Genetic interactions of hypomorphic mutations in the m⁷G cap-binding pocket of yeast nuclear cap binding complex: An essential role for Cbc2 in meiosis via splicing of MER3 pre-mRNA

Structure-guided mutations in the cap-binding pocket of yeast Cbc2

An alignment of the primary structures of yeast Cbc2 and the 156-aa human CBP20 homolog highlights 101 positions of amino acid side chain identity/similarity (Supplemental Fig. S1). Here, six of the conserved side chains in Cbc2 that comprise the m⁷Gppp binding site in the crystal structure of human CBC (Fig. 1) were mutated to alanine. These were as follows: Tyr24, which stacks on the m⁷G nucleobase and engages in hydrogen bonds with the β-phosphate of m⁷Gppp and the ribose 2′OH; Asp120, which accepts a hydrogen-bond from the m⁷G N1 atom; Asp122, which accepts a hydrogen-bond from the m⁷G N2 atom; Arg129, which donates hydrogen bonds from its Nɛ and NH2 atoms to the cap ribose O3′ and O2′ atoms, respectively (and which also makes a bidentate salt bridge to Asp122); Arg133, which engages the β-phosphate of m⁷Gppp; and Phe91, which is adjacent to the cap ribose. The alanine mutations were introduced into a plasmid-borne yeast CBC2 gene under the control of its native promoter and then tested for activity in vivo in a cbc2Δ yeast strain, in parallel with a wild-type CBC2 plasmid and an empty CEN vector. The cbc2Δ cells grew well on YPD agar at 34°C but were slow growing at 25°C, 30°C, and 37°C, as gauged by colony size. cbc2Δ cells failed to grow at 18°C and 20°C (Fig. 2A,B). Growth was restored at all temperatures after transformation of cbc2Δ with the wild-type CBC2 plasmid. Each of the alanine mutants also supported growth of cbc2Δ cells at all temperatures tested (Fig. 2A,B).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 2.

Effects of mutating the Cbc2 cap-binding pocket on yeast growth. (A) The phenotypes of cbc2Δ cells or cbc2Δ tgs1Δ cells harboring CEN LEU2 plasmids with wild-type CBC2 or the indicated mutant alleles were compared by spotting 3-μL aliquots of 10-fold serial dilutions of cells (from liquid cultures grown to mid-log phase at 34°C and adjusted to A₆₀₀ of 0.1) to SD−Leu agar and incubating the plates at the indicated temperatures. Suppression of the tgs1Δ cs phenotype is denoted by ● symbols at right. (B) The growth phenotypes of cbc2Δ and cbc2Δ tgs1Δ cells bearing LEU2 plasmids with the indicated genes were assessed as described in panel A. Suppression of the tgs1Δ cs phenotype is denoted by ◀ symbols at right. The N-terminal amino acid sequence of Cbc2 is shown at the bottom, with the two m⁷G-stacking tyrosines shaded gray and the margins of the N-terminal truncation mutants indicated by arrows.

In light of the structural evidence that the conformation of the N-terminal peptide segment of CBP20 is critical for formation of a proper cap binding site in human CBC (Calero et al. 2002; Mazza et al. 2002), we also tested the effects of deleting 21 or 42 amino acids from the N terminus of yeast Cbc2 (Fig. 2B). We found that the NΔ21 and NΔ42 alleles complemented the growth defects of cbc2Δ cells (Fig. 2B). These results suggest that an intact cap-binding pocket in yeast CBC is not required for apparently normal vegetative growth.

The benign effects of the above alanine and deletion mutations contrast with the overt growth defects of cbc2Δ cells (Fig. 2), which are similar to the growth defects of sto1Δ cells that lack the large subunit of yeast CBC (Fortes et al. 1999). An unexpected observation that a sto1Δ cbc2Δ double mutant grew better than either single mutant prompted the suggestion that production of either CBC subunit alone had a dominant negative effect on cell growth (Fortes et al. 1999). If that is the case, it complicates even further the issue of whether CBC subunits have cap-dependent and cap-independent functions. Thus, we revisited the issue, by mating STO1 cbc2Δ and sto1Δ CBC2 haploids, sporulating the resulting heterozygous diploids, and dissecting the tetrads to score growth and genotype the haploid progeny. We obtained the expected outcomes of a two-gene cross, which yielded parental ditype (all four spores being slow-growing), nonparental ditype (two fast-growing wild-type progeny and two slow growing sto1Δ cbc2Δ progeny), and tetratype (one fast growing wild-type, two slow growing single mutants and one slow-growing double mutant) segregation patterns (Fig. 3A).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 3.

sto1Δ cbc2Δ double mutants give rise to spontaneous suppressors of slow growth. (A) Haploid progeny derived from tetrad dissection of a heterozygous diploid strain are shown. Spores from 11 tetrads were placed on YPD agar medium and incubated for 3 d at 30°C. The relevant genotype of the diploid is indicated at the top; A–D refers to colonies derived from sister spores; the segregation patterns of the markers for sto1Δ (hyg^R) and cbc2Δ (kan^R) are indicated below each tetrad: (T) tetratype, (PD) parental ditype, (ND) nonparental ditype. (B) The growth of wild-type cells was compared to that of sto1Δ cbc2Δ cells either taken from a colony immediately after germination or after continuous growth of individual haploids at 30°C and isolation of faster growing derivatives S1, S2, and S3. Cultures were adjusted to A₆₀₀ of 0.1 and aliquots (3 μL) of serial dilutions were spotted to YPD agar. Plates were photographed after 2 d of incubation at 30°C, 34°C, and 37°C, 3 d at 25°C, or 5 d at 20°C.

Freshly germinated sto1Δ cbc2Δ haploids cells were amplified briefly in liquid culture and then spotted in serial dilution on YPD agar in parallel with a wild-type sister strain: The double mutant was quite sick at 30°C, 34°C, and 37°C, as gauged by colony size, and did not grow at 20°C or 25°C (Fig. 3B). We noted that propagated cultures of sto1Δ cbc2Δ cells or sto1Δ cells plated on YPD agar yielded mixtures of small and large colonies. (This was not the case for cbc2Δ cells.) Whereas individually selected large colonies continued to yield homogeneously large colonies upon further propagation, the individually selected small colonies recapitulated the small/large dimorphism upon further growth. Three genetically independent large colony sto1Δ cbc2Δ derivatives (S1, S2, and S3, obtained from cultures of initially slow-growing sto1Δ cbc2Δ haploid progeny from different tetrad dissections) were tested for growth on YPD agar and were found to grow as well as wild type at 34°C and to display improved growth at all other temperatures vis à vis sto1Δ cbc2Δ, although S1, S2, and S3 were moderately cold-sensitive at 20°C compared to wild type (Fig. 3B). Similarly enhanced growth was seen for “large colony” derivatives of sto1Δ (data not shown). These findings indicate that sto1Δ cbc2Δ cells and sto1Δ cells readily acquire spontaneous mutations that suppress their vegetative growth defects. We suspect that the “better-growing” properties of the sto1Δ cbc2Δ strain used previously are attributable to an acquired suppressor mutation, which might complicate the interpretation of phenotypes observed in sto1Δ cbc2Δ or sto1Δ strains as being directly reflective of CBC function. This concern appears not to apply to cbc2Δ or the cbc2 hypomorphs that we study presently.

Cbc2 cap binding site mutations suppress the cold-sensitivity of tgs1Δ cells

Tgs1, the enzyme that synthesizes TMG caps, is inessential for vegetative growth of S. cerevisiae (Mouaikel et al. 2002; Hausmann et al. 2008). tgs1Δ cells have no detectable TMG caps on their snRNAs, signifying that there is no Tgs1-independent route to form TMG caps. Despite the absence of TMG caps on their U1, U2, U4, and U5 snRNAs, tgs1Δ yeast cells display apparently normal steady state snRNA levels (Mouaikel et al. 2002) and no overt aberrations in the RNA or protein contents of their spliceosomal snRNPs, except for the acquisition of the nuclear CBC as a stoichiometric component of the U1 snRNP (Schwer et al. 2011). These findings indicated that the residual m⁷G cap of the U1 snRNP in tgs1Δ cells is accessible to, and occupied by, nuclear CBC.

tgs1Δ cells fail to grow at 18°C and 20°C (Fig. 2A,B). We reported recently that the biologically active cbc2-Y24A allele completely suppressed the cold-sensitive growth defect caused by tgs1Δ (Schwer et al. 2011). This is evinced by transforming a viable cbc2Δ tgs1Δ double mutant with a plasmid bearing either wild-type CBC2 or cbc2-Y24A. Whereas the CBC2 tgs1Δ strain was profoundly cold-sensitive, the cbc2-Y24A tgs1Δ strain grew as well as TGS1 cells at 18°C and 20°C (Fig. 2B and data not shown). We inferred that the cold-sensitive phenotype of tgs1Δ is not caused by the lack of TMG caps per se but rather by the ectopic association, at low temperatures, of nuclear CBC with the m⁷G cap of a normally TMG-capped U1 snRNA (Schwer et al. 2011). A corollary inference is that the capacity of otherwise benign Cbc2 mutations to suppress tgs1Δ cold-sensitivity can provide a genetic readout of diminished cap binding ability.

Thus, we transformed cbc2Δ tgs1Δ cells with the five other cbc2-Ala mutants (Fig. 2A) and the two N-terminal truncation mutants (Fig. 2B). In every case, growth at 18° and 20°C was restored. We surmise that the alanine substitutions for these cap binding side chains and the subtraction of the N-terminal 42-aa peptide (which includes the cap binding Tyr24 side chain) elicit genetically hypomorphic defects in cap binding by yeast CBC. To assess whether Cbc2 mutations affected the steady-state levels of Cbc2 in vivo, we performed Western blot analysis of whole-cell extracts of cbc2Δ and cbc2Δ tgs1Δ strains expressing TAP-tagged wild-type Cbc2, Cbc2-Ala, Cbc2-NΔ21, and Cbc2-NΔ42. The levels of immunoreactive Cbc2 were not diminished by the alanine mutations or the N-terminal deletions (Supplemental Fig. S2). The NΔ21 and NΔ42 polypeptides displayed expected incremental increases in electrophoretic mobility (Supplemental Fig. S2).

Cbc2-Y24A suppresses the requirement for Tgs1 in PCH2 pre-mRNA splicing

A network of mutational synergies between tgs1Δ and yeast pre-mRNA splicing factors suggests that TMG caps facilitate early steps in spliceosome assembly and that the effects of eliminating TMG caps on vegetative growth are buffered by other components of the splicing machinery (Hausmann et al. 2008; Chang et al. 2012). However, TMG caps are essential for yeast meiosis, where the absence of TMG caps in tgs1Δ cells causes a specific defect in the splicing of a subset of meiotic pre-mRNAs, including PCH2 and SAE3 (Qiu et al. 2011a,b). The Tgs1-dependence of the splicing of the PCH2 pre-mRNA was attributed to a nonconsensus branchpoint sequence in the PCH2 intron and an exceptionally long 5′ exon preceding the PCH2 intron. Indeed, Tgs1-dependent expression of a HIS3 reporter gene in vegetative cells could be achieved by inserting the PCH2 intron at a distal site in the HIS3 ORF (Qiu et al. 2011a,b). Here, we used the HIS3-[PCH2] reporter to query whether Tgs1-dependence might be circumvented by cbc2-Y24A. To do this experiment, we replaced the chromosomal CBC2 gene with cbc2-Y24A in TGS1 and tgs1Δ strains that carried the HIS3-[PCH2] cassette inserted at the chromosomal HIS3 locus. The HIS3-[PCH2] gene was functional in TGS1 CBC2 cells, as gauged by growth on agar medium lacking histidine, but not in tgs1Δ CBC2 cells (Fig. 4A). The instructive finding here was that HIS3-[PCH2] function was restored in tgs1Δ cbc2-Y24A cells (Fig. 4A). We infer that ectopic binding of CBC to the U1 snRNA cap in tgs1Δ cells is responsible (at least in part) for the Tgs1-dependence of PCH2 splicing. However, we found that cbc2-Y24A did not override the Tgs1 requirement for expression of a HIS3-[SAE3] reporter gene (data not shown).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 4.

cbc2-Y24A suppresses the Tgs1-dependence of HIS3-[PCH2] expression and PCH2 meiotic splicing. (A) Serial dilutions of yeast cells harboring the chromosomal HIS3-[PCH2] reporter gene were spotted in parallel on synthetic drop-out (SD) medium containing histidine (complete) or lacking histidine (–His) and incubated for 2 d at 30°C. The genotypes of the cells are indicated on the left. The HIS3-[PCH2] reporter is depicted at the bottom. The HIS3 ORF is colored gray. The PCH2 intron is colored black, and the nonconsensus branchpoint sequence is shown. (B) RNAs isolated from wild-type, tgs1Δ, cbc2-Y24A, and cbc2-Y24A tgs1Δ diploid SKY cells sampled 8 h post-transfer to sporulation medium were reverse transcribed with an oligo(dT) primer, and the cDNAs were PCR-amplified with primers flanking the intron of the chromosomal PCH2 gene (depicted at the bottom). The PCR products were resolved by native agarose gel electrophoresis and visualized by staining with ethidium bromide (the negative image is shown). The left lane shows the product of PCR-amplification of genomic DNA with the PCH2 primers, which is the same size as the RT-PCR product derived from the intron-containing PCH2 pre-mRNA. The positions and sizes (bp) of linear duplex DNA markers are indicated on the left. The positions of the RT-PCR products of unspliced and spliced PCH2 transcripts are indicated on the right.

To extend the analysis to true meiotic splicing, we replaced the chromosomal CBC2 locus with cbc2-Y24A in yeast SKY cells that were either TGS1 or tgs1Δ. (SKY, a derivative of SK1, is a strain of choice for studies of meiosis in light of its high efficiency and synchrony of sporulation.) Total RNA isolated from diploid wild-type, tgs1Δ, cbc2-Y24A, and cbc2-Y24A tgs1Δ SKY cells 8 h after transfer to sporulation medium was used as a template for oligo(dT)-primed reverse transcription (RT), followed by PCR amplification of the cDNA with oligonucleotide primers corresponding to exon sequences flanking the PCH2 pre-mRNA intron (Fig. 4B). Agarose gel electrophoresis resolved the 214-bp DNA derived by RT-PCR amplification of the spliced PCH2 mRNA from the 327-bp DNA derived from the unspliced PCH2 transcript (Fig. 4B). As reported previously (Qiu et al. 2011b), PCH2 meiotic splicing was defective in tgs1Δ cells, and the unspliced precursor was the predominant species seen after RT-PCR (Fig. 4B). The key finding was that cbc2-Y24A (which did not by itself affect meiotic splicing of the PCH2 transcript) restored the wild-type pattern of efficient PCH2 splicing in tgs1Δ cells undergoing sporulation (Fig. 4B). In contrast, the defect in SAE3 pre-mRNA splicing in tgs1Δ cells undergoing meiosis was not reversed in cbc2-Y24A tgs1Δ cells (data not shown). Thus, not all effects of the loss of TMG caps on mRNA splicing can be ameliorated by weakening the Cbc2 cap binding site.

Synthetic genetic interactions of Cbc2-Y24A with pre-mRNA splicing factors

The benign effect of Cbc2 cap binding site mutations on vegetative growth raises the prospect that the impact of these lesions, on splicing, for example, might be buffered by other actors in the splicing pathway. If so, then screening null alleles of vegetatively optional splicing factors for synthetic lethality or sickness with the hypomorphic allele cbc2-Y24A might yield genetic insights more narrowly focused on cap-dependent functions of CBC than was the original screen for synthetic interactors with a sto1Δ cbc2Δ double mutant (Fortes et al. 1999).

We focused first on the U1 snRNP subunit Nam8 and the spliceosome assembly factor Mud2, mutations of which synergize with sto1Δ cbc2Δ (Fortes et al. 1999). nam8Δ and mud2Δ null mutations have no overt impact on yeast vegetative growth but are synthetically lethal with tgs1Δ (Hausmann et al. 2008). Here, we constructed haploid yeast strains amenable to tests of synthetic lethality by either plasmid shuffle or mating. For example, a haploid strain with chromosomal cbc2-Y24A and mud2Δ loci bearing a MUD2 gene on a CEN URA3 plasmid was unable to form colonies at 30°C or 37°C on agar medium containing FOA (5-fluoroorotic acid), a drug that selects against the CEN URA3 MUD2 plasmid (Fig. 5A). Transformation of the cbc2-Y24A mud2Δ cells with CEN LEU2 plasmids bearing either MUD2 or CBC2 enabled colony formation on FOA agar (Fig. 5A). These results signify that the otherwise benign Cbc2-Y24A mutation is lethal in the absence of Mud2.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 5.

cbc2-Y24A is synthetic lethal with mud2Δ and nam8Δ and synthetic sick with mud1Δ, swt21Δ, brr1Δ, and ist3Δ. (A) Yeast cbc2-Y24A mud2Δ p360-MUD2 (URA3 CEN MUD2) or cbc2-Y24A nam8Δ p360-NAM8 (URA3 CEN NAM8) cells were transformed with CEN LEU2 plasmids harboring wild-type CBC2, MUD2, or NAM8 as specified. Leu⁺ transformants were selected at 30°C and then streaked to agar medium containing 5-FOA (0.75 mg/mL). Cells transformed with the empty LEU2 vector served as a negative control. The plates were photographed after 3 d of incubation at 30°C or 37°C. (B) Aliquots (3 μL) of serial 10-fold dilutions of haploid yeast strains of the specified genotypes were spotted on YPD agar medium. The plates were photographed after incubation for 2 d (30°C, 34°C, and 37°C), 3 d (25°C), or 5 d (20°C) as specified.

A haploid cbc2-Y24A nam8Δ strain harboring a CEN URA3 NAM8 plasmid was unable to form FOA-resistant colonies at 30°C but did at 37°C (Fig. 5A). Growth of cbc2-Y24A nam8Δ cells on FOA agar at 30°C was rescued by prior transformation with CEN LEU2 plasmids bearing either NAM8 or CBC2 (Fig. 5A). Viable cbc2-Y24A nam8Δ cells selected on FOA at 37°C were then tested for growth on YPD agar; the double mutants grew slower than either single mutant at 37°C and did not grow at 30°C or 25°C (data not shown). Thus, the Cbc2-Y24A mutation is conditionally lethal in the absence of Nam8.

Mud1 is an inessential subunit of the yeast U1 snRNP. A mud1Δ null mutation has no vegetative growth phenotype, yet mud1Δ is synthetically lethal with mud2Δ and nam8Δ, and mud1Δ is synthetically sick with tgs1Δ (Abovich et al. 1994; Chang et al. 2010; Qiu et al. 2011a). Also, Fortes et al. (1999) identified MUD1 in their screen for mutational synergy with sto1Δ cbc2Δ. Here, we mated cbc2-Y24A MUD1 and CBC2 mud1Δ strains, sporulated the diploids, and dissected tetrads to recover cbc2-Y24A mud1Δ double mutants that germinated and grew at 30°C. We compared the growth of cbc2-Y24A mud1Δ cells on YPD agar to that of CBC2 MUD1, cbc2-Y24A MUD1, and CBC2 mud1Δ cells derived from the same cross. Whereas the single mutants grew as well as wild-type yeast at all temperatures tested, the cbc2-Y24A mud1Δ strain formed smaller colonies at 30°C and 34°C and failed to thrive at higher (37°C) or lower (20°C or 25°C) temperatures (Fig. 5B, top panel).

Swt21 is an inessential splicing factor that displays mutational synergies with Tgs1, the U1 snRNP subunit Prp40, and the CBC subunit Sto1 (Murphy et al. 2004; Hausmann et al. 2008; Hage et al. 2009). We mated cbc2-Y24A SWT21 and CBC2 swt21Δ strains, sporulated the diploids, and dissected tetrads to recover viable cbc2-Y24A swt21Δ double mutants that were synthetically sick (at 25°C to 37°C) and cold-sensitive (at 20°C) compared to the single mutants, which grew as well as wild-type cells at 20°C–37°C (Fig. 5B, middle panel).

We also detected a synthetic growth defect when cbc2-Y24A was combined with brr1Δ or ist3Δ (Fig. 5B, bottom panel). Brr1 is an inessential splicing factor associated with U snRNPs that facilitates snRNP manufacture at low growth temperatures (Noble and Guthrie 1996). Yeast brr1Δ cells grow normally at 30°C–37°C but are slow-growing at 20°C. In contrast, the brr1Δ cbc2-Y24A double mutant failed to grow at 20°C, was extremely sick at 25°C and 30°C, and grew slower than either single mutant at 37°C (Fig. 5B, bottom panel). Ist3 is an inessential component of the U2 snRNP. An ist3Δ mutant grows well at 20°C–30°C but has a strong ts growth defect at 37°C. The ist3Δ cbc2-Y24A double mutant was sick at 20°C–34°C and inviable at 37°C (Fig. 5B, bottom panel).

Similar matings of cbc2–Y24A were performed with deletion mutants of yeast splicing factors Lea1 (a subunit of the U2 snRNP), Isy1 (a component of the NineTeen Complex that activates the spliceosome for catalysis), Cwc21 (a protein associated with the NineTeen Complex), and Swm2 (“synthetic with Mud2”). Viable double mutants were recovered after sporulation in all four instances, and we observed no synthetic growth defects of these double mutants when growth was tested at 20°C–37°C in parallel with the respective single mutants (data not shown).

We also tested for synthetic interactions of Cbc2-Y24A with proteins involved in RNA transactions other than pre-mRNA splicing, i.e., Lsm1 (mRNA decay), Pat1 (mRNA decay), Srb2 (RNA polymerase II transcription), and Rpn4 (transcription). We found no synthetic growth defects of the lsm1Δ cbc2-Y24A, pat1Δ cbc2-Y24A, srb2Δ cbc2-Y24A, or rpn4Δ cbc2-Y24A strains when growth was tested at 20°C–37°C in parallel with the respective single mutants (data not shown). Taken together, these results fortify the inferences that early steps of pre-mRNA splicing are the principal biological pathway impacted by the cap binding activity of yeast nuclear CBC.

Synthetic genetic interactions of Cbc2-Y24A with the yeast branchpoint binding protein Msl5

Saccharomyces cerevisiae Msl5 (branchpoint binding protein) orchestrates spliceosome assembly by binding the intron branchpoint sequence 5′-UACUAAC and establishing cross intron-bridging interactions with other components of the splicing machinery (Abovich et al. 1994; Abovich and Rosbash 1997; Rain et al. 1998; Rutz and Seraphin 1999; Wang et al. 2008; Chang et al. 2012). Unlike the optional splicing factors discussed above, Msl5 (a 476-aa polypeptide) is essential for yeast vegetative growth. The central branchpoint RNA binding domain of Msl5—composed of KH and QUA2 modules—is flanked by N- and C-terminal domains that have imputed functions in protein-protein interactions. By gauging the ability of Msl5 mutants to complement msl5Δ, we recently reported that the Mud2-binding (aa 35–54) and putative Prp40-binding (PPxY¹⁰⁰) elements of the Msl5 N-terminal domain are inessential, as are a C-terminal proline-rich domain (aa 382–476) and two zinc-binding CxxCxxxxHxxxxC motifs (aa 273–286 and 299–312) (Chang et al. 2012). A subset of conserved branchpoint RNA-binding amino acids in the central KH-QUA2 domain (aa 146–269) are essential, whereas other RNA-binding residues are dispensable. We used our collection of viable Msl5 mutants to illuminate synthetic genetic interactions between Msl5 and Mud2, Nam8, and Tgs1. The results suggested a network of important but functionally buffered protein–protein and protein–RNA interactions between the Mud2-Msl5 complex at the branchpoint and the U1 snRNP at the 5′ splice site (Chang et al. 2012). Here, we queried the genetic interactions of Cbc2-Y24A with our collection of Msl5 mutants that grow as well as wild-type MSL5 cells.

The instructive findings were that cbc2-Y24A was synthetically lethal with otherwise benign mutations of amino acids Asn163, Val165, Val195, Lys196, Thr265, Arg267, Lys252, and Arg253 in the Msl5 KH-QUA2 domain (Table 1). The NMR structure of human branchpoint binding protein (SF1) bound to an RNA (5′-AUACUAACAA) containing the consensus yeast branchpoint sequence 5′-UACUAAC (Liu et al. 2001) revealed an extensive network of contacts between many of these amino acids and the RNA nucleobases and sugars. The KH module engages the CUAACAA-3′ segment of the RNA (the branchpoint adenosine is in bold). The QUA2 module binds to the proximal 5′-AUACU segment of the RNA. The KH residue Asn163 contacts the cytosine base and Val165 contacts the adenine base preceding the branchpoint adenosine. These contacts are inessential for msl5Δ complementation by the N163A-V165A mutant but are essential for viability in the cbc2-Y24A background at all temperatures tested (Table 1). The contacts of Lys196 with the adenine base preceding the branchpoint adenosine and of Val195 with the branchpoint adenine are also inessential for msl5Δ complementation, but the combination of msl5-(V195A-K196A) and cbc2-Y24A resulted in smaller colonies at 37°C (scored as ++), microcolonies at 30°C (scored as +), and no growth at 25°C or 18°C (Table 1). The RNA interactions of QUA2 residues Thr265 (a van der Waals contact with the ApC phosphate) and Arg267 (with the ribose 2′-OH and 3′-phosphate of the 5′-terminal adenosine and with the N6, C5, and N7 atoms of the downstream adenine) are not essential for msl5Δ complementation. However, the msl5-(T265A-R267A) cbc2-Y24A double mutant was barely viable at 37°C (+ growth) and failed to grow at 30°C, 25°C, or 18°C (Table 1). The QUA2 double mutant K252A-R253A was unconditionally synthetic lethal with cbc2-Y24A. Whereas the single mutant K252R was also synthetic lethal with cbc2-Y24A at 18°C–30°C (and synthetic sick at 37°C), the single mutant R253K had no mutational synergy with cbc2-Y24A (Table 1). These results highlight how CBC-cap- and Msl5-branchpoint interactions make genetically overlapping contributions to spliceosome assembly in vivo.

In contrast, there was no equivalent synthetic lethality of cbc2-Y24A with paired cysteine-to-alanine mutations in the putative zinc-binding residues of the two zinc-knuckle domains of yeast Msl5. The C273A-C276A and C299A-C302A mutations that disrupt the proximal and distal knuckles, respectively, had no effect on cell growth in the cbc2-Y24A background (scored as +++ in Table 1).

Otherwise benign incremental C-terminal deletions of Msl5 (1-458, 1-437, 1-425, and 1-401) elicited a gradient of worsening synthetic sickness in the cbc2-Y24A background, culminating in unconditional lethality in the case of the msl5-(1-312) allele when combined with cbc2-Y24A (Table 1). Whereas deleting 34 aa from the N terminus of Msl5 had no apparent impact on yeast growth in cbc2-Y24A, extending the N-terminal deletion to 54 aa results in cold-sensitive (at 18°C) and temperature-sensitive (at 37°C) synthetic phenotypes (Table 1). Further deletion of the Msl5 segment from aa 55–68 resulted in unconditional synthetic lethality with cbc2-Y24A (Table 1). Insofar as the Msl5-(55-476) and Msl5-(69-476) mutants should be unable to form a heterodimeric complex with Mud2 (Wang et al. 2008), the mutational synergies of Cbc2-Y24A with the Msl5 N-terminal deletions are consistent with the lethality of cbc2-Y24A in the mud2Δ background (Fig. 5). Combining deletions of the terminal segments 1–54 and 438–458 (neither of which, per se, affected the growth of cbc2-Y24A at 30°C) resulted in unconditional lethality in the cbc2-Y24A background (Table 1). Msl5 contains a ⁹⁷PPxY¹⁰⁰ motif recognized by WW domain proteins (Wiesner et al. 2002). Whereas a triple-alanine mutant msl5-(P97A-P98A-Y100A) fully complements msl5Δ, we find that this allele is unconditionally lethal in the cbc2-Y24A background (Table 1). In sum, these results reveal manifold genetic connections between CBC cap binding and the branchpoint binding protein that had not been appreciated previously.

Effects of Cbc2 cap binding site mutations on SUS1 pre-mRNA splicing

The growth defects of sto1Δ and cbc2Δ null mutants are caused, at least in part, by aberrant histone modification, specifically increased histone H2B ubiquitylation (Hossain et al. 2009). A contributing factor is that splicing of the SUS1 mRNA encoding a ubiquitin protease is reduced in sto1Δ and cbc2Δ cells. SUS1 is one of the few yeast genes that contain two introns (Fig. 6), and it is the splicing of the first intron (which has nonconsensus 5′ splice site and branchpoint sequences) that is selectively impaired in the absence of CBC (Hossain et al. 2009). Here, we examined the effects of Cbc2 cap binding site lesions Y24A and NΔ42 on the SUS1 mRNA splicing pattern. Total RNA was isolated from CBC2, cbc2Δ, cbc2-Y24A, and cbc2-NΔ42 cells grown in liquid culture at 30°C and used as a template for cDNA synthesis by reverse transcriptase primed by oligo(dT). The cDNA was then amplified by PCR with SUS1 gene-specific primers corresponding to the sequences of exon 1 and exon 3 (Fig. 6). As reported previously (Hossain et al. 2009), the SUS1 transcripts in wild-type CBC2 cells consist predominantly of mature mRNA and a minority intermediate species in which the second intron is excised but the first intron is not (Fig. 6). In cbc2Δ cells, there is little mature mRNA, and the singly spliced intermediate comprises the predominant species, along with a significant fraction of unspliced pre-mRNA (Fig. 6). The salient findings were that (1) cbc2-Y24A cells evinced a wild-type pattern of SUS1 splicing, and (2) cbc2-NΔ42 cells showed just a slight decrement in joining of the first and second exons (Fig. 6). These results underscore that mutations of the cap-binding pocket of Cbc2 exert hypomorphic effects on splicing that contrast with the more severe effects of ablating CBC.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 6.

Effects of Cbc2 Y24A and NΔ42 mutations on SUS1 pre-mRNA splicing. RNAs isolated from cbc2Δ, CBC2, cbc2-Y24A, and cbc2-NΔ42 cells were reverse transcribed with an oligo(dT) primer, and the cDNAs were PCR-amplified with primers in the first and third exons of chromosomal SUS1 gene (depicted at the bottom). The PCR products were resolved by native agarose gel electrophoresis and visualized by staining with ethidium bromide. The right lane shows the product of PCR-amplification of genomic DNA with the SUS1 primers. The positions and sizes (bp) of linear duplex DNA markers are indicated on the right. The positions of the RT-PCR products of unspliced, partially spliced, and fully spliced SUS1 transcripts are indicated at left.

Because the SUS1 mRNA splicing defect is a straightforward molecular marker of the CBC null phenotype, it was of interest to gauge whether this defect was reversed in the sto1Δ cbc2Δ strains S1, S2, and S3 that displayed enhanced vegetative growth. Thus, we analyzed the SUS1 splicing patterns by RT-PCR amplification of SUS1 transcripts isolated from wild-type, sto1Δ, sto1Δ cbc2Δ, and the S1, S2, and S3 suppressor strains (Supplemental Fig. S3). The S1, S2, and S3 strains maintained a defective SUS1 splicing pattern whereby the partially spliced RNA retaining the proximal intron was the predominant species (Supplemental Fig. S3). Yet, the S1, S2, and S3 strains did form slightly higher levels of the fully spliced SUS1 transcript compared to that seen in the sto1Δ single mutant and the sto1Δ cbc2Δ double mutant, said levels still being much less than the mature SUS1 transcript seen in wild-type CBC2 cells (Supplemental Fig. S3). We surmise that the suppression of the CBC-null growth defect does not negate the CBC dependence of SUS1 splicing.

The Cbc2 N-terminal peptide is critical for yeast sporulation and meiosis

The yeast meiotic developmental program entails a shift in the processing patterns of specific meiotic pre-mRNAs from a vegetative “off” state, in which single introns are included, to a meiotic “on” state, in which the target introns are removed. The efficiency of meiotic intron removal is either regulated or governed by the actions of splicing factors or RNA modifying enzymes that are inessential for vegetative growth (e.g., Mer1, Nam8, Tgs1) (Engebrecht et al. 1991; Spingola and Ares 2000; Munding et al. 2010; Qiu et al. 2011a,b,c). The yeast sae1-1 mutation causes a severe sporulation defect in an otherwise wild-type background that is associated with a delay and decrement in meiotic recombination and a transient meiotic prophase arrest (McKee and Kleckner 1997). The wild-type SAE1 gene was isolated from a genomic library by complementation of the sae1-1 sporulation defect and then identified as CBC2 (McKee and Kleckner 1997). Whereas the nature of the sae1-1 mutation was not defined in the original study, our colleague Scott Keeney has since sequenced the sae1-1 locus and found that it contains a single mutation at nucleotide +3 in the translation start codon: from AUG to AUA. The predicted impact of this change would be to shift the translation start site to the next available in-frame AUG encoding Met43 (Fig. 2), assuming that the scanning ribosome elides an intervening out-of-frame AUG at nucleotides 32–34 of the ORF. Having shown above that the Cbc2-(43-208) protein (NΔ42) sustains normal vegetative growth of haploid cells, we evaluated the effects of the cbc2-NΔ42 allele on yeast sporulation, with the intent to create a “cleaner” version of sae1-1, uncomplicated by translational frame issues.

The sporulation experiments were performed in the SKY strain background. We monitored the appearance of four-spore asci as a function of time after transfer of a culture of yeast SKY diploid cells to sporulation medium. Wild-type SKY efficiently and synchronously formed asci between eight and 14 h and attained 86% and 90% sporulation efficiencies at 14 and 24 h, respectively (Fig. 7). In contrast, an isogenic cbc2-NΔ42 diploid was defective in executing the meiotic program, with delayed onset and a slowed rate of ascus formation, to extents of only 5% and 19% after 14 and 24 h, respectively (Fig. 7). (The sae1-1 mutant yielded 20% four-spore asci at 24 h [McKee and Kleckner 1997].) Spore viability was gauged by tetrad dissection (at the 24-h time point) and quantified as the percent of spores germinating to form macroscopic colonies after incubation for 3 d at 30°C on YPD agar medium. The spore viability values were 98% (310/316 viable spores) for the wild-type strain and 61% (413/676 viable spores) for the cbc2-NΔ42 strain. We conclude that an intact cap binding site in Cbc2 is needed for yeast sporulation, presumably because a key meiotic RNA transaction is facilitated by CBC interactions with RNA caps.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 7.

cbc2-NΔ42 diploids are defective for sporulation. Homozygous wild-type, swt21Δ, mud2Δ, and cbc2-NΔ42 SKY diploids were examined by light microscopy at the indicated times after transfer to sporulation medium. The percentages of the cell population comprising four-spore asci are plotted as a function of time.

We also tested the effects on sporulation of null mutants of two splicing factors—Mud2 and Swt21—that displayed mutational synergy with the hypomorphic Y24A mutant of Cbc2 during vegetative growth. The kinetics and extent of sporulation of the swt21Δ diploid were similar to that of the isogenic wild-type diploid (Fig. 7). In contrast, the mud2Δ diploid was defective for sporulation, yielding 12% and 22% asci at 14 and 24 h, respectively (Fig. 7).

Defective meiotic splicing of SAE3 and MER3 pre-mRNAs in cbc2-NΔ42 cells

We isolated total RNA from CBC2 and cbc2-NΔ42 SKY diploids 4 h after transfer to sporulation medium. cDNA was prepared from each RNA sample by reverse transcription primed by a mixture of antisense primers complementary to the 3′ exons of 14 meiotically spliced yeast mRNAs (Table 2). The cDNA preparations were then used as templates for gene-specific PCR amplification (Qiu et al. 2011a,b). The sense and antisense PCR primers corresponded to sequences flanking the introns so that the longer products of amplification of cDNA derived from unspliced pre-mRNAs could be easily resolved by native gel electrophoresis from the shorter products of amplification of cDNAs copied from spliced mRNA. The sense strand primer was 5′ ³²P-labeled in each PCR reaction so that we could quantify the distributions of unspliced and spliced cDNAs for each gene of interest. The results are compiled in Table 2, wherein each datum for splicing efficiency—[spliced/(spliced + unspliced)] × 100—is the average of three independent sporulation experiments and RT-PCR analyses. We operationally defined a significant mutational effect on meiotic splicing as one that elicits a greater than or equal to twofold reduction in splicing efficiency compared to a wild-type control (Qiu et al. 2011a,b). SAE3 and MER3 were the only two transcripts that met our criterion for a meiotic splicing defect in the cbc2-NΔ42 strain. SAE3 splicing efficiency was 81% in CBC2 cells versus 20% in cbc2-NΔ42. MER3 splicing efficiency was 79% in the CBC2 cells versus 32% in cbc2-NΔ42 (Table 2). In contrast, the cbc2-NΔ42 mutation had no adverse effect on splicing of the AMA1, MER2, HOP2, REC114, REC102, DMC1, PCH2, SPO1, and SRC1 transcripts (Table 2) and elicited only modest decrements in the splicing of MEI4, SPO22, and MND1 that did not meet our criterion of significance (Table 2).

Rescue of the cbc2-NΔ42 sporulation defect by expression of intronless cDNAs

The RNA analysis in Table 2 suggested that SAE3 and MER3 comprise a novel meiotic splicing “regulon” governed by Cbc2's cap binding activity. As shown previously (Qiu et al. 2011c), the working definition of a “complete” meiotic splicing regulon is the ability to rescue the sporulation defect caused by mutation of the splicing “governor” by expressing intronless cDNA versions of the essential meiotic RNAs that are targeted by the governor (which, in this case, are SAE3 and MER3). Therefore, we introduced intronless cSAE3 and cMER3 (under the control of their native promoters) into the chromosomal leu2∷hisG locus of cbc2-NΔ42 SKY diploids and then assessed their sporulation efficiency. In parallel, we analyzed sporulation by CBC2 diploids (positive control), cbc2-NΔ42 diploids with no cDNAs (negative control), and CBC2 diploids coexpressing cSAE3 and cMER3. The ectopic cDNA genes had no negative impact on the kinetics of sporulation of CBC2 cells (Fig. 8A). The salient finding was that the rate and extent of sporulation of the cbc2-NΔ42 strain was restored fully to the wild-type pattern by coexpression of cSAE3 and cMER3 (Fig. 8A). The spore viability value after tetrad dissection was 95% for the cbc2-NΔ42 + cSAE3 + cMER3 strain (148/156 viable spores) compared to 61% for cbc2-NΔ42. These results suggest that defective splicing of specific meiotic transcripts underlies the failure of cbc2-NΔ42 cells to execute the meiotic program.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 8.

An ectopic MER3 cDNA rescues the cbc2-NΔ42 sporulation defect. (A) CBC2 and cbc2-NΔ42 diploids with or without the cSAE3 and cMER3 cDNAs integrated at the chromosomal leu2hisG locus and (B) cbc2-NΔ42 diploids and derivatives with either cSAE3 or cMER3 integrated at leu2hisG were examined by light microscopy at the indicated times after transfer to sporulation medium. The percentages of the cell population comprising four-spore asci are plotted as a function of time.

To further define the nature of the defect, we generated cbc2-NΔ42 diploids bearing only cMER3 or cSAE3 inserted at the chromosomal leu2hisG locus. The key finding was that expression of just the MER3 cDNA sufficed to restore both the wild-type pattern of sporulation kinetics in the cbc2-NΔ42 background (Fig. 8B) and a wild-type level of spore viability (92%; 218/236 viable spores). In contrast, cbc2-NΔ42 cells expressing just the SAE3 cDNA exhibited the same delay in the onset of sporulation seen in the parental cbc2-NΔ42 strain and a slower rate of accrual of four-spore tetrads than that seen for the CBC2 and cbc2-NΔ42 + cMER3 strains (Fig. 8B). Nonetheless, cSAE3 expression did enhance the rate of spore formation versus cbc2-NΔ42, such that the yield of four-spore asci at 24 h was 45% with cSAE3 versus 22% without (Fig. 8B). However, the spore viability of the cbc2-NΔ42 + cSAE3 strain was 60% (154/256 viable spores), i.e., the same as cbc2-NΔ42. Taken together, these experiments implicate MER3 splicing as a limiting transaction in cbc2-NΔ42 cells undergoing meiosis.

To evaluate whether the rescue of cbc2-NΔ42 meiosis by ectopic expression of cMER3 might simply reflect increased MER3 gene dosage, we replaced the endogenous intron-containing chromosomal MER3 locus of SK1 haploids with an intronless MER3 cDNA, mated them to yield cbc2-NΔ42 cMER3 diploids, and then analyzed their sporulation. The allelic replacement of MER3 by cMER3 fully restored wild-type sporulation kinetics (Fig. 9A), indicating that bypass of feeble MER3 splicing in the cbc2-NΔ42 background was responsible for the rescue of the meiotic defect.

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

FIGURE 9.

Nonconsensus MER3 intron features dictate the cbc2-NΔ42 sporulation defect. The nucleotide sequence of the MER3 intron of yeast strain W303 is shown, highlighting its nonconsensus 5′ splice site and branchpoint and the location of its Mer1 enhancer (shaded in cyan). The point mutations (5′SS and BP) that we introduced into the MER3 intron are indicated. A single C-to-T difference in the MER3 intron of the yeast SK1 strain (from which SKY is derived) is highlighted in red. (A) cbc2-NΔ42 diploids with the indicated chromosomal MER3 alleles were examined by light microscopy at the indicated times after transfer to sporulation medium. The percentages of the cell population comprising four-spore asci are plotted as a function of time. (B) RNAs isolated from the indicated diploid CBC2 or cbc2-NΔ42 SKY strains (with chromosomal MER3 alleles as specified) at 4 h post-transfer to sporulation medium were reverse transcribed with a MER3 antisense primer complementary to the 3′ exon. The cDNAs were then PCR-amplified with sense and antisense primers flanking the MER3 intron. The splicing efficiencies are plotted; each datum is the average of three separate experiments ±SEM.

Nonconsensus MER3 intron features dictate the cbc2-NΔ42 sporulation defect

MER3 splicing during yeast meiosis is known to be dependent on Mer1, a splicing enhancer protein that is produced only in meiotic cells (Engebrecht et al. 1991; Nakagawa and Ogawa 1999). MER3, together with MER2, AMA1, and SPO22, comprise a four-gene meiotic splicing regulon controlled by Mer1 and the vegetatively inessential U1 snRNP subunit Nam8 (Spingola and Ares 2000; Qiu et al. 2011a,c). Mer1 up-regulation of MER3 splicing relies on the binding of the Mer1 protein to an enhancer element, 5′-ACACCCUU, located in the MER3 intron next to the 5′ splice site (Fig. 9; Spingola and Ares 2000; Qiu et al. 2011a). The MER3 intron has a nonconsensus 5′ splice site that is the decisive factor in Nam8/Mer1-dependency (Qiu et al. 2011a); two other members of the regulon (MER2 and SPO22) also have nonconsensus 5′ splice sites that dictate their reliance on Mer1 and Nam8 (Nandabalan et al. 1993; Qiu et al. 2011a). In light of the present findings that MER3 is the only member of the Mer1/Nam8 meiotic splicing regulon that is acutely sensitive to the cbc2-NΔ42 mutation, we sought to identify the features of the MER3 transcript that dictate this sensitivity.

Inspection of the MER3 intron reveals two features not found in other Mer1/Nam8 targets: (1) a deviant 5′ splice site (GUAGUA) found in no other yeast intron; and (2) a rare nonconsensus branchpoint (GACUAAC) (Fig. 9). We initiated an analysis of the cbc2-NΔ42 sensitivity of MER3 splicing by installing consensus 5′ splice site (5′SS = GUAUGU) and branchpoint (BP = UACUAAC) signals in the MER3 gene, singly and in combination (Fig. 9). These intron mutations were constructed in the MER3 gene derived from yeast strain W303 (Qiu et al. 2011a). The W303 MER3 intron differs from that of the SK1 strain by a single nucleobase change (T in SK1 versus C in W303) at a site between the branchpoint and the 3′ splice site (Fig. 9). Consequently, we also constructed a new control MER3 SKY diploid strain (designated WT* in Fig. 9) in which the W303 version of the MER3 intron was introduced at the two chromosomal SK1 MER3 loci. The single-base intron strain variation had no significant effect on the defective sporulation pattern of the cbc2-NΔ42 strain (Fig. 9A). As expected, the viability of the cbc2-NΔ42 MER3-WT* spores was compromised (48%; 30/80 viable spores).

The instructive finding was that introducing a perfect GUAUGU 5′ splice site (via a single U insertion, as shown in Fig. 9) restored wild-type sporulation kinetics in the cbc2-NΔ42 strain (Fig. 9A) as well as spore viability (94%; 75/80 viable spores). The consensus BP change elicited a partial restoration of the sporulation pattern, albeit with a residual kinetic lag in the appearance of four-spore asci (Fig. 9A). Nonetheless, the viability of the spores derived from the cbc2-NΔ42 MER3-BP strain was 94% (75/80 viable spores). The effects of combining the MER3 5′SS and BP changes were virtually identical to those seen for the single 5′SS change (Fig. 9A). These results signify that the aberrant MER3 5′ splice site and branchpoint are independent determinants of the sensitivity of MER3 splicing to cbc2-NΔ42, with the 5′ splice site exerting an apparently greater influence in this regard.

To relate sporulation with MER3 splicing, we performed RT-PCR analysis of the MER3 transcripts in cells isolated 4 h after transfer to sporulation medium. As expected, control CBC2 strains displayed high efficiencies of splicing of the WT and WT* MER3 transcripts (93%), whereas splicing was inefficient (33%) in cbc2-NΔ42 cells (Fig. 9B). Installing a consensus 5′ splice site increased MER3 splicing in cbc2-NΔ42 cells to 99%. The consensus BP change elicited a lesser increase in MER3 splicing, to 66% (Fig. 9B). Thus, the sporulation phenotype of cbc2-NΔ42 cells correlates with the efficiency of MER3 splicing.