Fission yeast poly(A) polymerase active site mutation Y86D alleviates the rad24Δ asp1-H397A synthetic growth defect and up-regulates mRNAs targeted by MTREC and Mmi1

Abstract

Expression of fission yeast Pho1 acid phosphatase is repressed under phosphate-replete conditions by transcription of an upstream prt lncRNA that interferes with the pho1 mRNA promoter. lncRNA-mediated interference is alleviated by genetic perturbations that elicit precocious lncRNA 3′-processing and transcription termination, such as (i) the inositol pyrophosphate pyrophosphatase-defective asp1-H397A allele, which results in elevated levels of IP₈, and (ii) absence of the 14-3-3 protein Rad24. Combining rad24Δ with asp1-H397A causes a severe synthetic growth defect. A forward genetic screen for SRA (Suppressor of Rad24 Asp1-H397A) mutations identified a novel missense mutation (Tyr86Asp) of Pla1, the essential poly(A) polymerase subunit of the fission yeast cleavage and polyadenylation factor (CPF) complex. The pla1-Y86D allele was viable but slow-growing in an otherwise wild-type background. Tyr86 is a conserved active site constituent that contacts the RNA primer 3′ nt and the incoming ATP. The Y86D mutation elicits a severe catalytic defect in RNA-primed poly(A) synthesis in vitro and in binding to an RNA primer. Yet, analyses of specific mRNAs indicate that poly(A) tails in pla1-Y86D cells are not different in size than those in wild-type cells, suggesting that other RNA interactors within CPF compensate for the defects of isolated Pla1-Y86D. Transcriptome profiling of pla1-Y86D cells revealed the accumulation of multiple RNAs that are normally rapidly degraded by the nuclear exosome under the direction of the MTREC complex, with which Pla1 associates. We suggest that Pla1-Y86D is deficient in the hyperadenylation of MTREC targets that precedes their decay by the exosome.

INTRODUCTION

Expression of the fission yeast pho1 gene encoding a cell surface acid phosphatase is repressed under phosphate-replete conditions by synthesis of an upstream prt lncRNA that interferes with activation of the pho1 mRNA promoter by transcription factor Pho7 (Fig. 1A; Schwer et al. 2017; Shuman 2020). Two other genes of the fission yeast Pho7-dependent phosphate acquisition (PHO) regulon—pho84 (inorganic phosphate transporter) and tgp1 (glycerophosphodiester transporter)—are similarly repressed by upstream interfering lncRNAs, prt2 and nc-tgp1, respectively (Shuman 2020). lncRNA-mediated transcription interference is alleviated by genetic perturbations that elicit precocious lncRNA 3′-processing and termination in response to proximal poly(A) signals in the prt lncRNA, resulting in derepression of Pho1 (Fig. 1B). Conversely, maneuvers that diminish the probability of terminating precociously result in Pho1 hyper-repression (Fig. 1A). Thus, Pho1 expression in phosphate-replete cells provides a sensitive read-out of genetic influences on 3′-processing/termination and a tool for discovery of regulators of this phase of the RNA polymerase II (Pol2) transcription cycle (Henry et al. 2011; Shah et al. 2014; Schwer et al. 2015, 2020, 2021; Chatterjee et al. 2016; Sanchez et al. 2018b, 2019, 2023).

FIGURE 1.

Genetic modulation of transcriptional interference with pho1 expression. (A) lncRNA transcription across the pho1 mRNA promoter interferes with pho1 mRNA synthesis. Genetic maneuvers that diminish the probability of lncRNA 3′-processing/termination prior to the mRNA promoter increase interference and hyper-repress pho1 mRNA expression. Genetic backgrounds in which Pho1 expression is hyper-repressed are shown at right. (B) Precocious termination of lncRNA synthesis in response to poly(A) signals well upstream of the pho1 mRNA promoter can result in derepression of pho1 mRNA transcription. Genetic backgrounds in which Pho1 expression is derepressed are shown at right.

For example, we found that prt lncRNA termination is subject to metabolite control by inositol pyrophosphate 1,5-IP₈, which acts as an agonist of precocious 3′-processing/termination (Sanchez et al. 2019). Fission yeast Asp1, the principal agent of IP₈ dynamics, is a bifunctional enzyme composed of an amino-terminal inositol pyrophosphate kinase domain and a carboxy-terminal inositol pyrophosphate pyrophosphatase domain (Pascual-Ortiz et al. 2018; Dollins et al. 2020; Benjamin et al. 2022). Deletion of Asp1, or kinase-inactivating mutation Asp1-D333A, eliminates intracellular IP₈ and causes hyper-repression of Pho1, while pyrophosphatase-defective mutation Asp1-H397A increases intracellular IP₈ and results in Pho1 derepression (Pascual-Ortiz et al. 2018; Sanchez et al. 2019). Too much IP₈, as a consequence of certain Asp1 pyrophosphatase mutant alleles, or of combining Asp1-H397A with deletion of a second inositol pyrophosphate pyrophosphatase enzyme Aps1, is toxic to the point of lethality (Sanchez et al. 2019; Garg et al. 2020). IP₈ lethality is overcome by loss-of-function mutations of 3′-processing and transcription termination factors CPF (the 13-subunit cleavage and polyadenylation factor complex), Pin1, and Rhn1 (Sanchez et al. 2019, 2020; Garg et al. 2020). These results suggest that IP₈ lethality is the consequence of overzealous 3′-processing/termination affecting one or more essential fission yeast genes. By genetic screening for suppressors of inositol pyrophosphate toxicosis, we are identifying additional cellular proteins and pathways that connect IP₈ to gene expression (Schwer et al. 2022). These include fission yeast proteins that have an SPX domain implicated in inositol pyrophosphate sensing (Wild et al. 2016; Schwer et al. 2022).

A separate suppressor screen for mutations that derepressed Pho1 led to the identification of 14-3-3 protein Rad24 (a phosphopeptide-binding protein) as an antagonist of precocious Pol2 transcription termination (Garg et al. 2022). Production of full-length interfering prt lncRNA was squelched in rad24Δ cells, accompanied by increased production of pho1 mRNA and increased Pho1 activity. Epistasis analyses showed that pho1 derepression by rad24Δ depended on CPF, Pin1, Rhn1, and the Thr4-PO₄ mark of the Pol2 CTD (Garg et al. 2022).

Instructive findings emerged from attempts to combine the pho1 derepressive rad24Δ and asp1-H397A mutations. The genetic cross yielded double-mutant rad24Δ asp1-H397A haploids that germinated and grew out very slowly after sporulation, grew poorly in liquid YES medium, and did not form macroscopic colonies on YES agar (Garg et al. 2022). In effect, the combination of these two strongly pho1 derepressive alleles was synthetically near-lethal. To interrogate the basis for the severe growth defect of rad24Δ asp1-H397A cells, we isolated spontaneous SRA (Suppressor of Rad24 Asp1) mutants that gave rise to rare larger colonies when the rad24Δ asp1-H397A strain was plated on YES agar at 30°C. We initially picked two candidate suppressor mutants—SRA-1 and SRA-2—which, after isolation of a single colony and amplification in liquid culture, yielded a homogeneous population of larger-than-parental colonies. Whole-genome sequencing of the SRA-1 and SRA-2 strains identified a single, apparently causative, suppressor mutation in each strain that mapped to a subunit of the CPF complex. SRA-1 cells bore a nonsense mutation in the ctf1 gene that truncated the 363-aa Ctf1 subunit of CPF at Tyr260. SRA-2 cells had a nonsense mutation in the ssu72 gene that truncated the 197-aa Ssu72 subunit of CPF at Gln67 (Garg et al. 2022). That an unbiased suppressor screen returned loss-of-function mutations in two different subunits of the CPF complex fortified the conclusion that the toxic effect of rad24Δ asp1-H397A is exerted at the level of RNA 3′-processing/termination.

In the present study, we extended the SRA screen with the goal of identifying new alleles that impact 3′-processing/termination. This yielded SRA-3 as a novel suppressor attributable to a single missense mutation Y86D in the pla1 gene encoding the poly(A) polymerase subunit of CPF. We describe here the characterization of the Pla1 enzyme and the effects of mutating Tyr86 in vitro and in vivo.

RESULTS

SRA screen yields a novel mutation in the active site of poly(A) polymerase Pla1

The SRA-3 strain was isolated as a spontaneous suppressor of the near-lethality of rad24Δ asp1-H397A cells. SRA-3 was slower growing than the rad24Δ single mutant at all temperatures, as gauged by colony size on YES agar (Fig. 2A), signifying that the suppression was partial. The rad24Δ and asp1-H397A single mutations elicit derepression of pho1 gene expression, which is reflected in 21-fold and 10-fold increases in cell surface-associated Pho1 acid phosphatase activity, respectively, vis-à-vis wild-type cells (Fig. 2C). Pho1 expression in SRA-3 cells (15-fold greater than wild-type) was intermediate between that of rad24Δ and asp1-H397A, consistent with partial suppression of the altered phosphate homeostasis phenotype by SRA-3. [The severe growth defect of the parental rad24Δ asp1-H397A strain confounded the direct comparison of its Pho1 activity to that of SRA-3].

FIGURE 2.

SRA screen identifies a novel mutation in poly(A) polymerase Pla1. (A,B) Serial dilutions of the indicated fission yeast strains were spotted on YES agar and grown at the temperatures specified. (C,D) The indicted strains were grown to A₆₀₀ of 0.5–0.8 in liquid culture in YES medium at 30°C. Cells were then harvested, washed with water, and assayed for Pho1 acid phosphatase activity by conversion of p-nitrophenylphosphate to p-nitrophenol. Activity is expressed as the ratio of A₄₁₀ (p-nitrophenol production) to A₆₀₀ (input cells). (E) Aliquots (5 µg) of purified recombinant wild-type Pla1 and the indicated mutants were analyzed by SDS–PAGE. The Coomassie-blue stained gel is shown. The positions and sizes (kDa) of marker polypeptides are indicated on the left. (F) View of the active site of budding yeast poly(A) polymerase (D154A mutant) in complex with ATP (stick model with blue carbon and yellow phosphorus atoms), magnesium (green sphere), and ApA_OH primer (stick model with gray carbons) (pdb 2Q66). Selected amino acids are shown as stick models with beige carbons and numbered according to their conserved counterparts in fission yeast Pla1. Hydrogen bonds and contacts to magnesium are indicated by black dashed lines. Van der Waals contacts of Tyr86 to the ATP ribose are denoted by green dashed lines. The trajectory of the nucleophilic attack of the ApA primer O3′ on the ATP α-phosphorus is indicated by a magenta dashed line.

Whole-genome sequencing of the SRA-3 strain identified a single, apparently causative, suppressor mutation in pla1, the gene encoding the poly(A) polymerase subunit of the CPF complex (Ohnacker et al. 1996; Vanoosthuyse et al. 2014). SRA-3 cells bore a Y86D missense mutation in the 566-aa Pla1 protein. Insights into the nature of this mutation were gleaned by reference to the crystal structure of budding yeast poly(A) polymerase in complex with ATP•Mn²⁺ and an oligo(A) primer (Balbo and Bohm 2007). The fission yeast and budding yeast enzymes share 373 positions of amino acid side chain identity/similarity (Supplemental Fig. S1). Figure 2F shows the budding yeast poly(A) polymerase active site, with conserved amino acids labeled according to their positions in fission yeast Pla1. Tyr86 is a constituent of the active site: it makes a hydrogen bond from the phenolic hydroxyl to the ribose 2′-OH of the primer terminus and van der Waals contacts to the ATP ribose (Fig. 2F). The equivalent of this Tyr residue is a Phe in the mammalian, Candida, Yarrowia, and Hydra poly(A) polymerase orthologs. Its replacement by Asp (nominally isosteric with the C β, γ, and δ atoms of Phe) would subtract two of the atomic contacts and introduce a potentially repulsive negative charge.

Effect of active site mutations on fission yeast growth and Pho1 expression

Whereas deletion of the fission yeast pla1 gene is lethal (Kim et al. 2010), it is not clear whether the essentiality of Pla1 reflects a need for its catalytic activity in poly(A) tail formation or its role as a structural component of the 13-subunit CPF complex (Vanoosthuyse et al. 2014). To our knowledge, this question has not been addressed via targeted mutation of putative catalytic residues of the Pla1 active site. To clarify the issue, and to address structure–activity relations at Tyr86, we constructed fission yeast strains in which single missense mutations were introduced at one pla1 allele in a diploid strain along with a 3′-flanking drug resistance marker. A similarly marked wild-type pla1 allele was introduced as a control. Sporulation of the diploid and screening large populations of random haploid progeny for the marked pla1 allele provides an assessment of mutational effects on viability. To probe the need for catalysis, we introduced alanine in lieu of Asp153, which is predicted to coordinate the catalytic Mn²⁺ that engages the RNA primer O3′ and the ATP α-phosphate (Bard et al. 2000). (Note that this manganese is not present in the structure shown in Fig. 2F because the corresponding Asp154 of budding yeast enzyme was mutated to alanine to prevent catalysis [Balbo and Bohm 2007]). Our finding that the pla1-D153A mutation was lethal (i.e., we recovered no viable drug-resistant pla1-D153A haploid progeny after sporulation of the pla1⁺ pla1-D153A diploid strain) implies that poly(A) synthesis by Pla1 is necessary for viability of fission yeast.

We replaced Tyr86 with Asp (the original SRA-3 mutation, now in an otherwise wild-type background) as well as with Phe, Ala, and Leu. We recovered viable haploids in each case. Spot tests of growth on YES agar indicated that: (i) the Y86F and Y86L strains grew well at all temperatures; (ii) Y86A cells grew slower than wild-type at 30°C, 25°C, and 20°C; and (iii) the Y86D strain was sick and phenocopied growth of SRA-3 (Fig. 2B). Assay of the Tyr86 mutants for Pho1 expression under phosphate-replete conditions showed that Y86F was equivalent to wild-type, whereas Leu, Ala, and Asp substitutions elicited progressive and statistically significant hyper-repression of Pho1 (Fig. 2D).

Poly(A) polymerase activity of recombinant Pla1

Poly(A) polymerase from budding yeast has been extensively characterized biochemically and structurally (Lingner et al. 1991; Zhelkovsky et al. 1998; Bard et al. 2000; Balbo et al. 2005; Balbo and Bohm 2007, 2009). Comparatively little is known about the biochemical properties of fission yeast Pla1 beyond the fact that recombinant Pla1 possesses manganese-dependent poly(A) polymerase activity in vitro (Ohnacker et al. 1996; Kühn et al. 2017). Here we produced wild-type Pla1 and mutants Y86D, Y86A, Y86L, Y86F, and D153A in Escherichia coli and purified them from soluble bacterial extracts (Fig. 2E). As a prelude to assessing mutational effects on Pla1 activity, we performed a basic characterization of the purified wild-type enzyme by assaying its ability to extend a 5′ ³²P-labeled 12-mer RNA primer (pUUUCUGUGGCUA) in the presence of 1 mM manganese and 1 mM ATP. A kinetic profile of the reaction of 0.04 µM Pla1 with 0.1 µM primer is shown in Figure 3A. Pla1 effected near-quantitative elongation of the input primer (97% extension) and the products increased in length in lockstep manner consistent with a distributive process. Reference to a 5′-radiolabeled DNA marker ladder analyzed in parallel indicated that the 12-mer primer was extended to ∼90 nt by 2 min, ∼250 nt by 5 min, and >600 nt by 15 min (Fig. 3A). Manganese was the preferred divalent cation cofactor for Pla1 (Fig. 3B). Cobalt and magnesium supported near-quantitative primer utilization, but the polyadenylated products were shorter than those generated with manganese. Nickel and zinc enabled inefficient primer utilization, with only a few cycles of AMP addition. Calcium, cadmium, and copper were inactive (Fig. 3B). A metal mixing experiment was performed in which reactions containing 0.5 mM manganese were supplemented with 0.5 mM calcium, cadmium, copper, nickel, or zinc (Fig. 3C). Whereas calcium and nickel did not affect the extent of manganese-dependent poly(A) addition, cadmium, copper, and zinc were severely inhibitory. These results suggest that cadmium, copper, and zinc compete with manganese for one or both metal-binding site(s) on Pla1 and, whence engaged, are ineffective in supporting catalysis. Pla1 is extremely specific for ATP as the substrate for polymerization (Fig. 3D). Replacement of ATP with 1 mM GTP, CTP, or UTP allowed for inefficient primer utilization and only one or two cycles of NMP addition (Fig. 3D). dATP in lieu of ATP effected the elongation of 86% of the primer (vs. 97% with ATP), albeit predominant by a single dAMP step (Fig. 3D).

FIGURE 3.

Characterization of Pla1 poly(A) polymerase activity. (A) Time course of primer extension. A reaction mixture (85 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 1 mM ATP, 0.1 µM 5′-³²P labeled RNA primer, and 0.04 µM Pla1 were incubated at 30°C. At times specified, aliquots (10 µL) were withdrawn and quenched immediately by adjustment to 47.5% formamide/12.5 mM EDTA. (B) Divalent cation specificity. Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM ATP, 0.1 µM (1 pmol) 5′-³²P labeled RNA primer, and 0.04 µM Pla1 (400 fmol), and either no metal (−) or 1 mM of the indicated divalent cation (as the chloride salt or CdSO₄) were incubated at 30°C for 30 min and subsequently quenched. (C) Metal mixing. Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 0.5 mM MnCl₂, 1 mM ATP, 0.1 µM (1 pmol) 5′-³²P labeled RNA primer, either no enzyme (−E) or 0.04 µM Pla1 (400 fmol), and either no other metal (−) or 0.5 mM of the indicated divalent cation (as the chloride salt or CdSO₄) were incubated at 30°C for 30 min and subsequently quenched. (D) NTP specificity. Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 0.1 µM 5′-³²P labeled RNA primer, and 0.04 µM Pla1, and either no added NTP (−) or 1 mM ATP (A), GTP (G), CTP (C), UTP (U), or dATP (dA) as specified were incubated at 30°C for 30 min and subsequently quenched. Primer extension products for all assays shown in Figure 3 were resolved by 8% urea-PAGE and visualized by autoradiography. The size in nucleotides of a DNA ladder (MspI digest of pBR322 plasmid) is indicated on the right of each panel. The proportion of the primer extended by 1, 2, or 3 nt in D was quantified in ImageQuant-TL after scanning the gel with a Typhoon FLA7000 imager.

The dATP result could reflect either an inherent specificity for rATP as the incoming substrate or the inability to extend the primer after dAMP is incorporated at the 3′ terminus. To parse this issue, we compared a series of ³²P-labeled 12-mer primers of identical nucleobase sequence composed of all ribonucleotides (R12), all deoxynucleotides (D12), 11 deoxynucleotides and a single 3′-terminal ribonucleotide (D11R1), and 11 ribonucleotides and a single 3′-terminal deoxynucleotide (R11D1) (Fig. 4A). These primers (0.1 µM each) were reacted for 30 min with 0.1 µM Pla1 and 1 mM ATP. D12 was not an effective primer, insofar as only 14% of the input D12 was extended by a single AMP addition step and only 3% was converted into longer poly(A) tails (Fig. 4A). The D11R1 and R11D1 primers were polyadenylated, albeit with lower efficiencies of primer utilization, 77% and 72%, respectively (Fig. 4A).

FIGURE 4.

Primer specificity for RNA versus DNA. (A) Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 1 mM ATP, 0.1 µM (1 pmol) of the indicated 5′-³²P labeled primer (R12—RNA; D12—DNA; D11R1—DNA with terminal 3′ ribonucleotide; R11D1—RNA with terminal 3′ deoxynucleotide), and either no added enzyme (−) or 0.1 µM (1 pmol) Pla1 were incubated at 30°C for 30 min and subsequently quenched. The sequence of the primers is shown at the bottom where p marked with (•) indicates the position of ³²P-label. Primer extension products were resolved on 8% urea-PAGE and visualized by autoradiography. The size in nucleotides of a DNA ladder is indicated on the right. The radioactivity was quantified in ImageQuant-TL after scanning the gel with a Typhoon FLA7000 imager. (B) Reaction mixtures (85 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 1 mM ATP, 0.1 µM of the indicated 5′-³²P labeled primer, and 0.04 µM Pla1 were incubated at 30°C. At times specified, 10 µL aliquots (containing 400 fmol Pla1, and 1 pmol primer and/or primer extension products) were withdrawn and quenched immediately by adjustment to 47.5% formamide/12.5 mM EDTA. The proportion of the extended primer, either by 1 nt or >1 nt, was quantified in ImageQuant-TL after scanning the gel with a Typhoon FLA7000 imager and is plotted as a function of time (min) for the indicated primer. (C) Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 0.1 µM 5′-³²P labeled 12-mer RNA primer, 0.04 µM Pla1, and 0, 0.1, 1, 10, 100, or 1000 µM ATP as specified were incubated at 30°C for 30 min. The products were resolved by urea-PAGE, visualized by autoradiography, and the extents of primer extension were quantified by scanning the gel.

Deeper insights ensued from the temporal profiles of the reactions of 0.04 µM Pla1 with 0.1 µM 12-mer primer and 1 mM ATP (the same conditions used in Fig. 3A). At the 1 min time point, 87% of the input R12 primer was extended by more than one step, and 5% of the input was extended by a single step; 96% and 97% of the primers were polyadenylated by 2 and 5 min, respectively (Fig. 4B). In contrast, addition of AMP to the D12 primer was slow, inefficient (<6% primer utilization), and limited to just one round of catalysis (Fig. 4B). Providing a single 3′-terminal ribonucleotide in an otherwise all-DNA primer in D11R1 elicited a significant improvement in the rate and extent of primer utilization versus D12. The D11R1 primer underwent a burst of single AMP addition at 1 and 2 min, when 24% of the input primer was converted to a +1 product. The +1 species declined as polyadenylated primers accumulated between 2 and 10 min, at which point a total of 83% of the input primer had been extended (Fig. 4B). Placing a single 3′-terminal deoxynucleotide in an otherwise all RNA primer in R11D1 resulted in a profound decrease in the rate and extent of primer utilization compared to R12. The +1 extension product comprised 21% of the input primer at 5 min, when only 4% of the primers had been elongated by more than one step (Fig. 4B). Over 30 min, the +1 extension product persisted, and longer products accumulated slowly, such that 53% of the input primer was extended at 30 min (Fig. 4B). The D11R1 primer was clearly superior to the R11D1 primer.

From these results, we surmise that Pla1 requires a terminal ribose for effective primer engagement, which we presume is enforced by the atomic contacts to the terminal ribose 2′-OH seen in the crystal structure of the budding yeast enzyme, entailing hydrogen bonds from Asp102 and Tyr87, the equivalents of Pla1 Asp101, and Tyr86 (Fig. 2F). We suspect that Asp101 is the crucial discriminating contact, insofar as it will be shown below that a Y86F mutation has no effect on Pla1 activity.

Figure 4C shows the products formed during a 30 min reaction of 0.04 µM Pla1 with 0.1 µM 12-mer RNA primer and 0.1 µM, 1 µM, 10 µM, 100 µM, or 1 mM ATP. The percent of primers extended increased with ATP concentration, being 36%, 58%, and 65% at 0.1 µM, 1 µM, and 10 µM ATP, respectively. The lengths of the poly(A) tracts also increased in lockstep with ATP concentration, from ≤11 at 0.1 µM ATP to >600 at 1 mM ATP. From the product distribution at 0.1 µM ATP, we calculated that 80% of the available ATP was added as AMP to the RNA primer. Thus, Pla1 is very efficient at scavenging limiting amounts of ATP substrate.

Effect of active site mutations on poly(A) polymerase activity

Wild-type and mutant Pla1 preparations were assayed for ATP-dependent primer extension as a function of increasing enzyme. The percent primer utilization and the poly(A) tract length increased with input wild-type Pla1 in the range of 25–200 fmol enzyme; higher enzyme resulted in longer poly(A) tails (Fig. 5A). The titration profile of the Y86F mutant was virtually identical to wild-type (Fig. 5A). The specific activity of the Y86A protein was slightly less than wild-type (e.g., compare WT and Y86A tail lengths at 100 and 200 fmol enzyme). The specific activity of the Y86L mutant was about half that of wild-type Pla1, that is, tail lengths at 100 and 200 fmol wild-type Pla1 were comparable to those at 200 and 400 fmol Y86L (Fig. 5A). In contrast, the Y86D protein displayed feeble polymerase activity in vitro, whereby 54% of the primer was extended by 1 pmol enzyme, predominantly by a single AMP (Fig. 5B). Increasing Y86D to 2 pmol drove more efficient primer utilization (88% in toto) and the generation of a ladder of short poly(A) tails (Fig. 5B, comprising 66% of the input label) similar to that formed by 50 fmol of wild-type Pla1 (Fig. 5A). Thus, the Y86D enzyme is roughly one-fortieth as active as wild-type Pla1. The D153A mutant was less than half as active as Y86D, that is, product formation by 1 pmol of D153A was equivalent to that of 0.4 pmol of Y86D (Fig. 5B). At 2 pmol of D153A protein, only 18% of the primers were elongated by 2–8 steps of AMP addition (Fig. 5B).

FIGURE 5.

Mutational effects on poly(A) polymerase activity. Reaction mixtures (10 µL) containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 1 mM MnCl₂, 0.5 mM DTT, 1 mM ATP, 0.1 µM (1 pmol) 5′-³²P labeled RNA primer, and the indicated amounts of WT, Y86F, Y86A, and Y86L Pla1 (A) or mutants Y86D and D153A (B) were incubated at 30°C for 30 min and subsequently quenched by adjustment to 47.5% formamide/12.5 mM EDTA. Primer extension products were resolved by 8% urea-PAGE and visualized by autoradiography. The size in nucleotides of a DNA ladder is indicated. In B, the proportion of the extended primer, either by 1 nt or >1 nt, was quantified in ImageQuant-TL after scanning the gel with a Typhoon FLA7000 imager.

Effects of pla1-Y86D on the poly(A) tails of individual mRNAs

RNA isolated from wild-type and pla1-Y86D cells was annealed to DNA oligonucleotides (Supplemental Table S1) complementary to the sequences of the act1, adh1, ssa2, eno101, rps13, pgk1, and fba1 mRNAs 66–124 nt upstream of their mapped poly(A) sites. After digestion with E. coli RNase H, the RNAs were resolved by gel electrophoresis, transferred to a membrane, and subjected to northern blotting with ³²P-labeled DNA oligonucleotide probes (Supplemental Table S1) complementary to sequences of the mRNAs located 3′ of the site of RNase H cleavage. Parallel RNA samples annealed to gene-specific antisense DNA oligonucleotides and to oligo(dT) were subjected to northern blotting after digestion with RNase H to cleave the transcripts internally and remove the poly(A) tail (Fig. 6). By comparing the sizes of the 3′ mRNA segments after RNase H digestion in the absence and presence of oligo(dT), this procedure provides a gauge of poly(A) tail length of specific fission yeast mRNAs in wild-type versus pla1-Y86D cells.

FIGURE 6.

Effects of Pla1-Y86D on polyadenylation of mRNAs in vivo. Total RNA (10 μg) prepared from pla1 wild-type and pla1-Y86D cells was treated with RNase H in the presence of DNA oligonucleotides complementary to mRNAs either with (+) or without (−) oligo(dT). The reaction products were resolved by denaturing-PAGE along with an aliquot of radiolabeled DNA marker and transferred to a nylon membrane. Northern blot analyses of the RNase H reaction products for (A) act1, (B) adh1, (C) pgk1, (D) fba1, (E) ssa2, (F) eno101, and (G) rps13 are shown. The positions and sizes (nt) of the DNA markers are indicated in each panel. A schematic of the genomic locus and detected fragments is shown above each blot (drawn to scale). The end of the coding sequence is depicted by a yellow arrow and the 3′-UTR is shown as a blue extension in the direction of synthesis. The location of the complementary DNA oligonucleotide used to generate RNase H cleavage products and the 5′-³²P labeled oligonucleotide probe used for detection are shown. The gene-specific RNase H products downstream from the cleavage sites are shown as yellow wavy lines with poly(A) tails, or without poly(A) tails when treated with RNase H in the presence of oligo(dT). The poly(A) sites determined by 3′-end sequencing of wild-type cells are plotted as a function of position across the gene and are depicted as bars at single nucleotide resolution.

Although there were minor differences in the poly(A) tail length distribution, we were surprised to find that the sizes of poly(A) tails in Y86D cells were very similar to wild-type (Fig. 6), notwithstanding the severe defect elicited by the Y86D mutation on polyadenylation by Pla1 in vitro. It is conceivable that Y86D has much less impact on Pla1 polymerase activity in the context of the CPF complex, for example, because the other CPF subunits keep Pla1-Y86D engaged with the pre-mRNA 3′-OH end generated by nascent RNA cleavage. Alternatively, the putative short poly(A) tails that we expected to be synthesized by Pla1-Y86D (based on its in vitro activity) are subsequently extended by one or more of the alternative Cid-family poly(A) polymerases found in fission yeast (Saitoh et al. 2002; Stevenson and Norbury 2006).

Effects of pla1-Y86D on binding to an RNA primer

To explore the aforementioned scenario in which the Y86D mutation affects the engagement of Pla1 per se with the RNA primer, we used fluorescence polarization to measure the binding of wild-type Pla1 and Pla1-Y86D to a 5′-(6-FAM)-labeled 12-mer RNA (UUUCUGUGGCUA; the same RNA used as primer in the polyadenylation assays). Incubation of the FAM-labeled RNA with wild-type Pla1 resulted in a steady concentration-dependent increase in emitted polarized light with an estimated K_d of 456 nM (Fig. 7). Incubation with Y86D yielded no increase in emitted polarized signal (Fig. 7). This suggests that a principal catalytic defect of Pla1-Y86D is the engagement of the RNA 3′-OH end.

FIGURE 7.

The Y86D mutation impairs Pla1 binding to an RNA primer. Binding of the specified concentrations of Pla1 WT or Y86D enzyme to 10 nM 5′-(6-FAM)-UUUCUGUGGCUA was determined in 50 mM Tris-HCl, pH 7.4, 50 mM NaCl, and 1 mM MnCl₂ in a 30 µL reaction. Binding reactions were set up in a 384-well plate (Corning 4514) and incubated at room temperature in the dark for 20 min prior to measurement. Excitation of the ligand was performed with linearly polarized light at 485 nm, and emission was measured at 535 nm at planes that were parallel and perpendicular to the plane of excitation at room temperature using a SpectraMax iD5 plate reader (settings: endpoint measurement mode; detection with 400 msec integration, auto photomultiplier tube, and 1 mm read height). The data are averages (±SEM) of three independent assays, where each assay is the average of three technical replicates. The data were fit to a Hill slope equation in GraphPad Prism with an estimated K_d of 456 nM and goodness of fit correlation coefficient of 0.99 for the Pla1 wild-type curve.

Transcriptome profiling of the pla1-Y86D strain

We performed RNA-seq in parallel on poly(A)⁺ RNA isolated from pla1-Y86D and wild-type pla1⁺ cells. cDNAs obtained from three biological replicates (using RNA from cells grown to mid-log phase in YES medium at 30°C) were sequenced for each strain; 93%–95% of the sequence reads (22–27 million per replicate) were aligned to genomic loci. Read densities (RPKM) for individual genes were highly reproducible between biological replicates (Pearson coefficients of 0.986–0.989). A cutoff of plus or minus twofold change in normalized transcript read level and an adjusted P-value of ≤0.05 were the criteria applied to derive an initial list of differentially expressed annotated loci in the pla1-Y86D mutant versus the wild-type control. We then focused on differentially expressed genes with average normalized read counts ≥100 in either the mutant or wild-type strain to eliminate transcripts that were expressed at very low levels in vegetative cells. We thereby identified sets of 171 and 97 annotated protein-coding genes that were, respectively, up-regulated and down-regulated by these criteria in pla1-Y86D cells (Supplemental Table S2). A striking finding was that 70 mRNAs (P-value <6.5 × 10⁻⁶²) and 39 noncoding RNAs up-regulated in pla1-Y86D cells also accumulated by >2.8-fold in a red1Δ mutant (Fig. 8; Supplemental Table S2; Lee et al. 2013).

FIGURE 8.

MTREC and Mmi1 targets accumulate in pla1-Y86D cells. Venn diagrams depicting numbers of overlapping and nonoverlapping mRNAs or noncoding RNAs that were up-regulated in red1Δ cells (Lee et al. 2013) or mRNAs up-regulated at the nonpermissive temperature for mmi1-ts cells that are considered the “Mmi1 regulon” (Chen et al. 2011).

Red1 is a scaffolding protein essential for the assembly of the multisubunit MTREC complex that includes a Mtl1-Red1 core embellished by three submodules—Cbc1–Cbc2–Ars2, Pab2–Red5–Rmn1, and Mmi1–Iss10—that are implicated in the decay of (i) cryptic unstable transcripts (CUTs) that are transcribed from intergenic, antisense, and heterochromatic regions of the genome; (ii) unspliced pre-mRNAs; and (iii) meiotic mRNAs (Zhou et al. 2015). Red1 links each of the MTREC submodules to Pla1 and the nuclear exosome (Zhou et al. 2015). It is proposed that when these RNAs associate with the MTREC complex after initial cleavage and polyadenylation by the CPF, they are subsequently hyperadenylated by MTREC-associated Pla1 and then degraded by the nuclear exosome (Zhou et al. 2015; Soni et al. 2023). Pla1 action in the context of the MTREC complex is critical for the turnover of MTREC-targeted RNAs (Yamanaka et al. 2010; Sugiyama and Sugioka-Sugiyama 2011; Soni et al. 2023).

In a red1Δ mutant, the MTREC complex fails to assemble, resulting in the accumulation of MTREC's RNA targets. That a similar set of transcripts accumulated in the pla1-Y86D mutant implies that the Y86D missense change affected the function of the MTREC complex. A specific subset of these genes that were coherently affected were the meiotic mRNAs that are targeted to the MTREC complex by the RNA-binding protein Mmi1. Of the 30 genes previously defined as the Mmi1 regulon (Chen et al. 2011), 24 mRNAs and the sme2 noncoding RNA accumulated in the pla1-Y86D mutant (P-value <2.4 × 10⁻³³) (Fig. 8; Supplemental Table S2). Note that SPBC1921.04c was below our threshold for normalized read counts but was included in our analysis because it is part of the Mmi1 regulon and it accumulates in a red1Δ mutant (Chen et al. 2011; Lee et al. 2013). Other bona fide noncoding Mmi1 targets, nam1 and nam2 (Touat-Todeschini et al. 2017), also accumulated by 4.3-fold and 6-fold, respectively, in pla1-Y86D cells (Supplemental Table S2).

Among the down-regulated genes in pla1-Y86D cells were the PHO mRNAs pho1 (encoding cell surface acid phosphatase) and pho84 (encoding an inorganic phosphate transporter), which were lower by twofold and threefold, respectively (Supplemental Table S2; Supplemental Fig. S2A). This was commensurate with Pho1 acid phosphatase activity being approximately threefold lower in the Y86D mutant (Fig. 2D). pho1 and pho84 are repressed in phosphate-replete cells by transcriptional interference from neighboring upstream lncRNAs, prt(nc-pho1) and prt2, respectively, that initiate from upstream lncRNA promoters and terminate at the pho1 and pho84 poly(A) sites (Fig. 1A; Shuman 2020). Analysis of the read counts across the regions preceding the pho1 and pho84 mRNA start sites revealed that prt lncRNA was increased by fourfold (Supplemental Fig. S2A), but there was no change in the abundance of prt2 lncRNA. We found that there was a 2.3-fold increase in the reads across the tgp1 mRNA region (Supplemental Table S2), which was a surprising divergence from a uniform effect on the PHO regulon expression observed during phosphate starvation or in other mutants (Carter-O'Connell et al. 2012; Shuman 2020). Expression of tgp1, a PHO regulon gene that encodes a glycerophosphodiester transporter, is repressed in phosphate-replete cells by transcriptional inference from the upstream flanking nc-tgp1 lncRNA (Shuman 2020). Inspection of the reads across the nc-tgp1–tgp1 locus showed that the apparent increase in tgp1 mRNA was attributable to the increase in production of the nc-tgp1–tgp1 read-through lncRNA (evaluated by a 1.3-fold increase in read density over the nc-tgp1 lncRNA segment preceding the tgp1 transcript) and not due to an increase in the tgp1 mRNA (Supplemental Fig. S2B). It is noteworthy that both prt and nc-tgp1 lncRNAs have DSR (determinant of selective removal) elements that are binding sites for Mmi1 (Chen et al. 2011; Yamashita et al. 2012; Kilchert et al. 2015). The DSR elements are critical for the elimination of these PHO-regulatory lncRNAs by the exosome and for promoting lncRNA 3′-processing and termination (Chatterjee et al. 2016; Sanchez et al. 2018a, Schwer et al. 2020). That these RNAs also accumulated agrees with the idea that Pla1-Y86D affects the Mmi1-dependent turnover of MTREC targets.

Effect of Pla1-Y86D on the poly(A) tail length of the Mmi1-regulated gene mug8

Our findings that transcripts targeted by the MTREC complex accumulated in pla1-Y86D cells suggested that Pla1 might have distinct functions in the CPF and MTREC complexes. We envisioned that Pla1-Y86D, without the aid of RNA primer-binding components present in the context of CPF, might be defective in hyperadenylating MTREC targets, which would lead to their accumulation observed in pla1-Y86D cells. The challenge of evaluating poly(A) tail length of MTREC targets in wild-type cells is that they are eliminated upon hyperadenylation. Thus, to test this idea, we measured poly(A) tail length of mug8—an Mmi1-dependent MTREC target—which was the most abundant of the MTREC targets.

Total RNA (40 µg) isolated from wild-type and pla1-Y86D cells was annealed to a DNA oligonucleotide complementary to the segment 175 nt upstream of the mapped mug8 poly(A) site (Fig. 9A,B). Parallel samples were also annealed with oligo(dT). After digestion with RNase H, the RNAs were resolved by gel electrophoresis, transferred to the membrane, and subjected to northern blotting with a ³²P-labeled DNA oligonucleotide probe complementary to the mug8 region located 3′ of the site of RNase H cleavage. The blot was reprobed with a ³²P-labeled DNA oligonucleotide complementary to the 7SL RNA, thereby serving as a loading control. The salient findings were: (i) mug8 mRNA accumulated in the pla1-Y86D strain, affirming the RNA-seq analysis; and (ii) the distribution of mug8 poly(A) tail lengths in pla1-Y86D cells was shorter than that of the wild-type control (Fig. 9C). This indicates that mug8 is polyadenylated, but seemingly not hyperadenylated, in the pla1-Y86D strain, likely contributing to its accumulation.

FIGURE 9.

Effects of Pla1-Y86D on polyadenylation of mug8 mRNA in vivo. Total RNA (40 µg) prepared from wild-type and pla1-Y86D cells was annealed to DNA oligonucleotides complementary to the mug8 mRNA either with (+) or without (−) oligo(dT), then treated with RNase H. The reaction products were resolved by denaturing-PAGE along with an aliquot of radiolabeled DNA marker, then transferred to a nylon membrane for northern analysis. (A) Schematic of the mug8 locus and RNase H reaction products. (B) Poly(A) sites (counts/base/million) determined by 3′-end sequencing of Pla1 wild-type and Y86D cells are plotted as a function of position across the mug8 locus (shown in A) and are depicted as bars at single nucleotide resolution. (C) Northern blot analysis of the RNase H reaction products of mug8. The arrow indicates the position of a slower migrating distribution of polyadenylated mug8 RNase H fragments in Pla1 wild-type cells compared to Y86D. The blot was stripped and reprobed with a DNA oligonucleotide complementary to 7SL RNA.

DISCUSSION

In this study, we expanded our screen for spontaneous suppressors of the near-lethal growth defect of asp1-H397A rad24Δ and recovered a Y86D missense mutation in the active site of the essential poly(A) polymerase Pla1. To define the effects of this mutant, we first biochemically characterized the fission yeast Pla1 enzyme, which: (i) has a preference for manganese as the metal cofactor; (ii) exclusively uses ATP to add more than two nucleoside monophosphates to an RNA primer; (iii) requires at least one ribonucleotide at the 3′ end of an oligonucleotide for efficient catalysis; and (iv) efficiently scavenges limiting amounts of ATP.

Evaluation of Pla1-Y86D polyadenylation activity in vitro revealed that it was reduced by 40-fold vis-à-vis wild-type Pla1, only twofold higher than Pla1-D153A, which, according to studies of S. cerevisiae poly(A) polymerase (Balbo and Bohm 2007), is impaired due to the failure to bind a catalytic metal. Whereas the fission yeast pla1-D153A allele was lethal in vivo, pla1-Y86D cells were viable but slow-growing. We were initially surprised to find that the pla1-Y86D allele had little impact on mRNA poly(A) tail size in vivo, as gauged by Northern analysis of seven individual protein-coding transcripts. The catalytic defect of Pla1-Y86D in ATP-dependent primer extension in vitro could be attributed to lack of affinity for the RNA primer, as measured by fluorescence polarization (Fig. 7). The effect of Y86D on primer engagement accords with the crystal structure of yeast poly(A) polymerase in which the conserved Tyr86 equivalent makes direct atomic contacts to the terminal ribose of the RNA primer (Fig. 2F). We envision that other subunits of the CPF complex compensate to position the 3′-OH end in the Pla1-Y86D active site subsequent to cleavage of nascent mRNAs by the Ysh1 subunit of CPF.

RNA-seq identified a set of mRNAs (∼3% of total mRNAs) that were up-regulated by at least twofold in pla1-Y86D cells versus wild-type, among which were many targets of the MTREC complex that accumulate in red1Δ and mmi1-ts mutants. Pla1 is part of the MTREC complex, independent of its presence in CPF, and its function in hyperadenylating RNA targets is important for the decay of MTREC targets by the exosome (Sugiyama and Sugioka-Sugiyama 2011; Yamanaka et al. 2010; Lee et al. 2013; Egan et al. 2014; Zhou et al. 2015; Soni et al. 2023). Our evaluation of the MTREC target mug8⁺ suggested that it has a shorter poly(A) tail distribution in pla1-Y86D cells versus wild-type, though the extent of the length difference is difficult to quantify because the steady-state level of mug8 in wild-type cells is very low.

Several questions remain open. First, what is the cause for the slow growth defect of pla1-Y86D cells, notwithstanding that polyadenylation by CPF seems relatively normal? The untimely accumulation of the meiotic MTREC targets is a possible reason. An example of this toxicity is observed in an mmi1Δ mutant, which is lethal due to inappropriate expression of mei4 (the master transcription factor for the mid-meiotic gene expression program) during vegetative growth (Harigaya et al. 2006; Sugiyama and Sugioka-Sugiyama 2011). That mmi1Δ lethality was suppressed by inactivating mutations in Mei4 argues that a similar mechanism might be at play for the observed pla1-Y86D growth defect. This is apparently not the case, insofar as a pla1-Y86D mei4Δ double-mutant was just as sick as the pla1-Y86D (Supplemental Fig. S3).

Second, why does Pla1-Y86D affect the function of the MTREC complex and not the CPF? We hypothesize that manifestation of the Pla1-Y86D defect occurs in the MTREC complex, because (unlike the case of CPF) other members of the MTREC complex are not able to deliver the RNA primer terminus to the active site of Pla1-Y86D. Thus, when polyadenylation is delegated to MTREC, Pla1-Y86D is deficient in RNA 3′ end binding and adenylation.

Third, is hyperadenylation involved in the recruitment of the exosome to the MTREC complex? The poly(A)-binding protein Pab2, a member of the MTREC complex, physically associates with Pla1 (Lemieux and Bachand 2009). Pab2 also associates with the exosome subunit Rrp6, and this interaction is important for snoRNA processing. By eliminating Pab2, snoRNAs accumulate poly(A) tails of lengths up to 300 nt (Lemay et al. 2010). This suggests that hyperadenylation by Pla1 promotes Pab2 binding which then recruits the nuclear exosome. Thus, lack of hyperadenylation in pla1-Y86D cells might hinder the recruitment of the exosome, accounting for the accumulation of MTREC-targeted transcripts in the pla1-Y86D mutant. An analogous pathway has been described in mammalian cells, whereby hyperadenylation of certain nuclear RNAs by canonical poly(A) polymerases and nuclear poly(A)-binding protein (PABN1) targets the transcripts for decay by the nuclear exosome (Bresson and Conrad 2013; Bresson et al. 2015).

And fourth, how does pla1-Y86D suppress the growth defect of rad24Δ asp1-H397A? Both rad24Δ and asp1-H397A elicit precocious 3′-processing/termination of PHO-repressive lncRNAs (Sanchez et al. 2019; Garg et al. 2022). The phospho-ligand binding site of the 14-3-3 protein Rad24 is critical for its function in antagonizing precocious lncRNA termination (Garg et al. 2022). The IP₈ pyrophosphatase activity of Asp1—lack of which in asp1-H397A cells leads to accumulation of IP₈ (Pascual-Ortiz et al. 2018)—is critical for its role in avoiding precocious lncRNA termination, that is, because IP₈ is an agonist of 3′-processing/termination (Sanchez et al. 2019). Combining the rad24Δ and asp1-H397A mutations leads to a near-lethal growth defect that could be reversed by loss-of-function mutations of inessential CPF subunits Ssu72, Ctf1, Dis2, Ppn1, and Swd22 and termination factor Rhn1. Such suppression implies that the toxicity of rad24Δ asp1-H397A is mediated by overzealous 3′-processing/termination of at least one essential gene (Garg et al. 2022). We showed previously that the DSR elements in prt(nc-pho1), to which Mmi1 binds, are required for the precocious termination elicited by rad24Δ (Garg et al. 2022) or asp1-H397A (Sanchez et al. 2019). We envision that pla1-Y86D mutation suppresses rad24Δ asp1-H397A by alleviating (at least partially) the overzealous precocious termination. How might this work? Given the evidence herein, based on poly(A) tail analysis of exemplary mRNAs, that Y86D does not affect Pla1 activity in the context of canonical CPF-mediated termination, we speculate that the Y86D change might impact the interconnected transcription termination function(s) imputed to the nuclear exosome, Mmi1, and MTREC (Lemay et al. 2014; Chalamcharla et al. 2015; Touat-Todeschini et al. 2017; Vo et al. 2019; Shichino et al. 2020).

MATERIALS AND METHODS

SRA screen

The SRA-3 mutant was isolated by plating rad24Δ asp1-H397A cells on YES agar at 30°C. SRA-3 was selected as a large colony as compared to the background of tiny colonies after 6 d of incubation. The large colony was restreaked and was homogenously large compared to the parental strain. The SRA-3 strain was colony-purified, spot-tested for growth, and used for whole-genome sequencing analysis.

Spot tests of fission yeast growth

Cultures of Schizosaccharomyces pombe strains were grown in liquid YES (yeast extract with supplements) medium until A₆₀₀ reached 0.5–0.8. The cultures were adjusted to an A₆₀₀ of 0.1 and aliquots (3 µL) of serial fivefold dilutions were spotted to YES agar. The plates were photographed after incubation for 2 d at 34°C, 2.5 d at 30°C and 37°C, 4 d at 25°C, and 6 d at 20°C.

Whole-genome sequencing

After PicoGreen quantification and quality control by Agilent BioAnalyzer, 500 ng aliquots of genomic DNA were sheared using a LE220-plus Focused-ultrasonicator (Covaris catalog # 500569), and sequencing libraries were prepared using the KAPA HyperPrep Kit (Kapa Biosystems KK8504) with modifications. DNA libraries were subjected to size selection by mixture with 0.5 vol of aMPure XP beads (Beckman Coulter catalog # A63882) after post-ligation cleanup. Libraries were not amplified by PCR and were pooled equivolume for sequencing. Samples were run on a NovaSeq 6000 in a 150 bp/150 bp paired-end run using the NovaSeq 6000 SBS v1 Kit and an S1 flow cell (Illumina). The average number of read pairs per sample was 10 million.

Mapping the SRA-3 mutation

The FASTA file for the S. pombe genome was accessed from PomBase. The whole-genome sequencing reads from the parental rad24Δ and asp1-H397A strains and the SRA-3 mutant strain were aligned to the genome using Bowtie2 (Langmead and Salzberg, 2012). The resulting SAM files were converted to BAM files using Samtools (Li et al. 2009). Variants were identified by BCFtools (Li 2011) using the criteria of adjusted mapping quality = 40, minimum base quality = 20, and disabled probabilistic realignment for the computation of base alignment quality (BAQ) for considering variations or insertion–deletion events. The multiallelic caller protocol was used for variant calling in BCFtools. Variants were annotated using SnpEff, with its in-built genome version for S. pombe (Cingolani et al. 2012). Variants were further filtered by removing all variations with an average mapping quality ≤25 (Phred scale). All variants present in the parental strain were excluded as noncausal mutations.

Allelic exchange at the pla1 locus

The pla1 allelic exchange integration cassettes were generated in a bacterial plasmid by standard cloning procedures. The pla1 ORF was PCR amplified from S. pombe cDNA with oligonucleotide primers that introduced restriction sites for cloning. Two-stage PCR overlap extension with mutagenic primers was used to introduce missense mutations into the pla1 ORF. The integration cassettes consisted of the following elements, proceeding from 5′ to 3′: (i) a 529-bp segment of genomic DNA 5′ of the pla1⁺ start codon; (ii) an intron-less ORF encoding wild-type or mutant Pla1; (iii) a 258-bp segment of genomic DNA 3′ of the nmt1⁺ stop codon containing polyA/termination signals from the nmt1⁺ gene; (iv) a kanMX gene conferring resistance to G418; and (v) a 487-bp segment of genomic DNA 3′ of the pla1⁺ stop codon. All plasmids were sequenced to exclude the presence of unwanted mutations. The integration cassettes were excised from the plasmids and transfected into a diploid strain. G418-resistant diploids were selected and a segment of the pla1::kanMX allele was PCR amplified and sequenced to verify that the desired alleles were present. Correct integrations at the target locus were verified by Southern blotting. Confirmed heterozygous diploids were sporulated and subjected to random spore analysis (Escorcia and Forsburg 2018), whereby spores (∼1000) were plated in parallel on YES agar and on G418-containing medium selective for the kanMX-marked pla1 alleles and then incubated at 30°C. G418-resistant haploid pla1-WT, Y86F, Y86L, Y86A, and Y86D progeny were recovered and comprised about half of the total progeny that grew on YES agar. In the case of the pla1⁺ pla1-D153A diploid, all viable haploid progeny were G418-sensitive. Failure to recover G418-resistant progeny signified that the D153A mutation was lethal.

Recombinant Pla1 proteins

We constructed pET28b-His₁₀Smt3-Pla1 plasmids encoding various iterations of the wild-type or mutant Pla1 proteins fused to an amino-terminal His₁₀Smt3 module under the transcriptional control of a T7 RNA polymerase promoter. The plasmids were transfected into E. coli BL21(DE3) cells. Cultures (1 L) amplified from single transformants were grown at 37°C in Terrific Broth containing 50 µg/mL kanamycin until A₆₀₀ reached 0.8, and then placed on ice for 1 h. The cultures were adjusted to 2% ethanol (v/v) and 0.5 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) followed by incubation at 17°C for 18 h with constant shaking. The cells were harvested and frozen at −80°C. Frozen cells were thawed and resuspended in 25 mL of buffer A (50 mM Tris-HCl, pH 7.4, 500 mM NaCl, 20 mM imidazole, 10% glycerol) and half a tablet of cOmplete EDTA-free Protease Inhibitor Cocktail (Roche). All subsequent purification procedures were performed at 4°C. Cell lysis was achieved by adding lysozyme to 1.2 mg/mL and incubating for 45 min followed by sonication. The lysate was centrifuged at 38,000g for 30 min and the supernatant was mixed with 1 mL of Ni-NTA-agarose resin (Qiagen) that had been equilibrated in buffer A. After 1 h of mixing on a nutator, the resin was recovered by centrifugation and washed twice with 30 mL of buffer A. The resin was poured into a column and washed with an additional 20 mL of buffer A. The bound protein was eluted with buffer A-300 (50 mM Tris-HCl, pH 7.4, 500 mM NaCl, 300 mM Imidazole, 10% glycerol). The His₁₀Smt3 tag was cleaved by treatment with Ulp1 protease during overnight dialysis against buffer A. The Pla1 proteins were separated from the His₁₀Smt3 tag by a second round of Ni-affinity chromatography and were recovered in the flow-through fraction. The purification of the Pla1 proteins was monitored by SDS–PAGE. The tag-free Pla1 proteins were then subjected to gel filtration through a HiLoad Superdex 200 pg 16/600 column (Cytiva Life Sciences) for Pla1-Y86D, Pla1-Y86A, Pla1-Y86L, and Pla1-D153A, or a Superdex-200 Increase 10/300 GL column (Cytiva Life Sciences) for Pla1-WT and Pla1-Y86F, equilibrated in buffer B (50 mM Tris-HCl, pH 7.4, 200 mM NaCl, 10% glycerol). The peak fractions of each preparation were concentrated by centrifugal ultrafiltration, and aliquots of the protein were flash frozen in liquid nitrogen and subsequently stored at −80°C. Protein concentrations were determined by using the Bio-Rad dye reagent with BSA as the standard.

Poly(A) polymerase assays

Reaction mixtures containing 50 mM Tris-HCl, pH 7.4, 50 mM NaCl, and divalent metal, DTT, ribonucleotide, 5′-³²P-labeled pUUUCUGUGGCUA primer, and Pla1-WT or mutant at concentrations specified in the figure legends were incubated at 30°C. Reactions were initiated by addition of the enzyme and quenched at the times specified by adjustment to 47.5% formamide/12.5 mM EDTA.

Acid phosphatase activity

Aliquots of exponentially growing fission yeast cultures were harvested, washed with water, and resuspended in water. To quantify acid phosphatase activity, reaction mixtures (200 µL) containing 100 mM sodium acetate, pH 4.2, 10 mM p-nitrophenylphosphate, and cells (ranging from 0.01 to 0.1 A₆₀₀ units) were incubated for 5 min at 30°C. The reactions were quenched by addition of 1 mL of 1 M sodium carbonate, the cells were removed by centrifugation, and the absorbance of the supernatant at 410 nm was measured. Acid phosphatase activity is expressed as the ratio of A₄₁₀ (p-nitrophenol production) to A₆₀₀ (cells). The data are averages (±SEM) of at least three assays using cells from three independent cultures.

Northern blot analysis of poly(A) tail length

RNA was isolated from 17 A₆₀₀ units of S. pombe wild-type or pla1-Y86D cells that were grown in liquid YES medium at 30°C to an A₆₀₀ of 0.6–0.7. Cells were harvested by centrifugation and total RNA was extracted via the hot phenol method. Equal aliquots of total cell RNA were annealed to gene-specific oligonucleotides in the presence or absence of oligo-dT₁₈ by heating at 95°C for 3 min and then snap cooling on ice followed by treatment with 5 U RNase H (NEB) for 1 h at 37°C. Reactions were terminated by the addition of an equal volume of phenol:chloroform mix, and the RNA was subsequently recovered. The RNA and an aliquot of 5′-³²P DNA ladder (MspI digest of pBR322, NEB) were denatured and resolved by electrophoresis through a 1-mm-thick, 14-cm-long, 6% polyacrylamide gel containing 7 M urea in 80 mM Tris-borate, and 1 mM EDTA. The steps for electro-transfer of gel contents to a nylon membrane (Hybond XL) and northern blotting were conducted as described (Rio 2014) with the following modifications: (i) 25 mM phosphate buffer, pH 6.5, was used as the electro-transfer buffer; (ii) oligonucleotide probes (Supplemental Table S1) and DNA ladder were 5′-³²P end-labeled using T4 polynucleotide kinase and [γ³²P]ATP; (iii) ULTRAhyb-Oligo mix (Invitrogen) was used as prehybridization and hybridization buffer; and (iv) the blots were washed twice for 30 min with 2× saline sodium citrate (SSC) buffer/0.5% SDS. After detection, hybridized probes were stripped using boiling water, and the membranes were reused for probing other RNAs.

Transcriptome profiling by RNA-seq and mapping of poly(A) sites genome-wide

The integrity of total RNA (prepared as described above) was gauged with an Agilent Technologies TapeStation instrument. For transcriptome profiling, the Illumina TruSeq Stranded mRNA sample preparation kit was used to purify poly(A)⁺ RNA from 500 ng of total RNA and to carry out the subsequent steps of poly(A)⁺ RNA fragmentation, strand-specific cDNA synthesis, indexing, and amplification. Indexed libraries were normalized and pooled for paired-end sequencing performed by using an Illumina NovaSeq 6000-S1 flow cell. FASTQ files bearing paired-end reads of length 51 bases were mapped to the S. pombe genome (PomBase) using HISAT2-2.1.0 with default parameters (Kim et al. 2015). The resulting SAM files were converted to BAM files using Samtools. Count files for individual replicates were generated with HTSeq-0.10.0 (Anders et al. 2015) using exon annotations from PomBase (GFF annotations, genome-version ASM294v2; source “ensembl”). RPKM analysis and pairwise correlations were performed as described previously (Schwer et al. 2014). Differential gene expression and fold change analysis were performed in DESeq2 (Love et al. 2014). Cutoff for further evaluation was set for genes that had an adjusted P-value (Benjamini–Hochberg corrected) of ≤0.05 and were up or down by at least twofold in pla1-Y86D in comparison to wild-type. Genes were further filtered on the following criteria: (i) Greater than or equal to twofold up and the average normalized read count for the mutant strain was ≥100; and (ii) greater than or equal to twofold down and the average normalized read count for the wild-type strain was ≥100. For genome-wide mapping of poly(A) sites, sequencing libraries from 500 ng of total RNA were made using the Lexogen QuantSeq 3′ mRNA-Seq Library Prep Kit REV, according to manufacturer's instructions. Indexed libraries were normalized and pooled for paired-end sequencing performed by using an Illumina NextSeq 500 mid-output flow cell. FASTQ files bearing paired-end reads of length 75 bases were mapped to the S. pombe genome (PomBase) using HISAT2-2.1.0 defining the following arguments: ‐‐max-intronlen 1200 ‐‐no-mixed ‐‐no-discordant. The resulting SAM files were converted to BAM files using Samtools keeping only the first in pair reads (samtools flag: 64).

Deletion of mei4

The mei4 open reading frame (amino acids 1–517) was deleted in an S. pombe diploid strain and replaced with a natMX antibiotic-resistance cassette as follows. First, using PCR amplification and standard cloning methods, we constructed plasmid pKS-mei4Δ-natMX, in which the natMX cassette is flanked by 580 bp of genomic DNA 5′ of the mei4 translation start codon and 561 bp of DNA 3′ of the mei4 translation stop codon. The linearized integration cassette was transfected into pla1⁺/pla1-Y86D::kanMX diploids and correct gene targeting was confirmed by Southern blotting of nourseothricin-resistant transformants. The heterozygous diploids were sporulated and nourseothricin-resistant mei4Δ or nourseothricin- and geneticin-resistant mei4Δ pla1-Y86D haploids were selected.

DATA DEPOSITION

The RNA-seq data in this publication have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO Series accession number GSE221438.

SUPPLEMENTAL MATERIAL

Supplemental material is available for this article.