Volume 113, Issue 4, Pages 533-545 (May 2003) Computational Modeling and Experimental Analysis of Nonsense-Mediated Decay in Yeast  Dan Cao, Roy Parker  Cell  Volume 113, Issue 4, Pages 533-545 (May 2003) DOI: 10.1016/S0092-8674(03)00353-2

Figure 1 Nonsense Codons in PGK1 mRNA Show a Polar Effect (A) Comparison of decay rates of PGK1 transcripts after repression of transcription in wild-type strain. Time points are indicated above the panel in minutes. Half-lives are shown on the right in minutes. (B) Steady-state level comparison in wild-type strains. The levels of nonsense-containing PGK1 relative to normal PGK1 (set at 100) are shown at the bottom of the gel. (C) Steady-state level comparison in an upf1Δ strain. The levels of nonsense-containing PGK1 relative to normal PGK1 (set at 100) are shown at the bottom of the gel. (D) The observed effect of different nonsense codon in PGK1 on decay rates and steady-state levels. This is a chart for the results shown in (A) and (B). The half-lives and steady-state levels shown are the average of three independent measurements. (E) The modeling results of the leaky surveillance model. kDID is set at 0.01 s−1 for the graph shown here. The simulated half-life for normal PGK1 is 27 min and the steady-state level of normal PGK1 is set at 100%. Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 2 The Leaky-Surveillance Model for NMD (A) The cartooned model. In this model, the surveillance efficiency determines the relative level of transcripts entering normal decay pool or NMD pool. For 5′ nonsense codon, the efficiency is higher so more transcripts are in the NMD pool. For 3′ nonsense codon, the efficiency is lower so more transcripts are in the normal decay pool. (B) The modeling framework for the cartoon shown in (A). The boxed part is the NMD pool. The unboxed part is same as the framework used to simulate normal PGK1 before (Cao and Parker, 2001). FL represents full-length. FG represents fragment. Same abbreviations are used throughout this manuscript. Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 3 Decay of PGK1 with Different Positions of Nonsense Codon Is Consistent with the PIES Model and Shows Multiple Upf1p Dependent Mechanisms of Destabilization Shown are transcriptional pulse-chase analyses of PGK1 mRNA with different positions of nonsense codon in a wild-type strain (left side) or in an upf1Δ strain (right side) (A, A′) PGK1 with a nonsense codon at codon 22. (B, B′) PGK1 with a nonsense codon at codon 142. (C, C′) PGK1 with a nonsense codon at codon 225. (D, D′) PGK1 with a nonsense codon at codon 319. (E, E′) Normal PGK1 without a nonsense codon. dT indicates oligo dT/RnaseH cleavage to indicate the position of fully deadenylated species. The samples from 0 min time points are used for dT treatment. Pre indicates a time point taken prior to induction of transcription. Time points are indicated above the panel in minutes after transcriptional repression. The relative position of full-length species and decay fragment trapped by the poly G structure are indicated at right. Upper and lower parts in (A)–(E) are different exposures of the same gel to show the full-length and decay fragment. The blots for (A)–(D) were probed with ORP121, which is complementary to the poly G sequence (Muhlrad and Parker, 1994). Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 4 PGK1 with an Early Nonsense Codon Can Also Undergo Accelerated Deadenylation Shown are transcriptional pulse-chase analyses of PGK1C22 (A) or normal PGK1 (B) in a partial decapping mutant dcp1-2 strain at 30°C. Time points are indicated above the panel in minutes. The samples from 0 min time points are used for dT treatment. The smear shown in both gels especially at late time points is likely caused by incomplete shutoff of transcription. Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 5 The Accelerated Deadenylation Is Independent on the Position of Nonsense Codon (A) Shown is decay from steady state of different PGK1 mRNAs in a dcp2Δ strain yRP1358. Time points are indicated above the panel in minutes after transcriptional repression. The half-lives are determined by running the same samples on agarose gel. The half-lives shown are averages from two independent measurements. (B) Deadenylation rates measured from the acrylamide northern gels in (A). Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 6 The PIES Model for NMD (A) The cartooned model. (B) The modeling framework for the cartoon shown in (A). Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)

Figure 7 The PIES Model Can Apply to Other Yeast Transcripts in addition to PGK1. Shown are steady-state poly(A) distributions of PGK1 and HIS4. (A) The steady-state distribution for PGK1. dT indicates oligo dT/RnaseH cleavage to indicate the position of fully deadenylated species. 50 μg of C22 and C142, 40 μg of C225, and C319 and 10 μg of normal PGK1 were run for the blot shown. (B) The steady-state distribution for HIS4. 45 μg of each total RNA were run for the blot shown. The blot shown was probed with oRP1144. The levels of nonsense-containing HIS4 relative to normal HIS4 (set at 100) are shown at the bottom of the gel. The quantitation is done by normalizing to the levels of 7s rRNA on the agarose gel probed with oRP1143. Cell 2003 113, 533-545DOI: (10.1016/S0092-8674(03)00353-2)