ow0-4039m s6.00 + .oo .... RNA 6 show any conformational isomerism or self-cleavage which are ... spectrom) was calculated from the plot of In(K) vs. lfl. The Ea ...
ow0-4039m s6.00 + .oo Pcqpmm PressLJd
Temh&on Leuers. Vol. 34, No. 2A, pp. 3929-3932 1993 Primtedin Great Britain
The Self-cleavage of Lariat-RNA Peter Agback. Corine Glemarec, Lee Yin, Anders Sandstrlhn, Janez Plavec, Christian Sund, Shun-i&i Yamakage, Garimella Viswanadham, Bertrand Rousse, Nitin Puri, & Jyoti C!hattopadhyaya* Departmentof BioorganicChemistry,Box 581. Biomedicd Center, Univasity of Upps&
S-751 23 Uppsah Sweden
Abstmct: L&at-RNAs 3 and 4 undergo site-specific se!f-cleavuge reaclion al the G3 -+ &IV7 phosphodiester bond by the nucleophilic attack of 2’-OH o G3 sugar moiety to iis Y-phosphate to give Y-hydro~@ terminal at & or d ond 2: 1’-cyclic phosphodiester of G( whereas loriat-tetramer I, pentawbzr2, the cyclic-A(2*+YJG-tetramer 5 and the crclic-A(3’~S’)c-tetr~r b are completely stable. The kuiat-RNAs3 and 4 are rhe smallest RNA known IOumiergo self-cleavage which is reminiscent of the RNA-hammerhead (Ribozyme)acdvity. The geometry thecleavage-site in 3 and 4 has been d&ud byfull wnfornrarioM1 onolysis by NMR and molecular dynamics calculation in water.
of
Most of the catalytic natural RNA molecules, including viroids, virusoids and satellite RNAs, that infect plants are large and form complex protein encapsulated tertiary structure. They undergo efficient sitespecific self-cleavage’ in vitro. These RNAs share a small structural domain (hammerhead) consisting of about 30 nucleotides which has been shown to be necessary and sufficient for the self-cleavagel.
The
catalytic activity of the hammerhead RNA includes the site-specific self-cleavage of a phosphodiester bond * ‘1, which is clearly different from group I and to give a S-hydroxyl and a 2’, 3’-cyclic phosphcdiester ternurn II introns and RNase P, and is widespread in plant viruses and animals. We herein report on the site-specific self-cleavage reaction of small model lariat-RNAs2 3 and 4, which leaves a 2’,3’-cyclic phosphate at the 3’end of G3 and a S-hydroxy terminal at C%J7, mimicking the hammerhead self-cleavage pmductt. The lariat-hexamer 3 exhibits two sets of proton resonances3*5 attributed to two different forms (Aform @ B-form) in slow exchange (k = 0.1 s-1 at 25’ C) on the NMR time scale (K = Xg/XA = 0.1 at 5 “C and 0.75 at 60 ‘C, AH = 7.1 f 0.6 kcal.mol-1, B = 21 f 1.1 cal.mol-t.K-1, Ea = 23.8 f 1.8 kcal.mol-1, AG# = 18 kcal.mol-1 at 25 ‘C)3 whereas lariat-heptamer 4 shows only one average conformation on the NMR time scale. The lariat-hexamer 3 and heptamer 4 undergo self-cleavage reaction at room temperature whereas tetramer 1,pentamer 2, the cyclic-A(2’-+5’)G-tetramer 5 and the cyclic-A(3’+5’)G_tetramer
6 do not self-
cleave. The site of phosphodiester cleavage is specifidl and occurs at the 3’-phosphate of the guanosine residue (G3) to give a guanosine 2’. 3’-cyclic phosphate4 and a S-hydroxy terminal (5’-OH@ in 3 and 5’OH-UT in 4). At 25 ‘C, lariat-heptamer 4 (k = 0.16. 10-S min-1) cleaves six times faster than hexamer 3 (k = 0.25. lo”l min-1). The rate of cleavage is temperature-dependent. the rate of cleavage decnases
considerably
Note that at higher temperature (>45’C),
suggesting that the conformation
required for the cleavage
reaction only forms at narmw temperature range (10 ‘C < T < -45 ‘C). The addition of magnesium ions did not produce any noticeable changes in the conformation, as evident by unaltered 3Jm couplings, despite the fact that it increased the rate of cleavage by a factor two. Thus, the coordination of Mg2+ to the non-bridging phosphate-oxygens
might help self-cleavage by stabilizing the transition state but is not necessary for its
formation. It has also been found that the triethylammonium salt of lariat-hexamer 3 did not self-cleave. The
3931
CS
5
6
Figure 1: Synthetic lariat-RNAs 1 - 4 modelling the natural lariat counterpart which is formed as an intermediate in the RNA splicing reaction. Lariat RNAs 1 and 2 are sterically strained and are stable. Hexameric Lariat RNA 3 shows two conformations which are in slow exchange on the NMR time scale, and only one of them self-cleaves between G3 and C6 (shown by ” -+ “). Heptameric lariat RNA 4 self-cleaves between G3 and U7 (shown by ” + “) at a faster rate than 3. while neither cyclic 2’+5’ RNA 5 nor 3’+5’ RNA 6 show any conformational isomerism or self-cleavage which are characteristic of hexametic lariat RNA 3.
3932
MD generated geometry show that a simple rotation of the local phosphate backbone at the cleavage-site from e- (e_3’p = 3.8 A) + et (&_3’p = 2.8 A) and a rotation of c+ + {t would position the leaving Ytermini of U7 for a potential in-line displacement by 2’-OH of G3 (Note that in the latter geometrical transition, a+, fit and F and the South-sugar of U7 remain unchanged). Such a geometry at the cleavage site would produce an optimal locul structm
for a neighbouring
nucleophilic attack by 2’-OH to give the
trigonal-bipyramidal transition stat&ntermediate [ab inirio optimized geometry of guanosine 2’,3’cyclic oxyphosphorane, Gaussian 92 HPB21G basis set: e = +l lo’, c = -178’. a = 66’. b_3p = 1.8 AI,and the subsequent cleavage of the P-O(J’U7) bond (a-torsion) between G3 and U7. In an independent study12 by the ub initio calculation on the cyclic oxyphosphorane as model intermediates during splicing and cleavage of RNA, it has also been shown that the ground state geometry across the a-torsion favours the a+ mode which is consistent with our ground state geometry at the cleavage-site. We believe that this present state-ofthe-art understanding of the conformational requirement of the self-cleavage reaction of RNA will be useful in the design of RNA catalyst (ribozyme). AcknaalcdgcmentszAurholsthankSwedishBeardfor Technical Deve@ment (NUlEK) and Swedisb Natual Science Research council (NFR). Wallenbeqstiftelsen. University of Upplala, For&ning&knitmnden @RN) for financii sopport.
References 1.
2.
3.
4.
(a)Fcaer,A. C.; Symons. R. H. Cell, 1987,d9,211. (b) Fostex,A. C.; Symons, R. H. Cell, 1987.50, 9. (c) IJlhbeck. 0. C. Namre, 1987.32&S%. (d) Symons, R. H. Trends Bbchem. Sci. ,1989,14.445.(e) Miller, W. A.; Miller. S. L. Nucleic Acids Res.. 1991.19.5313. (a) Sund. C.. Agback, P.: Chattopadhyaya, J. Ten&&on, 1993,PZ, 649. K at 8 tcmpa between F’ and 4%. were obtained bum the relative inte.nsitiesof the H2A1 at 500 &fHx ~H-NMR ia D20. Ihe AH and AS of A to B transition (cbemicai exchange connectivities were found between A ad B forms in the NOESY spectrom) was calculated from the plot of In(K) vs. lfl. The Ea for the transition was obtained [rate (ka) obtained fkom saturation transfer expe?imentsl from tbe Arrbenias plot In ka vs. In: Ag = 23 kcal.mol-l and A$ = 15.6 cal.n~$~.Ic~. G3H2’ (6 5.15 in the selfcleaved pmduct of 3 & 66.30 in the selfdeavcd product of 4) and G3H3’ (S 4.% in tbe self-clcavad product of 3 & 65.11 in tbe self-cleaved product of 4) am conelated with 6 31P at 21.4 ppo~ (JH~*P = Jmp = 8.2 HZ). HPLC (see ref 2 for conditions) of the self-cleavage w Rt for 3 = 8.63 min and Rt for tbe self-cleavage prodoct of 3 = 10.11 min (not separable by PAGE). lH-NMR (6 ppm in D20) for the self&avage product of 3 : A*: 6.11 (HI’). 7.97 (IQ, 8.21 (Ha); U2: 5.70 (Hl’). 7.70 0.5.75 (H5); G3: 5.84 (I-H’), 7.71 (H8). U4: 5.81 (Hl?. 7.76 (I-K). 5.79 (H5); C5: 5.86 (HI’), 7.77 (Ha), 5.85 (H5); C6: 5.76 (Hl’). 7.84 (H6) and 5.99 (H5). *H-NMR (Sppm in D2O):for the aclfclcavage product
5. 6.
7.
a. 9. 10. :::
of4 : Al: 6.25 f&l’), 8.12 (H2). 8.35 @IQ; U2: 5.82 cHl3.7.810.5.91 (Hs): G3: 5.99 (H17.7.84 (H8). ti: 5.96 (HI?. 7.91 (H6), 5.93 (H5); C5: 6.02 (Hl’). 7.97 @I6), 6.ti (Hi); c6: i.Oi (Hl~,7.98 (H6) aadij.lj-(H5);-LJ7&2 (I-W); 7.98 (H6) and 5.91 (Hn. The ~f-ckxtvaae aoduct of 4 however was not semrable kom 4 bv HPLC or PAGE. The~f~a&nd analysis of 1 - ids performed in RO by mcasurhg tcmperat&&&l&&&&l shifty of proton, phoephclrcwro and scalar COI@Q -6 (IH- lH, 1H-31P and l3C-3lP by DQF-COSY, lH-13C HSQC) and NOmY . 00 XiuUuwA A1 (North-suear.WI-glycxsyl bond, ? and B% U2 (South-sugar,anti- glycosyl bond, UC,fP and e-); G3 (~~-WC Syn-glMl bond, and es); d (North-Wgm. Mti-glycosyl bond, e and et): U7 (Sooth-sugar, (uui-glycosyl bond and y+). which is quite simifar lo rhe cot@mation ofe-fom ofhemtw 3. (b) ~-formofhwramu~: ~1 (s~~th-s~gar, d-gWosy1 bond and e-); 3 (Sauth-sugar~anti- glycosyl bond. y+ and e-1; G3 (bdmgar. mi-glycosyl bond, y+, f9 and e?; c6 @bth-sugar. aati-glycosyl bond and et). (a) Lsriat-tetmmer 1 [A1 (S, anti p. Bt, e3; U2 (S. anti. y+. pt. e3; G3 (S, syn, anti, y or r’. pt, e- ). (b) lariat penmmer 2 [A* (S. anti,Y+,pt. e-); U2 (S. an& p. pt. e-k G3 (S, anti, $ IX y,p9. (c) Cyclic A(2’+J’)G-te-trama 5 [AI (S, unri’.pt. e3: U2 (S. mui, pt. e?; G3 0% syn. ‘Y+.pt. E% C6 (N. unti, y+, pt. et). (d) Cyclic A(3’+5’)G-tetrame~ 6 [A* (S. wtti, y or+. pt. etk U2 (S. anti. Y+. B!k G3 0% syn. UC.8% & 0% tami,P, Bt, e?. (a) aa TOL % B-W. J. I% Feldgein.P. A.; kkstein. F; Bmcaiug,G. Nuckic Acids Res.. 1990.18.1971. (b) Slim, 0.; Gait. hf. J. Nucleic Acids Res.. 1991,19,1183. (c) Koizwni, M.; Ohtda, E. Biochemisoy , l991,30,5145. (a) Heus, H. A.: Pardi. A. J. hoof. Biol, 1991,217, 113. (b) Odai. 0.; Kc&ma, H.; Him&i. H.; s&am, T.; T~II&& T.; Ucsugi. S. Nucleickidr Res.. 1990.18,S955. (c) Pease. A. C.; Wanmu, D. E. Biochct&try, l990.39,9039. (a) Cotdter, C. L. 1. Am. Chem. .k., 1973.95.570. (b) Spenea, W. et ol. Acta Crystaflogr. 1978. B 34.1520. (c) Smgcr, M. et al. Acto Crystallogr., 1978. B 34.2803. W.; B&stein. F. J. Am. Ck Sot.. 1970,92.4712. (d) Sm, Kollman. P et al.. AMBER 4.0. University of Califania, San Ranciaco 1991. (a) Taba, K. er al Nucleic Aci& Res. l989,17,3699. (b) Uchbnaru. T. et al FASEEJ. 1993,7,137.
(Received in UK 21 April 1993)
