Immunological Characterization and Intracellular Localization of trans ...

Vol. 267, No. 28, Issue of ‘October 5, pp. 20159-20163,1992 Printed in U.S.A.

of BIOLOGICAL CHEMISTRY THE JOURNAL E 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Immunological Characterization and Intracellular Localization of trans-Spliceosomal Small NuclearRibonucleoproteins in Trypanosoma brucei” (Received for publication, June 8, 1992)

Zsofia Palfi and Albrecht Bindereif$ From the Max-Planck-Institut fur Molekulure Genetik, Otto- Warburg-Laboratorium, Ihnestrasse 73, 0-1000 Berlin 33 (Dahlem), Federal Republicof Germany

and U4/U6, the SL RNA is complexed with proteins in the Polyclonal antibodies were raised against purified protein components of the U2 small nuclear ribonucle- SL RNP (Michaeli et al., 1990; Cross et al., 1991) and can be oprotein (snRNP) from Trypanosoma brucei. Through considered a snRNP (Bruzik et al., 1988; Van Doren and immunoblot and immunoprecipitation analyses three Hirsh, 1988), although it contributes its 5”terminal SL seantisera were characterized that reacted specifically quence in trans-splicing. In comparison to the cis-spliceosowith U2 snRNP proteinsof molecular weights 40,000 mal snRNPs, we still know little about the biochemical com(anti-40K) and 16,500 (anti-16.5K),and with each of position, domain structure, andfunctional interactions of the fourproteinsofmolecular weights 14,000, 12,500, trans-spliceosomal snRNPs. We have recently identified sta10,000, and 8,500 (anti-CP).Anti-40K antibodies spe- bly bound protein components of the trypanosomal SL, U2, cifically immunoprecipitated theU2 snRNP fromtry- and U4/U6 snRNPs; a set of five proteins of molecular masses panosomal extracts, whereas anti-CP antibodies rec- 15.0, 14,0, 12.5, 10.0, and 8.5 kDa was detected in each of the ognized several snRNPs, including the SL RNP and three snRNPs. Based on their electrophoretic mobility, these the U2 and U4/U6 snRNPs; in addition, minor RNAs were detected, suggestingthat a family of snRNPs with proteins therefore appear to be common components; in adcommon or related protein componentsexists in tryp- dition, there areseveral snRNP-specific proteins (Palfi etal., anosomes. None of these antibodies cross-reactedsig- 1991). Here we describe the development and characterization of nificantly with total mammalian snRNP proteins, inpolyclonal antibodies specific for several trypanosomal dicating thatthetrypanosomalsnRNPproteinsare immunologically distinct from their mammalian coun-snRNP proteins; they enabled us to conclude that at least four proteins are common to theSL, U2, and U4/U6 snRNPs terparts. Using immunofluorescence microscopy, the snRNP proteinsexhibited a differential cellular distri- of Trypanosoma brucei, and they provide evidence that there bution. Whereas the40-kDa protein is localized exclu- may be additional, minor snRNPs carrying these common sively inthenucleus, with thenucleolusbeing ex- protein components. Interestingly, no immunological relacluded, a fractionof the common proteins also resides tionship was found between trypanosomal and mammalian in the cytoplasm. snRNP proteins. Based on immunofluorescence microscopy, snRNP proteins in trypanosomes appear to be predominantly nuclear-localized, with a fraction of the common proteins Processing of pre-mRNAs in trypanosomes proceeds residing in the cytoplasm. through trans-splicing of two precursor RNAs containing the EXPERIMENTALPROCEDURES 40-nucleotide SL’ miniexon andthe protein-coding exon, Growth of Trypanosomes and Preparation of Cell Extracts-The respectively (for recent reviews, see Laird (1989) and Agabian (1990)). The two precursor RNAs are joined together by a procyclic form of T.brucei brucei strain 427 (Cross, 1975) was cultivated at 28 “C in SDM-79 medium (Brun and Schonenberger, 1979) two-step process formally analogous to thecis-splicing mech- supplemented with 5% (v/v) fetal calf serum and 5 mg/liter hemin. anism of other eucaryotic systems. snRNPsare essential Total cell extracts were prepared as described (Cross et al., 1991) cofactors of both cis-splicing (reviewed by Steitz etal. (1988), from a 32-liter culture at a cell density of 2 x lo7 cells/ml. DEAE Luhrmann et al. (1990), and Green (1991)) and trans-splicing chromatography was carried out essentially as described (Cross et al., 1991), except that the MgC1, concentration of the extract and of all (Tschudi and Ullu, 1990; Hannon et al., 1991). In trypanosomes, U2, U4, and U6 RNAs have been found (Tschudi et buffers used for chromatography was adjusted to 10 mM. 1 ml of the DEAE fraction corresponds to 7 X 10” cell equivalents. al., 1986, 1988; Mottram et al., 1989) and are in the form of Immunization of Rabbits with Affinity-purified U2 snRNP ProRNA-protein complexes, the U2 andthe U4/U6 snRNPs teins-The T. brucei U2 snRNP was purified from 10 ml of DEAE (Michaeli et al., 1990; Gunzl et al., 1992). In addition to U2 fraction according to Palfi et al. (1991), using 100 pg of biotinylated

* This work was supported by Deutsche Forschungsgemeinschaft Grant SFB 344/C5. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom all correspondence should be addressed Max-PlanckInstitut fur Molekulare Genetik, Otto-Warburg-Laboratorium, Ihnestr. 73, D-1000 Berlin 33 (Dahlem), Germany. Tel.: 49-30-8307351; Fax: 49-30-8307-384. The abbreviations used are: SL, spliced leader; sn, small nuclear; RNP, ribonucleoprotein.

2”O-methyl RNA oligonucleotide, U2-5’, bound to streptavidinagarose beads. Urea-eluted proteins (150 pg)were separated on a preparative 15% SDS-polyacrylamide gel (Laemmli, 1970) and transferred onto nitrocellulose. Protein bands were excised from the blot and dissolved overnight at 4 “C in Me,SO (100 pl/cm2 of nitrocellulose). The dissolved protein samples weremixed with complete Freund’s adjuvant, and rabbits were injected subcutaneously at multiple dermal sites on days l , 14, and 28 with 20pgof each of the following proteins: 40-kDa protein (R1641), 16.5-kDa protein (R1642), and a mixture of the 14-, 12.5-, lo-, and 8.5-kDa proteins (R1643). The three rabbits received booster injections of approximately 10-20 pg of the respective snRNP proteins monthly. Blood

20159

20160

Trypanosoma1 forAntibodies Specific

was withdrawn prior to immunization (non-immune control),on day 35, and 1 week after each monthly boost. Immunoblotting Analysis-Rabbit antisera were first depleted of antibodies against nitrocellulose by an overnight incubation at 4 "C with nitrocellulose (10 cm'/ml). Blots of affinity-selected SL, U2, and U4/U6 snRNP proteins (Palfi etal., 1991) and anti-mBGimmunoaffinity-purified total human snRNP proteins (kind gift of Reinhard Luhrmann) were reacted with rabbit sera a t a dilution of1:250 in pre-blocking solution (10 mM Tris-HCI, pH 7.4, 1% bovine serum albumin, 5% fetal calf serum, 100 mM MgCI', 0.1% Triton X-100, 0.5% Tween 20, and 5 mM NaN3). Immunocomplexes were visualized withgoat anti-rabbit IgG-alkaline phosphatase conjugate (Fluka, diluted 1:8,000 in solution A) as described (Lehmeier et al., 1990). Immunoprecipitation Assays-Protein A-Sepharose-bound rabbit IgGs (100 pl of serum/50 p1 of packed protein A-Sepharose beads) were incubated with 100 pl of DEAE fraction for 1.5 hr at4 "C in 500 pl of IPP (10 mM Tris-HCI, pH 7.4, 500 mM NaCI, 0.1% Nonidet P40).Afterwashing the immunoprecipitates extensively with IPP, RNA was released from the beads by phenol-chloroform extraction in NET-2 buffer (50 mM Tris-HCI, pH 7.4, 150 mM NaCl, 0.05% Nonidet P-40) containing0.176 SDS. RNAswere ethanol-precipitated and analyzed by denaturing polyacrylamide gel electrophoresis and silver staining. When immunoprecipitated RNAs were analyzed by primer extension (Cross et al., 1991), 20 pl of DEAE fraction and 20 pl of antiserum were used. Immunofluorescence Microscopy-Procyclic form trypanosome cells (1-2 X IO7 cells) were collected by centrifugation and, after two washes in PBS, 10% fetal calf serum, fixed in PBS, 3% paraformaldehyde for 15 min. After centrifugation, cells were incubated in 0.1 M potassium phosphate, 0.1 M glycine buffer, pH 7.4, for 15 min and permeabilized by the addition of an equal volume of PBS, 0.2% Triton X-100. After 5 min, cells were washed in PBS, 20% fetal calf serum and incubated with rabbit sera (diluted 1:50 in PBS, 10% fetal calf serum) for 1 h a t 37 "C. Cells were then washed three times in PBS, 10% fetal calf serum and reacted with Texas Red-conjugated secondary antibodies (1:lOO-diluted goat anti-rabbit IgGs, Dianova). Finally, cells were extensively washed in PBS, 10% fetal calf serum, 0.1% sodium azide. An aliquot of the suspension was smeared onto a coverslip, allowed to dry, and fixed in 100% ethanol. For conventional microscopy, slides were mounted in PBS/glycerol(l:l) containing0.5 pg/ml Hoechst 33258 dye and visualized at magnification X 1000, using a Zeiss Axioplan fluorescence microscope. For confocal microscopy, samples were processed and examined a t the EMBL (Heidelberg) as described (Carmo-Fonseca et al., 1991).

snRNPs u2

40

SL U416

"

I

~

4

4

4

4

FIG. 1. Immunoblot analysis of T. brucei U2. U4/U6, and SL RNP proteins. U2, SL, and U4/U6 snRNP proteins were affinity-purified from DEAE fraction, separated by SDS-polyacrylamide gel electrophoresis, and probed with rabbit antibodies against the following U2 snRNP proteins: 40 kDa (lane anti-40K, A), the 16.5 kDa (lane anti-16.5K, A ) ,or a mixture of four proteins (14, 12.5, 10.0, and 8.5 kDa; lane anti-CP, A). As controls, the corresponding nonimmune sera were used (lanes N ) . For comparison, Ponceau S-stained total U2 snRNP proteins are also shown (lane U2 proteins). The proteins detected by immunostaining are marked with dots, and their sizes (in kDa) are indicated on the left. M, molecular mass markers (94, 67, 43, 30, 20.1, and 14.4 kDa).

immunoprecipitation assays were performed with a snRNPenriched DEAE fraction (Fig. 2). The relative abundance of the trans-spliceosomal snRNAs in totalRNA from the DEAE fraction and in immunoprecipitates was determined by primer extension, usinga mixture of primer oligonucleotides specific RESULTS for SL, U2, U4, and U6 RNAs; in addition, RNA B, another mnG-capped small nuclear RNA of T.brucei with unknown Six proteins of the T. brucei U2 snRNP were usedfor immunizing three rabbits: the 40-kDa protein, the 16.5-kDa function (Mottram et al., 1989), was assayed. Fig. 2A shows protein, and a mixture of four proteins of molecular masses that the anti-40K antibodies specifically immunoprecipitated U2 snRNA with an efficiency of greater than 50% and, in 14,000, 12,500, 10,000, and 8,500 Da, respectively. Antibody responses were tested by immunoblot analysis and immuno- addition, only very minor amountsof U4 and U6 RNAs (lane precipitation assays. Fig. 1 shows a representative immuno- anti-40K,A ) . In contrast, reactions with the anti-CP antibodies resulted in precipitation of not only U2, but also SL, U4, blot of affinity-purified U2 snRNP, SL RNP, and U4/U6 snRNP proteins of T.brucei. Each of the three rabbit anti- and U6 RNAs, all with efficiencies of around 50% (laneantibodies specifically recognized their respective proteins (lanes CP,A ) . Finally, the anti-16.5K antibodies did not immunoU2, A ) , whereas the corresponding control non-immune rab-precipitate any of these RNAs (laneanti-I6.5K, A ) . RNA B bit sera did not detect any trypanosoma1 snRNP proteins was not present in any of these threeimmunoprecipitates. As (lanes U2, N ) . When crude DEAE fraction was used instead expected, control reactions from DEAE fraction with nonof affinity-purified snRNP proteins, the anti-40K antiserum immune sera (lanes N ) and from total deproteinized RNA was still able to specifically stain the 40-kDa protein (data with antisera (data not shown) were negative, indicating that the antibodies are specific for snRNPs and that RNA comnot shown). Next, we determined whether these antibodies produced against U2 snRNP proteinswere also reactive with ponents are notrecognized. protein components of the othertrans-spliceosomal snRNPs, After determining the specificities of the rabbit antisera the SL RNP and the U4/U6 snRNP. Neither the anti-40K towardsthe knowntrans-spliceosomal snRNPs, we asked nor the anti-16.5K antibodies stained any of the known SL whether these are theonly RNPs recognized. Therefore total and U4/U6 snRNP proteins of T.brucei (data not shown); RNA present in immunoprecipitatesfrom DEAE extractwas however, the anti-CP antibodies recognized four proteins of analyzed directly by denaturing gel electrophoresis and silver the SL RNP and of the U4/U6 snRNP that had the same staining (Fig. 2B). As markers for U2, U4, and U6 RNAs, an molecular mass as those in the U2 snRNP (14-, 12.5-, lo-, anti-m3G immunoprecipitate was run in parallel (lane antiand 8.5-kDa proteins; Fig. 1,lanes S L and U4/U6).Therefore rn:,G). Only some high molecular weight RNAs and none of the three knowntrans-spliceosomal snRNPs (SL, U2, and the trans-spliceosomal snRNAs were precipitated by the nonimmune sera (lanes N ) . The anti-40K antibodiesspecifically U4/U6) share a t least four proteins. T o further characterize thespecificities of these antibodies, precipitated an RNA of approximately 150 nucleotides (lane

Antibodies Specific Trypanosoma1 for

snRNPs

20161

teins were probed with each of the three antisera against trypanosomal snRNP proteins (anti-40K, -16.5K, and -CP) in immunoblot analysis (Fig. 3). In a positive control, Y-12 monoclonal antibodies stained the mammalian Sm proteins B, B', and D (lane Y12),as expected (Lerner et al., 1981). In contrast,none of thethree polyclonal antibodiesagainst trypanosomalsnRNPproteins exhibited any significant B (lanesA ) . In cross-reaction with mammalian snRNP proteins w addition, we used immunoprecipitation to search in HeLa cell 5 nuclear extract for RNPs with cross-reactive protein compo0 nents; however, no specifically immunoprecipitated small RNAs were detected (data not shown). Thus, noimmunological relationship was found between trypanosomal and mamu4 malian snRNP proteins, indicating that trypanosomal snRNA B RNPs are immunologically distinct from their higher eucarU6 yotic counterparts. c Finally, the subcellular localization of snRNP proteins in trypanosomes was investigated by immunofluorescence miI croscopy, usingantibodiesagainstthe U2-specific 40-kDa u2 protein (anti-4OK) and againstfour common snRNP proteins SL (anti-CP) (Fig. 4). Procyclic form T. brucei cells were immunostained with these antibodies, and, as a control, with the U4 respective non-immune sera. In each experiment, the same U6 cells were also stained with a DNA-specificdye, which allows localization of the nucleus and the kinetoplast;significantly, the nucleolar structure can be localized as a region of lower staining intensity in thenucleus (panels A , C, E , and G)(see also Rudenko et dl. (1991)). The controls with non-immune is only weak, diffuse FIG. 2. Immunoprecipitation analysis of T.brucei snRNPs. sera gave nonuclearstaining;there snRNPs were immunoprecipitated from T. brucei DEAEfraction, background staining throughout thecytoplasm (panels B and using anti-4OK, anti-CP, or anti-16.5Kantibodies (lanes A ) , and D).Labelingcellswith anti-40Kantibodies clearly shows RNA was purified from the immunoprecipitates. As controls, reac- bright nuclear staining, excluding nucleoli (panel F ) . Simitions were done with the corresponding non-immune sera (lanes N ) . A , primer extension assays. Following immunoprecipitations from 20 larly, antibodies against the common snRNP proteins gave an intense nucleoplasmic staining; however, significant fluopl of DEAE fraction, the relative concentrations of SL, U2, U4, U6, detected in the cytoplasmic regions and RNA B in the immunoprecipitates (lanes anti-40K, anti-CP, and rescence could also be anti-16.5K) and in 10 pl of DEAE fraction (lane DEAE) were assayed (panel H). I _.I

by primer extension. The positions of the SL-, U2-, U4-, U6-, and RNA B-specific primer-extension products are indicated on the left. M, HpaII-digested pBR322 marker fragments. B , RNA analysis by silver staining. In addition to RNA immunoprecipitated from 100 p1 of DEAE fraction each with anti-40K, anti-CP, anti-16.5K, or antimnG antibodies (lanes anti-40K, anti-CP, anti-16.5K, and anti-msC), RNA from 20 pl of DEAE fraction has been analyzed (lane DEAE). The positions of the SL, U2, U4, and U6 RNA as well as several unidentified RNAs (arrows) are indicated on the right.

94 61 43

-

-

anti-40K, A ) , corresponding to the size of U2 RNA (148 nucleotides) and consistent with the primer extension assay - €5' 30 (Fig. 2 A , lane anti-40K, A ) ; the anti-16.5K antibodies did not TB yield any specific RNAs above background (Fig. 2B, lane anti201 16.5K, A ) ; finally, reactions with anti-CP antibodies resulted in several major species of small RNAs, which on the basisof -0 the primer extension assays (Fig. 2A, lane anti-CP, A ) and 14.4sizes were identified as SL, U2, U4, and U6 RNAs (Fig. 2B, lane anti-CP, A ) . Surprisingly,severaladditional,minor RNAs of approximately 190, 165,85, and80 nucleotides were reproducibly and specifically immunoprecipitated by the antiFIG. 3. Immunoblot analysisof HeLa snRNP proteins. Total CP antiserum. HeLa snRNP proteins purified by anti-mnG immunoaffinity chroIn sum, these results establish the40-kDa protein as a U2 matography were separated by SDS-polyacrylamide gel electrophosnRNP-specific component, and the four U2snRNP polypep- resis and probed with antibodies against the following U2 snRNP tides of molecular masses 14 kDa, 12.5 kDa, 10 kDa, and 8.5 proteins of T. brucei: 40-kDa protein (lane anti-40K, A ) , 16.5-kDa kDa as common proteins of the SL, U2, and U4/U6 snRNP, protein ( l a n e anti-16.5K, A ) , or a mixture of four proteins (14, 12.5, but notof an RNA B-containing complex. Interestingly, there 10.0, and 8.5 kDa; lane anti-CP, A ) . As controls, reactions with the N ) and with an anti-Sm correspondingnon-immunesera(lanes appear to be additional, minor small nuclear RNPs with these monoclonal antibody (lane Y12) are also shown. In addition, Ponceau common protein components. S-stained total HeLa snRNP proteins are displayed (lane T ) .On the We have also searched for potential immunological relaright, the positions of the HeLa snRNP proteins B, B', and D are tionships between trypanosomal and mammalian snRNP promarked. M, molecular mass markers (sizes in kDa indicated on the teins. Anti-m,lG immunoaffinity-purified human snRNP pro- left).

20162

Antibodies Specific for Trypanosoma1 snRNPs DISCUSSION

Previous studieshave indicated that thetrans-spliceosomal snRNPs of trypanosomes differ in several important features from the highly conserved snRNPs of other eucaryotes, both with respect to theirRNA and protein components(Michaeli et al., 1990; Cross et al., 1991) (for a review, see Agabian (1990)). T o obtain specific tools for the studyof the structure and function of trypanosomal snRNPs, we have raised antibodies against several T. brucei U2 snRNPproteinsand determined their specificities; the anti-40K and anti-16.5K antisera recognized two U2 snRNP proteins of molecular weights 40,000 and 16,500, respectively. However, only the FIG. 4. Subcellular localization of snRNP antigens in proanti-40K antibodies function in immunoprecipitation,recogcyclic form T. brucei cells by conventional immunofluoresnizingselectively the U2 snRNP in trypanosomal extract. cence microscopy. Cells were probed with anti-40K (panel F ) and anti-CP (panel H ) antibodies and with the corresponding non-im- This result confirms our initial assignment of the 40-kDa mune sera (NIS; panels R and D,respectiuely), followed by reaction protein as a U2-specific component (Palfi et al., 1991); no with Texas Red-conjugated secondary antibodies. The nucleus and other major R N P appears to share this protein component. kinetoplast were visualized by specifically staining DNA in the same The second antiserum, anti-16.5K, did not react with the U2 cells (panels A, C, E, and G). Magnification: X 1000. snRNP in solution,which suggests that thereactive epitopes of this protein are notaccessible within the U2 snRNP. The third antiserum, anti-CP, contains antibodies against four U2 snRNP proteins of molecular weights 14,000, 12,500, 10,000, and 8,500. Proteins of these sizes have recently been identified not only in affinity-purified U2 but also in U4/U6 and SL RNPs (Palfi al., et 1991). Significantly,the anti-CP antibodies recognize the same set of four proteins inimmunoblots of U2, U4/U6, and SL proteins, and in addition, all three transspliceosomal snRNPs (SL, U2, and U4/U6) could be efficientlyimmunoprecipitatedfromtrypanosomalextract. In sum, we have obtained strongimmunological evidence that at least these four proteins are common to the SL, U2, and U4/ U6 snRNPs, consistentwith our initial proteinidentification. Interestingly, several other small RNAs were detected in antiCPimmunoprecipitates,raisingthe possibility thatthere exists a family of RNPs carrying the same or related protein components; alternatively, these minor RNAs may be associated with SL, U2, or U4/U6 snRNPs. It will be important to characterize these additional small RNAs, because they may reveal more insight into a general RNA binding motif for the common proteins, which has yet tobe identified, and because they may be involved in trans-splicing or other RNA B, however, an m,G-capped processingreactions.RNA snRNA with unknown function (Mottram et al., 1989), was found not tobe complexed with thecommon proteins. FIG. 5. Confocalfluorescencemicroscopy of procyclic T. Using theseantibodies,no immunologicalrelationship brucei cells. snRNP proteins were localized by labeling with anti- could be detected between the trypanosomal and the mam40K (panel R ) and anti-CP (panel D )antibodies, and, as a control, using antibodies with the corresponding non-immune sera ( N I S ;panels A and C ) .The maliansnRNPproteins.Earlierstudies, against mammalian snRNP proteins, also failed to establish bar indicates 5 Mm. such correlations (Michaeli etal., 1990; Palfi et al., 1991). As T o obtain a higher resolution of the nuclear structure, we snRNPs are highly conserved general splicing factors, this is a surprising result. For example, proteins immunologically then applied confocalfluorescence microscopy (Fig. 5). As related to the mammalian Sm proteins are present in yeast seen by conventional immunofluorescence microscopy, nonand plants(Tollervey and Mattaj,1987; Siliciano et al., 1987; immunecontrols gave a weak backgroundstaining in the Palfi etal., 1989), and many elements of the RNA secondary cytoplasm, but no staining of the nucleus (panels A and C). structure have been conserved throughout evolution as well, For both the 40-kDa and the common snRNP proteins, a such as the binding motif for the common Sm proteins (rerather uniform distribution throughout thenucleoplasm was viewed by Guthrie and Patterson(1988)). confocal revealed (panels B and D).In sum, conventional and The immunolocalization data presented demonstrate that fluorescence microscopy revealed a predominantlynuclear in T. brucei cells most of the snRNP antigens are concentrated localization of U2-specific and common snRNP proteins in in the nucleus, as we know from higher eucaryotic systems trypanosome cells, excluding the nucleolus. In contrast to the (reviewed by Zieve and Sauterer (1990)). Based on the asU2-specific protein, a fraction of the common snRNP proteinssumption that most of the snRNP proteins are complexed also appears in the cytoplasm. We conclude that both U2- with RNA, our results are consistent with the nuclear localspecific and common proteins are predominantly nuclearization of trans-splicing. Nucleolar exclusion of components localized, with the nucleolus excluded and a fraction of the of the trans-splicing machinery in trypanosomes is in line common proteins residing in thecytoplasm. with previous observations in higher eucaryotic cells that cis-

Antibodies Specific for Trypanosoma1 snRNPs spliceosomal snRNAs and snRNPproteins are notdetectable in the nucleolus (see, for example, Reuter et al. (1984) and Spector et al. (1991)). Whereasthe UP-specific 40-kDa protein resides predominantly in the nucleus, a fraction of the common proteins was detected in thecytoplasm. This differential distribution might reflect an ordered assembly pathway for the U2 snRNP involving both cytoplasmic and nuclear stages, similar to the cytoplasmic-nuclear assembly pathway proposed for the mammalian U2 snRNP (reviewed by Zieve and Sauterer (1990)). Recently ithas been shown that RNA polymerase Ican mediate transcription of protein-coding genes in the nucleolus, yielding trans-spliced and polyadenylated mRNA (Rudenko et al., 1991; Zomerdijk et al., 1991). Our inability to detect any snRNP antigens in the nucleolus suggests that RNA polymerase I-mediated transcription of protein-coding genes and trans-splicing may not colocalize. The lack of antigenic relatedness clearly distinguishes the trypanosomal snRNPs from other eucaryotic snRNPs and may be based not only on their evolutionary origin, but also on trans-splicing specific features of the trypanosomal snRNPs. Therefore these antibodiesshould become very useful tools for the analysis of the RNA-protein structure of trypanosomal snRNPs, for cloning cDNAs of T. brucei snRNP proteins, and for searching for functional equivalents in other trans-splicing organisms. Acknowledgments-We acknowledgethe excellent technical assistance of Tse-I Lin and Susanne Freund, Karl-Heinz Rak for immunization of rabbits, and Michael Hearne for oligonucleotide synthesis. We thank Reinhard Luhrmann andIain Mattaj for gifts of antibodies and materials, Christine Clayton and Piet Borst for unpublished information, and Arthur Gunzl and Mike Cross for helpful comments on the manuscript. We are grateful to Maria Carmo-Fonseca and Angus Lamond for invaluable help and materials for immunofluorescence.

20163

REFERENCES Agabian, N. (1990)Cell 61,1157-1160 Brun, R., and Schonenberger,M. (1979)Acta Tropica 36,289-292 Bruzik, J. P., Van Doren, K., Hirsh, D., and Steitz, J. A. (1988)Nature 335, 559-562 Carmo-Fonseca, M., Tollervey, D., Pepperkok, C., Barabino, S. M. L., Merdes, A,, Brunner, C., Zamore, P. D., Green, M. R., Hurt, E., and Lamond, A. I. (1991)EMBO J. 10. 195-206 Cross, G.A.~M.(1975)'Parasitology 71, 393-417 Cross, M., Gunzl, A,, Palfi, Z., and Bindereif, A. (1991)Mol. Cell. Biol. 11, 5516-5526 Green, M. R. (1991)Annu. Reu. Cell Biol. 7,559-599 Gunzl, A., Cross, M., and Bindereif, A. (1992)Mol. Cell. Bid. 12, 468-479 Guthrie, C., and Patterson, B. (1988)Annu. Reu. Genet. 22,387-419 Hannon, G. J., Maroney, P. A,, and Nilsen, T. W. (1991)J. Biol. Chem. 266, 22792-22795 Laemmli, U. K. (1970)Nature 227,680-685 Laird, P. W. (1989)Trends Genet. 5 , 204-208 Lehmeier. T., Foulaki. K.. and Luhrmann, R. (1990)Nucleic Acids Res. 18. fi475-6484 ... - - - - -

Lerner, E. A., Lerner, M. R., Janeway, C. A., Jr., and Steitz,J. A. (1981)Proc. Natl. Acad. Sei. U.S. A. 78,2737-2741 Luhrmann, R., Kastner, B., andBach, M. (1990)Biochim. Biophys. Acta1087, 7 ~ -"e ~ 9 0 9 I "

Michaeli, S., Roberts, T. G., Watkins, K. P., and Agabian, N. (1990)J. Biol. Chem. 265, 10582-10588 Mottram, J., Perry, K. L., Lizardi, P. M., Luhrmann, R., Agabian, N., and Nelson, R. G. (1989)Mol. Cell. Biol. 9, 1212-1223 Palfi, Z., Bach, M., Solymosy. F.. and Luhrmann. R. (1989)Nucleic Acids Res. -17. ., 1445-1 - - -458 -Palfi, Z., Gunzl, A,, Cross, M., and Bindereif, A. (1991)Proc. Natl. Acad. Sci. U.S. A . 88,9097-9101 Reuter, R., Appel, B., Bringmann, P., Rinke, J., and Luhrmann, R. (1984)Exp. Cell Res. 154,548-560 Rudenko, G., Chung, H.-M. M., Pham, V. P., and Van der Ploeg, L. H. T. (1991)EMBO J. 10.3387-3397 Silkiano, P. G., Jones,". H., and Guthrie, C. (1987)Science 237,1484-1487 Spector, D. L., Fu, X.-D., and Maniatis, T. (1991)EMBO J. 10,3467-3481 Steitz, J. A,, Black, D. L. Gerke, V., Parker, K. A,, Kramer, A., Frendewey, D., and Keller, W. (1988)in Small Nuclear Ribonucleoprotein Particles: Structure and Functions (Birnstiel, M. L., ed) pp. 115-154,Springer-Verlag, Berlin Tollervey, D., and Mattaj, I. W. (1987)EMBO J. 6,469-476 Tschudi, C., and Ullu, E. (1990)Cell 61, 459-466 Tschudi, C., Richards, F. F., and Ullu, E. (1986)Nucleic Acids Res. 14,8893-I

I

8902 Tschudi, C., Krainer, A. R., and Ullu, E. (1988)Nucleic Acids Res. 16,

11375 Van Doren, K., and Hirsh, D. (1988)Nature 335, 556-559 Zieve, G. W., and Sauterer, R. A. (1990)Crrt. Reu. Bzochem. Mol. Biol. 25, 146 Zomerdijk, J. C.B. M., Kieft, R., Shiels, P. G., and Borst, P. (1991)Nucleic Acids Res. 19,5153-5158