Figure 1. Types of RNA processing and modifications. The phosphodiester backbone can be cleaved (arrows) to leave either 5¢ or 3¢ phosphate groups. The ribose sugars can be methylated on the 2¢ positions. The four canonical bases (G, C, A, U) can be modified in various ways (examples shown on the right). Watson–Crick hydrogen-bonding groups are indicated by dark gray (donor) or light gray (acceptor) circles. Replacement of oxygen by sulfur results in a weaker hydrogen bond acceptor. m7G: 7-methyl guanosine; 2,2,7 tmG: 2,2,7-trimethyl guanosine; I: inosine; rT: ribothymidine; : pseudouridine.

Figure 2. 7-Methyl guanosine inverted cap structure. The cap is shown with a 2¢-O-methyl group on the first templated base, which has also been methylated at the adenine N6 position.

Figure 3. Consensus sequences and reaction of pre-mRNA splicing. Consensus splice site sequences. Exons are shown as boxes and the intron as a line. The invariant GU and AG dinucleotides at the ends of the intron and the branchpoint A are shown in bold. R: purine; Y: pyrimidine; BP: branchpoint; PPT: polypyrimidine tract. Splicing occurs by two successive transesterification reactions. In step 1, the 2¢ OH of the BP attacks the phosphate at the 5¢ splice site, forming the reaction intermediates. In step 2, the 3¢ OH of the 5¢ exon attacks the 3¢ splice site forming the spliced mRNA (left) and the intron, which is released as a loop or ‘lariat’ structure.

Figure 4. Pre-mRNA 3¢ processing. 3¢ end processing occurs in two steps. First the pre-mRNA is cleaved and then a tail of ~250 A residues is added. The reactions depend upon the highly conserved AAUAAA sequence and a GU/U-rich sequence, which flank the site of cleavage.

Figure 5. mRNA editing. (a) A to I (inosine) and C to U editing occur by chemically identical hydrolytic deamination reactions at the N6 of adenosine and N4 of cytosine. (b) A to I editing is specified by intramolecular secondary structure in which an intronic ‘exon complementary sequence’ (ECS) base pairs with the exon containing the A to be edited. The exon–ECS interaction is stabilized by additional base pairing within the intron.

Figure 6. Pre-rRNA processing. 18S, 5.8S and 28S rRNAs are synthesized as a 45S pre-rRNA precursor with 5¢ and 3¢ external transcribed spacers (ETS) and internal transcribed spacers (ITS). Extensive ribose 2¢-O-methylations (CH₃) and pseudouridylations () are carried out by Box C/D and Box H/ACA snoRNPs. Subsequent endonuclease cleavages are also guided by snoRNPs, including U3, 8 and 22 and RNase MRP. Finally, exonuclease trimming events produce the mature rRNAs.

Figure 7. Pre-rRNA modifications are specified by snoRNA : pre-rRNA base-pairing.

Figure 8. Pre-tRNA processing. tRNA is generated from pre-tRNA by removal of 5¢ leader and 3¢ trailer sequences, addition of the 3¢ end CCA, splicing, ribose methylations and various base modifications, including invariant pseudouridines and dihydrouridines, and frequent A to I modification at the first position of the anticodon.

Figure 9. Generation of proteomic complexity by alternative pre-mRNA splicing and editing. (a) Types of simple alternative splicing event. In each case, constitutively spliced exon sequences are in gray and alternatively spliced sequences in black. Splicing patterns are indicated by diagonal dashed lines. (i) Retained intron; (ii) competing 5¢ splice sites; (iii) competing 3¢ splice sites; (iv) alternative promoters and 5¢ end exons; (v) alternative 3¢ end exons and poly(A) sites; (vi) cassette exon; (vii) mutually exclusive exons. (b) Combinations of these basic events allow individual genes to generate multiple isoforms. A combination of three translation start codons in exon 1, a U to C editing event in exon 6, alternative splicing of exon 5, and competing 5¢ splice sites on exon 9 allows the Wilms tumor 1 gene to generate up to 24 isoforms. Misregulation of the exon 9 alternative splice is the basis of Frasier syndrome, a urogenital developmental defect.