The novel variants of mb-1 and B29 transcripts generated by alternative mRNA splicing

Immunology Letters, 47 ( 1995) 15 1 - 156 0165.2478/95/$09.50 0 1995 Elsevier Science B.V. All rights reserved IMLET 2408

The novel variants of mb-1 and B29 transcripts generated by alternative mRNA splicing Mariko Koyama a, Tetsuya Nakamura av*, Masaaki Higashihara a, Bettie Herren ‘, Shoji Kuwata b, Yoichi Shibata b, Ko Okumura d and Kiyoshi Kurokawa a ‘First Department of Internal Medicine and ’ Department of Transfusion Medicine and Immunohematology, Faculty of Medicine. Uniwrsity of Tokyo, 7-3-l Hongo. Bunkyo-ku, Tokyo 113, Japan: ’ Department of Microbiology, Division of Developmental and Clinical Immunology, Uniuersity of Alabama at Birmingham, Birmingham, AL 35294, USA; d Department of Immunology, Juntendo Uniuersity School of Medicine, 2-I-l Hongo, Bunkyo-ku, Tokyo 113, Japan (Received

26 March 1995; accepted

Key words: B-cell receptor; Ig-a;

Is-p; Alternative

1. Summary The Ig-a/ Ig-p heterodimers encoded by mb-I and B29 genes, respectively, are crucial for the constitution of the B-cell receptor (BCR). We report here novel variants of mb-1 and B29 transcripts produced by alternative mRNA splicing. The proteins encoded by these variants are predicted to conserve transmembrane and cytoplasmic portions of Ig-a and Ig-p but lack a part of the extracellular portions containing cysteine residues which are required for intramolecular and intermolecular S-S bonds. Transfection studies revealed that the variant mb-1 and B29 did not contribute to the BCR expression on cell surfaces. Although peripheral B cells contain small amounts of the variant mb-1 and B29 transcripts, treatment with an anti-IgM antibody, LPS or IL-4 induces a significant increase in amounts of the variant transcripts. These observations suggest that Bcell activation induces alternative splicing of mb-1 and B29 transcripts which encode proteins unable to constitute the BCR.

2. Introduction The B-cell antigen receptor (BCR) is composed of Ig and disulfide-linked heterodimers consisting of Ig-a and Ig-p glycoproteins [l-5] which are encoded by the

Corresponding author: Tetsuya Nakamura, First Department of Internal Medicine, Faculty of Medicine, University of Tokyo, 7-3-l Hongo, Bunkyo-ku, Tokyo 113, Japan. l

SSDI 0165-2478(95)00071-2

18 April 1995) splicing; B-cell activation

B cell-specific genes mb-1 [6] and B29 [7], respectively. Molecular heterogeneity of Ig-o and Ig-p has been observed as a function of B-cell development (8-151. The heterogeneity is partly explained by the differential glycosylation of Ig-a and Ig-p [lo], but recent works have suggested the existence of post-transcriptional mechanisms [ 14,151. We report here novel variant mRNA of both human mb-1 and B29 genes which are generated by alternative splicing and lack a part of the extracellular domains. Since the deletion results in an in-frame shift of mRNA sequences, the products of the variant mb-1 and B29 are predicted to conserve the transmembrane and intracellular portions of Ig-a and Ig-p. The function of these variant products and the physiological significance are discussed.

3. Materials

and Methods

3.1. Cells and antibodies Human B lineage cell lines (Nalm6, Daudi, Ramos and IM-9) were described previously [lo]. PBMCs (5 X lo6 cells/ml) were isolated from healthy volunteers by Ficoll-Hypaque density gradient centrifugation and cultured with or without 2 pug/ml of SA-DA-4.4, 50 wg/ml of lipopolysaccharide (LPS) (Sigma, St. Louis, MO) or 20 U/ml of rIL-4 (Genzyme, Boston, MA) at 37°C for 48 h. The monoclonal antibodies (mAbs1 to human IgM @A-DA-4.4) and Ig-p (CB3- 1) were described previously [10,16,17]. Phycoerythrin (PE)labeled streptavidin was purchased from Southern Biotechnology Associates (Birmingham, AL). 151

3.2. Analysis of mRNA transcripts

The first strand cDNA was synthesized from 10 pg of total RNA [IS] with an AMV reverse transcriptase (Boehringer Mannheim, Indianapolis, IN) and oligo(dT) and used as templates in polymerase chain reaction (PCR). The used primers which are located at S- or 3’-end of mb-1 and B29 coding regions are an mb-lspecific sense primer S-ATGAAGCTTTAACCAACCCACTGGGAGAA-J, an mb- 1-specific antisense primer 5’-CGGGATCCTCACGGCTTCTCCAGCTGGAC-J, a B29-specific sense primer 5’-CCCAAGCTTATGGCCAGGCTGGCGTTGTCT-3’ and a B29-specific antisense primer 5’-CGGGATCCTCACTCCTGGCCTGGGTG-CTC-3’. DNA fragments derived from the reverse transcription (RT)-PCR were electrophoresed on agarose gels and transferred to Hybond Nf nylon membranes (Amarsham, Buckinghamshire, UK). The membranes were hybridized for 16 h at 50°C in the hybridization buffer containing the full-length of mb-1 or B29 cDNA labeled with DIG DNA Labeling Kit (Boehringer Mannheim) according to manufacturer’s instructions. The hybridized probes were detected using DIG Luminescent Detection Kit (Boehringer Mannheim). For DNA sequencing, the RT-PCR derived products were purified and DNA sequencing reactions were carried out using a Taq DyeDeoxy Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA) and analyzed on an Applied Biosystems Automated Sequencer Model 373A. 3.3. Transfection

assay

The EcoRI fragment of mb-1 cDNA (a kind gift from Dr. Peter Burrows, University of Alabama at Birmingham) [19], XbaI fragment of B29 cDNA (a kind gift from Dr. Randolph Wall, University of California at Los Angeles) [20], and variant mb-1 and B29 cDNA obtained from Ramos B cells by RT-PCR were introduced into an expression vector pEF-BOS (a kind gift from Dr. Shigekazu Nagata, Osaka Bioscience Institute, Japan) [21]. Vectors containing inserts with the correct orientation were selected and used for transfection studies. Murine fibroblast L cells were first introduced with ~468 (a kind gift from Dr. Michel Nussenzweig, The Rockefeller University, New York) [22] containing the constructs of human p heavy-chain (HC), K light-chain (LC) and a neomycin-resistant gene with the Lipofectin method (GIBCO) and designated L~K 0. One or both of the variant or wild-type mb-1 and B29 introduced in pEF-BOS were then transfected to L~KO together with l/10 molar of pSV2bsr (Funakoshi, Tokyo, Japan). Cells growing in the presence of 5 pg/ml of blastocidin S (Funakoshi) were selected and 152

designated L F.K MF (transfected with wild-type mb- I), L~KBF (wild-type B29), LFKMFBS (wild-type mb-1 and variant B29), L~KMSBF (variant mb-1 and wildtype B29), L~K MSBS (variant mb-1 and B29) and L~KMFBF (wild-type mb-1 and B29). 3.4. Immunojluorescence

analysis

Cells (1 X 106) were incubated with biotinylated SA-DA-4.4, CB3-1 or isotype-matched control mAb with irrelevant specificity (50 pg/ml) for 20 min at 4°C and the bound mAbs were detected with PE-labeled streptavidin (20 pg/ml). Cells were analyzed on a FACScan using Lysis II software (Becton Dickinson, San Jose, CA).

4. Results and Discussion 4.1. Analysis of mb-1 and B29 transcripts

in human B

cells

Total RNA was prepared from B-lineage cell lines, reverse-transcribed to cDNA and amplified by PCR using mb-l- or B29-specific primers located at the 5’and 3’-ends of the respective coding region. As shown in Fig. lA, 2 bands of 700 bp and 590 bp were hybridized with an mb-1 cDNA probe. The larger fragment corresponds to the full-length mb-1 transcript observed in a control PCR using the cloned mb-1 cDNA as a template (lane 4). The smaller band was detected in all B-lineage cell lines examined (lanes l-3), but was not detected in the control PCR (lane 4). Fig. 1B shows the result of RT-PCR analysis of B29 transcripts. Two fragments of 690 bp and 380 bp were

” E < i

bp

1018-

506 -

*

1 1

2

.. 3

4

1

2

3

4

of mb-I and B29 transcripts in B cells. Total RNA was extracted from Nalm6 (lane I), Ramos (lane 2) and IM9 (lane 3) and subjected to reverse-transcriptasereaction.The RT-products (A,B,

Fig. 1. Analysis

lanes I-3). mb-I cDNA (A, lane 4) and B29 cDNA (B, lane 4) were used as templates for PCR with mb-l- (A) or 829. (B) specific primers. The PCR-products were electrophoresed and subjected to Southern blot analysis using mb-1 (A) or B29 (B) cDNA probes. The solid arrows indicate the full-length mb-1 and B29, and the open arrows their smaller forms.

hybridized with a B29 cDNA probe. As is the case with the mb-1 , control PCR using the cloned B29 cDNA showed only a 690 bp fragment (lane 41, indicating that the larger fragment was derived from full-length B29 transcripts. The smaller mb-1 and B29 transcripts are not caused by incorrect primer-annealing due to crosshybridization, since 3 different sets of primers on both mb-1 and B29 sequences yield 2 distinct PCR products whose differences in size are identical regardless of the primer combinations (data not shown). 4.2. Sequences

of the smaller mb-I and B29 transcripts

The smaller mb-1 and B29 fragments designated mb-1S and B29S, respectively, were amplified from Daudi B cells, purified and subjected to further sequence analysis. Comparison of the mb-1S sequence with the mb-I cDNA [19] revealed that mb-1S lacked 114 nucleotides encoding a part of the extracellular domain (nucleotide position 266-379) without a frame shift (Fig. 2A). Thus, mb-IS encodes a protein whose extracellular portion is shorter by 38 amino acids than, but otherwise identical to, Ig-cy except that one amino acid (position 89) is changed from glycine to glutamic acid. As a result of the deletion, 2 out of 3 extracellular cysteine residues and 3 out of 6 potential N-glycosylation sites [19] are deleted. On the other hand, B29S lacked 312 nucleotides of B29 cDNA encoding an almost entire extracellular domain (nucleotide position 119-430) of Ig-/? (Fig. 2B). Since the deletion did not result in a frame shift, the predicted B29S amino acid

sequences consisted of a hydrophobic leader segment, a small extracellular portion of 30 amino acids and intact transmembrane and cytoplasmic portions of Ig-p [20]. The deletion results in loss of all 5 extracellular cysteine residues and 3 potential N-glycosylation sites in the extracellular portion of Ig-p. We have sequenced 9 independent mb-IS clones and 6 B29S clones derived from Daudi and Ramos and obtained identical results (data not shown). To investigate the molecular mechanism for generation of mb-1S and B29S, we compared the DNA sequences with the genomic structures of these genes. The human mb-1 gene consists of 5 exons and the extracellular portion is encoded by exon II and the proximal part of exon III [23]. The deletion of mb- 1S is from GT at position 266 (Fig. 2A) in exon II to AG at the 3’.end of the second intron. On the other hand, the comparison of the B29S sequence with the human B29 gene structure [24] reveals that the deletion is from CT at the S-end of the second intron to AG at the Y-end of the third intron including entire exon III. The fact that Sand 3’-ends of the deleted fragments in both cases have cryptic splice donor (GT) and acceptor sequences (AG) suggests that the variants are produced by alternative splicing of the mb-1 and B29 transcripts [25,26]. 4.3. Reconstitution

of the BCR with mh-IS and B29S

Since the deletion in both cases results in an in-frame shift, mb-1S and B29S products are predicted to conserve transmembrane and cytoplasmic portions identical

B 20 60

L S,.A. V Y,_L.G P,.C, C .Q .h L H M H K MB-1F Cl%Tf.&iXGTCTA~~~GG~~C~~TCTGGATGCACAAGGTCCCAGCA MB-IS CTCT~CCTGT~~~~~G~~C~~~CCTGTGCATCCACAACGTCCCACCA ;LMVSLGEDAHFP MB-11: TCATTCATGCTGAGCCTGCCCCAACACGCCCACTTCCA MB-K? TCATTCATGCTGAGCCTGCGCGAACACGCCCACTTCCA

V

P

A

40 I?0

run

H-N CGCACAATACCAGCAAC CGCACAATAGCACCAAC

GO 180

VA i&?-b % R V L H Cd% P P E F MB 1F ~ACGCCAACCTCACCTCGTGGCGCCTCCTCCATCCCAACTACACCTGGCCCCCTCAG~C 2:: MB-IS 4ACCCCAACGTCACCTGGTCCCGCGTCCTCCATCGCAACTACACGTGCCCCCCTCAG~C ix* CHO 100 I, G P GE D P N Ei L I I P N v N r L H MB-IF TTGGCCCCCCCCGACGACCCCAATGGTACGCTCATCATCCACAATGTGAACAAGAGCCAT300 MB-1S T7CGGCCC~~~C~A~~ACCCC~T~................................... YPE‘HESYQPS MB-1F b&C~TA;AC~T ....... MB-1S ..__..._.....,....,,_.....,._.._._..................

120 360

BZSF B29S B29F 029s BZSF B29S B29F B29S EYDEVPPPLILEIGRVIl B29F CAGATCGACGAGAATCCCCAGCAGCTGAAGCTGCA4AACG~~ ..._.. BZBS ...... ..._______............................... CH" NESLATLTIP‘IRFEL.~~:" B29F AACGAATCTCTCGCCACCCTCACCATCCAAGGCATCCGC~CAG~;~C~~T~~~ 4TCTAI 829s .........___....._.............__...........................

140 PPPRPFLDMCECT TYLRVR' MB-1F ACCTACCTCCGCGTGCGCCACCCGCCCCCCAGGCCCTTCCTCGACATGCGGGAGGGCAC~420 ~AGCCGCCCCCCACGCCCTTCCTGCACATCGGGCAGGCCAC~ MB-1S ......."'."".

B29F BZSS

I( N R MB-1F AAGAACCG4 MB-l.5 AAGAACCC4

R Y M, FSTLAQLKPRNTLKD B29F CCAGTCATCGCATICAGCACC~GGCACAGCTGAAGCACAGGAACACGCTGA4~G4l 8295 .........-CA~CAGCACCTTCGCACAGCTCAACCACACGAACACGCT~44tiG.4T

143 429

100 :ioo 120 360

:59 177

Fig. 2. Nucleotide sequences and predicted amino acid sequences of mb-I S and B29S. The nucleotide sequences of the extracellular ponions of mb-1S and B29S are aligned with those of the full-length mb-1 (mb-1F) and B29 (B29F). The deletions in mb-1S and B29S are represented by dots. The predicted amino acid sequences are represented by single letter codes. The marks and symbols in the sequences are shaded; leader peptides, boxes; cysteine residues and CHO; potential sites for addition of the N-linked glycosylation.

153

to wild Ig-cz and Ig-p and thus could be transmembrane molecules. However, both deletions cause loss of the cysteine residues in the extracellular portion proximal to the cell surface which are thought to be involved in disulfide-linkage between Ig-cl! and Ig-p [ 11. One of the essential functions of Ig-a and Ig-p is to associate with

Ig, stabilize its transmembrane portion and promote cell surface expression of the BCR [1,27-291. Thus, to investigate whether mb-1S and B29S products could contribute to BCR constitution, a transfection study was carried out. Various combinations of pEF-BOS containing mb-1, mb-lS, B29 or B29S were transfected into L

1. Daudi

100

10’

102

105

10'

loo

fluorescense(log)

4.

10'

102

10'

10'

loo

fluorescense(log)

5. L~KBF

LpcMF

10'

102

IO'

1w

fluorescense(log)

6. L~KMFBS

ii

t’ = 8 100

10'

102

101

104

18

fluorescense( log)

7. LpcMSBF

fluorescense(log)

"7 10'

10'

10'

fluorescense(log)

8.

LpicMSBS

fluorescense(log)

10'

loo

10'

101

fluorescense(

103

IO'

log)

9. L~KMFBF

fluorescense(log)

Fig. 3. Cell surface expression of the BCR on mb-l(S) and B29W transfectants. Daudi B cells (panel 1). L (21, L~KO (3), L~KMF (4). L~KBF (5), L&K MFBS (61, L FK MSBF (7), L ILKMSBS (8) and L~K MFBF (9) were incubated with biotinylated SA-DA-4.4 (thick lines) or CB3- 1 (thin lines) followed by PE-labeled streptavidin and analyzed on a FACScan. An isotype-matched control mAb served as a control (dotted lines).

154

B

A bp 1018-

12

3

4

5

b

X( Xa)ni

I

1018-

’ ’ “,

*.,>2

tB29F tmb-1F +mb-1S

506ggtgI

(:*. tB29S

Fig. 4. Activation of peripheral B cells induces alternative splicing of mb-l and B29. PBMCs obtained from healthy volunteers were cultured without stimulation (lane 1) or with anti-IgM mAb (lane 2). LPS (lane 3) or IL-4 (lane 4) for 48h. Reverse-transcribed RNA prepared from the harvested cells (A,B, lanes l-4). mb-1 cDNA (A, lane 5) or 829 cDNA (B, lane 5) was used as templates in PCR with mb-I -specific primers (A) or B29-specific primers (B). The PCR products were electrophoresedand subjected to Southern blot analysis using mb-1 (A) or B29 CR)-specific probes.

cells expressing human p HC and K LC (L~K 0) and designated as described in Materials and Methods. After the expression of transgenes was confirmed by RT-PCR (data not shown), the cell surface expression of the BCR on these transfectants was examined by flow cytometry using anti-IgM and Ig-p mAbs. L~KMFBF expressed cell-surface BCR to the degree similar to Daudi B cells (Fig. 3, panels I and 9>, but L~KMFBS did not show any surface BCR expression (panel 61, indicating that B29S transcripts did not support surface BCR expression. Introduction of B29 alone (L~KBF) unexpectedly allowed surface BCR expression (panel 51, but the BCR expression on L~KMSBF (panel 7) did not exceed that of L~K BF, indicating that the mb-1S product did not contribute to constitution of BCR.

in down-regulation activation.

of BCR in the course of late B-cell

Acknowledgements We thank Drs. Peter Burrows, Randolph Wall, Michel Nussenzweig and Shigekazu Nagata for providing reagents. We also thank Ms. Mari Takeuchi and Ms. Masami Yanagisawa for excellent technical assistance. This work is supported by grants-in-aid from the Ministry of Education, Culture and Science of Japan (05670896, 06770838).

References 4.4. B-cell activation mb-I and 829

induces

alternative

splicing

of

To investigate the physiological significance of the post-transcriptional regulation of mb-1 and B29 genes, we analyzed normal B cells activated with various stimuli. PBMCs from healthy volunteers contained mainly full-length mb-1 and B29 transcripts and the smaller forms were observed only after longer exposure (Fig. 4A,B, lane 1). However, when cells were stimulated with an anti-IgM mAb or lL4 for 48 h, the amounts of both mb-1s and B29S increased (Fig. 4A,B, lanes 2 and 4). LPS also induced the induction of B29S (lane 3). These results together with the results of transfection studies suggest that activation of B cells induces the alternative splicing of mb-1 and B29 genes which does not contribute to the constitution of the BCR. Therefore, this post-transcriptional regulation of mb-1 and B29 genes may have some physiological role

111Reth, M. (1992) Annu. Rev. Immunol. 10, 97. 121Hombach, J., Leclercq, L., Radbruch, A., Rajewsky,

K. and Reth, M. (1988) EMBO J. 7, 3451. 131Campbell, K.S. and Cambier, J.C. (1990) EMBO J. 9, 441. [41 Van Noesel, C.J.M., Borst, J., De Vries, E.F.R. and Van Lier, R.A.W. (1990) Eur. J. Immunol. 20, 2789. 151Van Noesel, C.J.M., Van Lier, R.A.W., Cordell, J.L., Tse, A.G.D., Van Schijndel, G.M.W., De Vries. E.F.R., Mason, D.Y. and Borst, J. (1991) J. Immunol. 146, 3881. [61 Sakaguchi, N., Kashiwamura, S., Kimoto. M.. Thalmann. P. and Melchers, F. (1988) EMBO J. 7, 3457. [71 Hermanson, G.G., Eisenberg, D., Kincade, P.W. and Wall, R. (1988) Proc. Natl. Acad. Sci. USA 85, 6890. [81 Venkitaraman, A.R., Williams, G.T., Dariavach. P.. and Neuberger, M.S. (1991) Nature 352, 777. [91 Hombach, J., Tubata, T., Lrclercq, L., Stappert. H. and Reth, M. (1990) Nature 343, 760. [lOI Nakamura, T., Kubagawa, H. and Cooper. M.D. (1992) Proc. Natl. Acad. Sci. USA 89, 8522. [ill Campbell, KS., Hager, E.J. and Cambier, J.C. (1991) J. Immunol. 147. 1575.

155

[12] Ishihara, K., Wood, W., J. Jr., Wall, R., Sakaguchi, N., Michnoff, C., Tucker, P.W. and Kincade, P.W. (1993) J. Immunol. 150, 2253. [13] Chen, J., Stall, A.M., Herzenberg, L.A. and Herzenberg, L.A. (1990) EMBO J. 9, 2117. [14] Friedrich, R.J., Campbell, K.S., and Cambier, J.C. (1993) J. Immunol. 150, 2814. [15] Goitsuka, R., Youn, H.Y., Watari, T., Tsujimoto, H. and Hasegawa, A. (1994) J. Immunol. 153.5127. [16] Maruyama, S., Kubagawa, H. and Cooper, M.D. (1985) J. Immunol. 135, 192. [17] Hata, D., Nakamura, T., Kawakami, T., Kawakami, Y., Herren, B. and Mayumi, M. (1993) Immunol. Lett. 40, 65. [18] Chomczynski, P. and Sacchi, N. (1987) Analytical B&hem. 162, 156. [19] Ha, H., Kubagawa, H. and Burrows, P.D. (1992) J. Immunol. 14 8, 1526. [20] Wood, W., J. Jr., Thompson, A.A., Korenberg, J., Chen, X.N., May, W., Wall, R. and Denny, CT. (1993) Genomics 16, 187.

156

[21] Mizushima, S. and Nagata, S. (1990) Nut. Acids Res. 18. 5322. [22] Costa, T.E., Franke, R.R., Sanchez, M., Misulovin, Z. and Nussenzweig, MC. (1992) J. Exp. Med. 175, 1669. [23] Ha, H., Bamoski, B.L., Sun, L., Emanuel, B.S. and Burrows, P.D. (1994) J. Immunol. 152, 5749. 1241 Hashimoto, S., Chiorazzi, N. and Gregersen, P.K. (1994) Immunogenetics 40, 145. [25] Cech, T.R. (1983) Cell 34, 713. [26] Ge, H., Ping, Z. and Manley, J.L. (1991) Cell 66, 373. [27] Lankester, A.C., van Schijndel, M.W., Fromme, J., Cordell, J.L., van Lier, R.A.W. and van Noesel, C.J.M. (1994) J. Immunol. 152, 2157. 1281 Pleiman, C.M., Chien, munol. 152, 2837. [29] Blum, J.H., Stevens, Chem. 268, 27236.

N.C. and Cambier, T.L. and DeFranco,

J.C. (1994) J. ImA.L. (1993) J. Bio.