- Email: [email protected]
Close relationship between certain nuclear and mitochondrial introns
J. Mol. Biol. (1983) 16/, 595-605
Close Relationship Between Certain Nuclear and Mitochondrial Introns Implications for the Mechanism of R N A Splicing R. B. WAR[• l, C. Sr
2, T. A. BROWN1 AND R. W . DAViF,S~
1Applied Molecular Biology Group, Department of Biochemistry University of Manchester Institute of Science and Technology P.O. Box 88, Manchester M60 1QD, U.K. 2Department of Biology, University of Essex Wivenhoe Park, Colchester C04 3SQ, Essex, U.K. (Received 15 November 1982, and in revised form 28 February 1983) We present the first indication of a direct relationship between a nuclear and a mitochondrial splicing system. The intron in the precursor of the large, nuclearly coded ribosomal RNA of two species of Tetrah:ymena possesses all the features of a class of fungal mitochondrial introns. Sequences conserved in mitochondrial introns of different funga] species are also found in the same order in these Tet~nh,ymena nuclear introns, and the intron RNA call be folded to form a secondalw structure similar to t h a t proposed for mitochondrial introns by Davies et al. (1982). Tiffs "core" secondary structure brings the ends of the intron together. Ftu'thermore, the first intron in the precursor of the large, nuclearly coded rRNA of Physarum polycephalum also has the characteristic conserved sequences and core RNA secondary structure. The limited sequence d a t a available suggest that the intron in the large rRNA of chloroplasts in Chlam,ydomonas reinh~,rdtii also resembles the mitochondrial introns. Tetrahymena large nuclear rP~NA introns also have an internal sequence that can act as an a d a p t o r by pairing with upstream and downstream exon sequences adjacent to the splice junctions to precisely align the splice junctions. These nuclear introns therefore fit the model of the role of intron RNA in the splicing process that was proposed by Davies et al. (1982), suggesting t h a t the mechanisms of splicing may be very similar in these apparently diverse systems. I t is therefore probable t h a t the RNA secondary structures for which there is good evidence in the case of mitochondrial introns will be found to form the basis of active site structure and precise alignment in splicing and cyclization of the Tetrahymena intron "ribozyme".
R N A s p l i c i n g is a n i m p o r t a n t f o r m o f R N A p r o c e s s i n g t h a t o c c u r s d u r i n g t h e e x p r e s s i o n o f m a n y genes in n u c l e a r g e n o m e s o f h i g h e r a n d l o w e r e u k a r y o t e s ( B r e a t h n a c h & C h a m b o n . 1981) a n d in t h e e x p r e s s i o n o f s o m e m i t o c h o n d r i a l genes in l o w e r e u k a r y o t e s ( B o r s t & Grivell, 1981) a n d p l a n t s ( F o x & L e a v e r , 1981); m o r e 595 0022-2836/83/190595-11 $03.00/0
9 1983 Academic Press Inc. (London) Ltd.
596
1%. B.
WARING ET AL.
recently, introns and thus RNA splicing have been found in chloroplasts (Allet & Rochaix, 1979; Koch et al., 1981). The nuclear and mitochondrial RNA splicing systems are generally thought to be distinct, largely because of the kinds of mutations obtained in Saccharomyces cerevisiae (Jacq et al., 1982) and the differences in the primary structure of nuclear and mitochondrial introns, e.g. mitochondrial introns do not obey the G - T . . . A-G rule of nuclear introns (Breathnach & Chambon, 1981) and some have open reading frames (Lazowska et al., 1980). We show that certain nuclear introns in the protozoal genus Tetrahymena are similar to some fungal mitochondrial introns that have been studied closely; they have key conserved sequences in common, and form closely related RNA secondary structures both within the intron and around the splice junctions. This strongly suggests that these nuclear introns fit a model of mitochondrial intron RNA structure and function developed by Davies et al. (1982). This is the first clear evidence of a relationship between nuclear and mitochondrial splicing systems. The nuclear introns concerned are in the genes for the cytoplasmic, nuclearly coded, large ribosomal RNA of Tetrahymena thermophila and Tetrahymena pigmentosa (Wild & Sommer, 1980; Kan & Gall, 1982). One of the introns in the (nuclearly coded) large rRNA of Physarum polycephalum (Nomiyama et al., 1981a) and the intron in the large rRNA of chloroplasts of Chlamydomonas reinhardtii (Allet & Rochaix, 1979) are also shown to possess characteristic features first defined for mitochondrial introns. We will refer to these introns as TtNrRNA, TpNrRNA, PpNrRNA1 and CrCprRNA, respectively. The Physarum intron is at an identical position in the rRNA to the mitochondrial rRNA intron of S. cerevi~iae (Dujon, 1980) and Aspergillus nidulans (Netzker et al., 1982) but the others are differently placed. Davies et al. (1982) recognized a series of key conserved sequences (Waring et al., 1982) and structural features of mitochondrial introns by comparing sequences of nine introns of two fungal species, A. nidulans and S. cerevisiae. The model of mitochondrial intron structure and fnnction that Davies et al. (1982) based on these comparisons is supported by a body of evidence detailed in that publication, and we will use it as the basis for the discussion in this paper. Four conserved sequences (P, Q, R and S in Fig. l(a)) are always found in the regions of the mitochondrial introns that have been shown to function in cis in the splicing reaction (Jacq et al., 1982; De La Salle et al., 1982) and they are always in the same order. The precise location and pairing of two further regions, named E and E' (Davies et al., 1982) is always conserved. Parts of these sequences that can base-pair are important in the formation of secondary structure within the intron, and all known cis-acting mutations affecting splicing are in these pairing regions and reduce the strength of the pairings (Davies et al., 1982). All these sequences are found in the standard mitochondrial order in the TtNrRNA, TpNrRNA, PpNrRNA1 introns (Table 1), and in the three nuclear introns the pairing capacity of these sequences is clearly conserved (Fig. l(b) and (c)). Burke & RajBhandary (1982) have also noted the presence of the P~ sequence in these introns. When these base-pairings are taken as a starting point for RNA secondary structure formation, a number (usually eight) of other conserved base-pairings
MITOCHONDRIAL AND NUCLEAR INTRONS
597
between similarly or identically placed but non-conserved sequences become evident, and all the mitochondrial introns form identical secondary structures despite wide variations in base sequence. We stress that if any one intron RNA sequence is considered alone, it is possible to find alternative ways of pairing the conserved sequences, but we have only considered pairings to be biologically meaningful if they occur between identical or identically placed sequences of all the introns studied. Furthermore, Michel et al. (1982) have subjected some of these introns to extensive computer analysis and derived very similar favoured secondary structures. A consensus intron sequence and secondary structure derived from nine fungal mitochondrial introns by Davies et al. (1982) is shown in Figure l(a). The nuclear rRNA introns of the Tetrahymena species can take up essentially the same secondary structure, as shown in Figure l(b); it is clear that a remarkable level of detail is conserved. The differences between the Tetrahymena and mitochondrial consensus structures all have precedents in one or other individual mitochondrial intron; the double rather than single P2/L2 stem and loop are found in the mitochondrial rP~NA intron (Yo~; Davies et al., 1982), the L5 loop, which normally contains the maturase-coding region, is very small in the Tetrahymena introns but also in two otherwise standard mitochondrial introns (Davies et al., 1982). The PQ pairing has a single base bulge in the Q sequence but this is compensated for by A" U and U ' G base-pairs being replaced by stronger C" G base-pairs. The P6 stem of the L4 loop is weaker than usual but the rest of the loop could be extensively base-paired (13 base-pairs from 40 bases). This is not shown in Figure l(b). Most of the conserved secondary structural features are also found in PpNrRNA1 (Fig. l(c)). A second intron is present in the nuclear gene coding for the large cytoplasmic rRNA of P. polycephalum (Nomiyama et al., 1981b); this intron also has P~-like and S-like sequences that could pair with one another. Possible P, Q and S sequences can be found in the fragmentary sequence of the chloroplast rRNA intron sequence, so this intron may have the core secondary structure of mitochondrial introns as well. The conserved folding of the mitochondrial introns brings the 5' and 3' splice junctions close together. Davies et al. (1982) proposed that the juxtaposition of the splice sites allowed the efficient formation of base-pairs between a sequence in a particular region of the intron and upstream and downstream exon bases adjacent to the splice sites. The internal intron sequence acts as an adaptor to appose exon sequences and was called an internal guide RNA sequence (IGS) by Davies et al. (1982). The IGS is found in a characteristic position within the intron (Fig. l(a)): in the same position in the large nuclear rRNA introns of the two Tetrahymena species, an equivalent sequence is found that can pair with upstream and downstream exon sequences next to the splice junctions. The structure formed by these pairings (P1 and P10 in Fig. l(a)) has conserved features in mitochondrial introns (Davies et al., 1982), which are also found in the Tetrahymena nuclear intron; notably, the base-pair next to the 5' splice junction in the IGS-upstream exon pairing (P1 ; Fig. l(a) and (b)) is between a G in IGS and a U, which is the last base of the upstream exon. A typical mitochondrial IGS-exon pairing cannot be identified for PpNrRNA1, although candidates for related structures can be suggested, e.g. a structure with good P1 and P10
5/ Exon P1 -xxxxxxxxxxxxgguXXXXX I I I l . If )
L1
xxxxxxxxxx~R~xxx] IGS
IIIII Exon 3t XGxxxxxxxxxxxxxxxxxX~ Pl0 X XXXXXI I I I I ) Optlona] loop XXXXXX XXXXXXXXXX P9 I I I I I I I ) L6 XXXXXXXXXX X S P7 x UAUAGUCC L4 A flirt / / G AUCAG X-X P8 A XAR A C-G P6 -XXXXXXXXXA G Y-GN L5 I l t i i I E~ Y ACU-A ~ -XXXXXXXXXXXXCG CGACUAAXXXXXXXXXX( IIIIIIII XXXXXXXXXX L2 P2 X (Optiona)) X *
ilfltt
ttti,
litqt
L3
* XXXXXXXXGCUXXAUGCUGGXAAXXXXXXXXXE P3 P P4 P5
(a) PI ~ a u -gggaguaacuaugacucucuAAAUUGCAAA
5/
t,tJt.tlI
)
UCCGGACUAUUGAAA~GGGAGGUUUCCAUUIJA ( IIIIIII.II IIIIII. 3~ GCACCUG~UAGCUAG UCGuaagguagccaaaugccucgucaucug U oA ~ PIO AGA~AACCAAAUUUC A U IIIIIII U UGCAUCGGUUUAA iCAGGUUUUGAUU~ ] UAAUAC Guuccu~xgaa L2 P2 Ua g flliltll ) ,III.I ) A AACUAAUAAUAAA uug'uaugoga
GV GGCCGCCGA~ ACCACA g~oaaca~o l.ttliI | lift1A IllI ) CUGGCGGAUG~AGUGGUGC aa~gaug
(a
ua:
GAGGGUgA P9 I t f . l ) L6 ACCUCUCCUU
S
G
UAUAGUCG
A IIIII L4 40 bases G AUCAG / / P8 A Ap A G-C P6 GUAU~CUUCUCAUA , A"C. U.G ~ 59 ( IIIIIII1"1 E U ACU-A A ~ u bases GAUA~AGAAGGGGCUGGCUG C~CCAAUCCUGGGACAGGCAGULs ,,,,, ,,,,, ,,,.., ,,,,. L3 * GGCAAGACCGUCAAAUUGCGG~A~GGGU~AcAECGUUCE p g P5
(b)
31 UUAGAGAAUCGGaacuggcuugu-
G AAAUGGGUGCGGGCU lillll C
CGCCCGC U 131 bases U ACCGGGGUAAGCGGGGCGGUGCCA P9
II1.111. IIIIIIII A CCUGGUCCCGAGuAG#CCGCCACUUC U L6 S C L4 91 bases UGCAGUCC / / G G
IIlll
C-G
GUCAG
G-C
P8 A G RC C-G P6 -CGUUUCGCGCGCGUUCA A A U-A L5 - . . l l l l I 1 1 . - . 1 1 E ' A ACU-A Q 270 bases -CUGGAGCGAGCGUGGGUGUGGA CGACUAUUCGGGGCUAG68 bases I.IIIII IIIII .I-.I L3 GCGACGCACCUGGAUGCUG~.3GUUCAAAGGG E p ~ L P5 u ~ 5 AGCC-199bases-CACACAUAGUUuaggacaccauu-
(c) 5~ PI -ggguucaaaacguAAAUA ,l,.t
)
UGACGUUCGUGUGCGUCACA ( II.IIIII III 3' AAUGUAAGCA U~gugagacaguuL2 P2 A AT PI0 UUU
Lcore
structure? J
FIG. 1. (a) Secondary structure of a generalized mitochondrial int,'on described fully by Davies el al. (1982). The exon sequence is shown in lower case. X denotes any base: Y a pyrimidine base. Conserved bases and conserved base-pairs are shown. P, Q, R and S are conserved base sequences (Table 1). E and E' are short sequences defined by their position relative to P and R, respectively, and by their ability to pair with each other. Other conserved stem and loop structures are marked. The internal guide sequence is boxed and forms the P1 and PI0 pairings with the end of the upstream and downstream exons, respectively. The secondary structure is broken at the point indicated by asterisks in order to show it in 2 dimensions, the bases next to the asterisks are thus contiguous. (b) The secondary structure of the nuclear large ribosomal RNA intron of Tetrahymena pigmentosa, with the variant bases and positions of the structure of Tetrahymena thermophila (shown in italicized lower case). The internal guide sequence is boxed and the layout is the same as in (a). The sequences were from Wild & Sommer (1980) and Kan & Gall (1982). (c) Secondary structure of the first intron of the nuclear large ribosomal RNA of Physarum polycephalum. No guide sequence is shown but the core structure is present with the layout as in (a). There are 222 bases between the upstream splice site and the start of the E sequence. The sequence is fl~)m Nomiyama el al. (1981a). (d) A possible internal guide sequence and precise alignment structure in the intron of the large rRNA of the chloroplast of Chlamydomonas reinlu~rdlii. Only part of the sequence is known (Allet & Rochaix, 1979). Possible P1, P l 0 and P2 pairings are shown. Further sequencing will verify whether there is a core structure. Bars have been used to indicate A' U and G "C base-pairs. Centre dots have been used to indicate G-U base-pairs. Hyphens have been omitted fl'om sequences for clarity.
9+~ ++
+r
.2,
'~+ +-
__
,
E
,~, e~
~
~t~
e~
++ + <__
, <_.
, <--
, <---
~+'~
e-.
+ .B~._-+ ~
o~
++++._m +
MITO('HONDRIAL AND NUCLEAR INTRONS
601
pairings but with the splice sites staggered by five base-pairs. The ends of the CrCprRNA intron have been sequenced (Allet & Rochaix, 1979): a possible IGS sequence is present in the expected position, and can pair with upstream and downstream exon bases to form a "mitochondrial" precise alignment structure (P1 and P1O in Fig. l(d)) with a potential P2/L2 stem loop in the standard position. Therefore, this chloroplast intron as well as the Tetrahymena nuclear introns, could use a similar method of aligning the splice sites to the mitochondrial introns. Thus introns in nuclearly coded and probably chloroplast rRNAs resemble some mitochondrial introns closely, suggesting that what is known about RNA splicing in one system may apply in the other. The nuclear intron sequences provide supporting evidence for the splicing model described by Davies et al. (1982): (1) the conservation of intron structure-forming capacity in primary sequence underlines its biological importance; (2) the R-S pairing of PpNrRNA1 provides another example of base changes in the pairing region of R being compensated for by changes in S to maintain base-pairing ; (3) comparison of the rRNA introns of the two Tetrahymena species provides evidence for the importance of several other pairing regions identified by Davies et al. (1982), in that once again base exchanges in one particular strand are compensated for by a change in the other strand that maintains the pairing, e.g. P2, P5 and the paired stem after P9 in Figure l(b); (4)the splice sites are known precisely in the Tetrahymena nuclear rRNA, and are situated within the precise alignment structure exactly as predicted on the mitochondrial model; (5) the different degrees of conservation of core structure and precise alignment structm'e in the Physarum introns suggests that these are, to an extent, separate domains of RNA structure that are subject to reassortment. A considerable amount of work has been carried out on RNA splicing in the Tetrahymena system in vitro, which has led to the remarkable conclusion that the excision of the intron from the precursor of the large rRNA of Tetrahymena, the ligation of the exons and subsequent cyclization of the free intron RNA can all occur at a significant rate in the complete absence of proteins (Cech et al., 1981 ; Zaug & Cech, 1982: Kruger et al., 1982; Zaug et al., 1983). The RNA requires only the presence of a guanosine cofactor, a divalent cation (Mg2+) and a monovalent cation (NH~) in order to undergo these reactions. Cech el al. (1981)have proposed that the lack of an energy requirement can be explained by a phosphoester transfer mechanism, in which each ligation step is linked to a cleavage step, and that the reaction is driven by the high concentration of guanosine cofactor. Therefore, all the structures essential for this series of precise cleavage and ligation reaction must be formed by the precursor RNA itself, and in the case of the autocyclization reaction, it has been shown clearly that only intron RNA sequences are involved (Kruger et al., 1982). The Tetrahymena intron RNA has been proposed (Kruger et al., 1982) to possess the following properties of enzymes; it lowers the activation energy for the reactions, its activity depends on a precise structure, it has a specific binding site for the guanosine cofactor, and two or more domains of the RNA form an active site or sites for the phosphoester transfer reaction.
602
R.B. WARING ET AL.
We propose that the secondary structure model developed for fungal mitochondrial introns by Davies et al. (1982), when applied as described here to the T e t r a h y m e n a rRNA intron sequence, provides a clear rationalization for the structures that make up the active site of the "ribozyme", and for the structure that presents the splice junctions correctly to the active site. As Kruger et al. (1982) pointed out, the active site of the ribozyme must be a cleft or hole that can exclude water. Even without invoking tertiary folding of the model structure of Figure l(a) (mitochondrial) and (b) (Tetrahymena), it is clear that the conserved pairings central to the "core structure" (R-S, P-Q and E-E') encircle a hole in the molecule, which is a good candidate for the active site. We propose that this forms the basis for the active site, and that the guanosine cofactor is held in place by base-pairing to one or more of the conserved unpaired bases in the conserved sequences. The precise alignment structure (Davies et al., 1982) could be folded into this hydrophobic pocket and held correctly in place for the attack of the guanosine cofactor on the phosphoester bond at the splice junction and subsequent reactions. Thus, the core structure may play an active role in the splicing reaction in addition to its role in folding up the intron. We have also shown that the splice functions of the Tetrahymena intron can be aligned by an internal guide sequence in a manner very similar to that proposed for a set of mitochondrial introns by Davies et al. (1982). This suggests that the splice junctions of these nuclear introns could be aligned by a different mechanism to that operative in the introns in nuclear messenger RNAs, where the small nuclear RNA U1 has been proposed to function as an external guide (Lerner et al., 1980; Rogers & Wall, 1980). Although the 5' and 3' cuts are not made simultaneously according to the reaction scheme of Cech and co-workers (Cech et al., 1981 ; Zaug et al., 1983), it is nevertheless crucial to hold the splice junctions closely together, particularly for the 3' cutting and ligation reaction in which the exposed 3' end of the upstream exon must attack the particular phosphoester bond at the 3' splice junction. The IGS model provides a plausible way of bringing these critical bonds close together and holding them together until the 3' end cleavage and exon ligation reactions have been completed. Successful completion of the first two cleavage and ligation reactions, resulting in release of the linear intron RNA, allows a secondary base-paired structure involving two short helices to form at the 5' end of the intron. Before splicing, this structure is weaker (according to the rules proposed by Tinoco et al. (1971)) than the P1 pairing. This secondary pairing is shown in Figure 2. Kruger et al. (1982) reported that a 15-base fragment is released from the 5' end of the intron RNA in the autocyclization reaction. This places the site of cyclization at the phosphoester bond between the last base-paired U residue in the first stem of the bottom structure of Figure 2 and the next unpaired A residue. The cyclization site is in the same position relative to the second short helix as the 3' splice site is with respect to the P10 pairing in the alignment structure. If this potential structure is actually relevant to the cyclization reaction, it is interesting that the mitochondrial introns of the class studied by Davies et al. (1982) are unable to form this structure at the 5' end of free intron RNA, and there is no evidence that these introns circularize, in contrast to another class of
MITOCHONDRIAL
AND NUCLEAR
INTRONS
603
Splicing G |
5
P ~
a -gggaguaacuaugacucucu~AAUUGCAAA I.I1[.111 ) UCCGGACUAUUGAAAA[GGGAGGUUUCCAUU[UA ( Illllll.ll IIIIIIGCACCUGAUAGCUAG U C
3~
UCGuaagguagccaaaugccuA PIO A
Cyclisation
5GAAAU~UGC rJJlt
(
UCCGGACUAUUGAAA IIIIIII.II GCACCUGAUAGCUAG U C AAUUU
)
\
\3' UCG A A UCAGG]
Core structure J
FI(;. 2. Possible seeondarr structure for cyc]ization of the nuclear L-rRNA intron of Tetrahymena. The splicing structure is as shown in Fig. l(b), but including the G base t h a t initiates splicing by attacking the flint base of the intron (Zaug & Cech, 1982). The cyclization structure is based on evidence provided by Zaug el al. (1983), who show that, after excision, the last base of the intron attacks the A base, 15 bases fi'om the start of the intron (numbered as from the G base attached during excision), l~sulting in circularization and zelease of a 15mer. They note the first of the hairpin structures. The second hairpin within the IGS sequence {boxed) is also possible. The diagram shows the T. piffmentosa sequence. Base changes in T. thermophila are shown in lower case italics. Bars have been used to indicate A" U and G-C base-pairs. Centre dots have been used to indicate G ' U base-pairs. Hyphens have been omitted from sequences for clarity.
mitochondrial introns, for which circular forms of excised introns have been demonstrated (Arnberg et al., 1980; Hensgens et al., 1983). I t is not clear whether all or part of the results of the experiments with the Tetrahymena intron will apply to mitochondrial RNA splicing. Most mitochondrial introns require the function of an mRNA maturase encoded in one or other intron within the mitochondrial system (Lazowska et al., 1980; Jacq et al., 1982), although one of the introns considered by Davies et al. (1982), the mitochondrial rRNA intron of S. cerevisiae, can be excised without mitochondrial protein synthesis (Tabak et al., 1981). Nuclear mutations have also been shown to affect specific splicing reactions in yeast mitochondria" (Dieckmann et al., 1982). The Tetrahymena splicing reaction occurs 60 times faster in vivo than in vitro (Kruger et al., 1982), perhaps due to the presence of accessory proteins. The
604
R. B. WARING ET AL.
possibility therefore remains t h a t m R N A maturases and nuclearly coded protein factors play some role (e.g. stabilization of secondary structure) t h a t enables the mitochondrial splicing reaction to proceed at an adequate rate, rather than actually providing excision/ligation activity. We are now engaged in testing whether mitochondrial introns will self-splice in vitro. A preliminary investigation of some introns in nuclear m R N A precursors (human E-globin (Baralle et al., 1980) and ovalbumin (Heilig et al., 1982)) has not so far provided a clear example of an intron with the features defined as mitochondrial. Thus, to date all the mitochondrial-type introns outside mitochondrial occur in rRNA genes. I t is interesting t h a t the genes of all the RNAs with introns fitting the model of Davies et al. (1982) are located extrachromosomally at high copy n u m b e r in specialized organelles. The close resemblance of the Tetrahymena nuclear rRNA introns and the fungal mitochondrial introns is likely to reflect mobility of the introns themselves or of parts of the genome containing them between the nucleus and organelle at some time in the history of these organisms rather than convergent evolution. We thank Dr T. R. Cech for communicating results before publication. We acknowledge support from the Medical Research Council by project grant G7904290CB (to R.W.D. and C.S.).
REFERENCES Allet, B. & Rochaix, J. (1979). Cell, 18, 55-60. Arnberg, A. C., Van Ommen, G. J. B., Grivell, L. A., Van Bruggen, E. F. J. & Borst, P. (1980). Cell, 19, 313-319. Baralle, F. E., Shoulders, C. C. & Proudfoot, N. J. (1980). Cell, 21,621-626. Borst, P. & Grivell, L. A. (1981). Nature (London), 285, 439-440. Breathnach, R. & Chambon, P. (1981). Annu. Rev. Biochem. 50, 349-383. Burke, J. M. & RajBhandary, U. L. (1982). Cell, 31,509-520. Ceeh, R. T., Zaug, A. J. & Grabowski, P. J. (1981). Cell, 27, 487-496. Davies, R. W., Waring, R. B., Ray, J. A., Brown, T. A. & Scazzocchio, C. (1982}. Nature (London), 30{), 719-724. De La Salle, H., Jacq, C. & Slonimski, P. (1982). Cell, 28, 721-732. Dieckmann, C. L., Pape, L. K. & Tzagoloff, A. (1982). Proc. Nat. Acad. Sci., U.S.A. 79, 1805-1809. Dujon, B. (1980). Cell, 20, 185-197. Fox, T. D. & Leaver, C. J. (1981). Cell, 26, 315-323. Heilig, R., Muraskowsky, R., Kloepfer, C. & Mandel, J. L. (1982). Nucl. Acids Res. IO, 4363-4382. Hensgens, L. A. M., Arnberg, A. C., Roosendaal, E., Van der Horst, G., Van der Veen, R., Van Ommen, G. J. B. & Grivell, L. A. (1983). J. Mol. Biol. 164, 35-58. Jaeq, C., Pajot, P., Lazowska, J., Dujardin, G., Claisse, M., Groudinsky, 0., De La Salle, H., Grandchamp, C., Labouesse, M., Gargouri, A., Guiard, B., Spyridakis, A., Dreyfus, M. & Slonimski, P. (1982). In Mitochondrial Genes (Slonimski, P., Borst, P. & Attardi, G., eds), pp. 155-183, Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Kan, N. C. & Gall, J. G. (1982). Nucl. Acids Res. 10, 2809-2822. Koch, W., Edwards, K. & KSssel, H. (1981). Cell, 25, 203-213. Kruger, K., Grabowski, P. J., Zaug, A. J., Sands, J., Gottsehling, D. E. & Cech, T. R. (1982). Cell, 31, 147-157. Lazowska, J., Jaeq, C. & Slonimski, P. (1980). Cell, 22, 333-348.
MITOCHONDRIAL AND NUCLEAR INTRONS
605
Lerner, M. R., Boyle, J. A., Mount, S. M., Wolin, S. L. & Steitz, J. A. (1980). Nature (London), 283, 220-224. Michel, F., Jacquier, A. & Dujon, B. (1982). Biochimie, 64, 867-881. Netzker, R., KLchel, H. G., Basak, N. & Kfintze[, H. (1982). Nucl. Acids Res. 10, 47834794. Nomiyama, H , Sakaki, Y. & Takagi, Y. (1981a). Proc. Nat. Acad. Sci., U.S.A. 78, 13781380. Nomiyama, H., Kuhara, S., Kutika, T., Otsuka, T. & Sakaki, Y. (1981b). Nucl. Acids Res. 9, 5507-5520. Rogers, J. & Wall, R. (1980). Proc. Nat. Acad. Sci., U.S.A. 77, 1877-1879. Tabak, H. F. J., van der Laan, J., Osinga, K. A., Schouten, J. P., van Boom, J. H. & Veeneman, G. H. (1981). Nucl. Acids Res. 9, 4475-4483. Tinoco, I.. Uhlenbeck, O. C. & Levine, M. D. (1971). Nature (London), 280, 362-867. Waring, R. B., Davies, R. W., Scazzocchio, C. & Brown, T. A. (1982). Proc. Nat. Acad. Sci., U.S.A. 79, 6332-6336. Wild, M. A. & Sommer, R. (1980). Natttre (London), 283,693-694. Zaug, A. J. & Cech, T. R. (1982). Nucl. Acids Res. 10, 2823-2838. Zaug, A. J., Grabowski. P. J. & Cech, T. R. (1983). Nature (London), 801,578-583..
Edited by S. Brenner
Close Relationship Between Certain Nuclear and Mitochondrial Introns Implications for the Mechanism of R N A Splicing R. B. WAR[• l, C. Sr
2, T. A. BROWN1 AND R. W . DAViF,S~
1Applied Molecular Biology Group, Department of Biochemistry University of Manchester Institute of Science and Technology P.O. Box 88, Manchester M60 1QD, U.K. 2Department of Biology, University of Essex Wivenhoe Park, Colchester C04 3SQ, Essex, U.K. (Received 15 November 1982, and in revised form 28 February 1983) We present the first indication of a direct relationship between a nuclear and a mitochondrial splicing system. The intron in the precursor of the large, nuclearly coded ribosomal RNA of two species of Tetrah:ymena possesses all the features of a class of fungal mitochondrial introns. Sequences conserved in mitochondrial introns of different funga] species are also found in the same order in these Tet~nh,ymena nuclear introns, and the intron RNA call be folded to form a secondalw structure similar to t h a t proposed for mitochondrial introns by Davies et al. (1982). Tiffs "core" secondary structure brings the ends of the intron together. Ftu'thermore, the first intron in the precursor of the large, nuclearly coded rRNA of Physarum polycephalum also has the characteristic conserved sequences and core RNA secondary structure. The limited sequence d a t a available suggest that the intron in the large rRNA of chloroplasts in Chlam,ydomonas reinh~,rdtii also resembles the mitochondrial introns. Tetrahymena large nuclear rP~NA introns also have an internal sequence that can act as an a d a p t o r by pairing with upstream and downstream exon sequences adjacent to the splice junctions to precisely align the splice junctions. These nuclear introns therefore fit the model of the role of intron RNA in the splicing process that was proposed by Davies et al. (1982), suggesting t h a t the mechanisms of splicing may be very similar in these apparently diverse systems. I t is therefore probable t h a t the RNA secondary structures for which there is good evidence in the case of mitochondrial introns will be found to form the basis of active site structure and precise alignment in splicing and cyclization of the Tetrahymena intron "ribozyme".
R N A s p l i c i n g is a n i m p o r t a n t f o r m o f R N A p r o c e s s i n g t h a t o c c u r s d u r i n g t h e e x p r e s s i o n o f m a n y genes in n u c l e a r g e n o m e s o f h i g h e r a n d l o w e r e u k a r y o t e s ( B r e a t h n a c h & C h a m b o n . 1981) a n d in t h e e x p r e s s i o n o f s o m e m i t o c h o n d r i a l genes in l o w e r e u k a r y o t e s ( B o r s t & Grivell, 1981) a n d p l a n t s ( F o x & L e a v e r , 1981); m o r e 595 0022-2836/83/190595-11 $03.00/0
9 1983 Academic Press Inc. (London) Ltd.
596
1%. B.
WARING ET AL.
recently, introns and thus RNA splicing have been found in chloroplasts (Allet & Rochaix, 1979; Koch et al., 1981). The nuclear and mitochondrial RNA splicing systems are generally thought to be distinct, largely because of the kinds of mutations obtained in Saccharomyces cerevisiae (Jacq et al., 1982) and the differences in the primary structure of nuclear and mitochondrial introns, e.g. mitochondrial introns do not obey the G - T . . . A-G rule of nuclear introns (Breathnach & Chambon, 1981) and some have open reading frames (Lazowska et al., 1980). We show that certain nuclear introns in the protozoal genus Tetrahymena are similar to some fungal mitochondrial introns that have been studied closely; they have key conserved sequences in common, and form closely related RNA secondary structures both within the intron and around the splice junctions. This strongly suggests that these nuclear introns fit a model of mitochondrial intron RNA structure and function developed by Davies et al. (1982). This is the first clear evidence of a relationship between nuclear and mitochondrial splicing systems. The nuclear introns concerned are in the genes for the cytoplasmic, nuclearly coded, large ribosomal RNA of Tetrahymena thermophila and Tetrahymena pigmentosa (Wild & Sommer, 1980; Kan & Gall, 1982). One of the introns in the (nuclearly coded) large rRNA of Physarum polycephalum (Nomiyama et al., 1981a) and the intron in the large rRNA of chloroplasts of Chlamydomonas reinhardtii (Allet & Rochaix, 1979) are also shown to possess characteristic features first defined for mitochondrial introns. We will refer to these introns as TtNrRNA, TpNrRNA, PpNrRNA1 and CrCprRNA, respectively. The Physarum intron is at an identical position in the rRNA to the mitochondrial rRNA intron of S. cerevi~iae (Dujon, 1980) and Aspergillus nidulans (Netzker et al., 1982) but the others are differently placed. Davies et al. (1982) recognized a series of key conserved sequences (Waring et al., 1982) and structural features of mitochondrial introns by comparing sequences of nine introns of two fungal species, A. nidulans and S. cerevisiae. The model of mitochondrial intron structure and fnnction that Davies et al. (1982) based on these comparisons is supported by a body of evidence detailed in that publication, and we will use it as the basis for the discussion in this paper. Four conserved sequences (P, Q, R and S in Fig. l(a)) are always found in the regions of the mitochondrial introns that have been shown to function in cis in the splicing reaction (Jacq et al., 1982; De La Salle et al., 1982) and they are always in the same order. The precise location and pairing of two further regions, named E and E' (Davies et al., 1982) is always conserved. Parts of these sequences that can base-pair are important in the formation of secondary structure within the intron, and all known cis-acting mutations affecting splicing are in these pairing regions and reduce the strength of the pairings (Davies et al., 1982). All these sequences are found in the standard mitochondrial order in the TtNrRNA, TpNrRNA, PpNrRNA1 introns (Table 1), and in the three nuclear introns the pairing capacity of these sequences is clearly conserved (Fig. l(b) and (c)). Burke & RajBhandary (1982) have also noted the presence of the P~ sequence in these introns. When these base-pairings are taken as a starting point for RNA secondary structure formation, a number (usually eight) of other conserved base-pairings
MITOCHONDRIAL AND NUCLEAR INTRONS
597
between similarly or identically placed but non-conserved sequences become evident, and all the mitochondrial introns form identical secondary structures despite wide variations in base sequence. We stress that if any one intron RNA sequence is considered alone, it is possible to find alternative ways of pairing the conserved sequences, but we have only considered pairings to be biologically meaningful if they occur between identical or identically placed sequences of all the introns studied. Furthermore, Michel et al. (1982) have subjected some of these introns to extensive computer analysis and derived very similar favoured secondary structures. A consensus intron sequence and secondary structure derived from nine fungal mitochondrial introns by Davies et al. (1982) is shown in Figure l(a). The nuclear rRNA introns of the Tetrahymena species can take up essentially the same secondary structure, as shown in Figure l(b); it is clear that a remarkable level of detail is conserved. The differences between the Tetrahymena and mitochondrial consensus structures all have precedents in one or other individual mitochondrial intron; the double rather than single P2/L2 stem and loop are found in the mitochondrial rP~NA intron (Yo~; Davies et al., 1982), the L5 loop, which normally contains the maturase-coding region, is very small in the Tetrahymena introns but also in two otherwise standard mitochondrial introns (Davies et al., 1982). The PQ pairing has a single base bulge in the Q sequence but this is compensated for by A" U and U ' G base-pairs being replaced by stronger C" G base-pairs. The P6 stem of the L4 loop is weaker than usual but the rest of the loop could be extensively base-paired (13 base-pairs from 40 bases). This is not shown in Figure l(b). Most of the conserved secondary structural features are also found in PpNrRNA1 (Fig. l(c)). A second intron is present in the nuclear gene coding for the large cytoplasmic rRNA of P. polycephalum (Nomiyama et al., 1981b); this intron also has P~-like and S-like sequences that could pair with one another. Possible P, Q and S sequences can be found in the fragmentary sequence of the chloroplast rRNA intron sequence, so this intron may have the core secondary structure of mitochondrial introns as well. The conserved folding of the mitochondrial introns brings the 5' and 3' splice junctions close together. Davies et al. (1982) proposed that the juxtaposition of the splice sites allowed the efficient formation of base-pairs between a sequence in a particular region of the intron and upstream and downstream exon bases adjacent to the splice sites. The internal intron sequence acts as an adaptor to appose exon sequences and was called an internal guide RNA sequence (IGS) by Davies et al. (1982). The IGS is found in a characteristic position within the intron (Fig. l(a)): in the same position in the large nuclear rRNA introns of the two Tetrahymena species, an equivalent sequence is found that can pair with upstream and downstream exon sequences next to the splice junctions. The structure formed by these pairings (P1 and P10 in Fig. l(a)) has conserved features in mitochondrial introns (Davies et al., 1982), which are also found in the Tetrahymena nuclear intron; notably, the base-pair next to the 5' splice junction in the IGS-upstream exon pairing (P1 ; Fig. l(a) and (b)) is between a G in IGS and a U, which is the last base of the upstream exon. A typical mitochondrial IGS-exon pairing cannot be identified for PpNrRNA1, although candidates for related structures can be suggested, e.g. a structure with good P1 and P10
5/ Exon P1 -xxxxxxxxxxxxgguXXXXX I I I l . If )
L1
xxxxxxxxxx~R~xxx] IGS
IIIII Exon 3t XGxxxxxxxxxxxxxxxxxX~ Pl0 X XXXXXI I I I I ) Optlona] loop XXXXXX XXXXXXXXXX P9 I I I I I I I ) L6 XXXXXXXXXX X S P7 x UAUAGUCC L4 A flirt / / G AUCAG X-X P8 A XAR A C-G P6 -XXXXXXXXXA G Y-GN L5 I l t i i I E~ Y ACU-A ~ -XXXXXXXXXXXXCG CGACUAAXXXXXXXXXX( IIIIIIII XXXXXXXXXX L2 P2 X (Optiona)) X *
ilfltt
ttti,
litqt
L3
* XXXXXXXXGCUXXAUGCUGGXAAXXXXXXXXXE P3 P P4 P5
(a) PI ~ a u -gggaguaacuaugacucucuAAAUUGCAAA
5/
t,tJt.tlI
)
UCCGGACUAUUGAAA~GGGAGGUUUCCAUUIJA ( IIIIIII.II IIIIII. 3~ GCACCUG~UAGCUAG UCGuaagguagccaaaugccucgucaucug U oA ~ PIO AGA~AACCAAAUUUC A U IIIIIII U UGCAUCGGUUUAA iCAGGUUUUGAUU~ ] UAAUAC Guuccu~xgaa L2 P2 Ua g flliltll ) ,III.I ) A AACUAAUAAUAAA uug'uaugoga
GV GGCCGCCGA~ ACCACA g~oaaca~o l.ttliI | lift1A IllI ) CUGGCGGAUG~AGUGGUGC aa~gaug
(a
ua:
GAGGGUgA P9 I t f . l ) L6 ACCUCUCCUU
S
G
UAUAGUCG
A IIIII L4 40 bases G AUCAG / / P8 A Ap A G-C P6 GUAU~CUUCUCAUA , A"C. U.G ~ 59 ( IIIIIII1"1 E U ACU-A A ~ u bases GAUA~AGAAGGGGCUGGCUG C~CCAAUCCUGGGACAGGCAGULs ,,,,, ,,,,, ,,,.., ,,,,. L3 * GGCAAGACCGUCAAAUUGCGG~A~GGGU~AcAECGUUCE p g P5
(b)
31 UUAGAGAAUCGGaacuggcuugu-
G AAAUGGGUGCGGGCU lillll C
CGCCCGC U 131 bases U ACCGGGGUAAGCGGGGCGGUGCCA P9
II1.111. IIIIIIII A CCUGGUCCCGAGuAG#CCGCCACUUC U L6 S C L4 91 bases UGCAGUCC / / G G
IIlll
C-G
GUCAG
G-C
P8 A G RC C-G P6 -CGUUUCGCGCGCGUUCA A A U-A L5 - . . l l l l I 1 1 . - . 1 1 E ' A ACU-A Q 270 bases -CUGGAGCGAGCGUGGGUGUGGA CGACUAUUCGGGGCUAG68 bases I.IIIII IIIII .I-.I L3 GCGACGCACCUGGAUGCUG~.3GUUCAAAGGG E p ~ L P5 u ~ 5 AGCC-199bases-CACACAUAGUUuaggacaccauu-
(c) 5~ PI -ggguucaaaacguAAAUA ,l,.t
)
UGACGUUCGUGUGCGUCACA ( II.IIIII III 3' AAUGUAAGCA U~gugagacaguuL2 P2 A AT PI0 UUU
Lcore
structure? J
FIG. 1. (a) Secondary structure of a generalized mitochondrial int,'on described fully by Davies el al. (1982). The exon sequence is shown in lower case. X denotes any base: Y a pyrimidine base. Conserved bases and conserved base-pairs are shown. P, Q, R and S are conserved base sequences (Table 1). E and E' are short sequences defined by their position relative to P and R, respectively, and by their ability to pair with each other. Other conserved stem and loop structures are marked. The internal guide sequence is boxed and forms the P1 and PI0 pairings with the end of the upstream and downstream exons, respectively. The secondary structure is broken at the point indicated by asterisks in order to show it in 2 dimensions, the bases next to the asterisks are thus contiguous. (b) The secondary structure of the nuclear large ribosomal RNA intron of Tetrahymena pigmentosa, with the variant bases and positions of the structure of Tetrahymena thermophila (shown in italicized lower case). The internal guide sequence is boxed and the layout is the same as in (a). The sequences were from Wild & Sommer (1980) and Kan & Gall (1982). (c) Secondary structure of the first intron of the nuclear large ribosomal RNA of Physarum polycephalum. No guide sequence is shown but the core structure is present with the layout as in (a). There are 222 bases between the upstream splice site and the start of the E sequence. The sequence is fl~)m Nomiyama el al. (1981a). (d) A possible internal guide sequence and precise alignment structure in the intron of the large rRNA of the chloroplast of Chlamydomonas reinlu~rdlii. Only part of the sequence is known (Allet & Rochaix, 1979). Possible P1, P l 0 and P2 pairings are shown. Further sequencing will verify whether there is a core structure. Bars have been used to indicate A' U and G "C base-pairs. Centre dots have been used to indicate G-U base-pairs. Hyphens have been omitted fl'om sequences for clarity.
9+~ ++
+r
.2,
'~+ +-
__
,
E
,~, e~
~
~t~
e~
++ + <__
, <_.
, <--
, <---
~+'~
e-.
+ .B~._-+ ~
o~
++++._m +
MITO('HONDRIAL AND NUCLEAR INTRONS
601
pairings but with the splice sites staggered by five base-pairs. The ends of the CrCprRNA intron have been sequenced (Allet & Rochaix, 1979): a possible IGS sequence is present in the expected position, and can pair with upstream and downstream exon bases to form a "mitochondrial" precise alignment structure (P1 and P1O in Fig. l(d)) with a potential P2/L2 stem loop in the standard position. Therefore, this chloroplast intron as well as the Tetrahymena nuclear introns, could use a similar method of aligning the splice sites to the mitochondrial introns. Thus introns in nuclearly coded and probably chloroplast rRNAs resemble some mitochondrial introns closely, suggesting that what is known about RNA splicing in one system may apply in the other. The nuclear intron sequences provide supporting evidence for the splicing model described by Davies et al. (1982): (1) the conservation of intron structure-forming capacity in primary sequence underlines its biological importance; (2) the R-S pairing of PpNrRNA1 provides another example of base changes in the pairing region of R being compensated for by changes in S to maintain base-pairing ; (3) comparison of the rRNA introns of the two Tetrahymena species provides evidence for the importance of several other pairing regions identified by Davies et al. (1982), in that once again base exchanges in one particular strand are compensated for by a change in the other strand that maintains the pairing, e.g. P2, P5 and the paired stem after P9 in Figure l(b); (4)the splice sites are known precisely in the Tetrahymena nuclear rRNA, and are situated within the precise alignment structure exactly as predicted on the mitochondrial model; (5) the different degrees of conservation of core structure and precise alignment structm'e in the Physarum introns suggests that these are, to an extent, separate domains of RNA structure that are subject to reassortment. A considerable amount of work has been carried out on RNA splicing in the Tetrahymena system in vitro, which has led to the remarkable conclusion that the excision of the intron from the precursor of the large rRNA of Tetrahymena, the ligation of the exons and subsequent cyclization of the free intron RNA can all occur at a significant rate in the complete absence of proteins (Cech et al., 1981 ; Zaug & Cech, 1982: Kruger et al., 1982; Zaug et al., 1983). The RNA requires only the presence of a guanosine cofactor, a divalent cation (Mg2+) and a monovalent cation (NH~) in order to undergo these reactions. Cech el al. (1981)have proposed that the lack of an energy requirement can be explained by a phosphoester transfer mechanism, in which each ligation step is linked to a cleavage step, and that the reaction is driven by the high concentration of guanosine cofactor. Therefore, all the structures essential for this series of precise cleavage and ligation reaction must be formed by the precursor RNA itself, and in the case of the autocyclization reaction, it has been shown clearly that only intron RNA sequences are involved (Kruger et al., 1982). The Tetrahymena intron RNA has been proposed (Kruger et al., 1982) to possess the following properties of enzymes; it lowers the activation energy for the reactions, its activity depends on a precise structure, it has a specific binding site for the guanosine cofactor, and two or more domains of the RNA form an active site or sites for the phosphoester transfer reaction.
602
R.B. WARING ET AL.
We propose that the secondary structure model developed for fungal mitochondrial introns by Davies et al. (1982), when applied as described here to the T e t r a h y m e n a rRNA intron sequence, provides a clear rationalization for the structures that make up the active site of the "ribozyme", and for the structure that presents the splice junctions correctly to the active site. As Kruger et al. (1982) pointed out, the active site of the ribozyme must be a cleft or hole that can exclude water. Even without invoking tertiary folding of the model structure of Figure l(a) (mitochondrial) and (b) (Tetrahymena), it is clear that the conserved pairings central to the "core structure" (R-S, P-Q and E-E') encircle a hole in the molecule, which is a good candidate for the active site. We propose that this forms the basis for the active site, and that the guanosine cofactor is held in place by base-pairing to one or more of the conserved unpaired bases in the conserved sequences. The precise alignment structure (Davies et al., 1982) could be folded into this hydrophobic pocket and held correctly in place for the attack of the guanosine cofactor on the phosphoester bond at the splice junction and subsequent reactions. Thus, the core structure may play an active role in the splicing reaction in addition to its role in folding up the intron. We have also shown that the splice functions of the Tetrahymena intron can be aligned by an internal guide sequence in a manner very similar to that proposed for a set of mitochondrial introns by Davies et al. (1982). This suggests that the splice junctions of these nuclear introns could be aligned by a different mechanism to that operative in the introns in nuclear messenger RNAs, where the small nuclear RNA U1 has been proposed to function as an external guide (Lerner et al., 1980; Rogers & Wall, 1980). Although the 5' and 3' cuts are not made simultaneously according to the reaction scheme of Cech and co-workers (Cech et al., 1981 ; Zaug et al., 1983), it is nevertheless crucial to hold the splice junctions closely together, particularly for the 3' cutting and ligation reaction in which the exposed 3' end of the upstream exon must attack the particular phosphoester bond at the 3' splice junction. The IGS model provides a plausible way of bringing these critical bonds close together and holding them together until the 3' end cleavage and exon ligation reactions have been completed. Successful completion of the first two cleavage and ligation reactions, resulting in release of the linear intron RNA, allows a secondary base-paired structure involving two short helices to form at the 5' end of the intron. Before splicing, this structure is weaker (according to the rules proposed by Tinoco et al. (1971)) than the P1 pairing. This secondary pairing is shown in Figure 2. Kruger et al. (1982) reported that a 15-base fragment is released from the 5' end of the intron RNA in the autocyclization reaction. This places the site of cyclization at the phosphoester bond between the last base-paired U residue in the first stem of the bottom structure of Figure 2 and the next unpaired A residue. The cyclization site is in the same position relative to the second short helix as the 3' splice site is with respect to the P10 pairing in the alignment structure. If this potential structure is actually relevant to the cyclization reaction, it is interesting that the mitochondrial introns of the class studied by Davies et al. (1982) are unable to form this structure at the 5' end of free intron RNA, and there is no evidence that these introns circularize, in contrast to another class of
MITOCHONDRIAL
AND NUCLEAR
INTRONS
603
Splicing G |
5
P ~
a -gggaguaacuaugacucucu~AAUUGCAAA I.I1[.111 ) UCCGGACUAUUGAAAA[GGGAGGUUUCCAUU[UA ( Illllll.ll IIIIIIGCACCUGAUAGCUAG U C
3~
UCGuaagguagccaaaugccuA PIO A
Cyclisation
5GAAAU~UGC rJJlt
(
UCCGGACUAUUGAAA IIIIIII.II GCACCUGAUAGCUAG U C AAUUU
)
\
\3' UCG A A UCAGG]
Core structure J
FI(;. 2. Possible seeondarr structure for cyc]ization of the nuclear L-rRNA intron of Tetrahymena. The splicing structure is as shown in Fig. l(b), but including the G base t h a t initiates splicing by attacking the flint base of the intron (Zaug & Cech, 1982). The cyclization structure is based on evidence provided by Zaug el al. (1983), who show that, after excision, the last base of the intron attacks the A base, 15 bases fi'om the start of the intron (numbered as from the G base attached during excision), l~sulting in circularization and zelease of a 15mer. They note the first of the hairpin structures. The second hairpin within the IGS sequence {boxed) is also possible. The diagram shows the T. piffmentosa sequence. Base changes in T. thermophila are shown in lower case italics. Bars have been used to indicate A" U and G-C base-pairs. Centre dots have been used to indicate G ' U base-pairs. Hyphens have been omitted from sequences for clarity.
mitochondrial introns, for which circular forms of excised introns have been demonstrated (Arnberg et al., 1980; Hensgens et al., 1983). I t is not clear whether all or part of the results of the experiments with the Tetrahymena intron will apply to mitochondrial RNA splicing. Most mitochondrial introns require the function of an mRNA maturase encoded in one or other intron within the mitochondrial system (Lazowska et al., 1980; Jacq et al., 1982), although one of the introns considered by Davies et al. (1982), the mitochondrial rRNA intron of S. cerevisiae, can be excised without mitochondrial protein synthesis (Tabak et al., 1981). Nuclear mutations have also been shown to affect specific splicing reactions in yeast mitochondria" (Dieckmann et al., 1982). The Tetrahymena splicing reaction occurs 60 times faster in vivo than in vitro (Kruger et al., 1982), perhaps due to the presence of accessory proteins. The
604
R. B. WARING ET AL.
possibility therefore remains t h a t m R N A maturases and nuclearly coded protein factors play some role (e.g. stabilization of secondary structure) t h a t enables the mitochondrial splicing reaction to proceed at an adequate rate, rather than actually providing excision/ligation activity. We are now engaged in testing whether mitochondrial introns will self-splice in vitro. A preliminary investigation of some introns in nuclear m R N A precursors (human E-globin (Baralle et al., 1980) and ovalbumin (Heilig et al., 1982)) has not so far provided a clear example of an intron with the features defined as mitochondrial. Thus, to date all the mitochondrial-type introns outside mitochondrial occur in rRNA genes. I t is interesting t h a t the genes of all the RNAs with introns fitting the model of Davies et al. (1982) are located extrachromosomally at high copy n u m b e r in specialized organelles. The close resemblance of the Tetrahymena nuclear rRNA introns and the fungal mitochondrial introns is likely to reflect mobility of the introns themselves or of parts of the genome containing them between the nucleus and organelle at some time in the history of these organisms rather than convergent evolution. We thank Dr T. R. Cech for communicating results before publication. We acknowledge support from the Medical Research Council by project grant G7904290CB (to R.W.D. and C.S.).
REFERENCES Allet, B. & Rochaix, J. (1979). Cell, 18, 55-60. Arnberg, A. C., Van Ommen, G. J. B., Grivell, L. A., Van Bruggen, E. F. J. & Borst, P. (1980). Cell, 19, 313-319. Baralle, F. E., Shoulders, C. C. & Proudfoot, N. J. (1980). Cell, 21,621-626. Borst, P. & Grivell, L. A. (1981). Nature (London), 285, 439-440. Breathnach, R. & Chambon, P. (1981). Annu. Rev. Biochem. 50, 349-383. Burke, J. M. & RajBhandary, U. L. (1982). Cell, 31,509-520. Ceeh, R. T., Zaug, A. J. & Grabowski, P. J. (1981). Cell, 27, 487-496. Davies, R. W., Waring, R. B., Ray, J. A., Brown, T. A. & Scazzocchio, C. (1982}. Nature (London), 30{), 719-724. De La Salle, H., Jacq, C. & Slonimski, P. (1982). Cell, 28, 721-732. Dieckmann, C. L., Pape, L. K. & Tzagoloff, A. (1982). Proc. Nat. Acad. Sci., U.S.A. 79, 1805-1809. Dujon, B. (1980). Cell, 20, 185-197. Fox, T. D. & Leaver, C. J. (1981). Cell, 26, 315-323. Heilig, R., Muraskowsky, R., Kloepfer, C. & Mandel, J. L. (1982). Nucl. Acids Res. IO, 4363-4382. Hensgens, L. A. M., Arnberg, A. C., Roosendaal, E., Van der Horst, G., Van der Veen, R., Van Ommen, G. J. B. & Grivell, L. A. (1983). J. Mol. Biol. 164, 35-58. Jaeq, C., Pajot, P., Lazowska, J., Dujardin, G., Claisse, M., Groudinsky, 0., De La Salle, H., Grandchamp, C., Labouesse, M., Gargouri, A., Guiard, B., Spyridakis, A., Dreyfus, M. & Slonimski, P. (1982). In Mitochondrial Genes (Slonimski, P., Borst, P. & Attardi, G., eds), pp. 155-183, Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Kan, N. C. & Gall, J. G. (1982). Nucl. Acids Res. 10, 2809-2822. Koch, W., Edwards, K. & KSssel, H. (1981). Cell, 25, 203-213. Kruger, K., Grabowski, P. J., Zaug, A. J., Sands, J., Gottsehling, D. E. & Cech, T. R. (1982). Cell, 31, 147-157. Lazowska, J., Jaeq, C. & Slonimski, P. (1980). Cell, 22, 333-348.
MITOCHONDRIAL AND NUCLEAR INTRONS
605
Lerner, M. R., Boyle, J. A., Mount, S. M., Wolin, S. L. & Steitz, J. A. (1980). Nature (London), 283, 220-224. Michel, F., Jacquier, A. & Dujon, B. (1982). Biochimie, 64, 867-881. Netzker, R., KLchel, H. G., Basak, N. & Kfintze[, H. (1982). Nucl. Acids Res. 10, 47834794. Nomiyama, H , Sakaki, Y. & Takagi, Y. (1981a). Proc. Nat. Acad. Sci., U.S.A. 78, 13781380. Nomiyama, H., Kuhara, S., Kutika, T., Otsuka, T. & Sakaki, Y. (1981b). Nucl. Acids Res. 9, 5507-5520. Rogers, J. & Wall, R. (1980). Proc. Nat. Acad. Sci., U.S.A. 77, 1877-1879. Tabak, H. F. J., van der Laan, J., Osinga, K. A., Schouten, J. P., van Boom, J. H. & Veeneman, G. H. (1981). Nucl. Acids Res. 9, 4475-4483. Tinoco, I.. Uhlenbeck, O. C. & Levine, M. D. (1971). Nature (London), 280, 362-867. Waring, R. B., Davies, R. W., Scazzocchio, C. & Brown, T. A. (1982). Proc. Nat. Acad. Sci., U.S.A. 79, 6332-6336. Wild, M. A. & Sommer, R. (1980). Natttre (London), 283,693-694. Zaug, A. J. & Cech, T. R. (1982). Nucl. Acids Res. 10, 2823-2838. Zaug, A. J., Grabowski. P. J. & Cech, T. R. (1983). Nature (London), 801,578-583..
Edited by S. Brenner