This article was downloaded by: [New York University] On: 06 January 2015, At: 21:22 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
Modes of Binding of 2′-AMP to RNase T1 A Computer Modeling Study a
c
P. V. Balaji , W. Saenger & V. S.R. Rao a
a b
Molecular Biophysics Unit
b
Jawaharlal Nehru Center for Advanced Scientific Research , Indian Institute of Science , Bangalore , 560 012 , INDIA c
Institute for Crystallography Free University Berlin , Takustrasse 6, D-1000 , Berlin 33 , Germany Published online: 21 May 2012.
To cite this article: P. V. Balaji , W. Saenger & V. S.R. Rao (1992) Modes of Binding of 2′-AMP to RNase T1 A Computer Modeling Study, Journal of Biomolecular Structure and Dynamics, 9:5, 959-969, DOI: 10.1080/07391102.1992.10507970 To link to this article: http://dx.doi.org/10.1080/07391102.1992.10507970
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [New York University] at 21:22 06 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, ISSN 0739-1102 Volume 9, Issue Number 5 (1992), "'Adenine Press (1992).
Modes of Binding of 2'-AMP to RNase T 1 A Computer Modeling Study P.V. Balaji1, W. Saenger3 and V.S.R. Rao 1•2 * 1
Molecular Biophysics Unit & Jawaharlal Nehru Center for Advanced Scientific Research Indian Institute of Science Bangalore 560 012, INDIA
Downloaded by [New York University] at 21:22 06 January 2015
2
3
Institute for Crystallography Free University Berlin Takustrasse 6, D-1000, Berlin 33, Germany Abstract The modes ofbinding of adenosine 2' -monophosphate (2' -AMP) to the enzyme ribonuclease (RNase) T 1 were determined by computer modelling studies. The phosphate moiety of 2'AMP binds at the primary phosp4ate binding site. However, adenine can occupy two distinct sites- (1) The primary base binding site where the guanine of2'-GMP binds and (2) The subsite close to theN 1 subsite for the base on the 3' -side of guanine in a guanyl dinucleotide. The minimum energy conformers corresponding to the two modes of binding of 2' -AMP to RNase T 1 were found to be of nearly the same energy implying that in solution 2' -AMP binds to the enzyme in both modes. The conformation of the inhibitor and the predicted hydrogen bonding scheme for the RN ase T 1 - 2' -AMP complex in the second binding mode (S) agrees well with the reported x-ray crystallographic study. The existence ofthe first mode ofbinding explains the experimental observations that RN ase T 1 catalyses the hydrolysis of phosphodiester bonds adjacent to adenosine at high enzyme concentrations. A comparison of the interactions of2'-AMP and 2'-GMP with RNase T 1 reveals that Glu58 andAsn98 at the phosphate binding site and Glu46 at the base binding site preferentially stabilise the enzyme- 2'GMP complex.
Introduction
Ribonuclease (RNase) T 1 (EC 3.1.27.3) is an extracellular enzyme secreted by the fungus Aspergillus oryzae and has 104 amino acid residues of known sequence [1 ].It is the best characterised [2] of all the enzymes of the microbial RN ase family which show a high degree of sequence homology [3]. The three-dimensional structure of RNase T 1 [4] and its complexes with the inhibitors 2'-guanosine monophosphate (2'-GMP)[5,6] and guanyl-2',5'-guanosine (G-2'p5'-G) [7] has been determined at high resolution by x-ray diffraction. The small size and the availability of the threedimensional structure at high resolution has made RNase T 1 a model system not *Author to whom correspondence should be addressed.
959
960
Balaji eta/.
Downloaded by [New York University] at 21:22 06 January 2015
only for understanding the specificity and mechanism of action but also for studying the protein-nucleic acid interactions using 2-D nmr [8], protein engineering [9,10] and computer modelling [11-13] methods. RNase T 1 catalyses the hydrolysis of the P-05' phosphodiester bond adjacent to guanosine in single stranded RNA with high specificity. However, the specificity of RNase T 1 is only relative but not absolute [14]. It also catalyses the hydrolysis of phosphodiester bonds adjacent to xanthosine and inosine, although at considerably reduced rates [14,15]. At high enzyme concentrations, phosphodiester bonds adjacent to adenosine are also hydrolysed by RNase T 1 [16-20] suggesting that adenine binds to the same site of the enzyme as guanine. Both adenosine and adenosine 2'-monophosphate (2'-AMP) act as competitive inhibitors ofRNase T 1 catalysed reaction [21] implying that these two molecules associate with the enzyme in a position where they interfere with substrate binding. The position could be the specific guanine binding site, the catalytic site where the phosphate group is bound, or the base subsite. The latter was indicated in the 1.8 A resolution x-ray crystallographic study of the RNase T 1 - 2'-AMP complex [22] which revealed that the phosphate moiety of2' -AMP occupies nearly the same phosphate binding site as in the RNase T 1 - 2' -GMP complex [11 ]. Adenine, however is not in the guanine binding site but is stacked on His92 imidazole, so that it occupies a site previously identified as subsite for the base N in a substrate GpN [23]. In view of these different binding modes, the energetically favoured modes and the preferred sites ofbinding of 2' -AMP to RNase T 1 are investigated by computer modelling methods to rationalise the results obtained by experimental studies both in solid state and solution. Methods The coordinates for the atoms ofRNase T 1 were taken from the 1.8 A resolution xray crystallographic data of the RN ase T 1 -vanadate [4], RN ase T 1 - 2' -GMP [5] and the RNase T 1 - 2'-AMP complex [22] structures. All the CH, CH2 and CH3 groups in the protein were treated as united atoms. The coordinates for the polar hydrogen atoms (i.e., hydrogens attached toN and 0) of protein and the coordinates for the inhibitor 2' -AMP were generated using standard bond lengths and bond angles [2426]. In the x-ray diffraction study of the RN ase T 1 - 2' -GMP complex [5], most of the water molecules were found around the surface of the protein and sparsely in the active site, and the inhibitor binding site was found to be part of an apparently underhydrated surface portion. Hence instead of including the solvent molecules explicitly, the effect of solvent in damping the electrostatic forces was modelled by using a distance dependent dielectric constant which weighs the short range interactions more than the long range interactions. The phosphate moiety of both 2'-AMP and 2'-GMP occupy the same phosphate binding site comprising the amino acid residues Asn36, Tyr38, His40, Glu58, Arg77, His92 and Asn98, but the orientations and consequently the hydrogen bonding schemes differ. Hence with the phosphate moiety at the primary phosphate binding site, two sites (primary and subsite) accessible for the binding of adenine moiety were explored. In the first mode, adenine occupies the primary binding site similar
Binding of 2' -AMP to RNase T1
961
Table I Conformational Angles and Energy Components for some of the Low Energy Conformers of the RNase T 1 - 2'-AMP Complexes Primary Site
Adenine occupying
Downloaded by [New York University] at 21:22 06 January 2015
Conformer Number Torsion Angle (degrees) C4 -N9 -Cl' -04' C3' -C4' -C5' -05' C4' -C5' -05' -05'H C4' -C3' -03' -03'H Cl' -C2' -02' -P C2' -02' -P -P03 02' -P -P03 -POH Pseudorotation Phase Angle P CO) Ribose Pucker ENERGY (kcal/mol) Ligand Protein Interaction Total
*
X-ray* study
Sub site
PI
P2
P3
Sl
S2
S3
S4
S5
67 -63 -65 159 164 -89 26
121 -176 67 78 172 - 51 107
53 62 94 -170 -153 178 28
-110 56 158 -175 -164 90 -170
- 98 59 172 -177 -166 -164 -162
- 75 180 179 -165 90 180 -177
-137 56 149 147 108 - 59 96
-91 54 81 166 93 157 174
-93 80
43 C4'exo
-
C2'exo
157 C2'endo
157 C2'endo
159 C2'endo
28 C3'endo
151 C2'endo
151 C2'endo
131 Cl'exo
- 14.3 -191.0 - 97.1 -302.3
- 16.4 -172.4 -122.6 -311.5
- 15.4 -190.4 - 80.9 -286.7
- 17.2 -172.8 -104.2 -294.2
- 12.6 -180.1 -102.7 -295.5
- 11.2 -183.4 -103.7 -298.3
- 13.2 -195.0 -104.1 -312.3
- 11.3 -173.6 -116.4 -301.4
2
121
From the 1.8 A resolution RNase T 1 - 2'-AMP complex study [22).
to guanine of2'-GMP in the RNase T 1 - 2'-GMP complex [11,27). The amino acid residues Tyr42-Asn43-Asn44-Tyr45-Glu46 and Asn98 comprise this base binding site. In the second, adenine occupies a site which is close to the N1 [28). The conformers corresponding to these two modes of binding of 2'-AMP to RNase T1 are referred to by the prefixes P and S in discussion. In view of the severe restriction on the available computer time, the calculations were carried out in three steps - contact criteria, energy minimisation in torsion angle space and energy minimisation in Cartesian coordinate space. About 10-15 starting conformations were selected for minimisation for both modes of binding. The form of the potential energy function and the parameters used along with the details of the procedures for contact criteria and energy minimisation in both torsion angle and Cartesian coordinate spaces have been described elsewhere [11,27).
Results and Discussion The conformational angles of the bound inhibitor and the energy components in some of the low energy conformers of the RNase T 1 - 2'-AMP complex for the two modes of binding are given in Table I. The conformational angles of2' -AMP in the 1.8 A resolution x-ray crystallographic study of the RNase T 1 - 2' -AMP complex are also given in Table I. The proposed hydrogen bonding scheme in the corresponding complexes is given in Table II. Stereo diagrams of the active site residues ofRNase
05'H
05'
03'H
03'
Ribose 02'
N7
N6H2
N6Hl
N3
Adenine Nl
AMP
N43 1.7 E46 1.6 N43 2.3
0 137 OEI 171 H-N 174
PI
HOI 123 HE2 167
N98 0 1.7 166
N98 H02 1.8 140
N43 2.5 H40 2.3
P2
N98 001 1.8 139
0 144 OEI 168 H-N 168
P3
N43 1.7 E46 1.6 N43 2.3
Primary binding site
N36 H02 1.7 167
R77 HH21 2.5 146
872 OG 2.2 136
81
HE2 145 HOI 164
N36 ODI 1.8 148
H92 1.8 N98 1.8
82
N36 ODI 1.7 163
83
8ubsite
N36 001 1.8 141
HOI 162 0 126
84
N98 1.7 N98 1.8
Table II Proposed Hydrogen bonding scheme in the RNase T 1 - 2'-AMP Complexes*
Downloaded by [New York University] at 21:22 06 January 2015
N36 ODI 1.7 151
N98 HDI 1.7 169
G74 0 1.7 139 R77 HH21 2.2 124
85
N36 ODI
G74 0
X-ray
:--
I»
CD
':::::
-
f
iii"
~
co
E58 OEl 1.7 156
P-OH
R77 1.9 H92 1.9 N98 2.5 Y38 1.8 E58 1.9 R77 1.9 E58 1.8
HH21 148 HE2 147 HDl 116 HH 164 HE2 137 HE 163 OEl 135
P2
E58 OEl 1.7 151
R77 HE 1.7 154
HH 158 HE2 156
HH21 150 HE2 139
P3
Y38 1.7 H40 2.0
R77 2.0 H92 1.9
Primary binding site
HH 159 HE 152 HH21 143 HDl 140
H92 HE2 1.7 146
Y38 1.7 R77 1.8 R77 1.9 N36 1.9
Sl
E58 OE2 1.8 148
R77 HE 1.7 145
E58 HE2 1.8 134
Y38 HH 1.7 160 H40 HE2 1.9 144
S2
HE 141 HH21 137 HE2 171 HH 165 HE2 157
E58 OEl 1.7 168
R77 1.9 R77 1.9 H92 1.8 Y38 1.7 H40 1.8
S3
Subsite
HH 134 HE2 170
Y38 OH 1.8 149
R77 HH21 1.7 178
Y38 1.7 H40 1.8
R77 HE 1.9 144 H92 HE2 1.9 136
S4
HE 156 HE2 157
HH 166 HE2 145
E58 OEl 1.7 146
R77 1.9 H92 1.8
Y38 1.7 H40 1.9
S5
E58 OE2
R77 HH21 H92 HE2 E58 OEl
R77 HE
Y38 HH
H40 HE2
Y38 HH
X-ray
* For each hydrogen bond, the hydrogen bond length (hydrogen bonded hydrogen ... acceptor) and the hydrogen bond angle (Donor-Hydrogen-Acceptor) are given. The criterion for the selection of hydrogen bonds is as suggested by Baker and Hubbard (24).
P-03
HDl 142 HH 164 HE2 140 HE 156
N36 2.0 Y38 1.7 H40 2.1 R77 1.7
H92 HE2 1.9 144
PI
P-02
Phosphate P-Ol
AMP
Table II continued
Downloaded by [New York University] at 21:22 06 January 2015
~
co
0) ('I)
.........
CD
(I)
I
:a
0
:b.
!\)
....
CQ 0
~
ta
:;·
Downloaded by [New York University] at 21:22 06 January 2015
964 Balaji et a/.
~N36
965
Downloaded by [New York University] at 21:22 06 January 2015
Binding of 2'-AMP to RNase T1
N43 Figure l: Stereo diagrams showing the active site residues ofRNase T 1 and the bound inhibitor in the RNase T 1 - 2'-AMP complex; (a) P2 conformer; (b) P3 conformer; (c) Sl conformer; (d) S5 conformer.
T 1 and the bound inhibitor are shown in Figure 1 for the conformers P2, P3, S1 and S5. Table I shows that the glycosidic torsion angle (C4-N9-Cl'-04') of the bound nucleotide is dependent on the site occupied by the base moiety of the inhibitor. The nucleotide takes syn orientation when the base occupies the primary binding site (conformers PI, P2 and P3), and anti orientation when the base occupies the subsite (conformers Sl, S2, S3, S4 and S5). However, there is no correlation between the ribose pucker and the site at which the base binds as observed for the glycosyl torsion angle. In both the low energy conformers P3 and S5, the ribose moiety adopts the C2' -endo pucker form. It can also be seen from Table I that eventhough the difference in the total energy of the conformers P3 and S5 is very small, the interaction energy is better in conformer P3 and the internal protein energy in S5. The most favoured conformers (P3 and S5) of the RNase T 1 - 2' -AMP complex have nearly the same total energy irrespective of whether adenine occupies the primary binding site (P) or the subsite (S), suggesting that two modes are probable. The phase angle which defines the puckering of the ribose and the glycosyl torsion angle of2'AMP in conformer S5 are close to that found from x-ray crystallographic studies [22)(Table I). In this conformer (Figure 1d) adenine occupies a site which is close to the Nl subsite [29] and forms one hydrogen bond each with the backbone carbonyl ofGly74 and the side chain guanidinium group of Arg77 (Table II) in addition to the stacking interaction with His92 imidazole. The phosphate group occupies the phosphate binding site. The hydrogen bonding scheme predicted for the S5 conformer agrees well with that proposed by x-ray crystallographic study [22). In the S1 conformer adenine occupies the N1 subsite [29) forming a hydrogen bond with the side chain hydroxyl group of Ser72. However, the stacking interaction between adenine and His92 imidazole is absent (Figure 1c) as are hydrogen bonding interactions between the phosphate moiety and the amino acid residues His40 and Glu58 (Table II). Consequently, this conformer is energetically less favourable than the
966
Balaji et a/.
Downloaded by [New York University] at 21:22 06 January 2015
conformer S5.It is interesting to note that when guanyl dinucleosidephosphates bind to RNase TI,the base on the 3’-terminal of guanosine favours an orientation similar to that of adenine of 2’-AMP in the S1 confor-mer perhaps due to additional steric constraints present in binding the dinucleoside phosphates [ 121. In the minimum energy conformer P3 ofthe RNase Tl-2’-AMPcomplex,the amino acid residues Tyr42 and PhelOO form a hydrophobic pocket for the base. The hydrogen bond between the Asn98 side chain amide group and Tyr45 hydroxyl group orients the phenyl ring of Tyr45 so as to maximise the stacking interaction with adenine (Figure lb). The orientation of adenine in this complex is similar to that of guanine in the RNase TI- 2’-GMPcomplex [111.Although the hydrophobic interactions between the active site residues and adenine is similar to that found in the RNase TI - 2’-GMP complex, the number of hydrogen bonds formed by adenine with the enzyme are less compared to guanine. The position and orientation of adenine in the P1 and P3 conformers are very similar. However, the ribose adopts C4’-exo and C2’-endo pucker forms respectively in these two conformers. Consequently, the interaction between Asn98 side chain and the ribose moiety present in the P3 conformer will not be possible in the P1 conformer. The backbone conformation of the dipeptide fragment Asn43-Asn44 in these two conformers is similar to that in RNase TI with no guanine bound [4]. Consequently the base forms one hydrogen bond each with the backbone carbonyl and amide groups of Asn43 in addition to a hydrogen bond with the side chain carboxyl group of Glu46 (Table 11; Figure lb). When energy minimisation is carried out starting with the backbone of this dipeptide fragment in the same conformation as that observed in the RNase TI - 2’-GMP complex [5], where it is flipped 140” with respect to the guanine-free enzyme [4], adenine is pushed out of the binding site due to electrostatic repulsion between the 6-amino group of adenine and the backbone amide ofAsn44. As a result, in conformer P2 adenine forms two weak hydrogen bonds with the side chains of His40 and An43 (Table 11) and may not have stacking/hydrophobic interactions with the enzyme (Figure la) leading to higher energy complex. Thus the present studies show that 2’-AMP favours to bind to RNase TI in essentially two modes with adenine occupying two different binding sites and phosphate occupying the same site. The two binding modes have nearly the same energy indicating that in solution, 2’-AMP binds to RNase TI in the two modes with nearly equal probability.The S5 conformer where adenine is stackingwith the His92 imidazole is, in the crystalline state, additionally stabilised by one intermolecular contact with a symmetry related RNase TI molecule, 05’...OE2 Glu102 [22]. However, we can not exclude that this mode of binding is the one that prevails also in solution.The hydrolysis ofthe phosphodiesterbonds adjacentto adenosine at higher enzymeconcentrationscan be explained only by the binding of adenosine at the primary site. Recentlythe structureof the RNase TI- 2’-AMP complexwas predicted by a free-energy perturbation study starting from the structure ofthe RNase TI- 2’-GMP complex by Hirono and Kollman [29]. The conformation predicted for the enzyme -inhibitor complex from this study does not agree with that of either the x-ray crystallographic studies [22] or any of the low energy conformers reported from the present study.
Binding of P‘-AMP to RNase T,
967
Table In Energy components for the interaction between the active site residues of RNase T, and the inhibitors 2’-AMP and 2‘-GMP (in kcal/mol)
-
RNase TI - 2’-GMP
RNase TI 2’-AMP Amino Acid
Primary site (P3)
Downloaded by [New York University] at 21:22 06 January 2015
VDW ELE Asn 36 Tyr 38 His 40 Tyr 42 Asn 43 Asn 44 Tyr 45 Glu 46 Glu 58 Ser 72 Gly 74 Arg 77 His 92 Asn 98 Asn 99 Phel00
-1.74 -1.12 -2.92 -2.85 -0.91 -0.73 -6.37 -0.58 -1.70 -0.02 -0.13 -0.20 -1.76 -1.98 -1.45 -5.89
- 1.30 - 3.83 -20.23 - 0.37 - 2.62 0.08 0.26 0.72 - 0.19 0.01 1.26 -24.24 -20.62 1.13 0.10 - 0.23
HB -0.02 -5.98 -1.16 -0.01 -3.54 -0.00 -0.18 -3.10 -6.04 0.00 0.00 -4.11 -1.07 -3.27 -0.02 -0.02
Subsite (S5) TOT VDW ELE
- 3.06 -10.93 -24.31 - 3.23 - 7.07 - 0.65 - 6.29 - 2.96 - 7.93 - 0.01 1.13 -28.55 -23.45 - 4.13 - 1.37 - 6.14
-2.66 -0.61 -1.54 -0.18 -0.03 -0.01 -0.54 -0.06 0.08 -0.05 0.59 -3.57 -5.88 -3.17 -0.77 -1.79
- 0.50 -3.45 -20.96 - 0.14 0.10 0.04 0.06 3.96 - 0.63 0.08 - 0.73 -19.84 -21.53 - 0.90 0.61 - 0.22
HB -3.59 -5.99 -1.09 0.00 0.00 0.00 -0.03 0.00 -6.05 0.00 -2.72 -1.81 -0.66 -2.86 0.00 0.00
Primary site TOT VDW ELE - 6.75 -10.05 -23.59 - 0.32 0.07 0.03 - 0.51 3.90 - 6.60 0.02 - 2.86 -25.22 -28.07 - 6.93 - 0.16 - 2.01
-1.04 -0.55 -1.91 -4.75 -3.23 -0.38 -4.21 0.98 -1.70 -0.02 -0.15 -1.51 -1.72 -1.65 -1.02 -3.57
HB
TOT
0.97 - 0.02 - 2.03 3.66 - 5.97 -10.18 -14.13 - 3.06 -19.10 - 0.81 - 0.05 - 5.61 - 3.20 - 1.10 - 7.53 - 2.21 - 1.11 - 3.70 - 1.55 - 1.10 - 6.86 - 8.62 - 6.15 -13.79 - 6.08 -11.77 -19.55 0.08 0.00 0.06 1.42 0.00 1.27 -25.26 - 2.29 -29.06 -20.80 - 1.02 -23.54 - 7.11 - 8.06 -16.82 0.54 - 0.01 - 0.49 - 0.05 0.00 - 3.62 -
-
The failure to predict the observed structure for the RNase T I -2’-AMP complex by Hirono and Kollman could be due to the consideration of an improper startingconformation for the backbone of the Asn43-Asn44 dipeptide fragment. In fact, earlier molecular dynamics studies carried out both in vacuum and aqueous media have indicated the flipping of the backbone ofAsn43-Asn44 dipeptide fragment [30].The present study shows that this dipeptide fragment favours a different conformation in the RNaseT, - 2’-AMP complex from that in the RNaseT, - 2’-GMP complex and the internal protein energies are also different [l 11. It is essential that these changes are also considered for analysis in any theoretical calculation [see for example Ref. 311. In view of this, the apparent agreement between the experimentally determined and theoretically calculated values of AAGbind(differencein free energy of binding) reported by Hirono and Kollman appears to be fortuitous since the change in internal protein (enzyme) energy upon inhibitor binding has been left out of their analyses. The contributions from the van der Waals, electrostatic and hydrogen bond interactions of all the active site amino acid residues in the RNase T, - 2’-AMP complex are tabulated in Table I11 for conformers P3 and S5.The contributions of these residues in the RNase T, - 2’-GMP (C2’-endo)complex [ 111 are also given for comparison. It can be seen from this table that the contributions of the individual amino acid residues constituting the phosphate binding site i.e., Tyr38, His40, Arg77 and His92 are approximately the same in all the three complexes. However, interaction between Glu58 and the phosphate moiety is better in the RNase TI - 2’-GMP complex compared to either of the RNase T, - 2’-AMP complexes.TheAsn98 - ribose interaction
968
Balaji et al.
Downloaded by [New York University] at 21:22 06 January 2015
observed in the RNase Tl-2’-GMP complex is absent in the RNase TI - 2‘-AMP complex. Among the amino acid residues constituting the base binding site, the Glu46 contribution is significantly less in both the P3 and S5 conformers of the RNase T I- 2’-AMPcomplex compared to the RNase T I- 2’-GMP complex.In the S5 conformer, there is a net electrostatic repulsion between Glu46 and 2’-AMP and in the P3 conformer, this repulsive interaction is slightly offset by the hydrogen bonding interaction with the 6-amino group of adenine. Among the other primary base binding site residues, contribution from PhelOO is more in the P3 conformer of the RNase TI - 2’-AMP complex compared to the RNase T, - 2’-GMP complex. In conclusion, the present study has shown that the adenosine-ribose moiety of 2‘AMP can bind to the primary (specific) or secondary (non-specific) substrate binding sites with about equal probability. The conformation about the glycosyl bond adopted by2‘-AMP when bound to the enzyme is found to be dependent on the site occupied by adenine, syn in the primary and anti in the secondary site. An analysis of the interactions of the active site amino acid residues with 2’-AMP and 2’-GMP shows that Glu46, Glu58 and Am98 stabilise the enzyme complex with 2’-GMP more than that with 2‘-AMP.These results are in agreement with x-ray, competitive binding and spectroscopic data. References and Footnotes 1. K. Takahashi,J. Biochem. (Tokyo) 98,815 (1985). 2. U. Heinemann and U. Hahn. In Protein-Nucleic acid interaction, (W.Saenger and U.Heinemann, eds.), Topics in molecular and structural biology, Vol. 10. Macmillan Press Ltd., London, pp 111 (1989). 3. C. Hill. In Structure and chemistry qfribonucleases,Proceedings of the first international meeting, (kPavlovsky and K.Polyakov. eds.), Mowcow, 17 1 (1988). 4. D. Kostrewa, H.-W. Choe, U. Heinemann and W. Saenger, Biochemhtry 28,7592 (1989). 5. R. Ami, U. Heinemann. R. Tokuoka and W. Saenger. J. Biol. Chem. 263. 15358 (1988). 6. S. Sugio, T. Amisaki, H. Ohishi and K-I. Tomita,J. Biochem. 103,354 (1988). 7. J. Keopke. M. Maslowska. U. Heinemann and W. Saenger,J. Mol. Biol. 206,475 (1989). 8. E. Hoffmann, J. Schmidt, J. Simor and H. Rueterjans. Nucleosides and Nucleotides 7,757 (1988). 9. T. Hakoshima, S. Toda. S. Sugio, K.4. Tomita. S. Nishikawa, H. Morioka, K Fuchimura. T. Kimura, S. Uesugi, E. Ohtsuka and M. Ikehara, Prot. E n s . 2,55 (1988). 10. H.-P. Grunert. k Zouni. M. Beineke, R. Quaas, Y. Georgalis, W. Saenger and U. Hahn, Eur. J. Biochem. 197.203 (1991). 11. P.V. Balaji, W. Saenger and V.S.R. Rao, Curr.Sci. 60,363 (1991). 12. P.V. Balaji, W. Saenger and V.S.R. Rao,J. Biomol. Str. Dyn. 9,215 (1991). 13. P.V. Balaji and V.S.R. Rao, Znd. J. Biochem. Biophys. 28.358 (1991). 14. F. Egami. T. Oshima and T. Uchida, Mol. Biol. Biochem. Biophys. 32,250 (1980). 15. K Takahashi and S. Moore, The Enzymes XV, 435 (1980). 16. K Sat0 and F. Egami,J. Biochem. (Tokyo) 44,753 (1957). 17. AM.Michelson and C. Monny, Biochim. Biophys. Acta 166,294 (1968). 18. M. 1rie.J. Biochem. (Tokyo) 58,599 (1965). 19. FA. Neelon, M. Molinaro. H. Ishikura, L.B. Sheiner and G.L. Cantoni, J. Biol. Chem. 242, 4515 (1967). 20. R.C. Warrington, Biochim. Biophys. Acta 353.63 (1974). 21. M. Irie,J. Biochem. (Tokyo) 56.495 (1964). 22. J. Ding, G. Koellner. H.-P. Grunert and W. Saenger, . I Biol. Chem., (in press), (1991). 23. Lenz, A, Cordes, F.. Heinemann, U. and Saenger, W., J. Biol. Chem. 266,7661 (1991). 24. E.N.Baker and R.E. Hubbard. h o g . Biophys. Mol. Biol. 44,97 (1984). 25. R. Taylor and 0. Kennard, J. Mol. Str. 78, 1 (1982).
Binding of 2'-AMP to RNase T1 26. 27. 28. 29. 30.
969
W. Saenger., "Principles of nucleic acid structure", Springer Verlag, First edition, Berlin (1982). P.V. Balaji, W. Saenger and V.S.R Rao, Biopoly. 30,257 (1990). P.V. Balaji and V.S.R Rao,J. Biomolec. Str. Funct. 9, (in press) (1992). S. Hirano and P.A Kollman,J. Mol. Bioi. 212, 197 (1990). AD. MacKerell, L. Nilsson, R Rigler, U. Heinemann and W. Saenger, Proteins: Structure, Function and Genetics 6, 20 (1989). 31. C.L. Brooks III, M. Karplus and B.M. Pettitt. In "Proteins: A theoretical perspective of dynamics, structure, and thermodynamics", Adv. Chern. Phys. 71, 60 ( 1988).
Downloaded by [New York University] at 21:22 06 January 2015
Date Received: October 11, 1991
Communicated by the Editor Dina Moras
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
Modes of Binding of 2′-AMP to RNase T1 A Computer Modeling Study a
c
P. V. Balaji , W. Saenger & V. S.R. Rao a
a b
Molecular Biophysics Unit
b
Jawaharlal Nehru Center for Advanced Scientific Research , Indian Institute of Science , Bangalore , 560 012 , INDIA c
Institute for Crystallography Free University Berlin , Takustrasse 6, D-1000 , Berlin 33 , Germany Published online: 21 May 2012.
To cite this article: P. V. Balaji , W. Saenger & V. S.R. Rao (1992) Modes of Binding of 2′-AMP to RNase T1 A Computer Modeling Study, Journal of Biomolecular Structure and Dynamics, 9:5, 959-969, DOI: 10.1080/07391102.1992.10507970 To link to this article: http://dx.doi.org/10.1080/07391102.1992.10507970
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [New York University] at 21:22 06 January 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, ISSN 0739-1102 Volume 9, Issue Number 5 (1992), "'Adenine Press (1992).
Modes of Binding of 2'-AMP to RNase T 1 A Computer Modeling Study P.V. Balaji1, W. Saenger3 and V.S.R. Rao 1•2 * 1
Molecular Biophysics Unit & Jawaharlal Nehru Center for Advanced Scientific Research Indian Institute of Science Bangalore 560 012, INDIA
Downloaded by [New York University] at 21:22 06 January 2015
2
3
Institute for Crystallography Free University Berlin Takustrasse 6, D-1000, Berlin 33, Germany Abstract The modes ofbinding of adenosine 2' -monophosphate (2' -AMP) to the enzyme ribonuclease (RNase) T 1 were determined by computer modelling studies. The phosphate moiety of 2'AMP binds at the primary phosp4ate binding site. However, adenine can occupy two distinct sites- (1) The primary base binding site where the guanine of2'-GMP binds and (2) The subsite close to theN 1 subsite for the base on the 3' -side of guanine in a guanyl dinucleotide. The minimum energy conformers corresponding to the two modes of binding of 2' -AMP to RNase T 1 were found to be of nearly the same energy implying that in solution 2' -AMP binds to the enzyme in both modes. The conformation of the inhibitor and the predicted hydrogen bonding scheme for the RN ase T 1 - 2' -AMP complex in the second binding mode (S) agrees well with the reported x-ray crystallographic study. The existence ofthe first mode ofbinding explains the experimental observations that RN ase T 1 catalyses the hydrolysis of phosphodiester bonds adjacent to adenosine at high enzyme concentrations. A comparison of the interactions of2'-AMP and 2'-GMP with RNase T 1 reveals that Glu58 andAsn98 at the phosphate binding site and Glu46 at the base binding site preferentially stabilise the enzyme- 2'GMP complex.
Introduction
Ribonuclease (RNase) T 1 (EC 3.1.27.3) is an extracellular enzyme secreted by the fungus Aspergillus oryzae and has 104 amino acid residues of known sequence [1 ].It is the best characterised [2] of all the enzymes of the microbial RN ase family which show a high degree of sequence homology [3]. The three-dimensional structure of RNase T 1 [4] and its complexes with the inhibitors 2'-guanosine monophosphate (2'-GMP)[5,6] and guanyl-2',5'-guanosine (G-2'p5'-G) [7] has been determined at high resolution by x-ray diffraction. The small size and the availability of the threedimensional structure at high resolution has made RNase T 1 a model system not *Author to whom correspondence should be addressed.
959
960
Balaji eta/.
Downloaded by [New York University] at 21:22 06 January 2015
only for understanding the specificity and mechanism of action but also for studying the protein-nucleic acid interactions using 2-D nmr [8], protein engineering [9,10] and computer modelling [11-13] methods. RNase T 1 catalyses the hydrolysis of the P-05' phosphodiester bond adjacent to guanosine in single stranded RNA with high specificity. However, the specificity of RNase T 1 is only relative but not absolute [14]. It also catalyses the hydrolysis of phosphodiester bonds adjacent to xanthosine and inosine, although at considerably reduced rates [14,15]. At high enzyme concentrations, phosphodiester bonds adjacent to adenosine are also hydrolysed by RNase T 1 [16-20] suggesting that adenine binds to the same site of the enzyme as guanine. Both adenosine and adenosine 2'-monophosphate (2'-AMP) act as competitive inhibitors ofRNase T 1 catalysed reaction [21] implying that these two molecules associate with the enzyme in a position where they interfere with substrate binding. The position could be the specific guanine binding site, the catalytic site where the phosphate group is bound, or the base subsite. The latter was indicated in the 1.8 A resolution x-ray crystallographic study of the RNase T 1 - 2'-AMP complex [22] which revealed that the phosphate moiety of2' -AMP occupies nearly the same phosphate binding site as in the RNase T 1 - 2' -GMP complex [11 ]. Adenine, however is not in the guanine binding site but is stacked on His92 imidazole, so that it occupies a site previously identified as subsite for the base N in a substrate GpN [23]. In view of these different binding modes, the energetically favoured modes and the preferred sites ofbinding of 2' -AMP to RNase T 1 are investigated by computer modelling methods to rationalise the results obtained by experimental studies both in solid state and solution. Methods The coordinates for the atoms ofRNase T 1 were taken from the 1.8 A resolution xray crystallographic data of the RN ase T 1 -vanadate [4], RN ase T 1 - 2' -GMP [5] and the RNase T 1 - 2'-AMP complex [22] structures. All the CH, CH2 and CH3 groups in the protein were treated as united atoms. The coordinates for the polar hydrogen atoms (i.e., hydrogens attached toN and 0) of protein and the coordinates for the inhibitor 2' -AMP were generated using standard bond lengths and bond angles [2426]. In the x-ray diffraction study of the RN ase T 1 - 2' -GMP complex [5], most of the water molecules were found around the surface of the protein and sparsely in the active site, and the inhibitor binding site was found to be part of an apparently underhydrated surface portion. Hence instead of including the solvent molecules explicitly, the effect of solvent in damping the electrostatic forces was modelled by using a distance dependent dielectric constant which weighs the short range interactions more than the long range interactions. The phosphate moiety of both 2'-AMP and 2'-GMP occupy the same phosphate binding site comprising the amino acid residues Asn36, Tyr38, His40, Glu58, Arg77, His92 and Asn98, but the orientations and consequently the hydrogen bonding schemes differ. Hence with the phosphate moiety at the primary phosphate binding site, two sites (primary and subsite) accessible for the binding of adenine moiety were explored. In the first mode, adenine occupies the primary binding site similar
Binding of 2' -AMP to RNase T1
961
Table I Conformational Angles and Energy Components for some of the Low Energy Conformers of the RNase T 1 - 2'-AMP Complexes Primary Site
Adenine occupying
Downloaded by [New York University] at 21:22 06 January 2015
Conformer Number Torsion Angle (degrees) C4 -N9 -Cl' -04' C3' -C4' -C5' -05' C4' -C5' -05' -05'H C4' -C3' -03' -03'H Cl' -C2' -02' -P C2' -02' -P -P03 02' -P -P03 -POH Pseudorotation Phase Angle P CO) Ribose Pucker ENERGY (kcal/mol) Ligand Protein Interaction Total
*
X-ray* study
Sub site
PI
P2
P3
Sl
S2
S3
S4
S5
67 -63 -65 159 164 -89 26
121 -176 67 78 172 - 51 107
53 62 94 -170 -153 178 28
-110 56 158 -175 -164 90 -170
- 98 59 172 -177 -166 -164 -162
- 75 180 179 -165 90 180 -177
-137 56 149 147 108 - 59 96
-91 54 81 166 93 157 174
-93 80
43 C4'exo
-
C2'exo
157 C2'endo
157 C2'endo
159 C2'endo
28 C3'endo
151 C2'endo
151 C2'endo
131 Cl'exo
- 14.3 -191.0 - 97.1 -302.3
- 16.4 -172.4 -122.6 -311.5
- 15.4 -190.4 - 80.9 -286.7
- 17.2 -172.8 -104.2 -294.2
- 12.6 -180.1 -102.7 -295.5
- 11.2 -183.4 -103.7 -298.3
- 13.2 -195.0 -104.1 -312.3
- 11.3 -173.6 -116.4 -301.4
2
121
From the 1.8 A resolution RNase T 1 - 2'-AMP complex study [22).
to guanine of2'-GMP in the RNase T 1 - 2'-GMP complex [11,27). The amino acid residues Tyr42-Asn43-Asn44-Tyr45-Glu46 and Asn98 comprise this base binding site. In the second, adenine occupies a site which is close to the N1 [28). The conformers corresponding to these two modes of binding of 2'-AMP to RNase T1 are referred to by the prefixes P and S in discussion. In view of the severe restriction on the available computer time, the calculations were carried out in three steps - contact criteria, energy minimisation in torsion angle space and energy minimisation in Cartesian coordinate space. About 10-15 starting conformations were selected for minimisation for both modes of binding. The form of the potential energy function and the parameters used along with the details of the procedures for contact criteria and energy minimisation in both torsion angle and Cartesian coordinate spaces have been described elsewhere [11,27).
Results and Discussion The conformational angles of the bound inhibitor and the energy components in some of the low energy conformers of the RNase T 1 - 2'-AMP complex for the two modes of binding are given in Table I. The conformational angles of2' -AMP in the 1.8 A resolution x-ray crystallographic study of the RNase T 1 - 2' -AMP complex are also given in Table I. The proposed hydrogen bonding scheme in the corresponding complexes is given in Table II. Stereo diagrams of the active site residues ofRNase
05'H
05'
03'H
03'
Ribose 02'
N7
N6H2
N6Hl
N3
Adenine Nl
AMP
N43 1.7 E46 1.6 N43 2.3
0 137 OEI 171 H-N 174
PI
HOI 123 HE2 167
N98 0 1.7 166
N98 H02 1.8 140
N43 2.5 H40 2.3
P2
N98 001 1.8 139
0 144 OEI 168 H-N 168
P3
N43 1.7 E46 1.6 N43 2.3
Primary binding site
N36 H02 1.7 167
R77 HH21 2.5 146
872 OG 2.2 136
81
HE2 145 HOI 164
N36 ODI 1.8 148
H92 1.8 N98 1.8
82
N36 ODI 1.7 163
83
8ubsite
N36 001 1.8 141
HOI 162 0 126
84
N98 1.7 N98 1.8
Table II Proposed Hydrogen bonding scheme in the RNase T 1 - 2'-AMP Complexes*
Downloaded by [New York University] at 21:22 06 January 2015
N36 ODI 1.7 151
N98 HDI 1.7 169
G74 0 1.7 139 R77 HH21 2.2 124
85
N36 ODI
G74 0
X-ray
:--
I»
CD
':::::
-
f
iii"
~
co
E58 OEl 1.7 156
P-OH
R77 1.9 H92 1.9 N98 2.5 Y38 1.8 E58 1.9 R77 1.9 E58 1.8
HH21 148 HE2 147 HDl 116 HH 164 HE2 137 HE 163 OEl 135
P2
E58 OEl 1.7 151
R77 HE 1.7 154
HH 158 HE2 156
HH21 150 HE2 139
P3
Y38 1.7 H40 2.0
R77 2.0 H92 1.9
Primary binding site
HH 159 HE 152 HH21 143 HDl 140
H92 HE2 1.7 146
Y38 1.7 R77 1.8 R77 1.9 N36 1.9
Sl
E58 OE2 1.8 148
R77 HE 1.7 145
E58 HE2 1.8 134
Y38 HH 1.7 160 H40 HE2 1.9 144
S2
HE 141 HH21 137 HE2 171 HH 165 HE2 157
E58 OEl 1.7 168
R77 1.9 R77 1.9 H92 1.8 Y38 1.7 H40 1.8
S3
Subsite
HH 134 HE2 170
Y38 OH 1.8 149
R77 HH21 1.7 178
Y38 1.7 H40 1.8
R77 HE 1.9 144 H92 HE2 1.9 136
S4
HE 156 HE2 157
HH 166 HE2 145
E58 OEl 1.7 146
R77 1.9 H92 1.8
Y38 1.7 H40 1.9
S5
E58 OE2
R77 HH21 H92 HE2 E58 OEl
R77 HE
Y38 HH
H40 HE2
Y38 HH
X-ray
* For each hydrogen bond, the hydrogen bond length (hydrogen bonded hydrogen ... acceptor) and the hydrogen bond angle (Donor-Hydrogen-Acceptor) are given. The criterion for the selection of hydrogen bonds is as suggested by Baker and Hubbard (24).
P-03
HDl 142 HH 164 HE2 140 HE 156
N36 2.0 Y38 1.7 H40 2.1 R77 1.7
H92 HE2 1.9 144
PI
P-02
Phosphate P-Ol
AMP
Table II continued
Downloaded by [New York University] at 21:22 06 January 2015
~
co
0) ('I)
.........
CD
(I)
I
:a
0
:b.
!\)
....
CQ 0
~
ta
:;·
Downloaded by [New York University] at 21:22 06 January 2015
964 Balaji et a/.
~N36
965
Downloaded by [New York University] at 21:22 06 January 2015
Binding of 2'-AMP to RNase T1
N43 Figure l: Stereo diagrams showing the active site residues ofRNase T 1 and the bound inhibitor in the RNase T 1 - 2'-AMP complex; (a) P2 conformer; (b) P3 conformer; (c) Sl conformer; (d) S5 conformer.
T 1 and the bound inhibitor are shown in Figure 1 for the conformers P2, P3, S1 and S5. Table I shows that the glycosidic torsion angle (C4-N9-Cl'-04') of the bound nucleotide is dependent on the site occupied by the base moiety of the inhibitor. The nucleotide takes syn orientation when the base occupies the primary binding site (conformers PI, P2 and P3), and anti orientation when the base occupies the subsite (conformers Sl, S2, S3, S4 and S5). However, there is no correlation between the ribose pucker and the site at which the base binds as observed for the glycosyl torsion angle. In both the low energy conformers P3 and S5, the ribose moiety adopts the C2' -endo pucker form. It can also be seen from Table I that eventhough the difference in the total energy of the conformers P3 and S5 is very small, the interaction energy is better in conformer P3 and the internal protein energy in S5. The most favoured conformers (P3 and S5) of the RNase T 1 - 2' -AMP complex have nearly the same total energy irrespective of whether adenine occupies the primary binding site (P) or the subsite (S), suggesting that two modes are probable. The phase angle which defines the puckering of the ribose and the glycosyl torsion angle of2'AMP in conformer S5 are close to that found from x-ray crystallographic studies [22)(Table I). In this conformer (Figure 1d) adenine occupies a site which is close to the Nl subsite [29] and forms one hydrogen bond each with the backbone carbonyl ofGly74 and the side chain guanidinium group of Arg77 (Table II) in addition to the stacking interaction with His92 imidazole. The phosphate group occupies the phosphate binding site. The hydrogen bonding scheme predicted for the S5 conformer agrees well with that proposed by x-ray crystallographic study [22). In the S1 conformer adenine occupies the N1 subsite [29) forming a hydrogen bond with the side chain hydroxyl group of Ser72. However, the stacking interaction between adenine and His92 imidazole is absent (Figure 1c) as are hydrogen bonding interactions between the phosphate moiety and the amino acid residues His40 and Glu58 (Table II). Consequently, this conformer is energetically less favourable than the
966
Balaji et a/.
Downloaded by [New York University] at 21:22 06 January 2015
conformer S5.It is interesting to note that when guanyl dinucleosidephosphates bind to RNase TI,the base on the 3’-terminal of guanosine favours an orientation similar to that of adenine of 2’-AMP in the S1 confor-mer perhaps due to additional steric constraints present in binding the dinucleoside phosphates [ 121. In the minimum energy conformer P3 ofthe RNase Tl-2’-AMPcomplex,the amino acid residues Tyr42 and PhelOO form a hydrophobic pocket for the base. The hydrogen bond between the Asn98 side chain amide group and Tyr45 hydroxyl group orients the phenyl ring of Tyr45 so as to maximise the stacking interaction with adenine (Figure lb). The orientation of adenine in this complex is similar to that of guanine in the RNase TI- 2’-GMPcomplex [111.Although the hydrophobic interactions between the active site residues and adenine is similar to that found in the RNase TI - 2’-GMP complex, the number of hydrogen bonds formed by adenine with the enzyme are less compared to guanine. The position and orientation of adenine in the P1 and P3 conformers are very similar. However, the ribose adopts C4’-exo and C2’-endo pucker forms respectively in these two conformers. Consequently, the interaction between Asn98 side chain and the ribose moiety present in the P3 conformer will not be possible in the P1 conformer. The backbone conformation of the dipeptide fragment Asn43-Asn44 in these two conformers is similar to that in RNase TI with no guanine bound [4]. Consequently the base forms one hydrogen bond each with the backbone carbonyl and amide groups of Asn43 in addition to a hydrogen bond with the side chain carboxyl group of Glu46 (Table 11; Figure lb). When energy minimisation is carried out starting with the backbone of this dipeptide fragment in the same conformation as that observed in the RNase TI - 2’-GMP complex [5], where it is flipped 140” with respect to the guanine-free enzyme [4], adenine is pushed out of the binding site due to electrostatic repulsion between the 6-amino group of adenine and the backbone amide ofAsn44. As a result, in conformer P2 adenine forms two weak hydrogen bonds with the side chains of His40 and An43 (Table 11) and may not have stacking/hydrophobic interactions with the enzyme (Figure la) leading to higher energy complex. Thus the present studies show that 2’-AMP favours to bind to RNase TI in essentially two modes with adenine occupying two different binding sites and phosphate occupying the same site. The two binding modes have nearly the same energy indicating that in solution, 2’-AMP binds to RNase TI in the two modes with nearly equal probability.The S5 conformer where adenine is stackingwith the His92 imidazole is, in the crystalline state, additionally stabilised by one intermolecular contact with a symmetry related RNase TI molecule, 05’...OE2 Glu102 [22]. However, we can not exclude that this mode of binding is the one that prevails also in solution.The hydrolysis ofthe phosphodiesterbonds adjacentto adenosine at higher enzymeconcentrationscan be explained only by the binding of adenosine at the primary site. Recentlythe structureof the RNase TI- 2’-AMP complexwas predicted by a free-energy perturbation study starting from the structure ofthe RNase TI- 2’-GMP complex by Hirono and Kollman [29]. The conformation predicted for the enzyme -inhibitor complex from this study does not agree with that of either the x-ray crystallographic studies [22] or any of the low energy conformers reported from the present study.
Binding of P‘-AMP to RNase T,
967
Table In Energy components for the interaction between the active site residues of RNase T, and the inhibitors 2’-AMP and 2‘-GMP (in kcal/mol)
-
RNase TI - 2’-GMP
RNase TI 2’-AMP Amino Acid
Primary site (P3)
Downloaded by [New York University] at 21:22 06 January 2015
VDW ELE Asn 36 Tyr 38 His 40 Tyr 42 Asn 43 Asn 44 Tyr 45 Glu 46 Glu 58 Ser 72 Gly 74 Arg 77 His 92 Asn 98 Asn 99 Phel00
-1.74 -1.12 -2.92 -2.85 -0.91 -0.73 -6.37 -0.58 -1.70 -0.02 -0.13 -0.20 -1.76 -1.98 -1.45 -5.89
- 1.30 - 3.83 -20.23 - 0.37 - 2.62 0.08 0.26 0.72 - 0.19 0.01 1.26 -24.24 -20.62 1.13 0.10 - 0.23
HB -0.02 -5.98 -1.16 -0.01 -3.54 -0.00 -0.18 -3.10 -6.04 0.00 0.00 -4.11 -1.07 -3.27 -0.02 -0.02
Subsite (S5) TOT VDW ELE
- 3.06 -10.93 -24.31 - 3.23 - 7.07 - 0.65 - 6.29 - 2.96 - 7.93 - 0.01 1.13 -28.55 -23.45 - 4.13 - 1.37 - 6.14
-2.66 -0.61 -1.54 -0.18 -0.03 -0.01 -0.54 -0.06 0.08 -0.05 0.59 -3.57 -5.88 -3.17 -0.77 -1.79
- 0.50 -3.45 -20.96 - 0.14 0.10 0.04 0.06 3.96 - 0.63 0.08 - 0.73 -19.84 -21.53 - 0.90 0.61 - 0.22
HB -3.59 -5.99 -1.09 0.00 0.00 0.00 -0.03 0.00 -6.05 0.00 -2.72 -1.81 -0.66 -2.86 0.00 0.00
Primary site TOT VDW ELE - 6.75 -10.05 -23.59 - 0.32 0.07 0.03 - 0.51 3.90 - 6.60 0.02 - 2.86 -25.22 -28.07 - 6.93 - 0.16 - 2.01
-1.04 -0.55 -1.91 -4.75 -3.23 -0.38 -4.21 0.98 -1.70 -0.02 -0.15 -1.51 -1.72 -1.65 -1.02 -3.57
HB
TOT
0.97 - 0.02 - 2.03 3.66 - 5.97 -10.18 -14.13 - 3.06 -19.10 - 0.81 - 0.05 - 5.61 - 3.20 - 1.10 - 7.53 - 2.21 - 1.11 - 3.70 - 1.55 - 1.10 - 6.86 - 8.62 - 6.15 -13.79 - 6.08 -11.77 -19.55 0.08 0.00 0.06 1.42 0.00 1.27 -25.26 - 2.29 -29.06 -20.80 - 1.02 -23.54 - 7.11 - 8.06 -16.82 0.54 - 0.01 - 0.49 - 0.05 0.00 - 3.62 -
-
The failure to predict the observed structure for the RNase T I -2’-AMP complex by Hirono and Kollman could be due to the consideration of an improper startingconformation for the backbone of the Asn43-Asn44 dipeptide fragment. In fact, earlier molecular dynamics studies carried out both in vacuum and aqueous media have indicated the flipping of the backbone ofAsn43-Asn44 dipeptide fragment [30].The present study shows that this dipeptide fragment favours a different conformation in the RNaseT, - 2’-AMP complex from that in the RNaseT, - 2’-GMP complex and the internal protein energies are also different [l 11. It is essential that these changes are also considered for analysis in any theoretical calculation [see for example Ref. 311. In view of this, the apparent agreement between the experimentally determined and theoretically calculated values of AAGbind(differencein free energy of binding) reported by Hirono and Kollman appears to be fortuitous since the change in internal protein (enzyme) energy upon inhibitor binding has been left out of their analyses. The contributions from the van der Waals, electrostatic and hydrogen bond interactions of all the active site amino acid residues in the RNase T, - 2’-AMP complex are tabulated in Table I11 for conformers P3 and S5.The contributions of these residues in the RNase T, - 2’-GMP (C2’-endo)complex [ 111 are also given for comparison. It can be seen from this table that the contributions of the individual amino acid residues constituting the phosphate binding site i.e., Tyr38, His40, Arg77 and His92 are approximately the same in all the three complexes. However, interaction between Glu58 and the phosphate moiety is better in the RNase TI - 2’-GMP complex compared to either of the RNase T, - 2’-AMP complexes.TheAsn98 - ribose interaction
968
Balaji et al.
Downloaded by [New York University] at 21:22 06 January 2015
observed in the RNase Tl-2’-GMP complex is absent in the RNase TI - 2‘-AMP complex. Among the amino acid residues constituting the base binding site, the Glu46 contribution is significantly less in both the P3 and S5 conformers of the RNase T I- 2’-AMPcomplex compared to the RNase T I- 2’-GMP complex.In the S5 conformer, there is a net electrostatic repulsion between Glu46 and 2’-AMP and in the P3 conformer, this repulsive interaction is slightly offset by the hydrogen bonding interaction with the 6-amino group of adenine. Among the other primary base binding site residues, contribution from PhelOO is more in the P3 conformer of the RNase TI - 2’-AMP complex compared to the RNase T, - 2’-GMP complex. In conclusion, the present study has shown that the adenosine-ribose moiety of 2‘AMP can bind to the primary (specific) or secondary (non-specific) substrate binding sites with about equal probability. The conformation about the glycosyl bond adopted by2‘-AMP when bound to the enzyme is found to be dependent on the site occupied by adenine, syn in the primary and anti in the secondary site. An analysis of the interactions of the active site amino acid residues with 2’-AMP and 2’-GMP shows that Glu46, Glu58 and Am98 stabilise the enzyme complex with 2’-GMP more than that with 2‘-AMP.These results are in agreement with x-ray, competitive binding and spectroscopic data. References and Footnotes 1. K. Takahashi,J. Biochem. (Tokyo) 98,815 (1985). 2. U. Heinemann and U. Hahn. In Protein-Nucleic acid interaction, (W.Saenger and U.Heinemann, eds.), Topics in molecular and structural biology, Vol. 10. Macmillan Press Ltd., London, pp 111 (1989). 3. C. Hill. In Structure and chemistry qfribonucleases,Proceedings of the first international meeting, (kPavlovsky and K.Polyakov. eds.), Mowcow, 17 1 (1988). 4. D. Kostrewa, H.-W. Choe, U. Heinemann and W. Saenger, Biochemhtry 28,7592 (1989). 5. R. Ami, U. Heinemann. R. Tokuoka and W. Saenger. J. Biol. Chem. 263. 15358 (1988). 6. S. Sugio, T. Amisaki, H. Ohishi and K-I. Tomita,J. Biochem. 103,354 (1988). 7. J. Keopke. M. Maslowska. U. Heinemann and W. Saenger,J. Mol. Biol. 206,475 (1989). 8. E. Hoffmann, J. Schmidt, J. Simor and H. Rueterjans. Nucleosides and Nucleotides 7,757 (1988). 9. T. Hakoshima, S. Toda. S. Sugio, K.4. Tomita. S. Nishikawa, H. Morioka, K Fuchimura. T. Kimura, S. Uesugi, E. Ohtsuka and M. Ikehara, Prot. E n s . 2,55 (1988). 10. H.-P. Grunert. k Zouni. M. Beineke, R. Quaas, Y. Georgalis, W. Saenger and U. Hahn, Eur. J. Biochem. 197.203 (1991). 11. P.V. Balaji, W. Saenger and V.S.R. Rao, Curr.Sci. 60,363 (1991). 12. P.V. Balaji, W. Saenger and V.S.R. Rao,J. Biomol. Str. Dyn. 9,215 (1991). 13. P.V. Balaji and V.S.R. Rao, Znd. J. Biochem. Biophys. 28.358 (1991). 14. F. Egami. T. Oshima and T. Uchida, Mol. Biol. Biochem. Biophys. 32,250 (1980). 15. K Takahashi and S. Moore, The Enzymes XV, 435 (1980). 16. K Sat0 and F. Egami,J. Biochem. (Tokyo) 44,753 (1957). 17. AM.Michelson and C. Monny, Biochim. Biophys. Acta 166,294 (1968). 18. M. 1rie.J. Biochem. (Tokyo) 58,599 (1965). 19. FA. Neelon, M. Molinaro. H. Ishikura, L.B. Sheiner and G.L. Cantoni, J. Biol. Chem. 242, 4515 (1967). 20. R.C. Warrington, Biochim. Biophys. Acta 353.63 (1974). 21. M. Irie,J. Biochem. (Tokyo) 56.495 (1964). 22. J. Ding, G. Koellner. H.-P. Grunert and W. Saenger, . I Biol. Chem., (in press), (1991). 23. Lenz, A, Cordes, F.. Heinemann, U. and Saenger, W., J. Biol. Chem. 266,7661 (1991). 24. E.N.Baker and R.E. Hubbard. h o g . Biophys. Mol. Biol. 44,97 (1984). 25. R. Taylor and 0. Kennard, J. Mol. Str. 78, 1 (1982).
Binding of 2'-AMP to RNase T1 26. 27. 28. 29. 30.
969
W. Saenger., "Principles of nucleic acid structure", Springer Verlag, First edition, Berlin (1982). P.V. Balaji, W. Saenger and V.S.R Rao, Biopoly. 30,257 (1990). P.V. Balaji and V.S.R Rao,J. Biomolec. Str. Funct. 9, (in press) (1992). S. Hirano and P.A Kollman,J. Mol. Bioi. 212, 197 (1990). AD. MacKerell, L. Nilsson, R Rigler, U. Heinemann and W. Saenger, Proteins: Structure, Function and Genetics 6, 20 (1989). 31. C.L. Brooks III, M. Karplus and B.M. Pettitt. In "Proteins: A theoretical perspective of dynamics, structure, and thermodynamics", Adv. Chern. Phys. 71, 60 ( 1988).
Downloaded by [New York University] at 21:22 06 January 2015
Date Received: October 11, 1991
Communicated by the Editor Dina Moras
