Apoptosis Triggers Specific, Rapid, and Global mRNA Decay with 3′ Uridylated Intermediates Degraded by DIS3L2

Apoptosis is a tightly coordinated cell death program that damages mitochondria, DNA, proteins, and membrane lipids. Little is known about the fate of RNA as cells die. Here, we show that mRNAs, but not noncoding RNAs, are rapidly and globally degraded during apoptosis. mRNA decay is triggered early in apoptosis, preceding membrane lipid scrambling, genomic DNA fragmentation, and apoptotic changes to translation initiation factors. mRNA decay depends on mitochondrial outer membrane permeabilization and is amplified by caspase activation. 3′ truncated mRNA decay intermediates with nontemplated uridylate-rich tails are generated during apoptosis. These tails are added by the terminal uridylyl transferases (TUTases) ZCCHC6 and ZCCHC11, and the uridylated transcript intermediates are degraded by the 3′ to 5′ exonuclease DIS3L2. Knockdown of DIS3L2 or the TUTases inhibits apoptotic mRNA decay, translation arrest, and cell death, whereas DIS3L2 overexpression enhances cell death. Our results suggest that global mRNA decay is an overlooked hallmark of apoptosis.

Apoptotic RNA decay occurs in many cell types responding to diverse stimuli, and is specific to mRNAs. Related to Figure 2.
(A-C) HCT116 cells were treated with thapsigargin, or left untreated, for 12 hrs, then cells were harvested for annexin V staining (A) or RNA was harvested and the indicated mRNAs (B) or ncRNAs (C) were assayed by qRT-PCR. mRNAs consistently declined, while ncRNAs were more stable.
(D-F) HCT116 cells were treated with tunicamycin, or left untreated, for 12 hrs, then cells were harvested for annexin V staining (D) or RNA was harvested and the indicated mRNAs (E) or ncRNAs (F) were assayed by qRT-PCR. mRNAs consistently declined, while ncRNAs were more stable.
(G-J) The indicated reporters (G) were transfected into HeLa cells, 48 hr later the cells were treated with STS or left untreated. GFP RNA expression was analyzed by Northern blot (H). The Northern blot probe targeted the GFP coding region. All of the constructs produced GFP RNA when introduced into cells. GFP protein expression was assayed by immunoblot (I) and flow cytometry (J). There was high protein expression from transcripts driven by a polymerase II (Pol II) promoter. When the polyadenylation signal sequence (PAS) was replaced by a self-cleaving hammerhead ribozyme (HR), protein expression was lost, but rescued by the addition of 60 adenine or 60 uridine nucleotides before the HR sequence.

Figure S3
Apoptotic mRNA decay is dependent on the mitochondrial proteins BAX and BAK. Related to Figure 3. HeLa cells were transfected with a control siRNA (CTL) or pooled siRNAs targeting BAX and BAK. 72 hr later, the cells were treated with STS±zVAD and harvested for immunoblot of the indicated proteins (A) or qRT-PCR for the indicated RNAs. Paired BAX/BAK knockdown rescued mRNA levels to a greater extent than zVAD, suggesting that mRNA decay depends of full activation of the mitochondrial apoptotic pathway. Error bars represent SEM of at least 3 independent experiments. * p<0.05. (D-G) EEF1A mRNAs were amplified by cRACE with a forward primer in the ORF. PCR products were run on a gel (D) and the fragments were isolated by gel purification. Some EEF1A decay products with intact 5´ ends had nontemplated tails, which were rich in uridylates (E). As with the ACTB mRNA decay products, most clones derived from TAP-treated RNA had intact 5´ ends (F), while all clones from TAP-untreated RNA had evidence of 5´ to 3´ decay (G). Clones in E-G were all isolated in one experiment.   (C,D) HeLa cells were transfected with CTL or DIS3L2 siRNAs, then treated with STS for 4 hr. Total RNA was harvested and subjected to U-tailing assays. U-tailed ACTB decay intermediates (arrow) accumulated in STS-treated HeLa cells after DIS3L2 knockdown as measured by RT-PCR (C) and cRACE (D). U-tailed intermediates also accumulated in untreated living cells after DIS3L2 knockdown (C). A summary of the cRACE results with statistical analysis is presented in Figure 7C.
(E-G) DIS3L2 knockdown in HeLa cells partially restored mRNA levels (E) and reduced effector caspase activation (F) and caspase 3 cleavage (G) following 4 hr of treatment with STS. Error bars represent SEM of at least 3 independent experiments. * p<0.05; **p<0.01.  (H,I) HeLa cells were transfected with the indicated siRNA pools and RNA was analyzed by qRT-PCR relative to 7SL at the indicated times after adding α-amanitin+/-STS (H). The mRNA half-lives in STS-treated Hela cells were determined using the one-phase exponential decay equation (I). DIS3L2, but not DIS3L1, siRNAs increased mRNA half-life after STS treatment. (J-N). HCT116 cells were transfected with CTL, DIS3L1 or DIS3L2 siRNAs, then treated with TRAIL or left untreated for 3 hrs. Knockdown was confirmed by qRT-PCR (J), mRNA levels were measured by qRT-PCR relative to 7SL (K), apoptosis was assayed by annexin V staining (L), and caspase activation was measured by immunoblot (M) and a luminescent assay (N). DIS3L2, but not DIS3L1 siRNAs rescued mRNA levels, apoptosis, and caspase 3 cleavage and activation.   Table S1. Taqman miRNA qRT-PCR was performed according to the manufacturer's instructions (Life Technologies N808-0234).

Ambion siRNAs
Gene   Table 1: Primers used in this study, related to Experimental Procedures