Signals controlling alternative splicing of major histocompatibility complex H-2 class I pre-mRNA
Diane E. Handy 1, James McCluskey z' *, Andrew M. Lew l' **, John E. Coligan 1, and David H. Margulies 2
1 - - 2 Blologxcal Resources Branch and Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, USA
Abstract. The use of alternative splice acceptor sites dur- ing the removal of intron 7 in pre-mRNA splicing produces two forms of H-2K b protein: the predominant form, derived from a transcript that has spliced at the upstream splice acceptor site for exon 8 (long exon 8), and a K b molecule derived from a transcript that has spliced at the downstream acceptor site for exon 8 (short exon 8). We have identified a potential lariat branch point adenosine for the upstream acceptor splice site. This adenosine is found 28 bp from the splice junction and is contained in the sequence AGTGATGG. D-region genes, which use only the downstream splice site, have the sequence AGTGGTGG. We have used in vitro mutagene- sis to change this A of the H - 2 K b gene to G and have made the reciprocal change in H - 2 D d. Elimination of this adenosine in H - 2 K b alters the pattern of pre-mRNA splicing and results in a predominance of the K b molecules with short exon 8 encoded sequences. However, the addition of an adenosine in H - 2 D a is not sufficient to direct splicing to the upstream site.
Introduction
Class I antigens of the major histocompatibility complex (MHC) are polymorphic, membrane-bound glycopro- teins, noncovalently associated with beta-2 microglobulin (B2m) (Coligan et al. 1981, Nathenson et al. 1981), that
* Presentaddress:DepartmentofPathologyandImmunology, Monash University, Commercial Road, Victoria 3181, Australia
** Present address: The Walter and Eliza Hall Institute of Medical Research, Post Office, The Royal Melbourne Hospital, Victoria 3050, Australia
Address correspondence and offprint requests to: D. E. Handy, BRB, NIAID, NIH, Bldg. 5, Rm. 422, Bethesda, MD 20892, USA
consist of three extracellular domains (known as N, C 1, and C2), a transmembrane domain, and a cytoplasmic tail. The gene structure reflects the domain organization of the class I molecules. There are eight exons: a leader exon, one exon for each of the extracellular domains and the transmembrane region, and three exons that encode the cytoplasmic region (Steinmetz and Hood 1983). A hallmark of class I molecules is their high degree of diver- sity and genetic polymorphism (Kimball and Coligan 1983, Klein 1986). Multiple loci encode class I molecules, and there are more than 100 alleles for each locus (Klein 1986). In addition to the diversity generated by sequence differences, another potential source of diversity is creat- ed by alternative splicing of primary transcripts that results in multiple forms of H-2 protein derived from the same gene (Lew et al. 1986a).
Exon-encoded regions of pre-mRNA are joined by a two-step process that involves (1) formation of an inter- mediate lariat structure through the phosphodiester link- age of the 5' (donor) splice site to an adenosine (lariat branch point) 18-35 bp upstream from the 3' (acceptor) splice site and (2) excision of the lariat when the 5' splice site joins the 3' splice site (reviewed by Padgett et al. 1986). In yeast, the lariat branch point adenosine is found within the highly conserved sequence TACTAA_C (Dom- dey et al. 1984, Rodriguez et al. 1984). In higher eukary- otes, this sequence is less conserved and follows the con- sensus PyXPyTPuAPy for the globin genes (Reed and Maniatis 1985).
Previously, we suggested that an adenosine contained within the octamer AGTGATGG may be important in the choice of acceptor splice sites for the removal of intron 7 from H-2K b and H-2D a transcripts (Lew et al. 1986b). This octamer is found twice in intron 7 of H-2K b, and splicing occurs at splice acceptor sites downstream of either octamer. In the splicing of H-2D d transcripts, only one of the splice acceptor sites is used. Upstream from the unused acceptor splice site is the variant octamer AGTGGTGG.
82 D.E. Handy et al. : Alternative splicing of class I pre-mRNA
To evaluate the effects of the A or G nucleotide at posi- tion - 2 8 from the upstream acceptor site ( -28u) 1 on the regulation of pre-mRNA sPlicing of H-2K b and H-2D d transcripts, we have used site-directed mutagenesis to construct mutants of H - 2 K ~ and H - 2 D d genes and have examined the mRNA, protein, and functional expression of these genes following DNA-mediated gene transfer into mouse L cells.
Materials and methods
Cell lines. I-3, an L-cell transfectant that expresses H-2K b (Bluestone et al. 1985), was kindly provided by Dr. Bluestone, Immunology Branch, NCI, NIH. T4.8.3, an L-cell transfectant that expresses H-2D d, was previously described (Margulies et al. 1983).
Convention for numbering of nucleotides in intron 7. Nucleotides 5' of the splice aeceptor site for the short exon 8 (T in Fig. 1) are numbered from - 1 to - 2 8 and are denoted by the subscript D (downstream splice acceptor). Thus, the adenosine at position - 2 5 ( 0 ) is referred to as -251> Similarly, nucleotides 5' of the splice acceptor site for the long exon 8 (V in Fig. 1) are numbered from - 1 to - 5 6 and are denoted by the subscript U (upstream splice acceptor). Thus, the adenosine at position - 2 8 (o) is referred to as - 2 8 u.
Mutant H-2 genes. The strategy for constructing mutant H-2K b and H-2D a genes is outlined in Figure 2. Bam HI fragments that included intron 7 were subcloned into M13 vectors. Their orientation was estab- lished by analysis of restriction endonuclease digestions. Single-stranded templates were made from subclones that carried the H-2 coding strand on the + strand of the phage genome. In vitro mutagenesis was carried out as follows: 2 gg of single-stranded template was annealed to 10 ng of the 17-mer universal sequencing primer, 100 ng of the hybridization probe primer (BRL, Gaithersburg, Maryland), which anneals upstream of the M 13 polylinker, and 100 ng of the 18-mer mutagenic primer that had one mismatch with the template. An overnight incubation at 14 °C of the annealed mixture in the presence of 0.3 mM each of dATP, dCTP, dTTP, dGPT, and ATP, 6 units Klenow (BRL), and 400 units T4 DNA ligase (New England Biolabs, Beverly, Massachusetts) produced dou- ble-stranded molecules that were used to transform BMH71-18mutS, an Escheriehia coli strain deficient in DNA mismatch repair (Kramer et al. 1984). Mntagenized phage were identified by differential hybrid- ization to mutagenic primers that had been end-labeled with [7-32p]ATP by T4 polynucleotide kinase (Zoller and Smith 1983). Since the mutagenic oligomer for H-2K b is complementary to the wild type of H-2D a and the mntagenic oligomer of H-2D d is complementary to the wild type H-2K b, clones of mixed phenotype could be eliminated by screening duplicate dot blots of phage DNA that were selected from the initial screening. Sequence analyses by the dideoxy method (Sanger et al. 1980) confirmed these changes. Mutated Bam HI fragments were subcloned into Bluescribe m13 vectors (Vector Cloning Systems, San Diego, California). Xba I fragments that contained the 5' end of the H-2K b (1.7 kb) or H-2D d (1.8 kb) gene were inserted upstream of the Bam HI inserts to reconstruct the geue. Fragment orientation was checked by analysis of restriction endonuclease digestion. The final reconstructed genes had deletions of approximately 0.6 kb in intron 3.
Oligonucleotides. Mutagenic oligomers were made with one mismatch near the center of an 18-mer. The mutagenic oligomer for H-2K b was
t Our convention for the description of nucleotides in intron 7 is given in detail in Materials and methods
AACAACCCACCACTCACT, for H-2D d AACAACCCATCACT- CACT. The oligomer ATTGTCTGTCACCAAGTCCAC was used to sequence intron 7 and exon 8, as it represented a common stretch of sequence in the 3' untranslated region of H-2K b and H-2D d. All oligomers were synthesized on an Applied Biosystems DNA synthesizer and purified by polyacrylamide gel electrophoresis (PAGE) in prepara- tive 20 % gels containing 6 M urea.
DNA-mediated gene transfer. L cells (H-2 k) were transfected by the calcium phosphate method as previously described (McCluskey et al. 1986a).
Flow microfluorimeuy. Flow microfluorimetry was performed as described by McCluskey and co-workers (1986a) using a Becton Dickin- son fluorescence-activated cell sorter with a logarithmic amplifier. The monoclonal antibodies 20.8.4 (Ozato and Sachs 1981) and 34.5.8 (Evans et al. 1982) were used to identify H-2K b and H-2D d antigens, respec- tively.
lL-2 stimulation of T-cell hybridomas. The H-2Kb-reactive hybridoma, HTB 157.7, was the kind gift of Dr. M. Dorf (Minami et al. 1986). Assay for production of interleukin-2 (IL-2) using the IL-2-dependent ceil line CTL-L was carried out as described (McCluskey et al. 1986a). The monoclonal antibody Y-3 (Jones and Janeway 1981) was used to block stimulation of the hybridoma by H-2K b molecules.
RNase protection. The H-2K b probe has been previously described (Lew et al. 1986b). Briefly, a 345 bp Pst I-Pvu II fragment which spans from the 3'-third of exon 6 to just beyond exon 8 was directionally cloned into pGEM 2 (Promega Biotec, Madison, Wisconsin). The H-2D d probe was derived from a Sal I/Pvu II fragment of a partial cDNA of the normal H-2D d transcript and was directionally cloned into pGEM 2 at the Sal I and Sma I sites. These vectors were linearized at the Hind III site, and RNA was synthesized from the SP6 promoter in the presence of [c~-32p] uridine triphosphate. Duplexes were formed between the RNA probe and RNA from the transfected cell lines by overnight hybridization at 45 °C. Single-stranded RNA was digested by a mixture of RNases A and T1 at 34 °C for 1 h. Protected fragments were visual- ized in a denaturing 5% polyacrylamide sequencing gel. The products of dideoxy-sequencing reactions of a known template were elec- trophoresed in parallel as size markers.
Surface-labeling of transfectants. Transfectants were labeled by lactoperoxidase catalyzed radioiodination as previously described (Lew et al. 1986b). Cells were washed three times with phosphate-buffered saline and lysed at 4 °C for 1 h with 0.5 ml of a lysis buffer: 50 mM Tris buffer, pH 7.4, 150 mM NaC1, 0.02% NAN3, 0.1 mM phenytmethylsulfonyl fuoride (PMSF), 0.5% Triton X-100.
Immunoprecipitation. For preclearance, the entire labeled-lysate (>_ 108 cells in 0.5 ml) was incubated with 100 ~tl normal rabbit serum overnight in the cold. To remove immune complexes, the lysate was added to 100 gl of packed protein-A agarose beads (PAA-beads) and after mixing, incubated 1 h at room temperature. After centrifugation, the lysate was added to packed beads and the procedure repeated. Precleared lysates were transferred to clean tubes. To separate K b molecules from endogenous H-2 molecules of the L cells, 100 gl of EH144 (Geier et al. 1986), a monoclonal antibody specific for K b determinants, was used. Packed PAA-beads (250 gl) were used to precipitate the specific immune complex. To isolate D d molecules, the same procedure was used except with a Dd-speciflc monoclonat anti- body, 34.5.8. The agarose beads were washed four times in wash buffer (50 mM Tris buffer, pH 7.9, 0.45 M NaC1, 5 mM ethylenediamine tetra- acetate, 0.1 mM PMSF, 0.5% Triton X-100). Agarose beads were washed two times with 0.1 M NaPQ, pH 8.6. Immune complexes were disrupted by boiling for 3 min in 0.1 M NaPO4, 0.5% sodium
D.E. Handy et al.: Alternative splicing of class I pre-mRNA 83
dodecyl sulfate (SDS), and the beads were removed by centrifugation. After the addition of NP-40 to a final concentration of 0.5-0.6% and 5 units of N-glycanase <TM> (peptide:N glycosidase F; Genzyme, Boston, Massachusetts), samples were incubated overnight at 37 °C. The N-glycanase (TM> digested material was diluted by the addition of 3 vol. of lysis buffer. Approximately one fourth of each lysate was in- cubated with 20 ~tl each antipeptide sera overnight at 4 °C. Antipeptide sera were prepared in rabbits (as described by Lew et al. 1986b) and were kindly provided by Dr. W. L. Maloy (see Table 1 for specificities). To precipitate specific immune complexes, 60 gl of a 50% protein-A agarose suspension was added. Samples were treated as above, washed four times in wash buffer, and analyzed by SDS-PAGE under reducing conditions.
Results
In vitro mutagenesis of H-2K b and H-2D d genes. To test the hypothesis that the A/G difference at position - 2 8 U is involved in controlling splicing of exon 8 in H-2K b and H-2D d pre-mRNAs, site-directed mutants were generated. To eliminate this adenosine (Fig. 1, o), the H-2K b sequence of AGTGATGG was changed to AGTGGTGG using the complementary mutagenic oligomer AACAACCCACCACTCACT. Conversely, the mutagenic primer AACAACCCATCACTCACT was used to change the octamer AGTGG-TGG of H-2D d to AGTGATGG. Functional genes were reconstituted as depicted schematically in Figure 2. For simplicity, we refer to the mutagenized H-2K b gene as H-2Kbm and the mutagenized H-2D a gene as H-2D,i.
Analysis oftransfectants. Parental or in vitro mutagenized H-2K b and H-2D d genes were introduced into mouse L
cells by the calcium phosphate method of DNA-mediated gene transfer. To evaluate the cell surface expression of the protein products of the H-2K~ and H-2D~ genes, cells were reacted with monoclonal antibodies specific for H-2K b (20.8.4) or H-2D d (34.5.8) and analyzed by flow microfluorimetry. Homogeneously stained populations of cells were isolated by preparative cell sorting of the primary transformant pools. After culturing, cells derived from the preparative sorting were further analyzed for surface expression of the transfected molecules (Fig. 3). I-3 (K b) and JT25.8 (Kbm) react with 20.8.4 (anti-K b) and not 34.5.8 (anti-Dd), whereas T4.8.3 (D d) and JT25.3 (Ddm) react with 34.5.8 and not 20.8.4. The amount of surface expression of the class I antigens encoded by the transfected genes in I-3, T4.8.3, and JT25.3 is equivalent to that of the H-2K k and H-2D k molecules on the same cell. The expression of H-2Kbm antigen on JT25.8 was approximately 50% lower. These data demonstrate the expression of proteins derived from each of the transfect- ed genes.
Patterns of mRNA splicing. RNase protection experiments were used to evaluate the patterns of mRNA splicing in the transfectants. Total cellular RNA was hybridized to an antisense RNA corresponding to an H-2K b transcript with long exon 8 sequences. If the H-2K b mRNA from the transfectant has a long exon 8, then a fragment of 350 bases will be protected following RNase digestion (Fig. 4a). However, if the H-2K b mRNA has a short exon 8, a fragment of about 312 nucleotides will be protected (Fig. 4a). In I-3 (H-2K b) cells, the majority of the protected
Intron VII Exon VIII
i Dd G ............ T ....................... G ....... ~ ...... ~- ............. T ........... ~ ....
Dp G ............ T .... A ................ C-G ................ ? ............. T ........... ~- ....
{ Dk G ............ T .... A .................. G ................ 7 .......... A--T .................
Db G ............ T .... A ................ C-G ................. ~ ............. T .................
Ld G ............ T .... A ................ C-G ................. - ............. T .................
long exon VIII I J
short exon VIII []
Fig. 1. Sequence comparison of exon 8 and palt of intron 7 from H-2 region genes. All eight genomic sequences that have been published are shown: K b (Weiss et al. 1983); D d (Sher et al. 1985); K d (Kvist et al. 1983); K k (Arnold et al. 1984); D p (Schepart et al. 1986); D b and D k (Watts et al. 1987); L a (Moore et al. 1982). Vertical lines include potential AG acceptor splice sites. The adenosine (o) 28 bp 5' to the upstream acceptor splice site ( - 2 8 u V) is the proposed lariat branch point for splicing at the long exon 8 of K b canconical. The bases at this putative lariat branch point are boxed. The adenosine (-) 25 bp upstream of the downstream acceptor splice site ( - 2 5 D V) may be the lariat branch point for the down- stream site. In H-2K b, each of these putative branch points are contained within the repeated octamer AGTGATGG. A guanosine substitutes for the branch point A in the upstream octamer of the D-region genes. Nucleotides 3' of the acceptor splice site that are part of the polypyrimidine tract are overlined; purine residues in the polypyrimidine tract are circled
84 D.E. Handy et al.: Alternative splicing of class I pre-mRNA
X
Kb I X
\ \
\ \
1.7 Kb
\ \
\ \ \x
Kb-m 1
X
\ \ \
\ \ \ \ \
B 4 5 6 i mm • I B I 2.3 Kb
8 1 I
~ B \
\ \ ~XB l
7 8
subclone into M13
A
in vitro mutagenesis
G
subclone into plasmid
"G
reconstruct the gene
G
B I ! 8
I a"
X B1 2 3 X B Dd I I • IIII IIIIII I I
I I I X X B
"k k
\ \ \ \ a
\ \
Dd-m~
4 5 6 7 8 I l l • l i b
2.65 Kb
subclone imo
M13 E B
.G I I in vitro mutagenesis E B
,A I I subclone into plasmid E B
,~ I , t reconstruct the gene E n
I I CELL S U R F A C E E X P R E S S I O N OF W I L D - T Y P E
A N D M U T A N T GENE P R O D U C T S
E B I I I I E B t 13"
Fig. 2. Strategy for the construction of H-2Kbm and H-2D d. Exons are represented by black bars on the Xba I and Bam HI restriction maps of the wild- type H-2K b and H-2D d genes. Wavy lines indicate vector sequences. For H-2K b the adenosine of the upstream lariat branch point in intron 7 was mutat- ed to a guanosine. For H-2D d the guanosine at this position was mutated to an adenosine. The final reconstruct- ed, mutated gene contains a 600-700 bp deletion in intron 3
300 A) I-3 (H-2K b)
200 no Ab f'~ (xH-2K b
z • -{, C) T4.8.3 (H-2D a) uJ (J 200 )~o Ab /°~H.2Dd
0 0 I00 2ClO
B) JT25.8 (H-2Kb,)
no Ab h ~H-2Dd/ /
O) JT25.3 (H-2Ddm)
~ n o A b h c ( H . 2 D d
" " 1(~0 200
FLUORESCENCE F i g . 3A-D. Cell surface expression of wild-type and mutant gene products. Cells were stained with monoclonal antibodies and fluorescent (FITC) conjugate and analyzed on the fluorescence-activated cell sorter. 20.8.4 is a Kb-specific monoclonal antibody. 34.5.8 is a Dd-specific monoclonal antibody. Shown are staining profiles for (A) I-3 (H-2Kb), (B) JT25.8 (H-2Kb, n), (C) T4.8.3 (H-2Da), (D) JT25.3 (H-2Ddm)
RNA probe was 350 bases in length (Fig. 4b). The pattern in the H-2KUm transfectant, JT25.8, was much different: greater than 70 %, as determined by densitometry, of the protected probe was 312 nucleotides in size. The re- mainder (350 base band) represented protection by a tran- script that contained long exon 8 sequences. RNA from H-2K b or H-2KUm negative cells, T4.8.3 (H-2D a) and JT25.3 (H-2Dd), was not significantly protected from RNase digestions by this probe.
A probe derived from a normal (exon 6, 7, 8) H-2D d transcript was used in the RNase protection experiments to examine the splicing patterns of H-2D d mRNA (Fig. 5a). A normal H-2D a transcript (short exon 8) would protect a 185 base fragment. If a long exon 8 were present in a transcript, then a protected fragment of 174 nucleo- tides would be found. Previously, we have observed an alternative form of H-2D d mRNA that lacks exon 7 (short exon 8 minus exon 7) (McCluskey et al. 1986b). This alternative form of mRNA would result in the protec- tion of a fragment of 132 bases. The 185 base and 132 base fragments were found in the protection experiments
D.E. Handy et al.: Alternative splicing of class I pre-mRNA 85
exon IV V
120 J 120
383
SP6 VI VH VIII 3'UT Promoter
I 33 I 39 I 29 K b probe
350 K b probe + Long exon VIII mRNA + RNase
312 "-'~/9 K b probe + -- Short exon VIII
K b mRNA + RNase I I III
Sho~ exon vm a
Fig. 4a and b. RNase protection of H-2K b mRNA in transfectants, a Schematic representation of the Pst I/Pvu II fragment of H2-K b cDNA that was cloned in pGEM 2, the H-2K b probe, and the expected fragments (size in nucleotides is indicated above the arrows) after digestion by RNases; b RNA from H-2K b (I-3), H-2Kb,, (JT25.8), H-2D d (T4.8.3), and H-2Ddm (JT25.3) was hybridized to the H-2K b probe. After diges- tion by RNases, fragments were separated on a denaturing 5 % polyacrylamide gel. C marks the control lane with probe alone plus RNases. The band at 383 is undigested full-length RNA probe. The doublet at 312 that is faintly visible in all lanes is the result of the duplex between K k mRNA and the K b probe. Several mismatches exist between K k and K b sequences in exon 5, and since the predominant form of H-2K k mRNA con- tains short exon 8 sequences incomplete RNase digestion of the Kb/K k duplexes could yield a 312 base fragment
from both T4.8.3 and JT25.3 (Fig. 5b). No evidence of a long exon 8 containing mRNA was found.
Cell surface expression by transfectants. To evaluate directly the translation products of the mRNA transcripts previously studied, we next examined the H-2 proteins expressed by the transfectants after surface labeling by lactoperoxidase-catalyzed iodination. A three-step proce- dure was used for studying the products of the exogenous genes. (1) To isolate K b and D d molecules distinct from the endogenous H-2K k and H-2D k molecules of the L cells, these H-2 molecules were first isolated by immunoprecipitation with specific monoclonal antibodies and protein A agarose. (2) To reduce heterogeneity due to variable glyeosylation, the solubilized immunoprecipi- tates were treated with N-glycanase (TM) to remove N- linked carbohydrates. (3) To distinguish the long and short forms of the K b and D d molecules, the deglycosylated material was precipitated with antipeptide sera that are capable of distinguishing long exon 8 products, short exon 8 products, and exon 6-8 products that lack exon 7 (Table 1). Figure 6 shows the SDS-PAGE analysis of the 125I- labeled, monoclonal antibody precipitated, N-gly-
canase(TM)-treated, peptide-specific antibody reprecipi- tated molecules. As shown (Fig. 6a), the predominant form of H-2K b expressed by 1-3 cells is precipitated by antibodies specific for products of long exon 8-containing transcripts. In contrast, the H-2Kbm-expressing cells, JT25.8, predominantly express a protein derived from a short exon 8 transcript, although some protein derived from long exon 8 transcripts is detectable. The results of the protein analysis confirmed the conclusions from the RNase protection experiments: the mutation of the adeno- sine at position -2 8 u in H-2K b alters the pattern of K b protein expression.
As expected from the failure to detect H-2D d tran- scripts with long exon 8 sequences (see Fig. 5b), no D d protein corresponding to a long exon 8 transcript was detectable in immunoprecipitation analysis (Fig. 6); however, D d protein that corresponded to both forms (short exon 8 and short exon 8 minus exon 7) of D d mRNA shown in Figure 5b were found in the H-2D d and H-2D d transfectants (Fig. 6b).
Immune function oftransfectants. Since the H-2K b trans- feetants had a higher ratio of K b molecules with short
86 D.E. Handy et al. : Alternative splicing of class I pre-mRNA
( " ~ D d probe + 132 ~ .Z.. Short exon VIII minus exon VII
L I II I D d mRNA + RNase
I I I I I Long exon
VIII
174 9 D d probe+ long exon VIII D d mRNA + RNase
it
Fig. 5a and b. RNase protection of H-2D d mRNA in transfectants, a Schematic representation of the Sal I/Pvu II fragment of H-2D a cDNA that was cloned into pGEM 2, the H-2D a probe, and expected frag- ments (size in nucleotides is indicated over the arrows) after digestion by RNases. b RNA from H-2K b (I-3), H-2Kb~ (JT25.8), H-2D d (T4.8.3), and H-2D~ (JT25.3) was hybridized to the D d probe. After removing single-stranded RNA by treatment with RNases, fragments were separated on a denaturing 5 % polyacrylamide gel. C marks the control lane with probe alone plus RNases
exon 8 sequences than H-2K b transfectants, we tested the effect of the size of the cytoplasmic carboxyl terminus of the K b molecule on functional recognition by allospecific T cells. A T-cell hybridoma that produces IL-2 in response to H-2K b was tested for the ability to be stimulated by the parental or mutant antigens. The hybridoma HTB157.7 (Minami et al. 1986) produced IL-2 in response to both 1-3 (predominantly long exon 8) and JT25.8 (pre- dominantly short exon 8), but not in response to the H-2D d and H-2D~ transfectants which serve as non- specific controls (Table 2). The specificity of these responses to H-2K b was confirmed by antibody blocking.
Monoclonal antibody Y-3 blocked stimulation of HTB 157.7 by I-3 and JT25.8. Thus, the induced mutation in H-2K b had no apparent effect on the ability of the encoded proteins to be recognized by T-cell hybridomas, even though the mutation affects the pattern of protein expression by favoring the production of the shorter K b molecules.
Discussion
Processing of primary H-2 transcripts can result in the production of multiple proteins by the use of alternative splice junctions defining the 5' boundary of exon 8. Two forms of H-2K d, K b, K s, K q, and K k molecules have been found: one with a long and the other with a short exon 8 (see review by Lew et al. 1987). All H-2K- and H-2D-region genes that have had their nucleotide se- quences determined have a region at the 3' end that could encode an exon 8 ten codons long (see Fig. 1); however, the D a and L d proteins have only one amino acid derived from exon 8. This pattern of exon 8 expression results from the preferred use of a downstream (V) acceptor splice site (short exon 8). The upstream (V) acceptor splice site (long exon 8) is used predominantly in the splic- ing of H-2K transcripts except those for H-2K k (Maloy et al. 1987). Previously, we have suggested that adenosine (o) at position - 2 8 U in H-2K b and other K-region genes may be the lariat branch point for the use of the long exon 8 (Lew et al. 1986b). The H-2D-region molecules have a guanosine at this position. To assess the role of the adenosine in the use of the upstream acceptor splice site, we have changed it by in vitro mutagenesis to a guanosine in H-2K b, made the reciprocal change in H-2D d, and studied the expression of these mutant genes after transfer into mouse L cells. Table 3 summarizes these results as well as relevant information from the literature.
The substitution of a guanosine for the adenosine at - 2 8 u in H-2K b alters the pattern of splicing. The predominant species of H-2K b mRNA in transfectants that contain the wild-type gene has long exon 8-derived sequences, whereas the major form of H-2K u mRNA in transfectants with the mutant gene (H-2Kb,~) has se- quences from the short exon 8 (Fig. 4). Although the mu- tation in H-2K b greatly alters the splicing pattern in favor of the downstream acceptor splice junction, it does not completely abolish splicing at the upstream acceptor splice junction.
A number of explanations are possible for the role of the adenosine at - 2 8 u in the removal of intron 7 from class I pre-mRNA transcripts. A common feature of the class I genes is the repeat octamer in intron 7 (AGT- G ATGG) upstream of the used splice sites and the absence of this octamer upstream of the unused upstream splice junctions in D d and L d. One explanation is that the
D.E. Handy et al. : Alternative splicing of class I pre-mRNA
Table 1. C-terminal amino acid sequences*
87
Exon 6 Exon 7 Exon 8
320 t 330 340
I I
H-2K b long exon 8 GGKGGDYALAP GSQTSDLSLPDC~ VMVHDPHSLA]
H-2D d long exon 8" . . . . . . . . . . . . . . S - - M . . . . . . . . . . . S- - - V
H-2D d short exon 8 . . . . . . . . . . . . . . S - - M . . . . . . V
H-2D d short exon 8 ~ . . . . . . . . . . ( ) -~ minus exon 7
* Boxed are C-terminal residues from which antipeptide sera were made ~ All sequences are aligned to H-2K u * The sequences for the predicted long form of H-2D d were derived from the genomic sequences of the H-2D J gene
Fig. 6. a Expression of H-2K b proteins in H-2K b (I-3) and H-2K~n (JT25.8) transfectants. After immunoprecipitation with a Kb-specific monoclonal antibody, EH144, and digestion by N-glycanase (TM) to remove glycosyl units, samples were reprecipitated with antisera to peptides that correspond to the short exon 8 of H-2K b or long exon 8 of H-2K b (see Table 1). Immunoprecipitates were analyzed on SDS- PAGE under reducing conditions, b Expression of H-2D d proteins in H-2D d (T4.8.3) and H-2Dd~ (JT25.3) transfectants. After immuno- precipitation with a Dd-specific monoclonal antibody, 34.5.8, and digestion by N-glycanase (TM), samples were reprecipitated with antisera to peptides that correspond to the C-terminal region of Dd-short exon 8 minus exon 7, Kb-short exon 8, or Kb-long exon 8 (see Table 1). An asterisk marks a nonspecific background spot between lanes
g u a n o s i n e in t h e v a r i a n t o c t a m e r d i s r u p t s t he R N A s t ruc -
t u r e o r t h e i n t e r a c t i o n o f f a c t o r s t ha t n o r m a l l y f a v o r sp l ic -
i ng at t he u p s t r e a m sp l i ce s i te , t hus r e s u l t i n g in t he
p r e d o m i n a n t u s e o f t he d o w n s t r e a m sp l i ce s i te . H o w e v e r ,
t he e x p l a n a t i o n w e f a v o r is tha t t h e a d e n o s i n e ( w i t h i n t he
o c t a m e r A G T G A T G G ) at - 2 8 u is t he p r e f e r r e d la r ia t
b r a n c h p o i n t fo r t h e r e m o v a l o f i n t r o n 7 f r o m H - 2 K b p r e -
m R N A t r a n s c r i p t s . T h u s , n o r m a l K b w h i c h h a s th is
a d e n o s i n e at - 2 8 u p r e f e r e n t i a l l y f o r m s a l a r ia t in t e r -
m e d i a t e at th is s i te a n d t h e r e f o r e u s e s t he u p s t r e a m a c c e p -
t o r sp l i ce s i te o f e x o n 8 a l m o s t e x c l u s i v e l y . H o w e v e r ,
w h e n th is a d e n o s i n e is e l i m i n a t e d , e i t he r b y m u t a t i o n as
w e h a v e p e r f o r m e d e x p e r i m e n t a l l y o r b y g e n e t i c v a r i a t i o n
as d e t e r m i n e d in t he D - r e g i o n g e n e s D d, L d, D b, D ~, a n d
D p ( w h i c h h a v e the v a r i a n t o c t a m e r A T G G T G G , see
T a b l e 3), o t h e r s i t es a r e e m p l o y e d fo r b r a n c h p o i n t f o r m a -
t ion tha t f a v o r t h e u s e o f t he d o w n s t r e a m a c c e p t o r sp l i ce
Table 2. T-cell stimulation by wild-type and mutant transfected L cells*
Stimulator cells HTB157.f
No antibody anti-H-2K b
I-3 (H-2K b) 10 < 0.1
JT25.8 (H-2Kbm) 11 < 1 T4.8.3 (H-2D d) <0.1 <0.1
JT25.3 (H-2D~,) < O. 1 < O. 1
* Stimulation of the H-2Kb-reactive T-cell hybridoma HTD157.7 by L-cell transfectants was performed as described in Materials and methods. IL-2 units/mi are expressed as the reciprocal of the dilution of the supernatant necessary to stimulate the responder cells to 50% of maximum stimulation
+ HTB157.7 is a T-cell hybridoma that produces IL-2 in response to H-2K b (Minami et al. 1986). Fifty microliters of monoclonal culture supernatant containing 5 p,g/ml of Y.3 antibody (anti-H-2K b) was added to the cultures in the blocking experiments
88
Table 3. Sequences affecting splicing of intron 7
D . E . Handy et al. : Alternative splicing of class I pre-mRNA
Polypyrimidine tract
( - 5 - - - 1 4 ) u - 5 6 U - 4 3 u - 2 8 u - 1 8 t j - 1 3 D Exon 8 t
K b + * A C A C C Long > short
K~ + A C G C C Short > long
D d - G T G G T Short
Dam - G T A G T Short
K d + A C A C C Long > short
K k - A C A C C Short > long
Kkm + A C A C C Long*
D p + G T G G T Short
L a + G T G G T Short
D b + G T G G T Short
D k + G T G G T Short
* + indicates presence of intact polypyrimidine tract; - indicates disrupted polypyrimidine tract due to an A at position - 9 u t Long indicates that the detectable protein has a C-terminus encoded by a long exon 8; short indicates that the detectable protein has a C-terminus
encoded by short exon 8 * Archibald and colleagues (1986) produced a K k mutant with an intact polypyrimidine tract. These investigators were unable to detect molecules
with the short C-terminus for either K b or the mutant/~m
site. Although the downstream splice site is preferentially used in H-2Kbm, some transcripts with a long exon 8 are found. These probably result from the use of less efficient lariat branch point adenosines at position -31 u or -36u. The use of cryptic lariat branch points after muta- tion of a canonical lariat branch point has been reported for removal of introns in/3-globin transcripts (Padgett et al. 1985, Ruskin et al. 1985).
Based on the above interpretation of the importance of position -28tj in lariat branch point formation, it was surprising that the presence of the Kb-like adenosine in H-2Dam (at position -28u) did not alter the pattern of D d pre-mRNA splicing. This may be explained by other differences between the H-2Ddm and H-2K b sequences, most notably a change within the polypyrimidine tract which is located 5-14 bp upstream of the long exon 8 splice junction (see Table 3). In H-2K b, eight of ten bases are pyrimidines [T(G)TT(G)TCTTC], whereas H-2D d has an additional purine [T(G)TT(G)(A)CTTC] 9 bp from the upstream splice site. Support for the impor- tance of this purine comes from analysis of H-2K k splic- ing patterns. H-2K k is the only known K region molecule that predominantly uses the short exon 8 (see Lew et al. 1986b), and it is the only K-region gene with an A at posi- tion -9~; (in the polypyrimidine tract; see Fig. 2 and Table 3). Archibald and colleagues (1986) suggested that this adenosine may be the lariat branch point for splicing at the,downstream acceptor splice site. Instead, we sug- gest that the A in the octamer AGTGATGG (filled dot in Fig. 1), at position --25D, is the lariai-branch point for the short exon 8. This is supported by our observation that this octamer is repeated 5' of the upstream acceptor
splice site in H -2K b, and the A (open dot, -28u) in the upstream octamer is crucial for normal processing of H-2K b transcripts. Furthermore, other D-region tran- scripts (such as D b, D k, and L ~) as well as K-region tran- scripts can use the downstream acceptor splice site in the absence of the A at - 9 ~ (see Table 3) (Maloy et al. 1987).
The extra purine (A) (at position -9ty) in H-2D d (as well as in H-2K k) probably influences the use of the up- stream or downstream acceptor splice junction by disrupt- ing the polypyrimidine tract of the upstream splice junc- tion, as the presence of this A creates a dipurine (AG) in the middle of the polypyrimidine sequence. Since recognition of the polypyrimidine tract is important for one of the first steps in splicing, the binding of the U2 snRNP, disruption of the tract may greatly reduce the effi- cient formation of a spliceosome (Frendewey and Keller 1985, Maniatis and Reed 1987, Reed and Maniatis 1985, Ruskin and Green 1985) and may favor binding of the U2 snRNP at the downstream splice site. This may explain why H-2DOm transcripts did not splice at the upstream acceptor site even after a potential lariat branch point was introduced. However, the other differences between H -2K b and H-2D d and other D-region genes in intron or exon sequences may affect the choice of splice junctions, perhaps by creating secondary structures that are unfavorable for splicing. Table 3 lists some positions in intron 7 ( -56u , -43u, -18tj , --13D) where D-region genes differ from K-region genes.
Several systems have been described where alternative splicing results in the production of different protein products (Breitbart et al. 1987, Left et al. 1986, Padgett
D.E. Handy et al. : Alternative splicing of class I pre-mRNA 89
et al. 1986). In some instances, alternative splicing appears to be controlled developmentally or in a tissue- specific manner. Although the role of the C-terminus of the H-2 antigens is not well understood, there is evidence that alternative splicing at the C-terminus may be regulat- ed in a tissue-specific manner (Lew et al. 1986b, McCluskey et al. 1986b). H-2D a exists in two forms: the canonical form which has a peptide backbone of 37 kd, and a form of 35.5 kd that results from the removal of exon 7 by alternative splicing. The ratio of the two forms differs in cells from different tissues and ranges from 90:10 (normal-to-alternative) in lymphocytes to nearly 50:50 in some L-cell transfectants (McCluskey et al. 1986b). Since exon 7 encodes a site of phosphorylation, the distinct protein products may regulate some intracellu- lar signaling pathways. Most of the H-2K b molecules produced from the normal gene have a protein backbone of 38 kd with a C-terminus derived from the long exon 8. A minor population contains a C-terminus derived from the short exon 8 and has a protein backbone of 37 kd. Although both forms of K b are found in spleen and some other cell lines, the short form has not been found in T cells (Lew et al. 1986b). If the different forms of H-2K b have specific functions, then the Kbn transfectant may be ideal for studying functional differences between the long and short form of H-2K b antigen, as the mutation produces a cell line where the short (37 kd) molecule is the predominant form.
Acknowledgments. J. McCluskey is an Arthritis Foundation Investiga- tor. We are grateful for the critical comments of T. Kindt, E. Max, and K. Ozato. We are indebted to W. L. Maloy for supplying antipeptide serum and to R. Valas for assistance with radioiodinations. We appreci- ate the assistance of Brenda Marshall and Rochelle Howard in preparing this manuscript.
References
Archibald, A. L., Thompson, N. A., and Kvist, S. : A single nucleotide difference at the 3' end of an intron causes differential splicing of two histocompatibility genes. EMBO J. 5: 957-965, 1986
Arnold, B., Burgest, H.-G., Archibald, A. L., and Kvist, S. : Complete nucleotide sequence of the murine H-2K k gene. Comparison of three H-2K locus alleles. Nucleic Acids Res. 12: 9473-9487, 1984
Bluestone, J.A., Foo, M., Allen, H., Segal, D., and Flavell, R.A.: Allospecific cytotoxic T lymphocytes recognize conformational determinants on hybrid mouse transplantation antigens. J. Exp. Med. 162: 268-281, 1985
Breitbart, R.E., Andreadis, A., and Nadal-Ginard, B.: Alternative splicing: a ubiquitous mechanism for the generation of multiple pro- tein isoforms from single genes. Annu. Rev. Biochem. 56: 467-495, 1987
Coligan, J. E., Kin&, T. J., Uehara, H., Martinko, J., and Nathenson, S. J.: Primary structure of a murine transplantation antigen. Nature 291: 35-39, 1981
Domdey, H., Apostol, B., Lin, R.J., Newman, A., Brody, E., and Abelson, J.: Lariat structures are in vivo intermediates in yeast pre- mRNA splicing. Cell 39: 611-621, 1984
Frendewey, D. and Keller, W.: Stepwise assembly of a pre-mRNA splicing complex requires U-snRNPs and specific intron sequences. Cell 42: 335-367, 1985
Geier, S. S., Zeff, R. A., McGovern, D. M., Rajan, T. V., and Nathen- son, S. G.: An approach to the study of structure-function relation- ships of MHC class I molecules: isolation and serologic character- ization of H-2K b somatic cell variants. J. Immunol. 137: 1239-1243, 1986
Jones, B. and Janeway, C. A.: Cooperative interaction orB lymphocytes with antigen specific helper T lymphocytes is MHC restricted. Nature 292: 547-549, 1981
Kimball, E. S. and Coligan, J. E.: Structure of class I major histocom- patibility antigens. Contemp. Top. Mol. ImmunoL 9: 1-63, 1983
Klein, J.: Natural History of the Major Histocompatibility Complex, John Wiley and Sons, New York, 1986
Kramer, W., Drutsa, V., Jansen, H.-W., Kramer, B., Pflugfelder, M., and Fritz, H.-J.: The gapped duplex DNA approach to oligonucleo- tide-directed mutation construction. Nucleic Acids Res. 12: 9441-9445, 1984
Kvist, S., Roberts, L., and Dobberstein, B.: Mouse histocompatibility genes: structure and organization of a K d gene. EMBO J. 2: 245-254, 1983
Left, S. E., Rosenfeld, M. G., and Evans, R.M.: Complex transcrip- tional units: diversity in gene expression by alternative RNA processing. Annu. Rev. Biochem. 55: 1091-1117, 1986
Lew, A.M., Lillehoj, E.P., Cowan, E.P., Maloy, W.L., van Schravendijk, M.R., and Coligan, J.E.: Class I genes and molecules: an update. Immunology 86: 3-18, 1986a
Lew, A.M., Margulies, D.H., Maloy, W.L., Lillehoj, E.P., McCluskey, J., and Coligan, J.E.: Alternative protein products with different carboxyl termini from a single class I gene, H-2K b. Proc. Natl. Acad. Sci. U.S.A. 83: 6084-6088, 1986b
Lew, A. M., McCluskey, J., Maloy, W. L., Margulies, D. H., and Coli- gan, J. E.: Multiple class I molecules generated from single genes by alternative splicing of pre-mRNAs. Immunol. Res. 6:117-132, 1987
Maloy, W. L., Lew, A. M., Anderson, R. W., and Coligan, J. E.: H-2K molecules have two different C-termini, one of which is K region specific. Mol. Immunol., in press, 1987
Maniatis, T. and Reed, R.: The role of small nuclear ribonuclear protein particles in pre-mRNA splicing. Nature 325: 673-678, 1987
Margulies, D.H., Evans, G.A., Ozato, K., Canerini-Otero, D., Tanaka, K., Appella, E., and Seidman, J.G.: Expression of H-2D d and H-2L d mouse major histocompatibility antigen genes in L cells after DNA-mediated gene transfer. J. Immunol. 130: 463-470, 1983
McCluskey, J., Boyd, L., Foo, M., Forman, J., Margulies, D. H., and Bluestone, J.: Analysis of hybrid H-2D and L antigens with recipro- caUy mismatched amino-terminal domnains: functional T cell recognition requires preservation of fine structural determinants. J. Immunol. 137: 3881-3890, 1986a
McCluskey, J., Boyd, L.F., Maloy, W.L., Coligan, J. E., and Mar- gulies, D. J.: Alternative processing of H-2D d pre-mRNAs results in membrane expression of differentially phosphorylated protein products. EMBO J. 5: 2477-2483, 1986b
Minami, M., Nariuchi, H., Kawasaki, H., and Dorf, M. E.: The role of class II MHC molecules in the activation of class 1-reactive T cell hybridomas. J. Immunol. 136: 3341-3345, 1986
Moore, K.W., Shu, B.T., Sun, Y.H., Eakle, K.A., and Hood, L.: DNA sequence of a gene encoding a balb/c mouse L d transplan- tation antigen. Science 215: 679-682, 1982
Nathenson, S. G., Uehara, H., Ewenstein, B., Kindt, T. J., and Coligan, J. E.: Primary structural analysis of the transplantation antigens of
90 D.E. Handy et al. : Alternative splicing of class I pre-mRNA
the murine H-2 major histocompatibility complex. Annu. Rev. Biochem. 50: 1025-1052, 1981
Ozato, K. and Sachs, D. S. : Monoclonal antibodies to mouse MHC anti- gens. III. Hybridoma antibodies reacting to antigens of the H-2K b haplotype reveal genetic control of isotype expression. J. Immunol. 126: 317-321, 1981
Padgett, R.A. , Konarsko, M. M., Aebi, M., Hornig, H., Weissman, C., and Sharp, P. A. : Nonconsensus branch-site sequences in the in vitro splicing of transcripts of mutant rabbit/3-globin genes. Proc. Natl. Acad. Sci. U.S.A. 82: 8349-8353, 1985
Padgett, R.A., Grabowski, P.J . , Konarsko, M.M. , Seiler, S., and Sharp, P. A.: Splicing of messenger RNA precursors. Annu. Rev. Biochem. 55: 1119-1150, 1986
Reed, R. and Maniatis, T.: Intron sequences involved in lariat formation during pre-mRNA splicing. Cell 41: 95-105, 1985
Rodriguez, J. R., Pikielny, C. W., and Rosbach, M.: In vivo character- ization of yeast mRNA processing intermediates. Cell 39: 603-610, 1984
Ruskin, B. and Green, M. R. : Specific and stable intron-factor inter- actions are established early during in vitro pre-mRNA splicing. Cell 43: 131-142, 1985
Ruskin, B., Greene, J. M., and Green, M. R.: Cryptic branch point acti- vation allows in vitro splicing of human/3-globin intron mutants. Cell 41: 833-844, 1985
Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H., and Rose, B.: Cloning in single-stranded bacteriophage as an aid to rapid sequencing. J. Mol. Biol. 143: 161-178, 1980
Schepart, B. S., Takahashi, H., Cozad, K. M., Murray, R., Ozato, K., Allen, E., and Frelinger, J. A.: The nucleotide sequence and com- parative analysis of the H-2D p class I H-2 gene. J. Immunol. 136: 3489-3495, 1986
Sher, B. T., Nairn, R., Coligan, J. E., and Hood, L.: DNA sequence of the mouse H-2D d transplantation antigen gene. Proc. Natl. Acad. Sci. U.S.A. 83: 6084-6088, 1985
Steinmetz, M. and Hood, L.: Genes of the major histocompatibility com- plex in mouse and man. Science 222: 727-733, 1983
Watts, S., Vogel, J. M., Harriman, W. D., Itoh, T., Strauss, H. J., and Goodenow, R. S.: DNA sequence analysis of the C3H H-2K k and H-2D k loci. J. Immunol. 139: 3878-3885, 1987
Weiss, E., Golden, L., Zakut, R., Mellor, A., Fahrner, K., Kvist, S., and Flavell, R.A.: The DNA sequence of the H-2K b gene: evi- dence for gene conversion as a mechanism for the generation of polymorphism in histocompatibility antigens. EMBO J. 2: 453-462, 1983
Zoller, M.J. and Smith, M. : Oligonucleotide-directed mutagenesis of DNA fragments cloned into M13 vectors. Methods Enzymol. 100: 468-500, 1983
Received January 5, 1988; revised version received February 24, 1988
