tand B lymphocytes and lymphocytic cell lines. Initial studies with CMRF35 suggested an unusual pattern of serological reactivity which did not correspond to any of the known leukocyte differentiation antigen clusters. We describe here the cloning and sequencing of a cDNA encoding the CMRF35 antigen by means of expression in COS cells and immunoselection with the CMRE735 monoclonal antibody. The cDNA encodes a novel integral membrane glycoprotein of 224 amino acids that represents a new member of the immunoglobulin (Ig) gene superfamily The molecule comprises (a) a single extracellular Ig variable domain remarkably similar to the Fc receptor for polymeric IgA and IgM, (b) a membrane-proximal domain containing a high proportion of proline, serine and threonine residues that was predicted to be heavily 0-glycosylated, (c) an unusual transmembrane anchor that contained a glutamic acid and a proline residue and (d) a short cytoplasmic tail.Transcripts encoding the CMRF35 protein were detected in early monocytic cell lines, in peripheral blood T cells and in some B lymphoblastoid cell lines, confirming the results of immunocytological staining. However, the level of CMRF35 expression on peripheral blood T cells was shown to decrease in response to mitogenic stimulation. The likelihood that the CMRF3.5 antigen shares a common evolutionary ancestor with the poly Ig Fc receptor- is discussed.

1 Introduction

superfamily members are thought to have evolved through a series of mutation and gene duplication events [1].

The Ig gene superfamily comprises a large number of cell surface glycoproteins that share sequence homology with the V and C domains of antibody heavy and light chains (reviewed in [1, 21). Many members of this family play important roles in regulating the complex cell interactions that occur within the immune system including molecules such as the MHC glycoproteins [3, 41, theT cell receptor for antigen [ 5 ] ,theT cell accessory molecules CD4 and CD8 [6] the lymphocyte adhesion molecules ICAM-1 [7], CD2 and LFA-3 [8] and the immunoglobulin Fc receptors (reviewed in [9]). The great majority of Ig superfamily members are relatively complex polydomain macromolecules containing multiple Ig V- and C-like domains (e.g. the neural cell adhesion molecule NCAM and the platelet-derived growth factor [PDGF] receptor, see [ l , 21). However, a small but significant subset are relatively simple structures containing only singlevor D domains.This latter group which includes CD8, CD28 [lo] and theThy-1 molecule [ l l ] isof particular interest because these single-domain molecules may have diverged at a relatively early stage from the putative ancestral “homology unit” from which all the other Ig

We have isolated a new member of the Ig gene superfamily that contains a single Ig V-related domain, and which is recognized by a novel monoclonal antibody CMRF35 [12]. This monoclonal antibody which was originally raised by immunizing mice with peripheral blood lymphocytes has an unusual pattern of reactivity ([12] and Daish et al. manuscript submitted) that does not correspond to any of the recognized leukocyte antibody clusters. The CMRF35 antibody stains predominantly myeloid cells (i. e. peripheral blood granulocytes, monocytes and myeloid cell lines such as U-937 and HL-60) but also stains a small proportion of circulating T cells (25 %) and B cells (15 %) and large granular lymphocytes (LGL; Daish et al. manuscript submitted). These features of the CMRF35 antigen suggested a unique structure and we set out to clone the cDNA by immunoselection from cDNA libraries expressed in mammalian COS cells [lo, 131. This report describes the sequence and pattern of expression of this unusual antigen which appears to share a common ancestral origin with the Fc receptor for polymeric IgA and IgM [14-161.

[I 101131 2 Materials and methods

*

This work was generously supported by the WellcomeTrust. Part of the Research was also funded by the Leukaemia Research Fund, The McClelland Trust and the Cancer Society of New Zealand.

Correspondence: David G. Jackson, Molecular Immunology Group*, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, Great Britain

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992

2.1 Production and purification of monoclonal antibodies The monoclonal antibody CMRF35 (IgGzb) was produced by immunizing RBFDN strain mice with a mixture of Ficoll-purified human PBMC (single donor) and Percollpurified human LGL obtained from a patient with a chronic CD3%D16+ LGL leukemia. Ig were purified from the

+

0014-2980/92/0505-1157$3.50 .25/0

I IS8

Eur. J. Immunol. 1992.22: 1157-1163

D. G. Jackson, D. N. J. Hart, G. Starling and J. I. Bell

supernatants of hybridoma cultures by adsorption to protein G. 2.2 cDNA libraries and the expression cloning of the CMRF35 cDNA

Three different cDNA libraries were used. The cDNA library constructed from PHA-stimulated PBMC in the vector pCDM8 has been described previously [17]. The cDNA libraries from the B cell line JYand from LAK cells were the kind gift of Dr. David Simmons (ICRF, Cell Adhesion Laboratory, Oxford). For the expression cloning procedure the cDNA libraries were mixed and transfected into COS-1 cell monolayers using DEAE-dextran. After 48 h, transfectants expressing CMRF35 antigen on the cell surface were immobilized on goat anti-mouse IgG-coated dishes by panning with the CMRF35 monoclonal antibody (1/10 diluted hybridoma culture supernatant) followed by retrieval of the plasmid and reintroduction into fresh COS-1 cells (by spheroplast fusion, E. coli MC1061-P3) for two further rounds of screening essentially as described by Seed and Aruffo ~31. Plasmids selected by this method were finally prepared from individual colonies of E.coli and re-transfected into COS-1 cells for further screening with the CMRF35 monoclonal antibody by indirect immunofluorescence staining with a goat anti-mouse FITC conjugate. Fixed (2 YO formaldehyde in PBS, 2 mg/ml BSA, 5 min, 5 "C) and stained cells were viewed and photographed under a Zeiss Axioskop microscope (Carl Zeiss, Thornwood, NY) using Kodak Ektachrome EPT 160 film.

400-bp Xba I fragments subcloned into M13mpl8 as templates. Both DNA strands were fully sequenced and in some cases re-sequenced after substitution of 7-deazadGTP for dGTP.

3 Results and discussion 3.1 Isolation of a cDNA encoding the CMRF35 antigen cDNA clones encoding the CMRF35 antigen were cloned from a mixture of human lymphoid cDNA libraries constructed in the plasmid vector pCDM8 by immunoselection of transfected COS cells with the monoclonal antibody CMRF35. This technique which has previously been described in detail [lo, 131 uses antibody panning to capture COS cells expressing the relevant antigen followed by rescue of the plasmid and its reintroduction into fresh COS cells by spheroplast fusion for further rounds of immunoselection. After three such rounds of selection, the enriched plasmid population was amplified in E. coli and plated out yielding approximately 10000 bacterial colonies. Of ten randomly picked colonies, six contained plasmids which, after transfection into COS cells, directed the surface expression of the CMRF35 antigen as assessed by indirect immunofluorescence microscopy (Fig. 1).Restriction enzyme digestion confirmed the presence of the same 1.4-kb insert in each of the six positive clones (not shown) and one of these, termed CMRF 35.1, was chosen for further analysis. Inspection of the nucleotide sequence (Fig. 2) revealed a large open reading frame beginning with the first ATG triplet at position 224 and encoding a protein of 224 amino

2.3 Mitogenic stimulation of lymphocytes PBMC were stimulated in vitro for 72 h with PHA prior to harvesting as described in [17]. 2.4 Northern blot hybridization

For Northern blot analyses total RNA was prepared from cell lines using the one-step acid guanidinium isothiocyanate extraction method [18] and electrophoresed on formaldehyde-agarose gels (1.2 % agarose) prior to transfer to nitrocellulose membranes (Hybond-C extra, Amersham Int., Amersham, GB) according to standard methods [19]. Filters were hybridized (overnight, 42 "C, 50 YOformamide, 3X SSC) with either an 800-bp Xba I fragment containing the majority of the coding sequence of the CMRF35.1 clone or with a full-length Xho I fragment. Probes were radiolabeled with 32Pby means of random hexanucleotide priming. Blots were washed at high stringency (30 min, 65 "C, 0.2X SCC) prior to autoradiography. 2.5 DNA sequence determination The CMRF35 cDNA clonc was sequenced by the di-deoxy nucleotide chain termination method [20] using both the full-length clone in pCDM8 vector and the 800-bp and

A

B

Figure 1. Expression of the CMFW35 antigen on the surface of transfected COS cells. COS cell monolayers transfected with the CMRF3S.1 cDNA in the pCDM8 vector or with vector alone (mock-transfected control) were treated with EDTA (1 mM, in PBS, pH 7.5, 1 h) to generate a single-cell suspension prior to indirect immunofluorescent antibody staining with the CMRF35 monoclonal antibody as described in Sect. 2.2. The figure shows both the phase contrast image (A) and the fluorescence image (B) for the CMRF35 transfectants. Mock-transfected controls (not shown) showed no fluorescence.

CMRF35, a new Ig superfamily member related to the poly Ig receptor

Eur. J. Immunol. 1992. 22: 1157-1163 1 90

CTCTAAAGGCCACTAGCACCCATCCCAGAGCTGTCAGCACCGGCCTCAGCCCAGGCGGCTCTCTCCCTGAGCTTCCTGTAGCCCTGACC CTCTCCAGCCTCAGACCTGAGACAGGGCTGGACAAGGAAGCAGAGAGCAGAAG~GCAGAAGCGAAGCTCAGATCTGCTGGGAGGAAG

1 180

M T A R A W A S W R ATTACATTTTGTCCCCTCCTGGGGTCTTGCACAGTGGCAGGTGACATTCGTGTTACAGGAATGACTGCCAGGGCCTGGGCCTCGTGGCGG

11 270

S S A L L L L L V P G Y F P L S H P M T V A G P V G G S L S TCTTCAGCTCTGCTCCTCCTGCTTGTCCCAGGCTATTTTCCTCTGAGCCACCCCATGACCGTGGCGGGCCCCGTGGGGGGATCCCTGAGT

41 360

V

Q

C

R

Y

E

K

E

H

R

T

L

N

K

F

W

C

R

P

P

Q

I

L

R

C

D

K

I

V

1159

a

E

GTGCAGTGTCGCTATGAGAAGGAACACAGGACCCTCAACTTCTGGTGCAGACCACCACAGATTCTCCGATGTGACAAGATTGTGGAG

T

71 450

K

G

S

A

G

K

R

N

G

R

V

S

I

R

D

S

P

A

N

L

S

F

T

V

T

L

E

T

ACCAAAGGGTCAGCAGGGAGGAATGGCCGAGTGTCCATCAGGGACAGTCCTGC~CCTCAGCTTCACAGTGACCCTGGAG~TCTC -

101 540

X

131 630

V

E

E

D

A

G

T

Y

W

C

G

V

D

T

P

W

L

R

D

F

H

D

P

I

V

E

V

E

V

S

ACAGAGGAGGACGCAGGCACCTACTGGTGTGGGGTGGATACACCGTGGCTCCGAGACTTTCATGATCCCATTGTCGAGG~GAGGTGTCC

F

P

A

G

T

T

T

A

S

S

P

Q

S

S

M

G

T

S

G

P

P

T

K

L

P

V

H

T

W

GTGTTCCCGGCCGGGACGACCACAGCCTCCAGCCCCCAGAGCTCCATGGGCACCTCAGGTCCTCCCACG~GCTGCCCGTGCACACCTGG

P

S

V

T

R

K

D

S

P

E

P

S

P

H

P

G

S

L

F

S

N

V

R

F

L

L

L

V

L

L

161 720

CCCAGCGTGACCAGAAAGGACAGCCCCGAACCCAGCCCACACCCTGGCTCCCTGTTCAGCAATGTCCGCTTC~GCTCCTGGTCCTCTTG

191 810

E L P L L L S M L G A V L W V N R P Q R S S R S R Q N W P K GAGCTGCCCCTGCTCCTGAGCATGCTGGGTGCCGTCCTCTGGGTGAACAGACCTCAGAGAAGCTCTAGAAGCAGGCAG~TTGGCCCAAG

221 900 990 1080

G E N Q * GGTGAGAACCAGTAGCATCTGCTGTCCATCAAGGCCCTGTGCTGCAACAGAGCCCCTCTGGGGACTGGAATGACCTCCTGACCATCAAGG

The nucleotide and derived amino acid seaence of the QIRF-35 antiaen cDNA.

2

1

0

-1 -2 -7

-3

7

Figure 2. The nucleotide and deduced amino acid sequence of the CMRF35 cDNA. The complete sequence of the 1.2-kb insert within the Bst XI site of pCDM8 is shown in (a). The sites of potential N-glycosylation are boxed. The predicted C-terminal hydrophobic transmembrane region is underlined and a possible cleavage site for the N-terminal leader is marked with an arrow. (b) shows the Kyte-Doolittle hydropathy plot for the full CMRF35 coding sequence using a window size of 12 residues.

B

C

C'

POLY IgR V1 POLY IgR V4 CMRF35 TCR Ve lg KAPPA V,

46 49 47 49 45 C'

'

D

F 95 95 95 99 93

POLY laR V1 POLY I$ V4 CMRF35 TCR Vo lg KAPPA V,

POLY IgR V1 POLY IgR V4 CMRF35 TCR Ve lg KAPPA V,

E

E N - - VlfiG

G

Agure 3. Alignment of the Ig-like domain of the CMRF35 antigen with other V domains from the Ig gene superfamily. The dashes indicate positions where gaps have been introduced to optimize the alignments and boxes indicate positions where residues are identical bctwecn the individual sequences. The positions predicted for the strands B-F are depicted by the lines above the sequences and are based upon the assignments in [l].The position of the second conserved disulfide bond within the poly Ig receptor V domains is also shown.The identities of the individual sequences and their NBRFdatabase codes are as follows; poly IgRV1 + 4, human polymeric IgA Fc receptor V domains 1 + 4 (QRHUGS); Ig n V, human Ig n light chainV domain (K4RBF2); TcR Vb, human T cell receptor p chain V domain (D27552).

1160

D. G. Jackson, D. N. J. Hart, G. Starling and J. I. Bell

acids. A second in-frame ATG triplet is found at position 325; however, unlike the first ATG, the context of this second triplet is a poor fit to the Kozak consensus sequence [21] and is, therefore, unlikely to form a functional initiator. The initiating methionine is followed by a region encoding a stretch of predominantly hydrophobic amino acids that is predicted to function as an N-terminal leader sequence. A possible site for proteolytic cleavage of the leader at proline 20 was predicted using the rules of Von Heijne [22] yielding a mature protein of 204 amino acids containing a single C-terminal hydrophobic transmembrane anchor and a short, 19-residue cytoplasmic tail.These features are shown graphically in the Kyte-Doolittle hydropathy plot (Fig. 2b) and suggest that the CMRF35 antigen is a conventional type-I integral membrane protein. The CMRF35 cDNA also contained an almost perfect 43-44-bp repeat within the 3' untranslated sequence which lacked a polyadenylation signal (Fig. 2a).

3.2 The CMRF35 antigen is a new member of the Ig gene superfamily containing a single V domain Comparison of the amino acid sequence of CMRF35 with existing protein sequence databases using the FASTA analysis program [23] revealed the presence of a single Ig-related domain between residues 21-127 at the N terminus of the extracellular domain. The prediction that the CMRF35 antigen contains a V domain is based upon the extent of sequence homology with the Ig V-like domains of other members of the Ig gene superfamily [ l , 21 as assessed by comparison of individual sequence alignments (Fig. 3) and by quantitative measurement of sequence homology using the ALIGN program of Dayhoff [24]. The ALIGN program scores similarities between pairs of sequences by measuring the degree of identity both before and after 100 random shufflings of each amino acid sequence. The degree of similarity is defined in terms of the number of SD that the similarity score deviates from that expected if two random sequences were compared. In general scores of >2-3 SD indicate significant homology (discussed in [ l , 21). The sequence alignments (Fig. 3) illustrate two key structural features that distinguish the CMRF3.5 V domain from the smaller and more distantly related C domain. First, the CMRF35 V domain displayed the characteristically long spacing (66 amino acids) between the two conserved cysteine residues of the Ig fold (positions 19 and 86, p strands B and Fin Fig. 3), sufficient to accomodate the nine individual antiparallel p strands, including the extra C' and the short C" strand, that comprise the Ig V domains [l]. Second, the CMRF35 domain also contained the conserved arginine and aspartic acid residues (positions 57 and 80, respectively in Fig. 3) that form the salt bridge between p strands D and F and which adds to the stability of the Ig fold within Ig V domains [l,21. The ALIGN scores (Table 1) show the large values (5.7-11.7 SD) obtained when the CMRF35 V domain was compared with a range of other Ig V domains. These are in contrast to the low scores (0.13-3.2 SD) obtained for comparisons with Ig C domains from both the C1 and C2 subsets and clearly demonstrate that the CMRF35 Ig-like domain is a member of the V set of Ig superfamily sequences.

Eur. J. Immunol. 1992. 22: 1157-1163

3.3 The V domain of the CMRF35 antigen is closely related to the Fc receptor for polymeric Ig

The most striking feature of the data in Fig. 3 and Table 1 was the extent of sequence homology (30 YO-33 YO identity over 103 residues) between the CMRF35 antigen and the individual V domains of the human polymeric Ig Fc receptor (poly Ig receptor) [25]. Similar homology was also found with the rabbit poly Ig receptor [14] although the alignments have not been included so as to avoid duplication. The poly Ig receptor is expressed on glandular epithelial cells in the intestine, liver and lactating mammary gland where it transports polymeric IgA and IgM across the epithelium via transcytosis [14-161. The receptor-Ig complex, termed the secretory component is then released into the lumen via a discrete proteolytic cleavage event [26, 271. This unusual and interesting molecule is unique amongst the known Fc receptors in that it is assembled from Ig Vdomains rather than Ig C domains [14]. The poly Ig receptor comprises five conserved V domains and with the single exception of the V2 domain each contains a second conserved cysteine pair within p strands C and D that is disulfide linked [25]. Binding of dimeric IgA is known to involve a noncovalent interaction between the receptor V1 domain and the IgA C"2 domain [28]. Although disulfide bond formation between dimeric IgA and the receptor V5 domain has also been observed, this does not occur with all classes of IgA. Interestingly, the CMRF35 V domain shows the greatest homology with the IgA Table 1. Quantitative analysis of the similarity between the CMRF35 V domain and other Ig superfamily domains using the Dayhoff ALIGN programa)

V set

Poly IgR V domain no.

v1 CMRF35 V Poly IgR V1 Poly IgR V2 Poly IgR V3 Poly IgR V4 POly IgR V5

C set

c m 5 v

v3

v3

v4

8.5 11 7.7 15.5 10 - 14.5 9 7.8 7.7 14.5 7.8 15.5 9 9 1.8 3.1 9.25 11.7

-

Ig A C 0.92

5.7 10

(C1)

& CD3e 0.9 0.13

IpV TcRV~ v5 6.9 9 2.1 3.1 8.5

-

7.1 2.5 4.3 5.6 6.2 4

9.4 5.9 1.7 4.9 2.7 2.2

(a)

IgGFcR 3.2

a) The table lists the ALIGN scores (100 randomizations)in units of standard deviations for comparison of the CMRF35 V domain with each of the highest scoring Ig superfamily V domains from the FASTA database searches and a representative number of Ig C domains from the C1 and C2 sets. Assuming a normal distribution and no effect of sequence selection, scores of 3.L4.3 and 5.2 indicate chance probabilitiesof lC3, and lo-', respectively [l,21. By convention, the region compared in all cases begins 20 amino acids before the cysteine residue in p strand B of the Ig domain and extends to 20 amino acids after the cysteine residue in p strand F [l]. The identities of the Ig gene superfamily V domains and their NBRF database accession numbers are the same as in Fig. 3.The C domain sequences are IgG FcR, mouse IgGIRb Fc receptor [47]; CD38, human C D ~ chain E [48]; Pz-m, human @z-microglobulin(MGHUB2) and Ig h C, human Ig h chain C domain (L2HU).

Eur. J. Immunol. 1992. 22: 1157-1163

CMRF35, a new Ig superfamily member related to the poly Ig receptor

binding V1 domain of the poly Ig receptor (ALIGN score 11.7 SD, 30 Yo sequence identity over 103 amino acids) and with the closely related V4 domain (ALIGN score 11.0 SD, 33% identity over 103 amino acids). Furthermore the conserved disulfide-bonded cysteine residues characteristic of the poly Ig receptor V domains [25] can be seen also in the CMRF35 antigen surrounded by the conserved motifs K (F/Y) WC and QXXCXXVat the expected positions within fi strands C and C‘ (Fig. 3). Additional short blocks of sequence homology peculiar to the poly Ig receptor V domains are also found in fi strand E of the CMRF35 sequence (the FTVI/TL motif, Fig. 3). The ALIGN scores in Table 1 further indicate that the CMRF35 V domain is more closely homologous to the poly Ig receptor V1 and V4 domains than are either the receptor V2,V3 or V5 domains. In addition the CMRF35 V domain is a better match to the receptor V1 and V4 domains than are either the Ig x VL or Tcell receptor Vg sequences (Table l ) , which were previously thought to be the closest homologues of the poly Ig receptor domains [14]. These results suggest that the CMRF35 antigen and the poly Ig Fc receptor V domains represent a closely related pair within the Ig superfamily and are consistent with both genes having evolved from a common ancestor. Several ‘subfamilies’ within the Ig superfamily have recently been described that share not only close sequence homology, but also similarities in introdexon organization and chromosomal 1ocalization.These include the CD2LFA-3 pair on chromosome lp13 [29], the CD28lCTLA-4 pair on chromosome 3 2q33 [30], the CD4/LAG-3 pair on chromosome 1 2 ~ 1 [31] and the PDGF receptor/CSF-1 receptodc-Kit receptor group [1,21. Clearly such groupings reflect common evolutionary origins and it is to be expected that the discovery of further examples will lead to a more complete understanding of the evolutionary history of the Ig superfamily.We have not yet studied the genomic organization of the CMRF35 gene although preliminary genomic Southern blot analyses (data not shown) suggest a single locus in which the coding sequence may be split between six or more separate exons. It will be of interest to compare the precise introdexon organization and the chromosomal localization of the CMRF35 gene with that of the human poly Ig receptor in view of the similarities between the V domain sequences described above. 3.4 The remainder of the CMRF35 sequence encodes a proline-rich hinge-like region and an unusual transmembrane segment containing a glutamic acid residue Downstream of the Ig V-domain the predicted membrane proximal region of the CMRF35 antigen (residues 128-183, Fig. 2) contains a high proportion of proline (18 %), serine (20 YO)and threonine (13 %) and is predicted to adopt an extended open conformation by analogy with the hingelike sequences in the membrane-proximal regions of the T cell surface antigen CD8 [6, 321, the leukosialin CD34 molecule [33], the IL-2 receptor [34,35] and the membrane cofactor protein CD46 [36]. Interestingly, the membraneproximal region of the CMRF35 protein is similar to the heavily 0-glycosylated hinge region of the CD8 a chain

1161

which is thought to extend the CD8 V domain some 10 nm or more from the surface of the bilayer based on recent calculations [32]. It is possible that the role of the CMRF35 hinge region may be purely structural by analogy with CD8. Alternatively, heavy 0-glycosylation of the hinge region may protect the CMRF35 antigen from proteolytic degradation [37]. The membrane-proximal domain is immediately followed by an unusual 22-amino acid transmembrane anchor (Fig. 2) that contains a glutamic acid residue. The presence of charged amino acids within the lipid bilayer is energetically unfavorable and is relatively rare among the membrane-spanning segments of membrane proteins with the exception of specialized transporters that have multiple membrane-spanning segments encoding ion translocating channels (reviewed in [38]). However, a small group of lymphocyte cell membrane proteins including the IgE Fc receptor 1a and y chains [39,40], the CD3 y, 6 and E chains [41] and the recently described rat OX-47 molecule [42] each contain single glutamic or aspartic acid residues within their respective transmembrane segments and these segments display some sequence homology with each other [42]. Comparison of the CMRF35 transmembrane sequence with these other molecules however showed no obvious similarity (not shown). The tendency for membrane proteins with charged transmembrane segments to associate with one another in the lipid bilayer has led others to suggest that the charged amino acids may function directly in the formation of protein complexes [42]. This idea is supported by the experimental observation that each of the CD3 subunits must be present for transport and expression of the T cell receptor at the cell surface [39,41]. It is not yet clear whether the CMRF35 protein associates with other proteins within the membrane. However, the ability of COS fibroblasts to express the cDNA encoding this antigen of lymphoidmyeloid origin (see Sect. 3.5 below) argues that such interactions, if they occur in vivo, are not obligatory for membrane insertion. The putative membrane spanning segment of the CMRF35 sequence is followed by an 18-amino acid cytoplasmic tail that carries a net positive charge due to the presence of five basic amino acids (Fig. 2a). In the case of the poly Ig receptor a much longer (103 residue) cytoplasmic tail is present [14] and this functions in the transcytosis of receptor-Ig complexes across the epithelium [14-161. The CMRF35 cytoplasmic tail shares no obvious sequence homology with the latter structure. 3.5 Restricted expression of the CMRF35 mRNA Northern blot hybridization of total cellular RNA samples with the CMRF35 cDNA probe (Fig. 4) indicated the presence of a single sized transcript of approximately 1.4 kb in the myeloid cell lines U-937 (histiocytic lymphoma) and HL-60 (myeloid leukaemia), the B lymphoblastoid cell line IM-9, the EBV-transformed B cell lines Mann and Nell and in peripheral blood T cells. In contrast, only very low or negligible levels of the CMRF35 transcript were detected in the primitive B cell line JY, the early T cell leukemic lines MOLT-4, Jurkat and HPBALL; the mature B cell myeloma

1162

D. G. Jackson, D. N. J. Hart, G. Starling and J. I. Bell

B-CELL

MYELOID

Eur. J. Immunol. 1992. 22: 11.57-1163

T-CELL

lo3, RNA SIZE (bp)

Agure 4. Northern blot analysis of CMRF3.5 transcripts in different cell 1ines.Total cellular RNA was prepared from each of the cell lines shown in the figure and electrophoresed on a 1.2 % agaroseiformaldehyde gel followed by transfer to nitrocellulose membrane.The blot was probed with the 800-bp Xba I fragment of the cloned CMRF35 cDNA insert and exposed for autoradiography for 10 days at -80°C. The positions of RNA molecular weight calibration markers are shown at the right of the figure.

U-266 and the adherent macrophage-like cell line THP-1. Other non-lymphoid tissues including thyroid and small cell carcinoma of the lung were negative for CMRF35 antigen expression when tested by Northern blot analysis (data not shown). These results demonstrate that the CMRF35 antigen is expressed on cells from both the myeloid and lymphoid differentiation pathways and are consistent with the results of flow cytometric analyses with the monoclonal antibody CMRF35 (Daish et al., manuscript submitted). In these latter studies CMRF35 antigen expression was detected on monocytes, granulocytes, LGL and a proportion of T and B lymphocytes. Approximately 25 YO of peripheral blood T cells representing both the CD4 and CD8 subsets as well as the CD45RA and CD45RO lymphocytes are CMRF+ whereas few tonsil T cells (less than 5 Yo) express the antigen.The CMRF35 antigen is also present on a subpopulation of peripheral blood B cells (approximately 15 Yo).

The similarity between the CMRF35 V domain and the Fc receptor for polymeric IgA raises the possibility that the CMRF35 antigen may itself function as an Fc receptor. A number of candidate IgA Fc receptors have indeed been described on peripheral blood lymphocytes 143, 441, neutrophils 14.51 and macrophages [46]. However, few of these putative IgA receptors have been fully characterized and in the single example that has been cloned and sequenced to date [46] this resembles a conventional Fc receptor molecule containing two Ig-related C domains. It seems equally likely at this point that the CMRF35 antigen may function in cell adhesion or cell signalling by analogy with other single-V-domain Ig superfamily members such as CD28, CTLA-4 and CD8 [6, 101. Clearly these various possibilities will need to be tested in future experiments using cells transfected with the cloned CMRF35 cDNA in order to address this question directly.

The expression of the CMRF35 antigen on T and B cells may reflect the state of activation of these cells rather than the stage of differentiation. For example mitogenic stimulation of peripheral blood lymphocytes with PHAleads to a dramatic decrease in the level of the CMRF35 transcript within 72 h (Fig. 4) correlating with a similar decrease in T cell surface staining on flow cytometry (Daish et al., manuscript submitted).

We are grateful to Margaret Jonesfor generous and expert assistance with immunocytological staining, Jonathan Edwards for assistance with some of the figures and to Mark Inglis for photomicrography. We also thank Professor Alan Williamsfor many helpful discussions and for careful reading of the manuscript. Received November 22, 1991; in revised form January 4, 1992.

5 References 4 Concluding remarks In this report we have described the sequencing and expression of the CMRF35 antigen, a new member of the Ig gene superfamily that contains a single V-related domain. Our evidence demonstrates a very high degree of sequence homology between CMRF35 and the poly Ig Fc receptor for IgA and IgM and provides very strong evidence that these two superfamily members belong to a discrete evolutionary group derived from a common ancestral V domain gene.

1 Williams, A. F. andBarclay, A. N., Annu. Rev. Immunol. 1988. 6: 381. 2 Hunkapiller, T. and Hood, L., Adv. Zmmunol. 1989. 44: 1. 3 Lew, A. M., Lillehoj, E. F!, Cowan, E. P., Maloy, W. L., Van Schravendijk, M. R. and Coligan, J. E., Immunology 1986.57: 3. 4 Kaufman, J. F., Auffray, C., Korman, A. J., Shackleford, D. A. and Strominger, J., Cell 1984. 36: 1. 5 Ashwell, J. D. and Klausner, R. D., Annu. Rev. Zmmunol. 1990. 8: 139. 6 Littman, D. R., Annu. Rev. Immurzol. 1987. 5: 561.

Eur. J. Immunol. 1992. 22: 1157-1163

CMRF35, a new Ig superfamily member related to the poly Ig receptor

7 Simmons, D., Makgoba, M. W. and Seed, B., Nature 1988.331: 624. 8 Seed, B., Nature 1987. 329: 840. 9 Kinet, J. F!, Cell 1989. 57: 351. 10 Aruffo, A. and Seed, B., Proc. Natl. Acad. Sci. USA 1987.84: 8573. 11 Williams, A. F. and Gagnon, J., Science 1982. 216: 696. 12 Starling, G. C., PhD. Thesis University of Otago. 1989. 13 Seed, B. and Aruffo, A., Proc. Natl. Acad. Sci. USA 1987.84: 3365. 14 Mostov, K. E., Friedlander, M. and Blobel, G., Nature 1984. 308: 37. 15 Mostov, K. E., Kraehenbuhl, J. E! and Blobel, G., Proc. Natl. Acad. Sci. USA 1980. 77: 7257. 16 Brandtzaeg, I?, Clin. Exp. lmmunol. 1981. 44: 221. 17 Jackson, D. G. and Bell, J. I., J. Immunol. 1990. 144: 2811. 18 Chomczynski, F! and Sacchi, N., Anal. Biochem. 1987. 162: 156. 19 Sambrook, J., Fritsch, E. E and Maniatis, T., Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor 1989. 20 Sanger, F., Nicklen, S. and Coulson, A., Proc. Natl. Acad. Sci. USA 1977. 74: 5463. 21 Kozak, M., Nucleic Acids Res. 1987: 15: 8125. 22 Von Heijne, G., Nucleic Acids Res. 1986. 11: 4683. 23 Devereux, J., Haeberli, E! and Smithies, O., Nucleic Acids Res. 1984. 12: 387. 24 Dayhoff, M. O., Barker, W, C. and Hunt, L. T., Methods Enzymol. 1983. 91: 524. 25 Eiffert, H., Quentin, E., Decker, J., Hillemeir, S., Hufschmidt, M., Klingmiiller, D.. Weber, M. and Hilschmann, N., Hoppe Seyler's Z. Physiol. Chem. 1984. 365: 1489. 26 Mostov, K. E. and Blobel, G. J., J. Biol. Chem. 1982. 257: 11816. 27 Sztul, E. S., Howell, K. and Palade, G. E., J. Cell Biol. 1983. 97: 1582. 28 Frutiger, S., Hughes, G. J., Hanly,W. C., Kingzette, M. and Jaton, J.-C., J. Biol. Chem. 1986. 261: 16673. 29 Sewell,W. A., Palmer, R. W., Spurr, N. K., Sheer, D., Brown, M. H.. Bell,Y. and Crumpton. M. J., Zmmunogenetics 1988.28: 278. 30 Lafage-Pochitaloff, M. R., Costello, D., Couez, J., Simonetti, F!, Manoni, C., Mawas, C. and Olive, D., Immunogenetics 1990. 31: 198.

1163

31 Triebel, F., Jitsukawa, S., Baixeras, E., Roman-Roman, S., Genevee, C., Viegas-Pequignot, E. and Hercend, T., J. Exp. Med. 1990. 171: 1393. 32 Classon, B. J., Brown, M. H., Garnett, D., Somoza, C., Barclay, A. N. ,Willis,A. C. and Williams, A. F., Int. Immunol. 1991, (in press). 33 Killeen, N., Barclay, A. N.,Willis, A. C. and Williams, A. F., EMBO J. 1987. 6: 4029. 34 Leonard, W. J., Depper, J. M., Crabtree, G. R., Rudikoff, S., Pumphrey, J., Robb, R. J., Kronke, M., Svetlik, F! B., Peffer, N. J., Waldmann, T. A. and Greene, W. C., Nature 1984. 311: 626. 35 Nikaido, T., Shimizu, A . , Ishida, N., Sabe, H., Teshigawara, K., Maeda, M., Uchiyama, T. ,Yodoi, J. and Honjo, T., Nature 1984. 311: 631. 36 Liszewski, M. K. and Post, T. W. and Atkinson, J. I?, Annu. Rev. Immunol. 1991. 9: 431. 37 Reddy, l?, Caras, I. and Krieger, M., J. Biol. Chem. 1989.264: 17329. 38 Betz, H., Biochemistry 1990. 29: 3591. 39 Blank, U., Ra, C., Miller, L.,White, K., Metzger, H. andKinet, J. l?, Nature 1989. 337: 187. 40 Kinet, J. l?, Metzger, H., Hakimi, J. and Kochan, J., Biochemistry 1987. 26: 4605. 41 Clevers, H., Alarcon, B., Wileman, T. and Terhorst, C., Annu. Rev. Immunol. 1988. 6: 629. 42 Fossum, S., Mallett, S. and Barclay, A. N., Eur. J. Immunol. 1991. 21: 671. 43 Lum, L. G., Muchmore, A. V., Keren, D., Decker, J, Koski, I . , Strober, W. and Blaese, R. M., J. Immunol. 1970. 122: 65. 44 Gupta, S., Platsoucas, C. D. and Good, R. A., Proc. Natl. Acad. Sci. USA 1979. 76: 4025. 45 Mazengera, R. L. and Kerr, M. A., Biochem. J. 1990. 272: 159. 46 Maliszewski,C. R., March, C. J., Schoenborn, M. A . , Gimpel, S. and Shen, L., J. Exp. Med. 1990. 172: 1665. 47 Lewis, V. A . , Koch, T., Plutner, H. and Mellman, I., Nature 1986. 324: 372. 48 Gold, D. F!, Clevers, H., Alarcon, B., Dunlap, S., Novotny, J., Williams, A. F. and Terhorst, C., Proc. Natl. Acad. Sci. U S A 1987. 84: 7649.

Molecular cloning of a novel member of the immunoglobulin gene superfamily homologous to the polymeric immunoglobulin receptor.

The CMRF35 monoclonal antibody recognizes a cell membrane antigen present on the surface of monocytes, neutrophils, a proportion of peripheral blood T...
797KB Sizes 0 Downloads 0 Views