A set of nearest neighbor parameters for predicting the enthalpy change of RNA secondary structure formation

Abstract

A complete set of nearest neighbor parameters to predict the enthalpy change of RNA secondary structure formation was derived. These parameters can be used with available free energy nearest neighbor parameters to extend the secondary structure prediction of RNA sequences to temperatures other than 37 degrees C. The parameters were tested by predicting the secondary structures of sequences with known secondary structure that are from organisms with known optimal growth temperatures. Compared with the previous set of enthalpy nearest neighbor parameters, the sensitivity of base pair prediction improved from 65.2 to 68.9% at optimal growth temperatures ranging from 10 to 60 degrees C. Base pair probabilities were predicted with a partition function and the positive predictive value of structure prediction is 90.4% when considering the base pairs in the lowest free energy structure with pairing probability of 0.99 or above. Moreover, a strong correlation is found between the predicted melting temperatures of RNA sequences and the optimal growth temperatures of the host organism. This indicates that organisms that live at higher temperatures have evolved RNA sequences with higher melting temperatures.

Figure 1

(A) Free energy difference of RNA duplex CCGGUp. ΔG° (dashed line) was derived from Equation 3, where enthalpy and entropy were averaged from the optical melting curve fits, assuming that they were independent of the temperature. $\Delta G_{\text{T}}^{\text{o}}$ (solid line) was calculated from Equations 1–3, where the heat capacity was accounted. (B) Free energy difference is ΔΔG° = $\Delta G_{\text{T}}^{\text{o}}-\Delta G^{\text{o}}$ (62).

Figure 2

Improvement of prediction at optimal growth temperatures. The sequences are those from mesophiles (optimal growth temperature from 10 to 60°C) without organisms with optimal growth at 37°C. The lowest free energy secondary structures were predicted at the organims' optimal growth temperatures using two models. The previous model and parameters are those of Serra and Turner (24), which are widely used. The improved prediction uses the model and parameters presented in this work. The small and large subunits of rRNA sequences are divided into domains of <700 nt. The total sensitivity is the average of sensitivities of different types of RNA.

Figure 3

PPV for optimal structure and base pairs with different pairing probabilities. PPV equals the number predicted base pairs in that are in the known structure divided by total number of predicted base pairs. Pairs in the optimal structures are grouped by different thresholds of pairing probabilities. The pairing probabilities were calculated with a partition function calculation (22) at organisms' optimal growth temperatures, using the model and parameters presented in Materials and Methods. The small and large subunits of rRNA sequences are divided into domains of <700 nt. The sequences of different type of RNA are those from mesophiles (living from 10 to 60°C) without organisms living at 37°C.

Figure 4

Secondary structure prediction of Saccharomyces cerevisiae tRNA (RM4000) at optimal growth temperature (30°C) (B) and at 37°C (C) with the presented nearest neighbor parameters. Base pairs in the original structure (A) are derived from the comparative analysis database (–49,58,59). Structures are also color annotated to indicate predicted base pair probabilities (P_bp) for each helix: red, P_bp ≥ 0.95; yellow, 0.95 > P_bp ≥ 0.7; green, 0.7 > P_bp ≥ 0.3; blue, 0.3 > P_bp. The structures were drawn with XRNA () and Adobe Illustrator.

Figure 5

Experimental (Supplementary Data) (–31) versus predicted (T_m = ΔH°/ΔS° − 273.15) melting temperatures of hairpin stem–loop structures. The line shows the ideal location of points, predicted T_m = measured T_m. The root mean squared deviation (r.m.s.d.) of prediction compared to experiment is 5.86°C. The new enthalpy parameters provide improved T_m prediction compared to the previous compilation of parameters (24), which have an r.m.s.d. of 7.58°C as compared to experiment for this dataset.

Figure 6

Relationships of melting temperatures, nucleotide contents and optimal growth temperatures of different types of RNA in different organisms with optimal growth temperature from 10 to 90°C: (A) Predicted melting temperature; (B) G–C pair content; (C) G content; and (D) U content versus optimal growth temperature. Melting temperatures are predicted for different types of RNA sequences from comparative analysis databases (–49,58,59) with a two-state transition assumption.