Degeneracy of the genetic code and stability of the base pair at the second position of the anticodon (2024)

Degeneracy in the genetic code and interpretation of Lagerkvist's parameters

We address the rational behind the existence of the two main codon degeneracy families in the genetic code table (Fig. 1A). These two families are sufficient to characterize the symmetrical codes of the mitochondria and some primitive bacteria (Lehmann 2000).

The basic assumption on which the model is built is that wobbling was initially maximized. This maximization enables a reduction of the set of tRNAs necessary to translate small genomes. Why then are there not only fourfold degenerate codon families in the table? We show that the answer lies in the structural interpretation of Lagerkvist's parameters.

The first two parameters (strength of N36–N1 and N35–N2 base-pair associations) are self-explanatory. The structural interpretation of the third parameter (second position purine or pyrimidine) has not been clarified yet; we discuss it below.

In order to accommodate the decoding center of the ribosome, the anticodon loop of all tRNAs has to bear some standard features (Quigley and Rich 1976; Dix et al. 1986; von Ahsen et al. 1997; Sundaram et al. 2000; Stuart et al. 2003). The main characteristic of the loop relies on the presence of the U-turn motif (Quigley and Rich 1976), which requires several conserved interactions between residues of the loop. Among these, a critical hydrogen bonding pattern involves the 2′ OH group of U33 and the base at position 35. The result depends on whether N35 is a purine or a pyrimidine (Auffinger and Westhof 2001).

A purine enables the formation of a conventional donor–acceptor hydrogen bond between (U33)O2′-H and (R35)N7, bifurcated to (G35)O6 in the case of a guanine. The distance and angle are compatible with the formation of a strong hydrogen bond, i.e., ∼1.3 kcal mol⁻¹ (Jiang and Lai 2002).

With a pyrimidine at position 35, the interaction is much weaker since the approximate position of the (R35)N7 atom is occupied by a (Y35)C5-H group, which potentially only allows for the formation of a CH…O type of hydrogen bond, with (U33)O2′. The distance of 3.5 Å estimated from MD simulations is compatible with a hydrogen bond of ∼0.4 kcal mol⁻¹ (Jiang and Lai 2002).

Based on this analysis, the third parameter is defined as follows: We credit U33–N35 with S (“strong”) when N35 is a purine and with W (“weak”) when N35 is a pyrimidine. The structural significance of the three parameters can now be examined in Figure 1B,C. Figure 1B shows that the sum of three structural energies determines whether a codon belongs to a twofold or a fourfold degenerate family while Figure 1C shows that each of these energies contributes to the level of stability of the base pair at the second position, N35–N2:

• The first parameter (1) indicates the extent up to which N35–N2 is stabilized by a nearest neighbor, N36–N1 (Mathews et al. 1999).

• The second parameter (2) indicates the intrinsic N35–N2 base-pair stability.

• The third parameter (3) indicates whether N35 is further stabilized by U33.

The connection between the three parameters and the type of degeneracy allowed at the third position (N34–N3) requires an examination of the anticodon–codon interaction within the decoding center. Three residues of the ribosome (G530, A1492, and A1493, Escherichia coli numbering) interact with the minor groove of the anticodon–codon complex (Ogle et al. 2001). A1493 spans over the minor groove of the first base pair while A1492 and G530 span over the second base pair (Fig. 1C). When the base pair at the first position is G•C or C•G, the A-minor interaction realized by A1493 makes this base pair even stronger (V. Ramakrishnan, pers. comm.). It may explain why the contributions of the two first parameters to the stability of the base pair at the second position are identical (Fig. 1B).

Both N36–N1 and N35–N2 base pairs have to bear the Watson–Crick geometry to enable the optimal hydrogen-bonded interactions with the ribosome residues. Crystal structures however show that this type of monitoring does not occur at the third position: only one direct hydrogen bond is observed, between (G530)O6 and the 2′OH of the ribose of N3 (Ogle et al. 2001, 2002).

The fact that the ribosome does not directly check the base pair at the third position implies that it can only be sensed by its nearest neighbor, N35–N2.

Considering U, C, G, and A, three types of N34–N3 pairings are possible: Watson–Crick (WC), Wobble (Wb) (G•U or U•G), and base–base mismatch (Mm). Because the width of the base–base mismatches is larger or smaller than a regular base pair, they impose a stronger constraint on N35–N2 by distorting the double helix (Agris et al. 2007). In a crude approximation, N34–N3 Watson–Crick, and Wobble interactions are credited with S (“stabilizing”), while the base–base mismatches are credited with S-p, p standing for a mismatch penalty.

All local structural constraints in discussing the stability of the Watson–Crick geometry of N35–N2 have now been introduced. This parameter is critical since it enables the productive conformation of the decoding center of the ribosome (Ogle et al. 2002).

Table 1 shows that, below a total number of six consolidating hydrogen bonds defined by Lagerkvist's parameters, the base–base mismatch at the third position sufficiently destabilizes N35–N2 to prevent translation. It implies that two tRNAs are required for the translation of the four codons, which belong to two contiguous twofold degenerate codon families. With six and seven hydrogen bonds, a base at position N34 of a tRNA (often a U in mitochondria) is able to base pair with U, C, G, or A on the codon, and none of these interactions sufficiently perturb N35–N2 to prevent translation. These codons therefore belong to a fourfold degenerate codon family.

The twofold → fourfold transition establishes a threshold for translation at 6 + S-p (in hydrogen bonds units), which can be compared to 4 + S, the weakest interaction still allowing translation. While assuming that both are roughly equal, one gets p = 2 hydrogen bonds, which compares well with known experimental values (Mathews et al. 1999).

Effects of base modifications

Although the analysis shows that the organization of the degeneracy can be described without taking base modifications into account, these are necessary to fully explain two effects.

N34 modifications play a significant role in the accuracy of wobbling. In the simplest systems (such as the mitochondria), none of the tRNAs assigned to the fourfold degenerate codon families requires a N34 modification in order to wobble over the four bases (Suzuki 2005). In twofold degenerate codon families, however, N34 modifications (e.g., xm⁵s²U derivatives) are always present and have been shown to prevent “leaking” wobblings between contiguous families (Yokoyama et al. 1985; Yokoyama and Nishimura 1995). Thus, a sharp demarcation between the twofold and fourfold degenerate codon families could not be established solely on the basis of one hydrogen bond (Table 1).

Some codons cannot be translated if specific base modifications are not present on their cognate tRNAs (Takai et al. 1999; Takai and Yokoyama 2003; Agris 2004; Agris et al. 2007). For example, N37 modifications (often τ⁶A or m¹G) compensate for weak N36–N1 base pairs, in particular by increasing stacking (Grosjean et al. 1998; Konevega et al. 2004). The τ⁶A37 modification has also been shown to restore the otherwise absent canonical U-turn in some tRNAs (Sundaram et al. 2000).

In brief, base modifications remove coding ambiguities in contiguous twofold degenerate codon families and shape unstructured tRNA anticodon loops for translation.

Table 1 represents a situation found in some primitive bacteria and mitochondria, where N34–N3 base-pairing rules and degeneracy families match exactly (Bonitz et al. 1980; Inagaki et al. 1995; Suzuki 2005), meaning that the number of different tRNAs necessary for translation is at the minimum. As far as evolved genetic systems are concerned, some structural constraints require that more than one tRNA is usually necessary for the translation of all the codons in many codon families (Agris et al. 2007; Näsvall et al. 2007).

Supports to the model

The third of Lagerkvist's parameters, which we showed to be connected with the stability of N35 with respect to U33, is not directly involved in the anticodon–codon interaction. One could then ask whether this stabilization indeed affects translation. The earliest investigations on the U-turn already showed that it is the case (Uhlenbeck et al. 1982; Dix et al. 1986). For instance, direct binding experiments of tRNAs on P-sites programmed with complementary codons show that when U33 is replaced with dU33 in E. coli tRNA^Phe (an operation that removes the critical 2′-OH), the dissociation constant undergoes a 50-fold increase (Von Ahsen et al. 1997). A control experiment shows that the same substitution on tRNA^Asn has only a minor effect on the corresponding dissociation constant. Since tRNA^Asn has a U at position 35, no (strong) hydrogen bond is originally present, meaning that no effect is expected from the substitution.

Another type of data related to the N35 nucleotide supports our analysis. In a very small number of tRNAs, this position is occupied by pseudouridine (ψ). This base is similar to uridine, but the critical C5–H is replaced with N1–H, which allows for a stronger hydrogen bond with (U33)O2′ (Auffinger and Westhof 2001). In addition, RNA melting experiments have shown that the ψ·A base pair itself is stronger than U·A (Ward and Reich 1968). Our model therefore predicts a ψ35 → U35 substitution on a tRNA to correspond to (one to two) S → W substitutions on the diagram of Figure 1B, meaning a possible switch from a four-way wobble to a two-way wobble. This effect has been documented in the case of a plant cytoplasmic tRNA^Tyr, which has an anticodon ^3′GψA^5′ (Zerfass and Beier 1992). In the specific context of a tobacco mosaic virus (TMV) reading frame, this tRNA is able to suppress both UAG and UAA stop codons (which add to the regular translation of UAU and UAC). In vitro translation experiments showed that the single ψ35 → U35 mutation completely abolishes the suppression activity.

Influence of the N32–N38 base pair

Another fundamental feature of the anticodon loop, the N32–N38 closing base pair (see Fig. 1C), has also been shown to influence the extent of wobbling at the third position (Lustig et al. 1993). At the level of the gene, this base pair is usually C•A (48%), T•A (18%), T•T (11%), C•C (8%), or T•C (8%). With the exception of T•A, which can form a U•A Watson–Crick base pair in the tRNA (as observed in the unbound state), crystallographic data show that the other mentioned couples are usually connected with a single bifurcated hydrogen bond (Auffinger and Westhof 1999).

In vitro translation experiments have shown that if the U32–A38 closing base pair of E. coli tRNA₁^Gly (with a modified anticodon ^3′UCC^5′) is converted into C•A, the wobbling ability of this tRNA is extended from GGR to GGN (N = U, C, G, or A) (Lustig et al. 1993). Another study showed that the dissociation constant of E. coli tRNA₁^Ala bound to a programmed A-site is reduced by about one order of magnitude if the (rare) A32–U38 wild type is converted into U•U, A•A, or U•A, but it remains high if it is converted into C•G (Olejniczak et al. 2005).

The above examples show that the stability of the anticodon–codon interaction within the decoding center can be impaired when N32–N38 may form a Watson–Crick type of interaction (Olejniczak et al. 2005). In this case, the loop does indeed not adopt the (mandatory) U-turn geometry (Auffinger and Westhof 1999). With regard to the observed effect on the wobble (Lustig et al. 1993), a simple explanation is that the U32–A38 base pair adopts a Watson–Crick type of interaction in the free state. Upon codon binding, the U-turn geometry is restored only if the anticodon–codon interaction is strong enough to force the U32–A38 Watson–Crick to switch into the single hydrogen-bonded interaction. This would make any further penalization by an N34–N3 base–base mismatch prohibitive.

The fact that a N32–N38 parameter is not required to explain the organization of the degeneracy families in the genetic code table is consistent with our hypothesis of wobble maximization at the origin. It only implies that the N32–N38 base pair has to be non-Watson–Crick.

Article published online ahead of print. Article and publication date are at http://www.rnajournal.org/cgi/doi/10.1261/rna.1029808.

• Agris P.F. Decoding the genome: A modified view. Nucleic Acids Res. 2004;32:223–238. [PMC free article] [PubMed] [Google Scholar]
• Agris P.F., Vendeix F.A.P., Graham W. tRNA's wobble decoding of the genome: 40 years of modification. J. Mol. Biol. 2007;366:1–13. [PubMed] [Google Scholar]
• Auffinger P., Westhof E. Singly and bifurcated hydrogen-bonded base-pairs in tRNA anticodon hairpins and ribozymes. J. Mol. Biol. 1999;292:467–483. [PubMed] [Google Scholar]
• Auffinger P., Westhof E. An extended structural signature for the tRNA anticodon loop. RNA. 2001;7:334–341. [PMC free article] [PubMed] [Google Scholar]
• Bonitz S.G., Berlani R., Coruzzi G., Li M., Macino G., Nobrega F.G., Nobrega M.P., Thalenfeld B.E., Tzagoloff A. Codon recognition in yeast mitochondria. Proc. Natl. Acad. Sci. 1980;77:3167–3170. [PMC free article] [PubMed] [Google Scholar]
• Dix D.B., Wittenberg W.L., Uhlenbeck O.C., Thompson R. Effect of replacing uridine 33 in Yeast tRNA^Phe on the reaction with ribosome. J. Biol. Chem. 1986;261:10112–10118. [PubMed] [Google Scholar]
• Grosjean H., Houssier C., Romby P., Marquet R. Modulatory role of modified nucleotides in RNA loop-loop interaction. In: Grosjean H., Benne R., editors. Modification and editing of RNA. ASM Press; Washington, DC: 1998. pp. 113–133. [Google Scholar]
• Inagaki Y., Kojima A., Bessho Y., Hori H., Ohama T., Osawa S. Translation of synonymous codons in family boxes by Mycoplasma capricolum tRNAs with unmodified uridine or adenosine at the first anticodon position. J. Mol. Biol. 1995;251:486–492. [PubMed] [Google Scholar]
• Jiang L., Lai L. CH…O hydrogen bonds at protein–protein interfaces. J. Biol. Chem. 2002;277:37732–37740. [PubMed] [Google Scholar]
• Konevega A.L., Soboleva N.G., Makhno V.I., sem*nkov Y.P., Wintermeyer W., Rodnina M.V., Katunin V.I. Purine bases at position 37 of tRNA stabilize codon–anticodon interaction in the ribosomal A site by stacking and Mg²⁺-dependent interactions. RNA. 2004;10:90–101. [PMC free article] [PubMed] [Google Scholar]
• Lagerkvist U. “Two out of three”: An alternative method for codon reading. Proc. Natl. Acad. Sci. 1978;75:1759–1762. [PMC free article] [PubMed] [Google Scholar]
• Lagerkvist U. Unconventional methods in codon reading. Bioessays. 1986;4:223–226. [PubMed] [Google Scholar]
• Lehmann J. Physico-chemical constraints connected with the coding properties of the genetic system. J. Theor. Biol. 2000;202:129–144. [PubMed] [Google Scholar]
• Lustig F., Borén T., Claesson C., Simonsson C., Barciszewska M., Lagerkvist U. The nucleotide in position 32 of the tRNA anticodon loop determine ability of anticodon UCC to discriminate among glycine codons. Proc. Natl. Acad. Sci. 1993;90:3343–3347. [PMC free article] [PubMed] [Google Scholar]
• Mathews D.H., Sabina J., Zucker M., Turner D.H. Expanded sequence dependence of thermodynamic parameters improves prediction of RNA secondary structure. J. Mol. Biol. 1999;288:911–940. [PubMed] [Google Scholar]
• Näsvall S.J., Chen P., Björk G.R. The wobble hypothesis revisited: Uridine-5-oxyacetic acid is critical for reading of G-ending codons. RNA. 2007;13:1–14. [PMC free article] [PubMed] [Google Scholar]
• Ogle J.M., Brodersen D.E., Clemons W.M., Jr, Tarry M.J., Carter A.P., Ramakrishnan V. Recognition of cognate transfer RNA by the 30S ribosomal subunit. Science. 2001;292:897–902. [PubMed] [Google Scholar]
• Ogle J.M., Murphy F.M., IV, Tarry M., Ramakrishnan V. Selection of tRNA by the ribosome requires a transition from an open to a closed form. Cell. 2002;111:721–732. [PubMed] [Google Scholar]
• Olejniczak M., Dale T., Fahlman R.P., Uhlenbeck O.C. Idiosyncratic tuning of tRNAs to achieve uniform ribosome binding. Nat. Struct. Mol. Biol. 2005;12:788–793. [PubMed] [Google Scholar]
• Quigley G.C., Rich A. Structural domains of transfer RNA molecules. Science. 1976;194:796–806. [PubMed] [Google Scholar]
• Stuart J.W., Koshlap K.M., Guenther R., Agris P.F. Naturally-occurring modification restricts the anticodon domain conformational space of tRNA^Phe. J. Mol. Biol. 2003;334:901–918. [PubMed] [Google Scholar]
• Sundaram M., Durant P.C., Davis D.R. Hypermodified nucleosides in the anticodon of tRNA^Lys stabilize a canonical U-turn structure. Biochemistry. 2000;39:12575–12584. [PubMed] [Google Scholar]
• Suzuki T. Biosynthesis and function of tRNA wobble modifications. In: Grosjean H., editor. Fine-tuning of RNA functions by modification and editing. Vol. 12. Springer; Berlin: 2005. pp. 24–69. [Google Scholar]
• Takai K., Yokoyama S. Roles of 5-substituents of tRNA wobble uridines in the recognition of purine-ending codons. Nucleic Acids Res. 2003;31:6383–6391. [PMC free article] [PubMed] [Google Scholar]
• Takai K., Takaku H., Yokoyama S. In vitro codon-reading specificities of unmodified tRNA molecules with different anticodons on the sequence background of Escherichia coli tRNA₁^Ser. Biochem. Biophys. Res. Comm. 1999;257:662–667. [PubMed] [Google Scholar]
• Uhlenbeck O.C., Lowary P.T., Wittenberg W.L. Role of the constant uridine in binding of Yeast tRNA^Phe anticodon arm to 30S ribosome. Nucleic Acids Res. 1982;10:3341–3352. [PMC free article] [PubMed] [Google Scholar]
• Von Ahsen U., Green R., Schroeder R., Noller H.F. Identification of 2-hydroxyl groups required for interaction of a tRNA anticodon stem-loop region with the ribosome. RNA. 1997;3:49–56. [PMC free article] [PubMed] [Google Scholar]
• Ward D.C., Reich E. Conformational properties of polyformycin: A polyribonucleotide with individual residues in the syn conformation. Proc. Natl. Acad. Sci. 1968;61:1494–1501. [PMC free article] [PubMed] [Google Scholar]
• Yokoyama S., Nishimura S. Modified nucleosides and codon recognition. In: Söll D., RajBhandary U., editors. tRNA: Structure, biosynthesis, and function. AMS Press; Washington, DC: 1995. pp. 207–223. Chap. 12, [Google Scholar]
• Yokoyama S., Watanabe T., Murao K., Ishikura H., Yamaizumi Z., Nishimura S., Miyazawa T. Molecular mechanism of codon recognition by tRNA species with modified uridine in the first position of the anticodon. Proc. Natl. Acad. Sci. 1985;82:4905–4909. [PMC free article] [PubMed] [Google Scholar]
• Zerfass K., Beier H. Pseudouridine in the anticodon GΨA of plant cytoplasmic tRNA^Tyr is required for UAG and UAA suppression in the TMV-specific context. Nucleic Acids Res. 1992;20:5911–5918. [PMC free article] [PubMed] [Google Scholar]