Fig 1.
Domain structures of PUF60 and the AdML pre-mRNA 3’ intron sequence.
(A) PUF60 is a 559-amino acid protein, consisting of two RNA-recognition motifs (RRMs) and one U2AF homology motif (UHM). PUF60 resembles U2AF65 in the two RRMs and one UHM, but lacks the N-terminal Arg-Ser repeat-containing (RS) domain that U2AF65 possesses to associate with the branch point sequence (BPS). (B) The sequence of the 3’-adenovirus major late promoter (AdML3’) pre-mRNA 3’ splice site (ss) region used in these studies [10,22]. AdML3’ contains the whole BPS, and the poly-U tract, which represents the majority of the polypyrimidine tract (PPT). The 3’ terminal AG splice site is positioned on the far right.
Fig 2.
Binding of PUF60 domains to AdML pre-mRNA 3’ splice site region sequence.
(A) Binding curves of PUF60 domains, RRMs+UHM or RRMs, to *AdML3’ as determined by fluorescence anisotropy. (B) Reported parameters of RRMs+UHM and RRMs binding to *AdML3’ by curve fitting with a 2:1 stoichiometric model. (C) Fitting equation assuming two-site sequential binding of two PUF60 molecules to one *AdML3’.
Fig 3.
Overview of PUF60:dAdML3’ complex structure and its dimeric interface.
The overall structure of PUF60 RRMs:dAdML3’ complex is analogous to that of the FIR:FUSE complex (PDB ID: 2QFJ) in many structural features including the dimeric conformation of RRMs. Subunit A of PUF60 RRMs is shown in pink, and subunit B in cyan. The bound dAdML3’ is shown with carbon atoms in yellow. The opposing 5’– 3’ directionalities of the observed nucleotides indicate that the backbone loops out as it extends from one side of the dimer to the other. A possible path of the nucleotide backbone is shown in green.
Fig 4.
Identification of bound nucleotides.
Simulated-annealing omit maps are displayed, omitting all nucleotides (A,B) or only the nucleotides in the second binding site on each subunit (C,D). The identification of the first nucleotide on each site allowed for that to be included in map calculations, and the second base became apparent in the electron density maps. (A) The uracil in the first binding site on subunit A (U-17). (B) The uracil (from the poly-U tract) in the first binding site on subunit B. (C) A guanine (G-18) is bound in the second binding site on subunit A. (D) However, a uracil is found in the second site on subunit B. The color-coding of chains in all panels is the same as in Fig 3.
Fig 5.
Recognition of bases bound in the first position of the nucleic acid binding site on the PUF60 RRM domains.
(A) U-17 (immediately after the BPS) bound to subunit A. Lys-201 from subunit B donates a hydrogen bond to the nucleotide. (B) Poly-U tract uracil bound to subunit B.
Fig 6.
Recognition of second position bases by PUF60 RRMs.
Hydrogen-bonding helps identify the identity of bases bound to the second nucleotide binding site on each PUF60 RRM subunit. (A) A uracil in the poly-U region accepts a hydrogen bond from Asn-207 of subunit B in the second nucleotide binding site. (B) The equivalent Asn-207 of subunit A donates a hydrogen bond to G-18 in the second nucleotide position on that particular subunit.
Fig 7.
K201 residues are asymmetrically positioned, reaching across the dimerization interface to associate with bound nucleotides.
Nucleotides and the K201 residues from both the A and B-chain monomers of PUF60 are shown as sticks. The rest part of the PUF60 structure is shown in ribbon-style. One of the Lys-201 residues donates a hydrogen bond to U-17 of dAdML3’ on the other side of the dimeric interface.
Fig 8.
PUF60 dimerizes upon dAdML3’ binding.
(A) SEC-LS analysis of the dAdML3’:PUF60 RRMs+UHM complex. Each data point represents an average molecular weight measured at the apex of an eluent peak of the complex at a certain concentration. Expected MW (104.8 kDa) for a 2:1 (protein: Nucleic acid) complex is indicated by the dashed line, with its expected 5% deviation shown in dotted lines; (B) UV/RI analysis of the dAdML3’:PUF60 RRMs+UHM complex. The UV/RI ratios from the three data points at the highest concentrations from (A) are averaged, and shown in the filled bar to be compared to other standard and expected values as described; (C) SEC-LS analysis of the dAdML3’:PUF60 RRMs complex. Each data point represents an average molecular weight measured at the apex of an eluent peak of the complex at a certain concentration. Expected MW (55.6 kDa) for a 2:1 (protein: Nucleic acid) is indicated by the dashed line, with its expected 5% deviation shown in dotted lines; (D) UV/RI analysis of the dAdML3’:PUF60 RRMs complex. The UV/RI ratios from the three data points at the highest concentrations from (C) are averaged, and shown in the filled bar to be compared to other standard and expected values as described.
Fig 9.
PUF60 uses UHM to associate with SF1 and cooperatively interact with AdML3’ splice site region pre-mRNA.
(A) Representative isotherms from ITC experiments to detect the association between SF1 1–260 and either PUF60 RRMs+UHM or PUF60 RRMs. (Left panel) Titration of SF1 1–260 into PUF60 RRMs+UHM. The best fit of the binding stoichiometry (n) was one. All titrations were measured at 30°C. (Right panel) Titration of SF1 1–260 into PUF60 RRMs; (B) Reporting the thermal dynamic parameters measured in (A). The mean parameters and standard deviations of two experiments are reported. The standard state (°) for these experiments is defined as 1 M concentrations of each protein at 30°C. The dissociation constant (Kd) and stoichiometry of binding (n) are derived from curve fitting based on datasets subtracted with the reference titration of SF1 heat of dilution. a Calculated using the equation ΔG° = –RT ln(Kd-1). b Calculated using the equation ΔG° = ΔH°-TΔS°. (C) EMSA study of the interplay of SF1 1–260 and PUF60 RRMs+UHM with AdML3’ RNA. The bright-dark colors of the gel image are digitally reversed for clarity in presentation.
Fig 10.
Alternative strategies to structure the 3’ spice site region of pre-mRNA: Looping by PUF60 vs. Bending by U2AF65.
(A) Hypothetical drawing of the potential looping that can occur on pre-mRNA structure upon PUF60 binding; (B) Direct excerpt from the work of Kent et al. to show the current model for the bending of the pre-mRNA by U2AF65 and other factors [9]; (C) Direct excerpt from Sickmier et al. of the crystal structure of the U2AF65 RRMs in complex with a poly-U tract to demonstrate the bending mechanism supported by structural data [11]. Notice that, with either strategy, the directionality of the pre-mRNA will be modulated. The modulation of the directionality may bring distant components on the pre-mRNA into proximity.