Identification of multiple RNA features that influence CCR4 deadenylation activity.

The CCR4 family proteins are 3'-5'-deadenylases that function in the first step of the degradation of poly(A) mRNA. Here we report the purification to homogeneity of the yeast CCR4 protein and the analysis of its substrate specificities. CCR4 deadenylated a 7N+23A substrate (seven nucleotides followed by 23 A residues) in a distributive manner. Only small differences in CCR4 activity for different A length substrates were observed until only 1 A residue remained. Correspondingly, the K(m) for a 25N+2A substrate was found to be at least 20-fold lower than that for a 26N+1A substrate, although their V(max) values differed by only 2-fold. In addition, the total length of the RNA was found to contribute to CCR4 activity: up to 17 nucleotides (not necessarily poly(A)) could be recognized by CCR4. Poly(U), poly(C), and poly(G) were also found to be 12-30-fold better inhibitors of CCR4 compared with poly(A), supporting the observation that CCR4 contains a non-poly(A)-specific binding site. Surprisingly, even longer substrates (>/=45 nucleotides) stimulated CCR4 to become a processive enzyme, suggesting that CCR4 undergoes an additional transition in the presence of such substrates. CCR4 also displayed no difference in its activity with capped or uncapped RNA substrates. These results indicate that CCR4 recognition of its RNA substrates involves several features of the RNA that could be sites in vivo for controlling the rate of specific mRNA deadenylation.

The regulation of eukaryotic gene expression occurs at multiple levels, including the control of mRNA stability. The poly(A) tail at the 3Ј-end of eukaryotic mRNA plays a major role in controlling mRNA and protein levels (1). In many cases, deadenylation of the poly(A) tail is the first step required for decapping and 5Ј to 3Ј degradation of the mRNA sequence (2,3). Not only does rapid deadenylation lead to degradation of the mRNA and the shutoff of protein synthesis, but deadenylation itself reduces the efficiency of translation of mRNA (4). In yeast, mRNAs are deadenylated by first shortening the poly(A) tail to ϳ15 nucleotides, after which the 5Ј-cap is removed. Sequences within the RNA body play a role in controlling the rate of poly(A) removal as well as trans-acting proteins, including presumably the deadenylase itself (5,6).
In yeast, the CCR4 protein, as part of the CCR4-NOT complex, has been shown to be responsible for cytoplasmic deadenylation (7)(8)(9)(10). The CCR4 protein contains three major functional domains: an N-terminal activation domain, a central leucine-rich repeat (LRR) 1 domain that binds CAF1, and a C-terminal exonuclease III-like domain (9,11,12). The exonuclease III-like domain comprises the apparent CCR4 deadenylase function based on mutagenic studies (8,9). We (9) and others (8) have shown that CCR4 is a 3Ј-5Ј-RNase and a single-strand DNA-specific DNase that acts in a distributive manner with strong preferences for poly(A) substrates. The CCR4 protein is enzymatically active even in the absence of the CAF1 protein and other components of the CCR4-NOT complex. In addition, point mutations of the putative catalytic residues in the exonuclease domain of CCR4 abrogate its in vitro and in vivo activities (8,9). These observations strongly indicate that the CCR4 protein is the catalytic subunit of the mRNA deadenylase family of proteins, although CAF1 may display some deadenylase activity under certain conditions (10).
Model studies have indicated that alterations in deadenylation rates will have the greatest effects on changes in RNA degradation (13). We have therefore investigated the substrate preferences and enzymatic properties of CCR4 to understand the role CCR4 plays in selecting its mRNA substrates. We found that CCR4 prefers RNA substrates with at least 2 A residues at their 3Ј-ends. However, it was most efficient with substrates that were at least 17 nucleotides in length, although these sequences need not necessarily be poly(A). Inhibition studies further support a non-poly(A)-specific binding site for CCR4. Moreover, longer RNA substrates (at least 45 nucleotides) converted CCR4 to a processive enzyme. In addition, CCR4 by itself displayed no preference for capped substrates. These results suggest that the distributive-to-processive transition for CCR4 and the ability of CCR4 to contact a certain length or type of sequence may be sites by which other factors may regulate CCR4 deadenylase activity in vivo.

EXPERIMENTAL PROCEDURES
Plasmids and RNA Substrates-The full-length CCR4 open reading frame was cloned into a modified pYES2/CT vector (Invitrogen). The resultant CCR4 gene under the control of the GAL1 promoter contained an N-terminal FLAG tag and a C-terminal His 6 tag. Expression of CCR4 was in strain KY803-1a-1 (MAT␣ leu2-1 PET56 trp1-1 ura3-52 gal2 gcn4-1 ccr4::ura3). Commercially synthesized RNA oligonucleotides are described in Table I. The RNA substrates were generally labeled with T4 polynucleotide kinase and [␥-32 P]ATP and further purified by gel filtration on Sephadex G-25 spin columns. The capped and uncapped 50Nϩ5A and 50Nϩ30A substrates (pT3-L3(A 5 ) and pT3-L3(A 30 )) were synthesized in vitro as previously described (14). Poly(A) polymerase was used to attach radioactive 3Ј-AMP to the RNA sub-strates as described (7), and the resultant labeled RNAs were gel-purified.
Protein Extraction and Purification-FLAG-CCR4-His 6 purification was conducted by a few modifications as described previously (9). The yeast culture containing FLAG-CCR4-His 6 fusion protein was grown in a selective medium with 4% glucose to A ϭ 1.0 and then shifted to 4% galactose/raffinose for growth to A ϭ 1.5. The cells were washed and lysed in Tris acetate buffer (buffer A: 50 mM Tris HOAc (pH 7.9), 150 mM KOAc, 0.1 mM MgCl 2 , 10% glycerol, and 0.1% Nonidet P-40) plus a protease inhibitor mixture. After clarification of the crude lysate (15,000 ϫ g at 4°C for 10 min), the supernatants were ultracentrifuged at 100,000 ϫ g for 45 min at 4°C. The supernatants were subsequently incubated with washed Ni 2ϩ -agarose beads (with buffer A) for ϳ4 h at 4°C. After incubation, the Ni 2ϩ -agarose beads were washed extensively with buffer A containing 50 mM imidazole, and then the bound proteins were eluted with same buffer containing 250 mM imidazole. The eluted proteins were dialyzed against buffer A (with 5% glycerol) for 24 h with two changes of buffer. The dialyzed proteins were incubated with anti-FLAG M2 antibody-agarose (Sigma) at 4°C for 4 h. After extensive washes with buffer A, the FLAG beads were washed once with buffer B (buffer A without Nonidet P-40). To remove the background proteins, the beads were then incubated for 15 min at 4°C in buffer B containing 100 g/ml nonspecific hemagglutinin competitor peptide (YPYDVP-DYA, Research Genetics). After a pre-elution step, the bound FLAG fusion proteins were eluted twice with buffer B containing 200 g/ml FLAG peptide (Sigma). The eluted proteins were dialyzed against buffer B (with 5% glycerol) for 24 h with two changes of buffer. The purified proteins were then quantified using the Bradford assay. The CCR4 protein was further identified by SDS-PAGE and Western analysis.
In Vitro Deadenylation Assay-The following assay conditions were used for in vitro deadenylation: 50 mM Tris HOAc (pH 7.9), 0.01 mM MgCl 2 , 20 mM KOAc, 0.10 units of RNase inhibitor, 10% glycerol, 1-100 M RNA substrate, and the indicated amount of purified CCR4. The RNA substrates were radiolabeled as previously indicated (9). The reaction volume was 30 l, and the incubations were performed at 37°C unless otherwise indicated. Five-l aliquots of the reaction were removed, and the reaction was terminated by the addition of an equal volume of formamide/EDTA buffer. Reaction products were resolved by subsequent electrophoresis on 8 M urea and 16% polyacrylamide (19:1 (v/v) acrylamide/bisacrylamide) sequencing gels. The products were analyzed and quantified using PhosphorImager software.
Kinetic Assays for Distributive Activity of CCR4 -To determine the kinetic parameters for the distributive activity of CCR4, experiments were performed with radiolabeled 26Nϩ1A, 25Nϩ2A, and 22Nϩ5A RNA oligonucleotides using purified CCR4 protein. The enzyme concentration was adjusted to obtain linear reaction rates (at least up to 1 h). Five l of sample volume were removed at appropriate times following incubation at 37°C. The products were fractionated using a Tris borate/EDTA-polyacrylamide sequencing gel, on which the poly(A)excised products were distinguished as discrete bands. The reaction rate was determined by measuring the quantity of substrate remaining after 5 min for the 26Nϩ1A and 25Nϩ2A substrates, for which the predominant products were 26N and 25Nϩ1A, respectively. For the 22Nϩ5A substrate, the distributive rate was determined from the total amount of deadenylation that occurred in 5 min after quantitating the abundance of the 22Nϩ5A, 22Nϩ4A, 22Nϩ3A, 22Nϩ2A, and 22Nϩ1A products. For this analysis at low substrate concentration (see Fig. 3B, lower panel), we assumed that the relative rate of deadenylation of the 5A substrate to the 1A product was uniform based on our analysis showing uniform changes in abundance of each of these products with time (see Fig. 2 and Ref. 9; data not shown). At substrate concentrations Ͼ5 M, the predominant reaction was 22Nϩ5A to 22Nϩ4A (see Fig. 3B, lower panel). To determine K m and V max values, each kinetic experiment contained up to eight different substrate concentrations. Each substrate was analyzed in duplicate (sometimes in triplicate), and the experiments were conducted at least three times. The values obtained were fitted by linear regression analysis to the Michaelis-Menten equation and used to determine, from the intercepts, V max and K m constants (Table II).
The relative rates of the distributive deadenylation reaction were determined for the 2A substrates as described above, for the 5A substrates by quantitating the total amount of deadenylation that occurred within 5 min, and for the 20A substrates by determining the total amount of deadenylation that occurred within 1 min. For the 20A substrates, within 1 min, little of the completely deadenylated substrate is found; and a substantial amount of the 20A substrate remains, with only 19A, 18A, and 17A substrates being predominantly formed. Therefore, quantitating the total amount of deadenylation within 1 min provides a reasonable estimate of the 20A reaction rate, given that very little difference in 20A, 19A, and 18A reaction rates have been observed (see Fig. 2 and Ref. 9).

Conditions for in Vitro
Deadenylation-We optimized the conditions for CCR4 deadenylation in vitro using a radiolabeled 22Nϩ5A RNA substrate. We reported previously (9) that the exonuclease activity of CCR4 is dependent upon Mg 2ϩ . Here we found that CCR4 displayed a very low Mg 2ϩ optimum (0.01-0.1 mM) (Fig. 1). Other divalent cations such as Mn 2ϩ , Zn 2ϩ , Ni 2ϩ , and Co 2ϩ were inactive with CCR4 (data not shown). The monovalent K ϩ (17 mM), pH (7.9), and temperature (37°C) optima were also obtained ( Fig. 1). The Na ϩ optimum was also found to be low (20 mM) (data not shown). Moreover, we did not find any difference in these optimum conditions when we compared the partially purified CCR4 preparation (single-step FLAG purification) (9) with a pure CCR4 enzyme preparation (see below) (data not shown).
Distributive Action of CCR4 on RNA Substrates-We initially analyzed the activity of the CCR4 protein on a 7Nϩ23A RNA substrate. A time course of CCR4 deadenylation of this substrate ( Fig. 2) revealed that CCR4 clearly acted in a distributive manner. A plot of the relative abundance of each of the RNA species as a function of time showed that, as CCR4 catalyzed deadenylation, the abundance of the different species formed a nearly symmetric bell-shaped curve. This type of profile is consistent with CCR4 displaying no large differences in its binding and/or catalysis for substrates containing poly(A) tails from 23 A residues to at least 5 A residues (Fig. 2). Our previous results indicated, however, that CCR4 had major difficulty in deadenylating a substrate with only 1 terminal A residue (9). Careful analysis of the curves in Fig. 2 does indicate, however, a slight spreading of the abundance of the product toward shorter poly(A) lengths, suggestive of a slowing of the reaction rate with shorter substrate lengths (see below).

CCR4 Requires at Least 2 A Residues at the 3Ј-End to Effi-
ciently Deadenylate the RNA-To address whether CCR4 exhibited a binding or catalytic deficiency with the 1A substrate, we purified CCR4 to homogeneity prior to analyzing its kinetic parameters. To facilitate purification of the CCR4 protein from yeast, the CCR4 gene fused with FLAG and His 6 epitopes was overexpressed in a strain deleted for CCR4. The resultant fusion protein was purified to near homogeneity by a two-step affinity chromatographic procedure from yeast strain KY803-1a-1 (ccr4) (Fig. 3A, lane 2).
Using radiolabeled 26Nϩ1A and 25Nϩ2A RNA substrates, we determined the apparent K m and V max for each substrate with CCR4. Typical reactions as a function of substrate concentration are displayed in Fig. 3B. With the 25Nϩ2A RNA substrate, CCR4 displayed a K m of 5.4 M and a V max of 2.1 M/min. Whereas the V max value for the 26Nϩ1A substrate was comparable to that for the 25Nϩ2A substrate (2.2 M/min), its K m was increased by 20-fold. The inability of CCR4 to efficiently deadenylate an RNA substrate with 1 A residue (9) can be attributed to a decreased apparent binding ability of the 1A substrate. We also determined an apparent K m of 3.5 M and a V max of 0.86 M/min for the 22Nϩ5A substrate, which are similar to those found for the 25Nϩ2A substrate.
Increased Length of Non-adenylated Nucleotides of RNA Molecules Enhances the Distributive Activity of CCR4 -The substrate specificity of CCR4 was further investigated using different lengths of non-adenylated nucleotides bearing constant length poly(A) tails of 2, 5, and 20 A residues. The relative activity of CCR4 was determined using a concentration of the RNA molecules of 0.31 M, which was significantly below the K m for the 25Nϩ2A and 25Nϩ5A RNA substrates. A typical reaction is displayed in Fig. 4 (A-C) for RNA substrates containing 2, 5, and 20 A residues at their 3Ј-ends, respectively. Table III summarizes the relative reaction rates for the substrates analyzed. Several observations can be made from the changes in relative activity for CCR4 distributive deadenylation as a function of the length of the non-poly(A) region of the RNA substrate. First, for the 2A and 5A substrates, increasing the N length from 6 to 12 nucleotides substantially enhanced the reaction rate by 2-3-fold. In contrast, increasing the N length further from 12 to 25 nucleotides increased the relative activity by only 1.4 -1.7-fold (Table III). Second, once the total were incubated with 300 pg of purified CCR4 protein, and aliquots were removed at the times indicated. The lengths of the RNA substrates are indicated on the left. B, 5A RNA substrates. Reactions were conducted as described for A. C, CCR4 becomes a processive enzyme with long RNA substrates. Radiolabeled 25Nϩ20A, 12Nϩ20A, and 6Nϩ20A RNA substrates were incubated with 150 ng of purified CCR4. Reactions were conducted as described for A. The length of each of the RNA substrates is indicated on the left. The reaction product labeled 1A for each substrate denotes the locations of the completely deadenylated products. D, CCR4 does not act endonucleolytically. The 25Nϩ20A substrate was radiolabeled at its 3Ј-end with poly(A) polymerase to which ϳ3-33 A residues were added. The asterisks denote the range in positions from the 22A to 52A products that would have formed if endonucleolytic cleavage had occurred at the 1A-2A phosphodiester of the original 25Nϩ23A to 25Nϩ53A substrates.
RNA length became ϳ17-26 nucleotides, no major differences in the distributive reaction rate were observed for different N lengths irrespective of the poly(A) length (Fig. 5). Third, 12Nϩ5A was as active with CCR4 as 12Nϩ20A, indicating that the length of the poly(A) sequences per se does not alter CCR4 activity. These observations suggest that CCR4 is able to recognize a certain length of RNA substrate (at least 17 nucleotides) and that the A content is not necessarily important for this recognition. These results are also in agreement with those shown in Fig. 2, where a slight spreading of the distribution of 7Nϩ23A products indicated a slight slowing of the reaction rate at shorter substrate lengths.
CCR4 Protein Transitions to a Processive Enzyme with Substrates Greater than 44 Nucleotides-It is shown in Fig. 4C  (lanes 2-6) that a significant portion of the 25Nϩ20A substrate was converted directly to the 25Nϩ1A product. These data suggest that CCR4 can also act in a processive manner. Because it is also possible that the 25Nϩ1A product was created by an endonucleolytic cleavage occurring at the 1A-2A phosphodiester bond, we 3Ј-end-labeled the 25Nϩ20A substrate with poly(A) polymerase and reconducted the CCR4 deadenylation reaction. As shown in Fig. 4D, at time 0, an additional 3-33 A residues were added to the 25Nϩ20A substrate. Deadenylation of this substrate by CCR4 under conditions in which  CCR4 displayed a processive reaction with the 25Nϩ20A substrate ( Fig. 4C; data not shown) or a 50Nϩ10A substrate (data not shown) indicated that the poly(A) tail was being removed without the concomitant endonucleolytic cleavage at the 1A-2A phosphodiester that would have resulted in a range of 22-52 poly(A) products (marked with asterisks in Fig. 4D). Similarly, using an in vitro synthesized 50Nϩ30A substrate uniformly labeled with A (14), no 29A endonucleolytic product was observed following CCR4 processive deadenylation that resulted in a 50Nϩ1A product (data not shown). These data confirm that CCR4 can act in a processive manner. CCR4 behaved as a processive enzyme to a much greater extent with the 25Nϩ20A substrate than with the 12Nϩ20A substrate (Fig. 4C, lanes  7-12), indicating that the total length of the RNA substrate is important for CCR4 processivity and not just the A tail length.
No processivity was observed with the 7Nϩ23A substrate (Fig.  2), although for the 6Nϩ20 substrate displayed in Fig. 4C  (lanes 16 -18), a very small amount of 6Nϩ1A substrate was formed as a result of an apparently processive reaction. Similarly, CCR4 reacted with the 50Nϩ5A RNA substrate in a processive manner (Fig. 6, lane 6), indicating that it is not the length of the poly(A) sequence that dictates the distributive-toprocessive transition for CCR4. In fact, CCR4 acted primarily in a processive manner with the 50Nϩ5A, 50Nϩ10A, and 50Nϩ30A substrates ( Fig. 6; data not shown), suggesting again that the total length of the substrate, not its poly(A) length, dictates the distributive-to-processive transition for CCR4. We conclude that CCR4 can act as a processive enzyme in the presence of RNA substrates with lengths of at least 45 nucleotides. No difference in CCR4 reaction rate or ability to act in a processive manner was observed with capped and uncapped 50Nϩ5A RNA substrates (Fig. 6). 2 It should be noted that the apparent faster rate of conversion with uncapped versus capped substrates shown in Fig. 6 was not borne out by other experiments. Similar results were obtained with capped and uncapped 50Nϩ30A RNA substrates (data not shown).
CCR4 Activity in the Presence of Various Inhibitors-To investigate further the CCR4 deadenylase activity specificity for its RNA substrate, we analyzed the effect of various competitive inhibitors on CCR4 poly(A) removal. The above described deadenylation assay using the 22Nϩ5A RNA substrate was performed in the presence of increasing amounts of 5Ј-AMP, 5Ј-GMP, 5Ј-UMP, 5Ј-CMP, poly(A), poly(U), poly(G), and poly(C). Table IV shows that UMP was an ϳ2-fold better inhibitor than AMP, CMP, and GMP. We also observed that the K I of inhibition for ATP (1.2 pM) was much lower than that for AMP (data not shown). However, as was previously observed for the poly(A) ribonuclease (PARN) deadenylase (15), this inhibition was shown to result from the chelating effect of ATP with Mg 2ϩ ions, which thereby prevented the catalytic potential of the CCR4 enzyme (data not shown). Interestingly, we observed that poly(U), poly(C), and poly(G) were much more inhibitory than poly(A), consistent with our above conclusion that the CCR4 protein may be able to recognize a certain length of the RNA substrate and that the poly(A) sequence is of little importance to this recognition. DISCUSSION In this study, we report on the substrate specificities of the yeast CCR4 deadenylase. Several features of its substrate recognition were identified. First, as previously indicated (9), CCR4 acted in a distributive fashion to deadenylate a 7Nϩ23A substrate. With other substrates of comparable or shorter length, CCR4 was also a distributive enzyme; yet we observed that, with substrates 45 nucleotides or longer, CCR4 could act in both a processive and distributive manner. The ability to be a processive enzyme was also independent of the length of the poly(A) region of the substrate. One model that could account for this distributive-to-processive transition is that CCR4 binds better to longer substrates and therefore remains attached for a longer period of time. This model is unlikely, however, in that no difference was observed in the relative distributive activity with RNA substrates longer than 17 nucleotides (Figs. 4C and 5), implying that CCR4 even with longer substrates displays no differences in inherent affinity for the substrate. An alternative model is that CCR4 itself undergoes a transition in the pres-2 R. Parker, personal communication.  Table III).
FIG. 6. Comparison of deadenylation rates with capped and uncapped substrates. Capped and uncapped uniformly labeled 50Nϩ5A RNA substrates, as indicated, were incubated with 30 ng of partially purified CCR4 under standard conditions for the indicated times. The products were fractionated by electrophoresis on a denaturing 7% polyacrylamide gel. The resultant products are indicated on the left. ence of longer substrates such that a pool of CCR4 can now act in a processive fashion. The oligomeric structures of the highly processive -exonuclease and poly(A) ribonuclease (PARN) deadenylase (16,17) suggest that one such transition may be for CCR4 to become oligomeric. Perhaps longer RNA substrates provide the structure that accelerates this process by, for instance, providing significantly long enough sequences to which multiple CCR4 proteins can attach.
Second, we found that, as a distributive enzyme, CCR4 preferred substrates with lengths of at least 17 nucleotides. The decreases in reactivity with shorter substrates appeared to be a result of CCR4 decreased binding affinity, as the substrate concentrations used in these experiments were significantly below the K m of binding. The length of the poly(A) sequence did not alter these preferences, confirming observations of others 2 and ours 3 that CCR4 cannot specifically bind to poly(A). Whether this binding preference results from a sequence-specific recognition is not known. However, inhibition studies with poly(U) and poly(G) indicated that these compounds were 20 -40-fold better inhibitors of the CCR4 distributive reaction than was poly(A). CCR4 might display sequence specificity for nonpoly(A) sequences that might be enriched with U or G. Given that poly(A) sequences in yeast are up to 70 -80 nucleotides in length, one model suggests that only the extreme 3Ј-end of the poly(A) tail and the 3Ј-untranslated region sequences are bound by CCR4, with most of the poly(A) sequence looped out and not in contact with CCR4.
Third, either as a distributive or processive enzyme, CCR4 was deficient in removing a single terminal A residue from the mRNA (9). This requirement for 2 A residues at the 3Ј-end was shown to be the result of a much higher (20-fold) K m for the 1A substrate than for the 2A substrate. CCR4 can remove the terminal A residue and can also remove C or other nucleotides, albeit at very slow rates (9). Fourth, CCR4 itself did not display any difference with capped or uncapped substrates, suggesting that other proteins interacting with CCR4 aid its recognition of capped mRNA (8).
These substrate specificity analyses suggest the following model for CCR4 recognition of its RNA substrate. CCR4 binds to at least 2 terminal A residues, presumably in its active site. It also displays recognition for at least another 15 nucleotides; whether they are completely contiguous to these 2 A residues is not known. Finally, CCR4 may change its conformation or alter its oligomeric structure with substrates of even greater length (Ն45 nucleotides) to become a processive enzyme. Preliminary modeling of CCR4 binding to a 17-mer RNA based on the structure of the exonuclease III family protein HAP1 (18 -20) indicates that the C-terminal catalytic domain of CCR4 would not easily accommodate binding to RNA of such length. However, other regions of CCR4, especially its LRR, could be involved in RNA binding. Given that two other LRR-containing proteins require their LRR domain for contacting their cognate RNA substrates (21,22), it is quite possible that the LRR of CCR4 directly or indirectly facilitates CCR4 recognition of its substrate.
These features of CCR4 recognition of its substrates may, of course, be targets for control of the deadenylation process. Regulating the ability of CCR4 to bind specific mRNA or to transition to its processive state would greatly influence its enzymatic activity. The CAF1 protein and other components of the CCR4-NOT complex do not appear to be likely regulators of these features. For example, partially purified preparations of CCR4 that contain the complete CCR4-NOT complex (9) do not display significant differences in deadenylation compared with the purified protein in terms of requiring at least 17 nucleotides or in the transition to processivity. Other trans-acting proteins such as PUF3 and PAB1 could control these processes (7,8,23). The role of the other components of the CCR4-NOT complex may be to confer additional regulatory aspects to the deadenylation process. These could include binding to the mRNA cap structure or recognition of additional specific RNA sequences not present in our model RNA substrates used in this study.
Previous studies on the deadenylation of several mRNAs in vivo (5,7,8) clearly indicate that the rate of deadenylation can vary considerably for different mRNAs. Careful analysis of these in vivo deadenylation studies indicates that PGK1, which deadenylates slowly, does so in a distributive manner (5). In contrast, faster deadenylated mRNAs such as MFA2, STE3, COX17, and GAL10 deadenylate with a combination of both processive and distributive processes (5,7,23). This is an expected result given that processivity would allow the more rapid accumulation of product compared with the comparatively slow distributive reaction. Because CCR4 can display both distributive and processive activities, future studies will need to address how regulating these processes relates to the observed differences in mRNA deadenylation. In addition, the previously observed initially slow deadenylation of all mRNAs up to loss of the first 10 A residues (5) followed by more rapid deadenylation of the bulk of the poly(A) tail suggests an additional site at which CCR4 deadenylation may be controlled.