Origin of mutational effects at the C3 and G8 positions on hammerhead ribozyme catalysis from molecular dynamics simulations

A detailed understanding of the structure-function relationships in the hammerhead ribozyme^{1, 2} will aid in the understanding of other cellular RNA catalysts such as the ribosome,³ and have potential impact for rational drug design,⁴ and the development of new biotechnology.⁵ Much effort has been devoted to understand the details of the hammerhead ribozyme mechanism both experimentally and theoretically.^{1, 2, 6–8} Mutation experiments suggested that two conserved residues, G8 and C3, are critical for catalytic activity.⁷ The specific roles of these two residues, however, were not clear from the minimal sequence structures,^9–11 and it was not until the recent full length hammerhead structure¹² was solved that a structural basis emerged that could begin to reconcile these mutational effects and other aspects of the mechanistic debate.

This work reports a series of ten molecular dynamics (MD) simulations of the native and mutated full length hammerhead ribozymes in the reactant state and in an activated precursor state (2’OH deprotonated). Mutants simulations include the C3U, G8A, and G8I single mutants, and a C3U/G8A double-mutant that exhibits an experimental rescue effect.¹⁴ For each simulation, 60 ns of MD simulation was performed to ensure sufficient time for structural relaxation induced by the mutations. The results provide atomic level details into the origin of the observed mutation effects, and support a mechanism where the 2’OH of G8 acts as a general acid catalyst that is held in position through Watson-Crick hydrogen bonding between G8 and C3.

Simulations were performed with NAMD version 2.6¹⁵ using the Cornell et al.¹⁶ all-atom nucleic acid force field (parm94) provided in AM-BER9,¹⁷ and the TIP3P water model.¹⁸ The initial structure was taken from the 2.0 Å crystal structure with Mn²⁺ ions and solvent resolved (PDB: 2OEU) with Mn²⁺ ions replaced by native Mg²⁺ ions in the simulations. A Mg²⁺ implicated in catalysis was bound to A9:O_2P and G10.1:N₇ (Figure 1) in accord with available experimental data.^{13, 19}

Figure 1.

Active site of the full length hammerhead RNA using the canonical minimal sequence numbering scheme described in references 9 and 13.

Following 10 ns of ion/solvent equilibration, simulations were performed at 298 K and 1 atm in a cubic cell with PME²⁰ electrostatics in the presence of ~11,400 water molecules and 0.14 M NaCl, and carried out to 60 ns. Key active site structural parameters are provided in Table 1, and representative hydrogen-bond base pairing at the C3-G8 positions are shown in Figure 2. In addition a set of control simulations were performed on a benign U7C mutation, and of the wild type with the active site Mg²⁺ ion removed (see Supporting Information). An assumption herein is that the mutated sequences fold to a native-like structure. This assumption may not be valid as the mutations might affect the folding pathway, or preferentially stabilize inactive conformations.²¹

Table 1.

Characterization of the active site structure and fluctuations

	WT	C3U	G8A	C3U/G8A	G8I
d0	3.98(40)	7.80(66)	4.22(34)	3.94(39)	4.27(59)
rNu	3.98(34)	4.16(14)	3.2(10)	3.59(48)	3.77(44)
θinl	128.2(116)	127.1(67)	156.0(70)	141.3(169)	131.6(152)
F*	0.29(15)	0.23(4)	0.68(9)	0.49(23)	0.37(20)
r1	2.00(17)	–	–	–	–
θ1	162.9(89)	–	–	–	–
r2	2.01(10)	2.25(26)	–	2.04(21)	2.01(14)
θ2	160.9(90)	157.0(127)	–	160.7(104)	162.0(88)
r3	1.88(12)	1.90(13)	2.13(39)	2.06(22)	1.98(19)
θ3	164.2(85)	163.1(88)	149.4(151)	161.2(101)	163.6(90)
rNN†	2.97(9)	3.61(21)	5.33(66)	3.00(19)	2.98(13)
rHB	2.10(26)	3.51(66)	2.01(16)	2.38(46)	2.05(26)
θHB	152.1(138)	121.2(155)	164.2(85)	147.1(181)	155.6(124)
rHA	2.83(44)	7.82(58)	2.97(33)	3.27(64)	2.83(86)
θHA	118.2(168)	45.2(164)	119.3(150)	130.6(198)	118.5(344)
%‡	24.4	0.0	31.2	13.2	38.4
	d-WT§	d-C3U	d-G8A	d-C3U/G8A	d-G8I
d0	2.96(12)	2.94(13)	2.95(12)	2.93(12)	2.93(13)
rNu	3.54(17)	3.67(16)	3.80(18)	3.62(16)	3.58(18)
θinl	158.6(78)	154.2(74)	146.6(88)	155.5(73)	156.6(73)
F*	0.52(10)	0.45(8)	0.38(8)	0.48(8)	0.50(9)
r1	1.90(14)	–	–	–	–
θ1	160.1(103)	–	–	–	–
r2	1.98(11)	1.92(15)	–	2.08(17)	1.95(11)
θ2	163.1(84)	160.5(109)	–	160.8(101)	164.3(82)
r3	2.02(20)	1.93(15)	2.82(44)	1.96(17)	2.11(22)
θ3	163.3(93)	161.9(94)	106.9(116)	160.3(105)	163.2(95)
rNN†	2.96(10)	3.75(20)	6.58(46)	3.04(15)	2.93(11)
rHB	2.16(55)	1.94(14)	1.89(12)	1.95(14)	1.99(16)
θHB	152.0(151)	155.7(96)	158.5(90)	155.5(93)	156.0(99)
rHA	2.82(95)	2.99(99)	4.10(63)	2.61(91)	2.47(52)
θHA	126(398)	108.7(397)	82.8(215)	130.6(385)	138.6(286)
%‡	62.8	43.6	0.8	68.8	77.2

Analysis was performed over the last 25 ns (10 ps sampling). Distance and angles (Figure 1) are in Å and degrees, respectively. Standard deviations (SD) are listed in parenthesis divided by the decimal precision of the average (e.g., if the average is reported to two digits of decimal precision, the SD is divided by 0.01).

^*In-line fitness index.²⁴

^†The N₃…N₁ distance between nucleobase in the 3 and 8 position.

^‡The hydrogen bond contact percentage of the general acid with the leaving group defined as the percentage of the snapshots in which r_HA ≤ 3.0 Å and θ_HA ≥ 120°.

^§The notation ”d-” denotes the activated precursor state simulations having the C17:O₂′ deprotonated.

Figure 2.

Representative hydrogen-bonding of the C3:G8 base pair observed from the wild type, C3U, G8A, G8I and C3U/G8A and mutant simulations. Experimentally observed relative catalytic rates of mutant versus wild type minimal sequence ribozymes (k_mut/k_wt) are shown in parentheses (C3U from Ref. 22, G8A from Ref. 23, C3U/G8A from Ref. 14, and G8I from Ref. 7). The relative rates for the minimal sequence mutants may differ for the full length sequence.

Mg²⁺ can migrate to a bridging position in the activated precursor

In all reactant-state simulations, the Mg²⁺ stays near A9:O_2P and G10.1:N₇ and Table 1 shows that the overall active site integrity was preserved for all simulations but C3U, e.g., the distances between A9:O_2P and C1.1:O_2P are within reasonable deviation from the crystal value 4.3 Å. In all activated precursor state (deprotonated C17:O_2′) simulations, after few hundred ps, the Mg²⁺ migrates into a bridging position between A9:O_2P and C1.1:O_2P, and reduces the distance (d₀ in Table 1) by 1 Å relative to the reactant-state simulations. The C17:O_2′ has significant in-line fitness for nucleophilic attack on C1.1:P in all simulations. In all simulations except G8A, the G8:O_O2′ is hydrogen bonded to the leaving group, C1.1:_O5′ and positioned to act as the general acid.

C3U mutation disrupts the active site in the reactant

The C3U mutation reduces the catalytic rate by a factor ~3 × 10⁻⁴.²² The C3U mutation disrupts the normal Watson-Crick hydrogen bonding with G8 (Figure 2), causing a base shift that disrupts the active site structure in the reactant state. The distance between the A9 and scissile phosphate, d₀, increases more than 3.5 Å and breaks key hydrogen bonds between the O_2′nucleophile of C17 and N₁ of G12 (the implicated general base), and between the O_5′ leaving group of C1.1 and H_2′ of G8 (the implicated general acid). These perturbations of the active site in the reactant state would prevent activation of the nucleophile and progress toward the transition state. Experimental evidence shows that C3U indeed reduces the reaction constant by more than three orders of magnitude.²²

G8A mutation disrupts the positioning of G8:O_2′ as general acid in the activated precursor

The G8A mutation reduces the catalytic rate by a factor ≤ 0.004.²³ Simulation results indicate the G8A mutation considerably weakens the base pair with C3 with only one weak hydrogen bond that remains intact (Figure 2). In the reactant state simulation, G8A does not appear to dramatically alter the active site contacts relative to the wild-type simulation, with the exception of the A8:N₁…C3:N₃ distance which increases due to a shift in the hydrogen bond pattern (Table 1). In the activated precursor state, however, the hydrogen bond positioning between G8:H_2′ and C1.1:O_5′ is significantly disrupted relative to the wild-type simulation. The G8A mutation shifts the conserved 2’OH of G8 away from the ideal general acid position as in the wild-type simulation, and can possibly block the general acid step of the reaction. Mutation of G8 to 2-aminopurine (AP)^{7, 25} or to 2,6-diaminopurine (diAP),²⁵ which are expected to have similarly weakened hydrogen bonding as the G8A mutation, are observed to reduce the reaction rate by over three orders of magnitude.

G8I and C3U/G8A mutations are relatively benign

Whereas the relatively isosteric C3U and G8A mutations lead to considerably reduced catalytic rates, the G8I^{7, 25} and C3U/G8A¹⁴ mutations affect the rate by less than an order of magnitude. The C3U/G8A double-mutation and G8I single mutation simulations indicate that the hydrogen bond network retains the overall base positions relative to the wild-type simulation, although the number of base pair hydrogen-bonds is reduced from 3 to 2. Simulations results for C3U/G8A and G8I both in the reactant state and the precursor states suggest that these two mutations do not significantly alter the overall structure or any of the active site indexes that would affect activity relative to the wild-type simulation (Table 1).

Structural deviations that give rise to mutation effects can occur at different stages along the reaction path

In the present study, mutations may cause hindrance of the general base or acid steps, and/or otherwise compromise the active site structure. Although the canonical Watson-Crick hydrogen bond network is altered significantly in both C3U and G8A mutations, the simulations suggest that the origin of the mutational effect on the ribozyme kinetics can occur at different stages along the reaction path. In the reactant state, the Mg²⁺ ion is bound between the G10.1:N₇ and A9:O_P2. The large base-pair shift that occurs in the C3U mutation simulation results in compromise of the active site structure, including the loss of interactions between the proposed general base and nucleophile. In the activated precursor state, the Mg²⁺ ion occupies a bridging position between A9:O_P2 and the scissile phosphate. The G8A mutation, which is very weakly hydrogen bonded, does not sustain a catalytically viable position of the general acid.

Hydrogen bonding between nucleobases in the 3 and 8 positions is necessary but not sufficient to preserve active site structural integrity

The G8I and C3U/G8A mutations that largely preserve a stable base-pair hydrogen bonded scaffold lead to relatively benign mutations. A C3G/G8C base-pair switch mutation that preserves hydrogen bonded base pairing partially rescues activity relative to the single mutations, although still reduces activity 150–200 fold.^{12, 21} Recent analysis of all base-pair mutations indicate considerable variation in activity, but all of the non-native mutations at this position are considerably less active. ²¹ The present simulation results offer the prediction that whereas both C3U²² and G8diAP²⁵ single mutations are observed experimentally to reduce catalytic activity by several orders of magnitude, a correlated C3U/G8diAP double mutation, which retains base-pair hydrogen bonding, should exhibit a partial rescue effect in the hammerhead ribozyme.