The crystal structure and chemical-probing analysis of tRNA suggests that ... nearly 90% of the original self-cleavage activity of the ribozyme. By contrast ...
.:) 1994 Oxford University Press
3722-3727 Nucleic Acids Research, 1994, Vol. 22, No. 18
Identification of phosphate oxygens that are important for self-cleavage activity of the HDV ribozyme by phosphorothioate substitution interference analysis Yeon-Hee Jeoung+, P.K.R.Kumar, Young-Ah Suh§, Kazunari Taira and Satoshi Nishikawa* National Institute of Bioscience and Human Technology, Agency of Industrial Science and Technology, MITI, Tsukuba Science City, Ibaraki 305, Japan Received June 6, 1994; Revised and Accepted August 2, 1994
ABSTRACT A phosphorothioate substitution interference assay was used to investigate the role of the pro-Rp oxygens of phosphate groups in the self-cleavage reaction of the genomic human hepatitis delta virus (HDV) ribozyme. Incorporation of several different phosphorothioates (NTPctS) into the HDV ribozyme inhibited the self-cleavage activity. Incorporation of uridine 5' phosphorothioate or adenosine 5' phosphorothioate maintained 72% of the original selfcleavage activity whereas incorporation of guanosine 5' phosphorothioate or cytosine 5' phosphorothioate into the precursor reduced self-cleavage activity to about 20% in each case. Using partially substituted phosphorothioate-modified transcripts, we identified the pro-Rp oxygens that are important for the ribozyme activity, and they are located at positions 0, 1, 4, 5, 21, 24, 25, 27, 28, 30 - 34, 40, 43 and 75. In particular, the pro-Rp oxygens at positions 0, 1 and 21 are appear to be critical for the self-cleavage activity of the HDV ribozyme. INTRODUCTION The genome of human hepatitis delta virus (HDV) is a singlestranded circular RNA consisting of about 1700 nucleotides that includes two self-cleaving sequences, on the genomic and the antigenomic strands, respectively. It is believed that the selfcleaving ability of RNA (ribozyme activity) plays an important role in RNA processing during replication of the virus (1-3). As is the case for other self-cleaving RNAs, such as hammerhead and hairpin-type RNAs, the HDV ribozyme requires a divalent cation for self-cleavage activity and the cleavage results in production of a 2',3'-cyclic phosphate. The primary sequence of the HDV ribozyme does not resemble those of other known types of ribozyme. Thus, the HDV ribozyme represents a unique type of ribozyme. Several models of its secondary structure have
been proposed and, in order to elucidate the functional structure of the HDV ribozyme and to evaluate the role of nitrogen bases, we have used in vitro mutagenic analyses (4-8), as well as experiments with chemical probes (9). Our results indicate that important bases lie within the single-stranded regions (SSrA; 726-731, SSrB; 762-766 and SSrC; 708-715 in Figure 3) (5,6,8) that can be drawn in a pseudoknot-like model (10) of the secondary structure of the HDV ribozyme. Since a pseudoknot-like structure is not the most thermodynamically stable secondary structure for the ribozyme, this structure is thought to participate in some tertiary interactions (or interaction with protein) that is requires to generate the active conformation. The crystal structure and chemical-probing analysis of tRNA suggests that participation of the sugar-phosphate backbone in hydrogen bonding is important for stabilization of the tertiary structure of the RNA and that the backbone also provides a site for the recognition of protein. Therefore, it is reasonable to postulate that, in large RNAs in particular, the sugar-phosphate backbone might stabilize the tertiary structure. Furthermore, metal ions have been shown to be the real catalyst in ribozyme-catalyzed reactions (11-14). Recently, in attempts to evaluate the role of phosphate oxygens in catalytic RNAs, nucleoside phosphorothioates (NTPaS) have been incorporated in hammerhead ribozyme (15-17), pre-mRNA splicing (18), group I intron (19-22) and hairpin ribozyme (23). In phosphorothioate-substituted RNA, phosphate groups are replaced by phosphorothioate groups and, as a result, one of the peripheral oxygens of the relevant phophodiester is replaced by sulfur. Moreover, a chiral center is generated at the phosphate, with either an Rp or Sp configuration. However, only the Sp isomers of phosphorothioates are readily incorporated into RNA by polymerases. T7 RNA polymerase specifically recognizes the Sp isomer of NTPaS as a substrate and incorporates it, with inversion of its configuration, to generate the Rp isomer in the phosphodiester linkage (24). It is believed that secondary structure remains unchanged in a phosphorothioate-substituted nucleic acid
*To whom correspondence should be addressed Present addresses:
'Deparunent of Chemistry, Korea Advanced Institute of Science and Technology, Taejon and §Department of Life Science, Pohang
University of Science and Technology, Pohang, Korea
Nucleic Acids Research, 1994, Vol. 22, No. 18 3723 but the substitution can significantly alter the ability of the phosphate to form hydrogen bonds or to coordinate with metal ions. In the present study, we used phosphorothioate substitutions to locate specific phophodiester bonds that are important for the self-cleavage reaction of the HDV ribozyme. Incorporation of several different NTPaSs into the HDV ribozyme inhibited the self-cleavage activity. However, incorporation of only uridine 5' phosphorothioate or adenosine 5' phosphorothioate maintained nearly 90% of the original self-cleavage activity of the ribozyme. By contrast, incorporation of guanine 5' phosphorothioate or cytosine 5' phosphorothioate dramatically decreased the selfcleavage activity. Using partially substituted phosphorothioatemodified transcripts, we identified the pro-Rp oxygens that are important for the ribozyme activity.
MATERIALS AND METHODS Chemicals and methods Four pure nucleoside 5'-[x-thio] triphosphates (Sp form) were purchased from NEN Research Products (Dupont Inc., Boston, MA, USA). [a-32P]GTP and ['y-32P]ATP were from Amersham (Arlington Heights, IL, USA). Iodine was from Wako Pure Chemical Industries (Osaka, Japan). T4 polynucleotide kinase was from Nippon Gene (Toyama, Japan). T7 RNA polymerase and alkaline phosphatase were from Toyobo (Osaka, Japan). Plasmid The plasmid pUHD88 was used in the present study and its construction has been described previously (4). pUHD88 contains the genomic HDV sequence from nucleoside 683 to nucleoside 770 (numbering is based on that in ref. 1). All experiments were carried out using E. coli MV 1184 as host cells. Plasmid DNA was prepared from an overnight culture by the alkaline lysis procedure and subsequent treatment with QIAGEN-Tip 500 (QIAGEN Inc., Chatsworth, CA). Preparation of RNA transcripts pUHD88 was linearized with BamHI. The standard transcription reaction (10 1d) contained 40 mM Tris -HCl (pH 7.5), 6 mM MgCl2, 2 mM spermidine, 10 mM DTT, 0.01 % bovine serum albumin, 0.5 mM each of four NTPs, 95 units of T7 RNA polymerase and 1 ,ug of template DNA. The reaction was incubated for 1 hr at 37°C. For phosphorothioate substitution experiments, the transcription reaction was carried out with inclusion of one of the phosphorothioate analogs to give a substitution level of 20% per reaction except when stated otherwise. Transcripts were purified on a 10% polyacrylamide gel that contained 8 M urea. Uncleaved precursor RNA and the 3' self-cleaved product were isolated from the gel after visualization of bands by UV shadowing and extracted as described previously (4). End-labeling of RNA transcripts Purified transcripts were labeled at the 5 '-end in a 20 j1l reaction mixture that contained 20 pmoles of RNA, 20 pmoles of [y-32p] ATP (3000 Ci/mmol), 50 mM Tris-HCl (pH 7.6), 10 mM MgCl2, 5 mM DTT, 0.1 mM spermidine, 0.1 mM EDTA and 10 units of T4 polynucleotide kinase. The reaction was incubated for 1 h at 37°C and terminated by addition of EDTA to 25 mM. Terminal triphosphates of the uncleaved precursor RNA were dephosphorylated with alkaline phosphatase prior to treatment
with the kinase in 50 Atl of 50 mM Tris-HCI (pH 8.0), 1 mM MgCl2 and 0.02 unit of the enzyme for 30 min at 37°C, and then it was extracted with phenol/chloroform and precipitated with ethanol. Cleavage by iodine and mapping Cleavage by iodine was carried out by the method of Schatz et al. (25) with some modifications. Uncleaved precursor or the 3'self-cleaved product (104 cpm) was dissolved in 9 A1 of 10 mM HEPES (pH 7.0), 50% formamide (final concentration), and 1 1l of 10 mM iodine (dissolved in ethanol) was aded. The reaction mixture was incubated for 15 minutes at 65°C and placed on ice. One volume of loading dye solution (9 M urea, 20 mM EDTA, 0.1% bromophenol blue, 0.1% xylene cyanol) was added and the sample was subjected to electrophoresis on a 10% polyacrylamide gel that contained 8 M urea (sequencing gel). The gel was autoradiographed at -70°C under an intensify screen for 2 days and was also analyzed with a Bioimaging analyzer (BA100; Fuji Photo Film, Tokyo, Japan). Alkaline hydrolysis ladders were generated and reactions catalyzed by ribonuclease Ti were carried out as described previously (6).
Quantitative analysis The radioactivity of individual sequencing bands (Cx) of the uncleaved precursor and the 3'-product was determined. The intensity of bands was measured and compared for each phosphorothioate position that had been modified and cleaved by iodine. A ratio (Rx) was calculated for different phosphate positions by dividing the relative band intensity (%) for the uncleaved precursor by the relative band intensity (%) for the 3'-product. Rx is [CPrex / CPretotal] x 100 / [Cprox / CPrototal] x 100, where CPrex is the band intensity of uncleaved precursor at position x, CPretotal is the sum of band intensities for the uncleaved precursor and CProx is the band intensity for the 3'-product at position x. The ratio (Rx) provides a measure of the effect of substitution by NMPaS on the self-cleavage activity of the HDV ribozyme.
RESULTS In order to investigate the effect of phosphorothioate substitution by NTPaS, we analyzed self-cleavage activity during in vitro Table 1. Effects of phosphorothioate substitution on the activity of the HDV ribozyme HDV ribozyme
Average self-cleavage activity (% of control)
Unmodified
80 (100)
100% AcaS
72 (90)
100% UaS
72 (90)
100% AaS plus UaS
50 (63)
100% GcaS
23 (29)
20% GaS
33 (41)
100% CaS
20 (25)
20% CaS
47 (59)
Self-cleavage activity was analyzed during transcription in vitro in the presence of 6 mM MgC12 and NTPcSS (NaiS) instead of the corresponding NTPs.
3724 Nucleic Acids Research, 1994, Vol. 22, No. 18 transcription in the presence of 6 mM MgCl2 (Table 1) and one or more NTPaS instead of the normal NTPs. Since the HDV ribozyme is very active, almost all precursor RNAs were selfcleaved during transcription under our previous conditions (9). To isolate uncleaved precursor RNAs, we used much more T7 RNA polymerase (95 units/i utg of template DNA) and reduced the incubation time (from 4 h to 1 h). Even under these new conditions, 80% of transcription products were self-cleaved during transcription in vitro. When all four NTPaS were used during transcription, the HDV ribozyme activity was completely abolished. When GMPaS was substituted completely (100%) and
the self-cleavage activity fell to 23 % and 33 %, respectively (Table 1). In the case of CMPcaS, the self-cleavage activity decreased to 20% (after 100% substitution) and 47% (after 20% substitution), respectively. By contrast, the precursor HDV ribozyme synthesized with AMPctS or UMPaS retained 72% of the normal self-cleavage activity even at 100% substitution. Precursor HDV ribozyme prepared with both AMPcaS and UMPcaS still had more than 50% activity (Table 1). From these results it appears that phosphates 5' to A and U residues are not very critical for the self-cleavage activity of the HDV ribozyme
B
c 1E~~~~~~~~~~~~~~~~~~~~~~~7
32_
52SP!w#;?^t.-,W_
partially (20%) for GMP in the synthesis of the HDV ribozyme,
--
-xc
D
5; !lDeIed 13'-p'odJct ;
.4 :
Ohl Ti GE
-:.,w
2,__
. ~~~~~
.'7'
.,-"'. I..
0
Aia
F ....
7
.:
_
72 w_
4.
w
:: ::-.
...
25;
,-
-3
#N
J
I.
_L _
w
C; _
7. %* A.
-..
L
_
32 3 ,
-
4m "^''
.n I
I
----
_
te _
:w
2 214-
4'2
-
4w
*. A"
1 ',i".'
--
_2
s
48 47
3
37
...
27. `.:. R
..'
46_
27
,.
43 ...
4*1 0
*4
-.
''
,.,.
-.04-
4U1
32 3
I
1
._m
Figure 1. Phosphorothioate substitution interference assay of the HDV ribozyme. Transcription reactions were carried out in the presence of 20% GTPCSS or CTPaS with natural NTPs. Precursor and 3'-product RNAs of the self-cleavage reaction were labeled at their ends after isolation on a 10% denaturing polyacrylamide gel and treated with 1 mM iodine (see Materials and Methods). Numbers indicate positions of substituted phosphates. The phosphate at position 0 corresponds to the self-cleavage site. Arrows show the positions of bands with increased intensity in the case of the precursor and with diminished intensity in the case of the 3 '-product. The radioactivity was balanced for the transcript and 3'-cleaved product for both the GMPOIS (GaS) and CMPaS (CaS) lanes before the gel was loaded. A and C show precursors substituted with GaS or CcaS. B and D show 3'-products substituted with GaS or CaS. The band marked by an arrowhead in D is a minor self-cleaved product between C729 and A730 by intrinsic chemical cleavage not by iodine at Py-A-A sequence (data not shown, 31).
1. ~for
Nucleic Acids Research, 1994, Vol. 22, No. 18 3725
Table 2. Data from phosphorothioate substitution interference analysis of
and they also imply that at least one of the phosphates 5' to G
Phosp.# Ratio(Rx) Pre(%) Pro(%)
that reduce the self-cleavage activity of the HDV ribozyme,
phosphates 0-75 in the HDV ribozyme Nt # Phosp.# Ratio(Rx) Pre(%) Pro(%)
689 690 691
0 1 2
6.10 1.20
694
5
2.08
7
-
692 693
695
696 698 699 700 701 703 704 705 706 707
708
3 4
6
9
1.31
1.58
1.29
4.89 1.28 0.49 0.55
0.42 0.38
0.60
0.50
0.24
-
-
0.27
0.21
1.61
1.53 1.66 0.63 0.68 -
Nt #
731 732 733
42
736
47 48
734 735
737
1.12
0.52
0.67
0.44
738 740 741 742 743 745 746 747 748
-
-
-
750
1.05
10 11 12
1.02 0.90
1.69 0.57
14 15 16 17
1.01 1.24 1.18
0.69 -
19
18
0.21 0.41
0.56
0.45
0.60
749
43
44
45 46
49 51 52 53 54 55 57 58 59 60
61
714 715 716 717
25 26 27 28
1.76 2.12
0.72 0.72
0.41 0.34
758 759
62 63 64 66 67 68 69 70
719
30
1.43
0.60
0.42
761
72
721
32
0.46
763
723
34
709 710 711
712
718
720
722
724 725 726
727
728
729 730
20 21 22
6.04
2.90
0.48
751 752 753
23
1.624 1.50
0.72
0.44
7554 756
-
-
29 31
33
1.29 -
1.26 1.65
1.43
1.88 1.62
0.76
0.98 -
-
1.08
0.72
0.78
0.62
0.54
0.89
0.66
0.33
0.62 0.60
0.37
35 36 37
0.39 0.67
0.24 0.86
0.611.29
40
2.64 -
1.53 -
0.58 -
38 39
41
0.82 1.06
1.30
1.07 1.07
1.01
757
760
762
764 765
766 767 768
71
1.21 0.72 0.76 0.74 0.74 1
0.79 0.88
-
-
-
-
0.77
0.54 -
-
1.24 1.23
75
1.50
74 76
77 78 79
80
771
82
772
-
0.92 0.96
73
769
770
1.61 -
1.32 -
-
-
-
81
-
83
-
1.19 -
0.74 -
-
-
0.85
0.92
0.94
0.52 0.36 0.48 0.37 0.37 0 0.49 1.11 -
-
-
0.98
0.43 0.50 0.63 0.50 0.50 0 0.62 1.26 -
-
-
-
-
-
0.46
0.60
0.44 -
0.82 -
-
-
-
-
-
1.49 1.24
1.49
6-
x
substitution interference experiments were performed.
Transcription reactions were carried out using mixtures of NTPcS and NTP at a ratio of I to 4 (20%). After the transcription reaction in vitro, the precursor and the 3'-cleaved product were
separated and labeled at the 5' end. Representative autoradiograms after partial substitution and interference analysis of the transcript and 3'-cleaved product when 20% GMPCKS or CMPaiS was used, are shown in Figure 1. To identify the location of phosphorothioate hnkages, both 5'-end-labeled precursorsand 3'-cleaved products were treated with iodine. Since iodine is believed to result sufficient cleavage even when the transcripts possessless than 5% phosphorothioate linkages and moreover pses%plsuruoL it is highly reactive than epoxipropanol (25). The partially hydrolyzed products were separated by electrophoresis on a denaturing gel and the end-labeled products were detected with
the Bioimaging analyzer. Sites of cleavage by iodine in the HDV
ribozyme are on the 5'-sides of NMPcaS, while ribonuclease Ti cleaves on the 3'-side of guanosine nucleotides. Therefore
-
1.20
hydrolyzed products resulted from the above treatment migate
1.12
differently on denatuning PAGE. Radioactivity in individual bands or either in the precursor or the3'-cleaved product was measured with the Bioimaging analyzer and is given as a percentage (Pre or Pro) relative to total counts (CPretotal or in Table 2. The values shown are averages from three individual experiments. We calculated the ratio (Pre/Pro=R,)
1.01
1.53
1.02
-
-
-
-
-
-
-
-
-
-
-
-
Nt # refers to the position in the HDV genome, according to Kuo et al. (1). Phos. # refers to the position relative to the site of cleavage by iodine. Pre (%) indicates the relative intensity (%) of the band in the inactive precursor lane, Pro (%) indicates that of the band in the 3'-product, measured with the image analyzer. The Ratio (Rx) is Pre/Pro. The values represent the averages of three trials. A dash indicates that results were not determined because of the position of UTP and ATP substitutions, the position on the gel (79-83), or band
compression (60-65).
and C residues is important for activity of the ribozyme . To examine the precise location of phosphorothioate linkages
(CP'rx CPrx)
CPrmtotal)
of radioactivity of corresponding bands for the uncleaved
precursor and the 3'-product (see Materials and Methods and Figure 2). The intensities of bands were higher for the 1, 4, 5, 21, 24, 25, 27, 28, 30-34, 40, 43 and 75 phosphates in the
precursor compared to the 3'-cleaved product and gave ratios (Rx) of more than 1.42, average value. In particular, the ratios 1 a p o phosphate oxygens at positions I and 21 were quite higher than others, and probably these pro-Rp oxygen phosphates play
an important role in the self-cleavage reaction. The ratio for the
4 0
Cu3 2.........
......
..
.........
0
............................
-
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76
Phosp. No.
Figure 2. The ratio (R,) plotted against the phosphate number for the substitution interference assay. We took a ratio of more than 1.4 to mean that the phosphorothioate substitution interfered with the self-cleavage reaction. Dotted line indicates average value, 1.4.
3726 Nucleic Acids Research, 1994, Vol. 22, No. 18 phosphate oxygen at position 0 could not be measured since position 0 of the 3'-product did not appear on the sequencing gel (see Materials and Methods). However, in the case of the uncleaved precursor, the band intensity corresponding to phosphate oxygen 0 was more than five times higher than the average intensity of bands whose ratios were about 1, and more an 1.5 times higher than the band intensities that corresponded to phosphate oxygens 1 and 21. The importance of phosphates 60-65 could not be determined because of band compression on the sequencing gel and, similarly, data for phosphates 79-83 were obscured by the position of the bands (top of the gel).
DISCUSSION In tRNAs, a number of phosphate groups are known to contribute to tertiary interactions by participating in the formation of hydrogen bonds with nucleotides and also by participating in coordination with metal ions. Indeed, with NMPaYS as structural probes, several important phosphates have been identified in catalytic RNAs (15-23). Recently, in the HDV ribozyme, several bases were identified by use of chemical probes as tertiary-interacting residues, and these were found in the stem I and stem III regions (9). In an attempt to analyze the importance of the phosphates of the HDV ribozyme in self-cleavage activity, we used a phosphorothioate modification and interference assay in the present studies. As evidence from our results, the self-cleavage activity of the HDV ribozyme was only marginally affected by substitution with
auc33,
10 U30 U 30
0 9 " C-G C A-U
C-G
gG Co -
I
le
A0-6 5'cleavage
C0-6U
A766 SSC-G8 SSrB .6-C GC
cUW - 4~~~ 73A U c~~0 U - -U2o
site-b0O@g
7'u*G
G7
AAG SSrC
~ggcucgag
SrSSA750
UU2 UA7 cGU-AC7
76
C0G C-G
6v -C
G-C G-C A-U
CG Figure 3. Summary of data from phosphorotioate substitution interference assays on the pseudoknot-like model of the genomic HDV ribozyme. Dots represent ratios (R1) above 1.4 and the large dots indicate ratios of more than 6. Phosphates are numbered in italics and the cleavage site is shown by an arrowhead. Outline fonts indicate important bases from our mutagenic analyses (5,7,8).
UMPaS or AMPaS. It appears, therefore, that the phosphates (pro-Rp oxygens) 5' to A and U residues are not very important for the ribozyme activity. By contrast, in the case of CMPaSand GMPaS-substituted precursors, the self-cleavage activity was reduced to a low level (about 20%), indicating that at least one of the phosphates 5' to G and C residues has an important role in the ribozyme activity. In order to examine the precise locations of important phosphorothioate linkages in the HDV ribozyme, we made partial substitutions with only GMPCaS or CMPcKS to give an average substitution level of 20% NMPCKS residues per precursor. When the band intensities that corresponded to phosphorothioate-substituted positions were compared between the precursor and the 3'-cleaved product, seventeen higherintensity positions were found in the HDV ribozyme precursor. These phosphates had higher ratios (Rx) that were nearly 1.4 fold greater than those for other positions (Table 2). Among the important pro-Rp oxygen phosphates, numbers 1 and 21 had the highest ratio (RO, which was about 6 in both cases (Figure 2). Phosphate oxygen 21 is located in the SSrC region and this region has been shown to be involved in interactions that are important for catalysis by point mutation analysis (8,26). In particular, residue 709C is intolerant to substitutions by other bases and seems to be involved in tertiary interactions via the N3 of cytosine, as indicated by chemical probing (9). Phosphate 21 is on the 3' side of this nucleotide and this phosphate must be concerned with binding pocket of metal ions or tertiary folding interactions. Residue 690G which is 3' side of the phosphate 1 was also indicated as a tertiary interacted base from chemical probing (9). The band intensity for phosphate oxygen 0 was the highest of all the phosphates. Phosphate oxygen at the splicing site is particularly critical for the splicing reaction in group I introns (20,27) and in the splicing of pre-mRNA (28). Most of the relevant phosphates are located between the SSrC and SSrA regions (Figure 3), a result that is in good agreement with those of modified base interference analysis (29) and tertary interaction analysis that involved base modification with chemical reagents (9). When the 3'-cleaved product of HDV ribozyme was alkylated by treatment with ethylnitrosourea (ENU) under native conditions, the regions of phosphates between SSrA and SSrC were well protected. Phosphates 79-83 of stem II and 60-65 of stem IV were not analyzed because of our inability to visualize the relevant bands. The stem IV region can be deleted without any effect on the self-cleavage activity of the HDV ribozyme (4,29). From our recent lead-cleavage analyses, phosphate 75 seems to be near phosphate 21 (unpublished results), an indication that this phosphate may be involved in the tertiary interactions required for the active conformation. In order to rescue the ribozyme reaction, we have isolated the precursor RNA and cleavage reaction was carried out in the presence of Mn2+ but we have not observe any significant enhancement of the cleavage reaction. To evaluate the role of the above identified phosphates, a more detailed experimental analyses are underway by site specific incorporation of either Sp or Rp form thio-linkage in the ribozyme. Very recently tertiary structural model of HDV ribozyme was proposed (30). In the proposed model, when we examine the important phosphates 0, 1, 21 and 75 that are identified above, these phosphates are located close to each other and probably to form a part of the active site for the HDV ribozyme activity. We propose that phosphate oxygens in stem I and stem HI are
Nucleic Acids Research, 1994, Vol. 22, No. 18 3727 involved in the tertiary interactions required for the active conformation and, in particular, pro-Rp oxygens of phosphates 0, 1 and 21 are very important for ribozyme function, either via coordination with the divalent cation or via involvement in the tertiary interactions with other residues.
ACKNOWLEDGEMENT Y.-H.Jeoung and Y.-A.Suh are grateful to the Science and Technology Agency of Japan for their STA fellowships.
REFERENCES 1. Kuo, M., Sharmeen, L., Dinter-Gottlieb, G., and Taylor, J. (1988) J. Virol., 62, 4439-4444. 2. Sharmeen, L., Kuo, M.Y.P., Dinter-Gottlieb, G. & Taylor, J. (1988) J. Virol., 62, 2674-2679. 3. Wu, H.-N., Lin, Y.-J., Lin, F.-P., Makino, S., Chang, M.-F. and Lai, M.M.C. (1989) Proc. Natl. Acad. Sci. USA, 86, 1831-1835. 4. Suh, Y.-A., Kumar, P.K.R., Nishikawa, F., Kayano, E., Nakai, S., Odai, O., Uesugi, S., Taira, K. & Nishikawa, S. (1992) Nucleic Acids Res., 20, 747-753. 5. Kumar, P.K.R., Suh, Y.-A., Miyashiro, H., Nishikawa, F., Kawakami, J., Taira, K. & Nishikawa, S. (1992) Nucleic Acids Res., 20, 3919-3924. 6. Kumar, P.K.R., Suh, Y.-A., Taira, K. & Nishikawa, S. (1993) FASEB J., 7, 124-129. 7. Suh, Y.-A., Kumar, P.K.R., Kawakami, J., Nishikawa, F., Taira, K. & Nishikawa, S. (1993) FEBS Lett., 326, 158-162. 8. Kawakami, J., Kumar, P.K.R., Suh, Y.-A., Nishikawa, F., Kawakami, K., Taira, K. & Nishikawa, S. (1993) Eur. J. Biochem., 217, 29-36. 9. Kumar, P.K.R., Taira, K. & Nishikawa, S. (1994) Biochemistry, 33, 583-592. 10. Perrotta, A.T. & Been, M.D. (1991) Nature, 350, 434-436. 11. Uchimaru, T., Uebayashi, M., Tanabe, K. & Taira, K. (1993) FASEB J., 7, 137-142. 12. Piccirilli, J.A., Vyle, J.S., Caruthers, M.H. & Cech, T.R. (1993) Nature, 361, 85-88. 13. Steitz, T.A. & Steitz, J.A. (1993) Proc. Natl. Acad. Sci. USA, 90, 6498-6502. 14. Pyle, A.M. (1993) Science, 261,709-714. 15. Buzayan, J.M., Feldstein, P.A., Bruening, G. & Eckstein, F. (1988a) Biochem. Biophys. Res. Commun.,156, 340-347. 16. Buzayan, J.M., Feldstein, P.A., Segrelles, C. & Bruening, G. (1988b) Nucleic Acids Res.,16, 4009-4023. 17. Ruffner, D.E. & Uhlenbeck, O.C. (1990) Nucleic Acids Res., 18, 6025-6029. 18. Griffiths, A.D., Potter, B.V.L. & Eperon, I.C. (1988) J. Bio. Chem., 263, 12295-12304. 19. Waring, R.B. (1989) Nucleic Acids Res., 17, 10281-10293. 20. Suh, E. & Waring, R.B. (1992) Nucleic Acids Res., 20, 6303-6309 21. Christian, E.L. & Yarus, M. (1992) J. Mol. Biol., 228, 743-758. 22. Christian, E.L. & Yarus, M. (1993) Biochemistry, 32, 4475-4480. 23. Chowrira, B.M. & Burke, J.M. (1992) NucleicAcids Res., 20,2835-2840. 24. Griffiths, A.D., Potter, B.V.L. & Eperon, I.C. (1987) Nucleic Acids Res., 15, 4145-4162. 25. Schatz, D., Leberman, R. & Eckstein, F. (1991) Proc Natl. Acad. Sci. USA, 88, 6132-6136. 26. Thill, G., Vasseur, M. & Tanner, N.K. (1993) Biochemistry, 32, 4254-4262. 27. McSwiggen, J.A. & Cech, T.R. (1989) Science, 244, 679-683. 28. Moore, M. J. & Sharp, P.A. (1993) Nature, 365, 364-368. 29. Belinsky, M., Britton, E. & Dinter-Gottlieb, G. (1993) FASEB J., 7, 130-136. 30. Tanner, N.K., Schaff, S., Thill, G., Petit-Koskas, E., Crain-Denoyelle AM. & Westhof, E. (1994) Current Biol., 4, 488-498. 31. a) Kierzek, R. (1992) Nucleic Acids Res., 20, 5073-5077. b) idem 5079-5084.
