RNA Complexes

1436

IEEE/ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS,

VOL. 15,

NO. 5, SEPTEMBER/OCTOBER 2018

Identification and Analysis of Key Residues in Protein–RNA Complexes A. Kulandaisamy, Ambuj Srivastava, Pradeep Kumar, R. Nagarajan , S. Binny Priya, and M. Michael Gromiha Abstract—Protein–RNA complexes play important roles in various biological processes. The functions of protein–RNA complexes are dictated by their interactions, binding, stability, and affinity. In this work, we have identified the key residues (KRs), which are involved in both stability and binding. We found that 42 percent of considered proteins share common binding and stabilizing residues, whereas these residues are distinct in 58 percent of the proteins. Overall, 5 percent of stabilizing and 3 percent of binding residues serve as key residues. These residues are enriched with the combination of polar, charged, aliphatic, and aromatic residues. Analysis on subclasses of protein–RNA complexes based on protein structural class, function and RNA type showed that regulatory proteins, and complexes with single stranded RNA and rRNA have appreciable number of key residues. Specifically, Arg, Tyr, and Thr are preferred in most of the subclasses of protein–RNA complexes. In addition, residues with similar chemical behavior have different preferences to be KRs, such that Arg, Tyr, Val, and Thr are preferred over Lys, Trp, Ile, and Ser, respectively. Atomic level contacts revealed that charged and polar–nonpolar contacts are dominant in enzymes, polar in structural, and nonpolar in regulatory proteins. On the other hand, polar– nonpolar contacts are enriched in all these classes of protein–RNA complexes. Further, the influence of sequence and structural features such as conservation score, surrounding hydrophobicity, solvent accessibility, secondary structure, and long-range order in key residues are also discussed. We envisage that the present study provides insights to understand the structural and functional aspects of protein–RNA complexes. Index Terms—Protein–RNA complexes, binding site residues, folding and stability, key residues, contacts, propensity

Ç 1

INTRODUCTION

P

ROTEIN–RNA

interactions are vital for several cellular processes, which include translation, post-translational modifications, RNA transfer and gene regulation. ProteinRNA interactions are mediated by shape complementarity at the interface, electrostatic interaction between nucleotide backbone and side chain of amino acids, hydrogen bonds, cation-p interactions and van der Waals interactions [1], [2], [3], [4], [5]. Impairment in these interactions can result in several diseases such as neurological disorders, metabolic diseases, autoimmune pathologies and various types of cancers [6], [7], [8], [9], [10]. For example, structural changes upon binding in HuR, a RNA-binding protein (RBP) lead to several types of human cancer [11]. The loss of the DAZ gene which codes for a RBP in human causes several diseases related to spermatogenesis such as azoospermia and oligospermia [12], [13]. Investigations on factors influencing the stability and binding affinity of protein–RNA complexes provide deep insights to understand the recognition mechanism, structure-function relationship and the implications of diseases [5], [14], [15].




The authors are with the Department of Biotechnology, Bhupat and Jyoti Mehta School of BioSciences, Indian Institute of Technology Madras, Chennai, Tamilnadu 600 036, India. E-mail: {kulandai28, ambuj.88.in, i.pradeep92, nagabioinfostar, binny.bioinfo}@gmail.com, [email protected].

Manuscript received 28 July 2017; revised 3 Dec. 2017; accepted 21 Dec. 2017. Date of publication 14 May 2018; date of current version 5 Oct. 2018. (Corresponding author: M. Michael Gromiha.) For information on obtaining reprints of this article, please send e-mail to: [email protected], and reference the Digital Object Identifier below. Digital Object Identifier no. 10.1109/TCBB.2018.2834387

RNA binding proteins interact with RNA using different recognition motifs such as the K-homology domain (KH), zinc fingers and arginine-rich motifs [16], [17], [18]. The recognition mechanism of protein–RNA interactions has been extensively studied using binding propensity, amino acid– nucleotide pair preference, binding motifs and non-covalent interactions [1], [19], [20]. In addition, single amino acid mutations in RBP as well as base mutation in RNA at the interface can alter the structure, interaction pattern and binding affinity, leading to functional modification and some of them even cause diseases [10], [21], [22], [23]. The binding affinity of protein–RNA interactions are investigated using in vitro, in vivo and in silico methods. Experimentally, high throughput techniques such as HITSCLIP, CLIP-seq, RIP-seq and PAR-CLIP have been developed to determine the protein–RNA interactions in vivo on a genome-wide scale [24]. These information are stored in databases such as RAID and doRiNA [25], [26]. On the other hand in vitro techniques, RNAcompete [27], SEQRS [28], RNA-Map, EMSA, yeast one-hybrid assay and FRET [24] provide valuable information on protein–RNA binding affinity and specificity. Computational techniques are also widely used to understand the recognition mechanism of protein–RNA complexes by identifying the binding site residues and estimating their binding affinities [29], [30], [31]. Structure based prediction methods are mainly based on interface residue propensity [32], electrostatics, evolution and geometry [4], topology [33], [34] and energy functions [35], [36]. Sequence based methods utilize side chain pKa, hydrophobicity [37], PSSM profiles

1545-5963 ß 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See ht_tp://www.ieee.org/publications_standards/publications/rights/index.html for more information.

KULANDAISAMY ET AL.: IDENTIFICATION AND ANALYSIS OF KEY RESIDUES IN PROTEIN–RNA COMPLEXES

[38], [39], [40] solvent accessibility, amino acid properties [41] and motifs for predicting the binding sites and affinity of protein–RNA complexes [42], [43]. Recently, Barik et al. (2016) developed a Random Forest method to predict the change in binding energy upon mutation using structural and physiochemical features [44]. Tuvshinjargal et al. (2016) developed a method to predict both protein-binding nucleotides and RNA-binding residues from sequence data of the interaction partners [45]. Several investigations have been carried out to gain insights on the interplay between binding and stability of protein–protein complexes [46], [47]. However, this sort of bridging in protein–RNA complexes is poorly understood. The information available in thermodynamic database for proteins and mutants (ProTherm) and protein–nucleic acid interactions (ProNIT) has been effectively used to understand the stability and affinity of proteins and protein– nucleic acid complexes, respectively upon mutations [48], [49]. Reyes and Kollman (2000) and Zhao et al. (2006) analyzed the effect of amino acid mutations on the stability and binding affinity of U1A protein–RNA complexes using molecular dynamics simulations [50], [51]. Mittelberger et al. (2015) investigated the stability and affinity of a 34nucleotide RNA aptamer with interleukin 6 receptor and showed that the aptamers increase the stability without compromising its affinity for the target protein [52]. Recently, Pires et al. (2016) developed an integrated computational approach for evaluating the effects of mutations on protein stability and RNA binding affinity [53]. In spite of these studies, the importance of amino acid residues involved in binding and stability has not yet been completely explored. In the present work, we have identified the stabilizing (SRs), binding (BRs) and key residues (KRs) in 174 protein–RNA complexes using computational approaches. Among them, we found that in 103 complexes the stabilizing and binding residues are distinct and no key residues are identified. However in other complexes, 5 percent of stabilizing and 3 percent of binding residues are identified as key residues, and are dominated by polar, charged, aliphatic and aromatic residues. Interestingly, residues with similar chemical behavior have different preferences to serve as key residues; specifically Arg, Tyr, Val and Thr are dominant over Lys, Trp, Ile and Ser, respectively. Atomic level contacts show the enrichment of polar–nonpolar contacts in all classes of protein–RNA complexes whereas other contacts such as polar–polar, nonpolar–nonpolar depend on the type of the complex based on structure and function. Further, the influence of sequence and structure based features will be discussed.

2

MATERIALS AND METHODS

2.1 Dataset We have collected a non-redundant dataset of 174 protein– RNA complexes with a cutoff of 25 percent sequence identity from the information available in literature [31]. In addition, we have derived five subclasses of complexes based on: i) protein structural class, ii) RNA strand, iii) RNA conformation, iv) types of RNA and v) protein function. The complete information about different classifications and datasets are given in Supplementary information, which can be found on

1437

the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TCBB.2018.2834387.

2.2 Identification of Stabilizing, Binding and Key Residues Stabilizing residues (SRs) are identified using the procedure described earlier [54], [55]. In this method, a residue is identified as stabilizing if it satisfies the following conditions: surrounding hydrophobicity (Hp > 16); long-range order (LRO > 0.01); conservation score (CS > 6) and stabilization center (SC > 0). We used the tool, SRide (http://sride. enzim.hu/) for obtaining the stabilizing residues. Any residue/nucleotide is said to be at the interface if any of its heavy atoms is contacting with any heavy atom in  the partner chain. We used a distance cutoff of  3.5 A to identify the binding site residues [31]. The residues, which are common to both binding and stabilizing are termed as key residues [56]. 2.3 Sequence and Structure Based Parameters of SRs, BRs and KRs SRs, BRs and KRs are compared using sequence and structural features such as secondary structure, solvent accessibility, long-range order (LRO), surrounding hydrophobicity and conservation score. Secondary structure information and solvent accessibility are calculated using DSSP [57], which provides both information simultaenously and it is widely used in the literature. LRO is computed by comparing the contacts in primary and tertiary structures of a protein [58]. It is given byotherwise, N X nij ; nij ¼ 1; if ji  jj > 12; nij ¼ 0 otherwise; (1) LROi ¼ N j¼1 where, i and j are contacting residues. Surrounding hydrophobicity of a residue is computed using the experimental hydrophobicity of its surrounding  residues within a limit of 8 A [59]. HP ðiÞ ¼

20 X

nij hj ;

(2)

j¼1

where, nij is the total number of surrounding residues of type j around the ith residue of the protein; N is the total number of residues; and hj is the experimental hydrophobic index of residue type j in kcal/mol [60], [61]. To calculate the conservation score, the primary sequence of a PDB chain is aligned with homologous sequences present in Swiss-Prot database. The Consurf program is used to identify conserved residues in a protein sequence [62]. Furthermore, we have analyzed the distribution of contacting atoms between protein and RNA based on their chemical behavior: polar–polar, nonpolar–nonpolar, polar– nonpolar and charged–charged contacts.

2.4 Frequency of Occurrence of Amino Acid Residues The frequency of occurrence of SRs, BRs and KRs is computed using the equation FreqðiÞ ¼ nðiÞ=N; (3) where, n(i) is the number of amino acid residues of type i, and N is the total number of residues; i varies from 1 to 20.

1438

IEEE/ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS,

VOL. 15,

NO. 5, SEPTEMBER/OCTOBER 2018

Fig. 1. Venn diagram showing the number of stabilizing, binding, and key residues in non–redundant dataset of 174 complexes.

2.5

Occurrence of Disorder to Order Transition (DOT) Regions DOT regions are obtained by comparing the disorder residues in a free protein and its respective protein–RNA complex. The free proteins corresponding to each protein–RNA complex have been identified using sequence information and performing BLAST [63] search against PDB database with >99 percent sequence identity. Disorder to order transition regions are obtained by comparing the missing residues in a free proteins by locating “REMARK 450” in PDB file and ordered structures in corresponding protein–RNA complexes. DOT regions having at least 3 continuous residues are considered for the analysis.

3

RESULTS AND DISCUSSION

3.1 Key Residues in Protein–RNA Complexes In a non-redundant dataset of 174 protein–RNA complexes, we have identified the stabilizing, binding and key residues and the results are presented in Fig. 1. Among 43,441 residues in all complexes, 2,653, 4,152 and 135 residues are stabilizing, binding and key residues, respectively. This reveals that only few residues play a dual role of stabilizing and binding. Interestingly, the number of BRs is more than SRs in the considered set of complexes. However, we observed an opposite trend in PPIs that SRs are more than BRs [56]. Among SRs, 5 percent residues act as KRs, whereas it is 3 percent in the case of BRs. The statistical analysis showed that these observations are significant (p-value < 2:2  1016 ). It is noteworthy that only 73 out of 174 complexes contain KRs and the number of KRs varies from 1 to 6. The maximum of 6 KRs is present in HuD protein complex (1FXL) and the SRs, BRs and KRs are shown in Fig. 2. 3.2

Key Residues in Protein–RNA Complexes Based on Structural Classes and Protein Function Analysis of key residues in different structural classes (all-a, all-b, a þ b, a=b and others) showed that all-b class complexes have the highest number of SRs (8 percent) and BRs (12 percent). This observation might be due to the fact that b-strands have more stabilizing residues than other secondary structures [54]. The analysis of experimental data in ProTherm database supports this result that single amino acid substitution in b-strands destabilizes the structure compared with mutations in a-helices and coil regions. Further,

Fig. 2. Typical example showing the structure of HuD protein complex (1FXL). The stabilizing, binding, and key residues are marked in green, cyan and red spheres, respectively.

the residues in b-strands are enriched in hydrophobicity [64] and they have more number of long-range contacts [58]. These factors attribute to the enrichment of SRs in all-b proteins. The preference of BRs in all-b proteins might be due to the fact that these residues are involved in the formation of pocket for RNA binding [2]. On the other hand, all-a class has the lowest number of SRs (3 percent), BRs (9 percent) and KRs (0.2 percent) and the data are statistically significant (p < 1:96  108 ). In a þ b class, 8 percent of SRs and 4 percent of BRs act as KRs. Moreover, it has been observed that SRs are more likely to be KRs than BRs in all structural classes (Table 1). Based on protein function, we classified the complexes into enzymes, regulatory and structural proteins, and the SRs and BRs vary from 5 to 8 and 6 to 21 percent, respectively (Table 1). Among them, 3 to 8 percent of SRs and 3 to 5 percent of BRs act as key residues. The enzyme and regulatory complexes have similar contributions from both SRs and BRs (3 percent for enzymes and 6 percent for regulatory proteins). However in structural proteins, more number of SRs (8 percent) act as KRs compared to BRs (3 percent).

3.3 Key Residues in Protein–RNA Complexes Based on Strand, Conformation, and Types of RNA Protein–RNA complexes are classified into two groups based on the type of RNA strand: single and double. We observed that proteins bound with single stranded RNA have more than twice the percentage of KRs than complexes with double stranded RNA. The distribution of SRs and BRs in A, U and other conformation varies between 5 to 6 and 6 to 13 percent, respectively. In U and other (T and RH) conformations approximately 3 percent of BRs act as KRs. In U conformation 6 percent of SRs act as KRs, whereas in A and

KULANDAISAMY ET AL.: IDENTIFICATION AND ANALYSIS OF KEY RESIDUES IN PROTEIN–RNA COMPLEXES

1439

TABLE 1 Stabilizing, Binding, and Key Residues in Protein–RNA Complexes Type

% of SR

% of BR

% of

KR

% of KR in SR

% of KR in BR

Total no. of residues

Protein Structure

All–a (27) All–b (29) a/b (27) aþb (49) Others (11)

3.37 (248) 7.58 (605) 5.71 (556) 6.23 (660) 2.40 (61)

8.78 (646) 11.78 (941) 9.15 (890) 11.76 (1246) 10.25 (261)

0.16 (12) 0.46 (37) 0.25 (24) 0.48 (51) 0.18 (3)

4.84 (12) 6.12 (37) 4.32 (24) 7.73 (51) 4.92 (3)

1.86 (12) 3.93 (37) 2.7 (24) 4.09 (51) 1.15 (3)

7358 7986 9729 10597 2546

Protein Function

Enzyme (58) Regulatory (22) Structure (74)

5.38 (1215) 7.96 (287) 6.64 (721)

5.60 (1266) 8.71 (314) 20.70 (2248)

0.15 (33) 0.47 (17) 0.54 (59)

2.72 (33) 5.92 (17) 8.18 (59)

2.61 (33) 5.41 (17) 2.62 (59)

22595 3606 10860

RNA strand

Double (29) Single (156)

4.73 (318) 5.97 (2133)

8.11 (546) 11.61 (4145)

0.13 (9) 0.33 (118)

2.83 (9) 5.53 (118)

1.65 (9) 2.85 (118)

6729 35710

RNA conformation

A (25) U (140) Others (19)

4.66 (261) 6.13 (1734) 5.29 (450)

8.68 (486) 13.11 (3707) 5.80 (493)

0.16 (9) 0.37 (104) 0.16 (14)

3.45 (9) 6.00 (104) 3.11 (14)

1.85 (9) 2.81 (104) 2.84 (14)

5597 28278 8502

Type of RNA

rRNA (54) tRNA (28) Others (28)

6.56 (498) 5.66 (700) 5.74 (394)

26.17 (1985) 6.53 (807) 6.06 (416)

0.78 (59) 0.22 (27) 0.30 (21)

11.85 (59) 3.86 (27) 5.33 (21)

2.97 (59) 3.35 (27) 6.06 (21)

7586 12358 6862

Non redundant (174)

6.11 (2653)

9.56 (4152)

0.31 (135)

5.09 (135)

3.25 (135)

43441

Datasets

Non red

other classes half of them (3 percent) serve as KRs. We have estimated the statistical significance and the p-value is less than 7:59  107 . Protein–RNA complexes are classified into rRNA, tRNA and others (siRNA, miRNA, mRNA, etc.). There is no significant difference in SRs among them. However, rRNA has the highest BRs of 26 percent whereas tRNA and others have only 6 percent. This also confirms that rRNA is more prone to bind ribosomal proteins whereas tRNA only binds partially [13], [65], [66]. Accordingly, more number of KRs are observed in rRNA than other types.

3.4

Frequency of Occurrence of Stabilizing, Binding, and Key Residues The frequency of occurrence of stabilizing, binding and key residues in a set of non-redundant complexes is shown in

Fig. 3. Frequency of occurrence of 20 amino acid residues in stabilizing, binding and key residues: (a) protein–RNA complexes and (b) protein– protein complexes [56].

Fig. 3a. We noticed that hydrophobic residues such as Val, Leu, Ile, Ala and Gly are dominant among SRs (p < 2:28  1016 ), which reveals that hydrophobic interactions are important in stabilizing the protein–RNA complexes [15], [67]. In BRs, positively charged (Arg and Lys) and hydrophilic residues (Gly, Ser and Thr) are highly preferred (p < 2:28  1016 ), suggesting the dominance of electrostatic and hydrogen bond interactions in protein–RNA interactions [3], [68]. However, key residues are dominated by the combination of charged, polar, aliphatic and aromatic residues (Fig. S5, available online). Furthermore, residues with similar chemical behavior have different specificity to be KRs. For example, Arg, Tyr, Val, Thr are preferred over Lys, Trp, Ile, Ser, respectively. These results are found to be statistically significant (p-value < 2:28  1016 ). A Similar trend is observed in protein–protein complexes as seen in Fig. 3b. In addition, Asn prefers to serve as KRs and Ser has more preference than Thr in protein–protein complexes.

3.5 Frequency of Occurrence of Key Residues in Different Structural Classes and Functions We have analyzed the preference of specific amino acid residues to serve as KRs in various types of protein–RNA complexes and the preferred residues with an occurrence of more than 5 percent (random distribution) is presented in Table 2. We observed that Arg, Tyr and Thr are common to all structural classes. In addition, all-a, all-b, a þ b and a=b class of complexes have the specific preferences of Phe, Asp, Val and Ser, respectively. However, a few residues such as Met, Ala and Trp are not preferred in any of the classes (Table 2). Based on function, 12 residues are preferred in regulatory proteins with more than 5 percent occurrence whereas a set of 6 and 7 residues are preferred in enzymes and structural proteins, respectively. Specifically, Asp and Cys are preferred in enzyme and regulatory proteins. Arg is highly frequent in enzymes (40 percent), moderately preferred in structural (20 percent) and least likely in regulatory class (5 percent).

1440

IEEE/ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS,

VOL. 15,

NO. 5, SEPTEMBER/OCTOBER 2018

TABLE 2 Preferred Key Residues in Protein–RNA Complexes Type of protein–RNA complexes Structure All–a All–b aþb a/b Function Enzyme Regulatory Structural

Preferred residues Phe, Arg, Tyr, Thr, Ile, Glu, Gln, Gly, Val, Pro, Trp Arg, Tyr, Thr, Ile, Glu, Asp, Lys Arg, Thr, Lys, Tyr, Gln, Val Arg, Tyr, Thr, Gln, Glu, Ser, His, Asn Arg, Thr, Tyr, Glu, Asp, Cys Gln, Tyr, Ile, Phe, Arg, Thr, Glu, Asp, Cys, Ala, Ser, Trp Arg, Tyr, Lys, Thr, Gly, Ser, Val

RNA strand Double Single

Tyr, Asp, Gly, Phe, Ser, Thr, Val Arg, Tyr, Thr, Lys, Gln, Val

RNA conformation A U Others

Gly, Thr, Ile, Phe, Ser, Tyr, Val Arg, Thr, Lys, Tyr, Gln Arg, Tyr, Asp, Glu, Pro, Val

RNA type rRNA tRNA Others

Arg, Tyr, Lys, Asp, Gly, Ile, Val, His, Thr Arg, Glu, Tyr, Phe, Leu, Val Gln, Arg, Tyr, Phe, Thr

Preferred amino acid, which are common to all sub classes are shown in bold.

Interestingly, Lys acts as KR only in structural proteins and not preferred in regulatory proteins and enzymes. The residues Tyr, Thr and Val are preferred in both double and single stranded RNA. However, negatively charged residue Asp prefers to be KR in double stranded RNA whereas positively charged residues, Lys and Arg are preferred in single stranded RNA. Analysis of RNA conformation showed that only Tyr is common in all the conformations and the preference of other residues depends on the conformation of RNA. Similar results are observed in different types of RNA such as rRNA, tRNA and others. In summary, positively charged and aromatic residues are preferred as KRs in protein–RNA complexes, primarily through electrostatic and stacking interactions [2], [3].

3.6 Preference of Atomic Contacts in KRs The comparison of atomic contacts between protein and RNA for BRs and KRs showed no significant difference between them (Fig. 4a). However, KRs have more polar and less nonpolar contacts than BRs. Among the KRs, contacts between polar–polar and polar–nonpolar atoms are highly preferred in protein–RNA complexes. This result is consistent with our previous observation of KRs in protein– protein complexes [56]. Although there is no significant difference in the preference of contacting pairs in BRs and KRs (Fig. 4a) it varies among structural classes. Specifically, all-a class proteins have less polar–polar and more polar–nonpolar contacts in KRs than BRs (p-value : 8:5  104 ). The importance of electrostatic interactions in both protein–RNA binding and stability is supported by the presence of more number of charged contacts in KRs than BRs in all-b, a þ b and a=b classes of proteins (Fig. 4b).

Fig. 4. Atomic contacts in (a) non–redundant set of complexes and (b) structural class of proteins.

3.7 Influence of Protein Function on Atomic Contacts of KRs Based on protein function, we have analyzed the atomic contacts of BRs and KRs and the percentage of polar, nonpolar, polar–nonpolar and charged contacts are presented in Table 3. In all the three classes, polar, charged and polar–nonpolar contacts are preferred in both BRs and KRs. Enzymes have the highest charged pairs of 31 percent in KRs whereas regulatory proteins have only 11 percent of charged pairs. A reverse trend is observed for nonpolar contacts that these contacts are significantly reduced in KRs in enzymes (8 to 1 percent) and structural proteins (6 to 3 percent), whereas nonpolar contacts of KRs are increased in regulatory proteins (9 to 14 percent). There is no significant change of polar and polar–nonpolar contacts between BRs and KRs in any functional class of proteins. Comparison of BRs and KRs reveals that in enzymes the contribution of charged residues to be KRs is almost 2 fold higher than BRs, which shows that most of the BRs are involved in stability. A reverse trend is observed for regulatory proteins, where only 11 and 19 percent of charged residue pairs are observed in KRs and BRs, respectively. These results are statistically significant with a p-value 4:84  105 . 3.8 Sequence and Structural Analysis of Key Residues We have calculated conservation score, accessible surface area, surrounding hydrophobicity and long-range order to understand the sequence and structural features of KRs. Interestingly, 78 percent of KRs are found to have high conservation scores of 8 and 9. Based on secondary structure, most of the KRs (74.6 percent) are located in b-strands followed by coil (17.8 percent) and helix (7.6 percent). Since, KRs are also involved in binding, solvent accessibility analysis showed that 75 percent of KRs are buried inside protein–RNA complexes. Similar trend is observed in protein– protein complexes [56].

KULANDAISAMY ET AL.: IDENTIFICATION AND ANALYSIS OF KEY RESIDUES IN PROTEIN–RNA COMPLEXES

1441

TABLE 3 Preferred Atomic Contacts of Binding and Key Residues in Protein–RNA Complexes Based on Protein Functions Function

Type of residues

Polar (%)

Nonpolar (%)

Polar and Nonpolar (%)

Charged (%)

Enzyme

Binding Residues Key Residues Difference

30 24 6 (25%)

8 1 7 (700%)

46 44 2 (5%)

16 31 –15 (48%)

Regulatory

Binding Residues Key Residues Difference

28 34 –6 (18%)

9 14 –5 (36%)

44 41 3 (7%)

19 11 8 (73%)

Structural

Binding Residues Key Residues Difference

30 36 –6 (16%)

6 3 3 (100%)

44 43 1 (2.3%)

20 18 2 (11%)

Percentage change of the difference in the BRs and KRs is mentioned in parenthesis. BRs and KRs with more than 10 contacts are shown in bold.

The surrounding hydrophobicity of SRs, BRs and KRs, which provide insights of hydrophobic environment of these residues are shown in Fig. 5. The highest hydrophobicity of BRs, SRs and KRs were observed at