A flexible RNA backbone within the polypyrimidine tract is required for U2AF65 binding and pre-mRNA splicing in vivo

Abstract

The polypyrimidine tract near the 3' splice site is important for pre-mRNA splicing. Using pseudouridine incorporation and in vivo RNA-guided RNA pseudouridylation, we have identified two important uridines in the polypyrimidine tract of adenovirus pre-mRNA. Conversion of either uridine into pseudouridine leads to a splicing defect in Xenopus oocytes. Using a variety of molecular biology methodologies, we show that the splicing defect is due to the failure of U2AF(65) to recognize the pseudouridylated polypyrimidine tract. This negative impact on splicing is pseudouridine specific, as no effect is observed when the uridine is changed to other naturally occurring nucleotides. Given that pseudouridine favors a C-3'-endo structure, our results suggest that it is backbone flexibility that is key to U2AF binding. Indeed, locking the key uridine in the C-3'-endo configuration while maintaining its uridine identity blocks U2AF(65) binding and splicing. This pseudouridine effect can also be applied to other pre-mRNA polypyrimidine tracts. Thus, our work demonstrates that in vivo binding of U2AF(65) to a polypyrimidine tract requires a flexible RNA backbone.

FIG. 1.

Early stage of spliceosome assembly and box H/ACA RNA in directing RNA pseudouridylation. (A) At the early stage in spliceosome assembly, U1 recognizes the 5′ splice site of pre-mRNA, and U2AF⁶⁵ recognizes the polypyrimidine tract. The branch site (BS) and polypyrimidine tract (PolyPy) of adenovirus pre-mRNA are shown. U1 and U2AF (both the U2AF⁶⁵ and U2AF³⁵ subunits) are also indicated. The uridines in the polypyrimidine tract are sequentially numbered. (B) The hairpin-hinge-hairpin-tail structure of a box H/ACA RNA is shown. The internal loops of the hairpins serve as guides (or pseudouridylation pockets) that base pair with their respective substrates. The thick lines represent substrate RNAs, and the thin lines stand for box H/ACA RNA. The target nucleotides to be pseudouridylated are indicated (Ψ). Box H and box ACA are also indicated. The guide sequence of the 5′ pseudouridylation pocket is taken from PugU2-34/44, a Xenopus box H/ACA guide RNA targeting U2 at positions 34 and 44.

FIG. 2.

Mapping of the important uridines within adenovirus pre-mRNA. (A) pre-mRNA splicing in Xenopus oocytes. Lanes 1 and 2 are uninjected pre-mRNAs, regularly transcribed (U) and with pseudouridine substitution (Ψ), respectively. Lanes 3 and 4 are injected pre-mRNAs, regularly transcribed (U) and with pseudouridine substitution (Ψ), respectively. The RNA bands corresponding to the lariat intron, pre-mRNA, and mature mRNA are indicated. M, molecular size markers. (B) Construction of chimeric pre-mRNAs using RNase H site-specific cleavage directed by 2′-O-methyl RNA-DNA chimeras followed by mix-and-match two-piece ligation. The three cleavage sites are indicated, and the cleavage and ligation strategy, using cleavage site 1 as an example, is shown. Pre-mRNA with pseudouridine substitution is on the left, and the regularly transcribed (containing U) pre-mRNA is on the right. Upon cleavage at the same site (directed by the 2′-O-methyl RNA-DNA chimera), the 5′ and 3′ halves were mixed and matched. Two-piece ligation was then followed, generating a pair of chimeric pre-mRNAs. (C) Three pairs of chimeric pre-mRNAs (lanes 3 to 8) as well as regularly transcribed pre-mRNAs (lane 1; all U) and pre-mRNAs with pseudouridine substitution (lane 2; all Ψ) were assayed for splicing in Xenopus oocytes. 5′Ψ-3′U represents pre-mRNAs whose 5′ halves contained Ψ's and 3′ halves contained U's. 5′U-3′Ψ represents pre-mRNAs with reciprocal combinations.

FIG. 3.

Identification of two important pseudouridines in the polypyrimidine tract by RNA-guided RNA pseudouridylation. (A) Nine artificial box H/ACA guide RNAs targeting the nine uridines in the polypyrimidine tract (gRNA-U#1 to gRNA-U#9) were separately injected into U2-depleted oocytes (lanes 4, 5, 8, 9, and 13 to 17). Labeled pre-mRNA was then injected. Following pre-mRNA pseudouridylation, U2 snRNA was injected to rescue splicing activity. Lanes 1 and 10 are controls in which radiolabeled pre-mRNA was directly injected into intact oocytes. Lanes 2, 6, and 11 are negative controls in which radiolabeled pre-mRNA was injected into U2-depleted oocytes and no rescuing U2 was injected later. Lanes 3, 7, and 12 are positive controls in which radiolabeled pre-mRNA was injected into U2-depleted oocytes and rescuing U2 was injected later. Bands corresponding to the lariat intron, pre-mRNA, and rescuing U2 are indicated. (B) Relative splicing efficiency (relative ratio of lariat intron to the sum of lariat intron and unspliced pre-mRNA) of each reaction was quantified based on three independent experiments, with the control reaction (U2 depletion and reconstitution) being set as 1 (control-2). The numbers on the x axis correspond to the lane numbers in panel A. (C) Pre-mRNA containing a wild-type U₇ (lane 1) or a U₇-to-Ψ₇ change (lane 2) in the polypyrimidine tract was synthesized (also see the top panel), and splicing was examined in Xenopus oocytes. Bands corresponding to the 2/3 lariat intermediate, lariat intron product, and unspliced pre-mRNA are indicated. BP, branch point; M, molecular size markers.

FIG. 4.

Ψ-specific effect and its impact on the assembly of early splicing complexes. (A) Wild-type (WT) pre-mRNA (lanes 1, 2, 3, and 6) and a mutant pre-mRNA containing a U₇-to-G₇ change (lanes 4 and 5) were assayed for splicing in Xenopus oocytes. Lanes 1 and 4, splicing in intact oocytes; lane 3, splicing in U2-depleted oocytes; lane 2, splicing in oocytes in which U2 was depleted and later reconstituted; lanes 5 and 6, as in lane 2, with the exception that an artificial box H/ACA guide RNA, targeting position 7 of the polypyrimidine tract, was present. BP, branch point sequence. (B) Relative splicing efficiency (see legend to Fig. 3B). The numbers on the x axis correspond to the lane numbers in panel A. (C) Wild-type pre-mRNA (lanes 1 and 2) and mutant pre-mRNA containing a U₇-to-G₇ (lane 3) or a U₇-to-Ψ₇ (lane 4) change were assayed for spliceosome assembly in Xenopus oocytes. Lane 1 is a negative control in which spliceosome assembly was assayed in U2-depleted oocytes. Splicing complexes A, B, and C, as well as the heterogeneous complex H, are indicated. (D) Competitor adenovirus polypyrimidine tracts (with U [PolyPyU] on the left and with Ψ [PolyPyΨ] on the right) sandwiched by two stem structures are shown. (E) Wild-type pre-mRNA splicing was tested in the presence of competitor PolyPyU, with 15-fold (lane 2) and 45-fold (lane3) molar excess relative to pre-mRNA, or in the presence of competitor PolyPyΨ, with 15-fold (lane 4) and 45-fold (lane 5) molar excess. To monitor the stability of the competitors, a trace amount of radiolabeled PolyPyU or PolyPyΨ was mixed with the unlabeled competitor before injection. Lane 1, splicing without competitor. Bands corresponding to lariat intron, unspliced pre-mRNA, and spliced mRNA, as well as competitor PolyPyU or PolyPyΨ, are indicated.

FIG. 5.

Effect of U₇-to-Ψ₇ change on U2AF⁶⁵ binding. (A) Wild-type pre-mRNA (odd-numbered lanes) or pre-mRNA containing a Ψ₇ in the polypyrimidine tract (even-numbered lanes) was injected into the nuclei of Xenopus oocytes. Nuclear extracts were subsequently prepared and immunoprecipitated with anti-U2AF⁶⁵ (αU2AF⁶⁵) (lanes 1, 2, 5, and 6), or anti-Sm (αSm) (lanes 3, 4, 7, and 8) antibodies. Lanes 1 to 4 are the precipitated fraction (pellet), and lanes 5 to 8 are the unbound fraction (supernatant [Sup.]). (B) In vitro U2AF⁶⁵-polypyrimidine binding assay. The 5′-end radiolabeled U₇-containing polypyrimidine tract (U₇) (lanes 1 to 4) or Ψ₇-containing polypyrimidine tract (Ψ₇) (lanes 5 to 8; panel overexposed in order to visualize the shifted complex) was incubated with the indicated amount of pure U2AF⁶⁵ (lanes 1 to 8). After the reaction, the RNA-protein complex (indicated as U2AF-RNA) and unbound RNA (indicated as low mobility [Low Mobil] and high mobility [High Mobil]) were resolved on the native gel. Lanes 9 and 10 are controls in which the U₇ polypyrimidine tract and Ψ₇ polypyrimidine tract alone (in the absence of U2AF⁶⁵) were loaded, respectively, onto the native gel. Lanes 11 and 12 are controls in which the U₇ polypyrimidine tract and Ψ₇ polypyrimidine tract were analyzed on a denaturing gel, respectively. The sequences of U₇ and Ψ₇ are shown at the top.

FIG. 6.

U2AF⁶⁵ binding requires RNA backbone flexibility within the polypyrimidine tract. (A) Splicing of pre-mRNA containing a modified nucleotide in the polypyrimidine tract. Wild-type pre-mRNA (lane 1) and pre-mRNA in which U₃ (lane 2) or U₇ (lane 3) in the polypyrimidine tract was locked in the C-3′-endo configuration (LNA), or the 2′-OH group of U₇ was changed to 2′-H (lane 4) or 2′-F (lane 5) were assayed for splicing in Xenopus oocytes. (B) Relative splicing efficiency (relative ratio of the sum of lariat intron and spliced mRNA to the sum of lariat intron, spliced mRNA, and unspliced pre-mRNA) (see the legend to Fig. 3A). The numbers on the x axis correspond to the lane numbers in panel A.

FIG. 7.

The Ψ effect can be applied to different polypyrimidine tracts. (A) Pre-mRNA containing an adenovirus (Ad) (lanes 1 and 2) or β-globin (lane 3 and 4) polypyrimidine tract was assayed for splicing in Xenopus oocytes. In lanes 1 and 3, uridines were incorporated into the polypyrimidine tracts, whereas in lanes 2 and 4, pseudouridines were incorporated into the polypyrimidine tracts. Bands corresponding to the lariat intron, unspliced pre-mRNA, and spliced mRNA are indicated. The diagram of pre-mRNA and the sequences of polypyrimidine tract (both adenovirus and β-globin), as well as the branch point (BP) sequence, are shown at the top of the figure. The asterisk represents the substituted β-globin polypyrimidine tract. (B) As in Fig. 3A, except that six (instead of nine) uridines (see panel A) were targeted for pseudouridylation and function. The asterisks represent guide RNAs that are specific for uridines in the substituted β-globin polypyrimidine tract (see panel A). (C) Relative splicing efficiency (relative ratio of lariat intron to the sum of the lariat intron and unspliced pre-mRNA) (as in Fig. 3A). The numbers on the x axis correspond to the lane numbers in panel B.