Regulation of mrna Entry into Polysomes

Transcription

1 THE JOURNAL OF BOLOGCAL CHEMSTRY (C 1987 by The American Society for Biochemistry and Molecular Biology. nc. Vol. 262, No. 24. ssue of August 25, pp ,1987 Printed in U.S.A. Regulation of mrna Entry into Polysomes PARAMETERS AFFECTNG POLYSOME SZE AND THE FRACTON OF mrna N POLYSOMES* (Received for publication, September 12, 1986) Ellen M. Nelson and Matthew M. WinklerS From the Department of Zoology, Center for Developmental Bwlogy, University of Texas, Austin, Texas The kinetics of labeled histone mrna entry into polysomes was studied in nuclease-treated reticulocyte lysates. Added mrna rapidly bound 1 or 2 ribosomes. However, the formation of full size polysomes required at least 16 min. The amount of mrna bound to ribosomes reached a maximum (73%) within 2 min after mrna addition and then declined slowly for the remainder of the experiment. Two initiation inhibitors, aurintricarboxylic acid and 7-methylguanosine 5 -triphosphate, were found to affect polysome size and the fraction of mrna in polysomes in an opposite manner. These results suggest that initiation and reinitiation events may be intrinsically different. The relatively long time period required for the formation of large polysomes can be explained by large polysomes having higher initiation and/or reinitiation rates or slower elongation rates. These possibilities are not mutually exclusive. The results suggest that there exist several levels of control which can regulate polysome size and the fraction of mrna in polysomes. The basic steps in the pathway of protein synthesis in eucaryotic cells are now well understood. Almost completely defined in vitro translation systems can be assembled from purified components (see Jagus et al., 1981; Hershey, 1982 for reviews). However, the molecular mechanisms responsible for translational regulation are less well understood. The regulation of gene expression often occurs at the level of translation. Examples include modulation of protein synthesis in early invertebrate and vertebrate embryos, the shut-off protein synthesis during heat shock, serum deprivation, amino acid starvation and loss of substrate contact in tissue culture cells, heme deprivation in reticulocytes, and starvation or deciliation in Tetrahymena (see Jagus et al., 1981; Jackson, 1982 for reviews, and Farmer et al., 1978; Thomas et al., 1982; Calzone et al., 1983). The net effect in most of these cases is a reversible mobilization of existing mrna species. There appear to be at least two distinct pathways which regulate the assembly or disassembly of polysomes. One mechanism is explained by simple stochastic models in which no distinction is made between the binding of the first ribosome to an mrna and subsequent reinitiation events (Bergmann and Lodish, 1979; Godefrey-Colburn and Thach, 1981). This is exemplified by the effects of heme deprivation on reticulocyte protein synthesis. When a reticulocyte lysate is incubated in the absence of heme, protein synthesis shuts off after a * This work was supported by National nstitutes of Health Grant HD (to M. M. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed. short lag time. Polysomes become smaller until essentially all the ribosomes have run off the mrna. When heme is added back, ribosomes bind mrna to form polysomes again. n contrast to heme deprivation, the mechanisms which regulate the amount of mrna in polysomes in other examples appear to be fundamentally different. Cells in an active state (ie. unfertilized eggs or suspended tissue culture cells) have fewer ribosomes in polysomes than cells in the active state (Humphreys, 1971; Benecke et al., 1978). While there is less mrna in polysomes, the overall polysome size distribution is the same as under active physiological conditions. Thus, the mrna is not found in smaller than normal sized polysomes. This is an important observation because it implies that mechanisms exist for maintaining large polysome size even under conditions suboptimal for protein synthesis and that the fraction of mrna in polysomes can be regulated independently of polysome size. We use the term fraction of mrna in polysomes to refer to the distribution or partitioning of mrna between the nontranslated compartment (free mrnps) and the translated compartment (polysomes). A number of different mechanisms can be proposed to explain how polysome size could be regulated independently of the fraction of mrna in polysomes. One such mechanism would involve the binding of the first ribosome to an mrna being intrinsically different from the binding of subsequent ribosomes or reinitiation events. This type of situation has been discussed previously (Laskey et al., 1977; Asselbergs et al., 1978; Duncan and McConkey, 1982). Thus, mrnas which had bound a ribosome might bind subsequent ribosomes more efficiently than would free mrnas. This would result in the formation of fully loaded polysomes and free mrnas in the same environment. However, other mechanisms can also explain this phenomenon. Most studies of protein synthesis mechanism have focused on either the initial initiation events or the protein synthetic process after it has reached steady state. n order to obtain basic information about the process of polysome formation, we studied the kinetics of polysome formation in a cell-free translation system after the addition of exogenous mrna. Labeled histone mrna was added to nuclease-treated reticulocyte lysates, and the rate of ribosome binding and polysome formation was determined. Stochastic models predict that maximum polysome size will be reached rapidly, i.e. after one ribosome transit time. n contrast, we found that polysome formation is slow relative to average ribosome transit times. mrna bound to ribosomes appears to have a small competitive advantage over free mrna in that large polysomes are formed at the expense of free mrnas binding to ribosomes. The abbreviations used are: mrnp, messenger ribonucleoprotein; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; ATA, aurintricarboxylic acid; m GTP, 7-methylguanosine 5 -triphosphate

2 11502 Regulation of Polysome Size n addition we found that two different initiation inhibitors have opposite effects on the fraction of mrna in polysomes and on polysome size. These results indicate that the entry of mrna into polysomes is potentially subject to several levels of control so that polysome size and the fraction of mrna in polysomes can be regulated independently of each other. EXPERMENTAL PROCEDURES Materials-Nuclease-treated rabbit reticulocyte lysates were prepared essentially as described by Jackson and Hunt (1983). Lysates contained 150 mm potassium acetate, 0.5 mm MgC12, 10 mm phosphocreatine, 0.1 mm amino acids except methionine and serine, 5 pg/ ml creatine phosphokinase, 20 p~ hemin, and 50 pg/ml calf liver trna (Boehringer Mannheim). Labeled histone mrna was prepared by incubating embryos of the sea urchin Stronglyocentrotus purpuratus at 15 C with10 pci of [3H]uridine(40Ci/mmol,Schwarz/ Mann) added 5 h after fertilization and an additional 20 pci/ml [3H] uridine added 9 h after fertilization. Embryos were harvested at 10 h and histone mrna prepared as described previously (Winkler et al., 1985). A brief outline is as follows. Total RNA was isolated, sedimented on sucrose gradients, and fractions in the 9 S region of the gradient were assayed for their ability to program histone synthesis in a reticulocytelysate.peakfractions werepooledand ethanol precipitated. The mrna had a specific activity of about 12,000 cpm/ pg of RNA. While this mrna is contaminated with 18 S ribosomal RNA, on the basis of radioactivity, it is quite pure. Globin mrna waspreparedfrom salt-washed polysomes as described previously (Winkleretal.,1985).Aurintricarboxylicacid was obtained from Sigma, and 7-methylguanosine 5 triphosphate was from Pharmacia P-L Biochemicals. Methods-The entry of labeled histone mrna into polysomes was st.udied by preincubating nuclease-treated reticulocyte lysate for 5 min at 30 C to allow any fragments of polysomes remaining after the nuclease treatment to disaggregate. Histone mrna was added and, at appropriate intervals, 25-p1 aliquots were removed and added to 125 p1 of ice-cold sucrose gradient buffer (75 mm KCl, 5 mm MgC2, 10 mm HEPES, ph 7.4, 1 mm EGTA, and 1 mm dithiothreitol). n experiments withmore than sixtimepoints,thesealiquotswere frozen in liquid nitrogen until analyzed. Control experiments indicated that the freezing step did not affect results. Samples were then layered onto 15-40% sucrose gradients and centrifuged a in Beckman SW 50.1 rotor for 45 min, at 50,000 rpm, at 4 C. Gradients were analyzed by upward displacement with 60% sucrose with an SCO gradient fractionation apparatus. Approximately ml aliquots were collected directly into scintillation vials and counted with 4.5 ml of a 1:9 water Aquamix (West Chem Products) mixture. Data Reduction-The top four or five fractions of each gradient have variable counting efficiencies due to quenching by the large amounts of hemoglobin from the reticulocyte lysate. After scintillation counting the top five fractions of each gradient were spiked with 40,000 cpm of [ Hluridineandrecounted to determine the counting efficiency of eachvial.controls treated withedta or ribonuclease A indicate that there is a small amount of nonpolysomal material which creates a background of radioactive counts throughout the gradient. To correct for this 10cpmwere subtracted from the bottom fraction, 20 cpm from the next fraction, and 40 cpm from all subsequent fractions. These values are estimates extrapolated from several control experiments. After these corrections, the total number of cpm/gradient was determined and the cpm in each fraction normalized so that the total number of cpm/gradient was This value is the average of the total cpmfrom54 separate sucrose gradients for experiments which had the standard addition of 60 pg/ ml added histone mrna. n other experiments in which the amount of added mrna differed, the normalization was adjusted proportionally. The absorbance at 254-nm tracing was used to determine which RESULTS Entry of Labeled mrna into Polysomes-n vivo labeled histone mrna was used to study the kinetics of ribosome binding and polysome formation in nuclease-treated reticulocyte lysates. Labeled mrna was added to reticulocyte lysates, and at various time intervals aliquots were removed, diluted in ice-cold sucrose gradient buffer, centrifuged on sucrose gradients, and fractionated after pumping through a UV monitor. The amount of labeled mrna in each fraction was determined by scintillation counting. Correction factors were applied to compensate for differential quenching of radioactivity in different fractions and for the presence of nonribonuclease-sensitive radioactive material which sedimented in the gradients. These procedures are described in detail under Materials and Methods. A typical result is shown in Fig. 1A. Labeled mrna binds ribosomes and begins to move into polysomes. This can be seen both by shifts of radioactivity in the sucrose gradient and by the appearance of UV-absorbing material corresponding to increasingly large polysomes. Early time points reveal an initial rapid binding of mrna to 1 or 2 ribosomes. Later time points show continued but significantly slower binding of more ribosomes to form fully loaded polysomes. This differential movement of mrna into small polysomes followed by a gradual shift to larger polysomes is seen more clearly in Fig. 1B. This figure shows the relative amount of mrna in different polysome size classes at different times after mrna addition. Only after 16 min does the mrna become symmetrically distributed in large polysomes. The data from four experiments have been pooled and are shown in Fig. 1C. By 15 s after mrna addition, 59% of the mrna has bound at lease a single ribosome. The amount of mrna bound to ribosomes increases to 72% by 2 min after mrna addition. The amount of bound mrna then decreases slowly for the remainder of the experiment. This decrease is probably not caused directly by competition of mrnas for ribosomes since the level of added mrna is well below saturating levels. n most of the experiments described here, histone mrna was added at a level of 60 pg/ml. Fig. 1C also shows the results of an experiment where mrna was added at a level of 20 pg/ml. The ribosome binding kinetics, the fraction of mrna in polysomes, and polysome size are very similar to those seen with 60 pg/ml mrna. The total amount of ribosomes in polysomes at 20 gg/ml was understandably about one-third of that at 60 pg/ml. n addition, Fig. 2 shows the effect of varying mrna concentration across a wide range. Neither the total amount of mrna in polysomes nor the number of ribosomes in polysomes reaches saturation until a level of at least 240 pg/ml added histone mrna. However, both polysome size and the fraction of mrna in polysomes decreases with increasing mrna concentration. Thus, while ribosomes and probably most of the classical initiation factors are not limiting, it appears that there is competition for some other factor, perhaps an mrna binding factor. Thach s group has provided evidence for an mrna discrimination factor which would interact with mrnas prior to ribosome binding (Walden et al., 1981). The binding of such a factor could determine if a particular mrna would form polysomes or not. n addition, to demonstrate that these fractions or portions of fractions were in each particular polysome results are not complicated by one histone mrna having an size class. mrna in polysomes was defined as mrna in the 40 S, 80 intrinsic competitive advantage over others, we examined the S, or polysome region of the gradient. The voltage signal from the relative translational efficiencies of the histone mrnas in a UV monitor was run into a dual channel chart recorder, with one channel set at a high sensitivity, allowing for detection of even very competition assay. Histone mrna was translated at a numsmall amounts of polysomal material. n those few cases when distinct ber of different concentrations and the ratio of the translation polysome peaks were not visible, their positions were extrapolated products determined (data not shown). There was no evidence from parallel gradients which contained more UV-absorbing material. of differential competitive advantage. The decrease in the

3 Regulation Size of Polysome e 5 a O O O hoction Number a a 30 E : O a 0 Polysome Sue [ H Histone mrna (pg/rnl) FG. 2. Effect of varying histone mrna concentration. The fourcurvesshown are the fraction mrnainpolysomes (A), the number of cpm of labeled mrna in polysomes (A), median polysome size (O), and the relative number of ribosomes in polysomes (0). The relative number of ribosomes in polysomes was calculated as described in the legend to Fig. 1C. Aliquots were removed and processed for sucrose gradient analysis 15 min after the addition of histone mrna L Mlnutes FG. 1. Entry of labeled histone mrna into polysomes. A, distribution of histone mrna in polysomes as determined by measurements of absorbance at 254 nm and by scintillation counting of sucrosegradientfractions. The curue with the closedcircles (0) represents 3H cpm and the heavy line the A254 tracing. 60 pg/ml labeled histone mrna was added to a nuclease-treated reticulocyte lysate, and at the indicated times aliquots were removed and processed as describedunder ExperimentalProcedures. B, distribution of labeled histone mrna in different size classes of polysomes. The data in this figure were derived from the same experiment shown in A and were analyzed as described under Experimental Procedures, C, changesin the distribution of mrna,polysomesize, and the relative amount of ribosomes in polysomes. The data in this figure shown by the closed circles (0) represent the average values of up to four separate experiments, and the error bars show the standard deviation. The open circles (0) show the results of a single experiment in which 20 pg/ml histone mrna was used. Polysome size refers to the median polysome size. The relative number ofribosomeswas calculated by multiplying the number of cpm in each polysome size class by the number value of the polysome size class, i.e. 1 for 80 S, 2 for dimers, etc. and determining the sum of these values. amount of mrna bound to ribosomes after 2 min is not due to the early shut-off of protein synthesis in the system. Protein synthesis continues for at least an hour, and the fraction of ribosomes in polysomes increases until about 20 min. t is also unlikely that the decrease is due to degradation of mrna. Deproteinized mrna isolated from the region of sucrose gradients containing nontranslated globin mrnps is translationally active (data not shown). These experiments do not exclude the possibility that modifications of the globin mrnp occur during the experiment which alter the translational activity of the mrnp. The most striking observation is the relatively long period of time it takes to form full sized polysomes (Fig. 1C). Sto- chastic models predict that polysomes should become fully loaded within one ribosome transit time. The histone mrnas in these experiments code for the four core histones and H1. Their lengths range from 102 to 135 amino acids (H1 is 190 amino acids) as compared with 0- which have 141 and 146 amino acids, respectively. The ribosome transit time for globin at 31 C is 1.2 min (Howard et al., 1970). Thus, the average transit time for histones at 30 C should be similar. t is clear that the time required to form fully loaded polysomes is at least several transit times. This slow rate of loading implies that the rate at which a just-terminated ribosome reinitiates on an mrna must be faster than the rate at which new ribosomes bind the mrna. f this were not the case, then polysome size could not increase. There are a number of models which can explain these results, some of which are described in detail under Discussion. Effect of nitiation nhibitors on mrna Binding and Polysome Size-There are suggestions in the literature that primary initiation and reinitiation events may be intrinsically different. t has been shown that once polysomes are allowed to form in a nuclease-treated reticulocyte lysate that protein synthesis is quite resistant to effects the of 7-methylguanosine 5 -monophosphate (Asselbergs et al., 1978). We define primary initiation as the first time a ribosome binds an mrna and reinitiation as when a just-terminated ribosome rebinds the same mrna. Another consideration is that the first binding of a ribosome to an mrna may alter the mrna in such a way that subsequent initiation events differ. t seemed probable that if different types of initiation events exist, then there might be initiation inhibitors which would affect these processes differently. We screened a variety of inhibitors before finding two which gave clear-cut results. Aurintricarboxylic acid (ATA) and 7-methylguanosine 5 - triphosphate (m7gtp) were found to have different effects on polysome size and the fraction of mrna in polysomes. The results of these experiments are shown in Fig. 3. Levels of inhibitor were used which reduced the rate of protein synthesis to 62.5% of untreated controls. While protein synthesis was diminished to the same extent, these inhibitors had differing effects on polysome size and the amount of mrna in polysomes. ATA had only a small effect on the rate

4 -_ 1504 Regulation of Polysome Size Polysome Size 2 2 Minutes FG. 3. Effect of m GTP and ATA on the entry of labeled histone mrna into polysomes. A, distribution of labeled histone mrna in different polysome size classes. m GTP (23 p ~ and ) ATA (4.5 ~M) were added 2 min prior to the addition of the mrna. The data for the control represent average values from up to four separate experiments with the error bars showing the standard deviation. B, changes in the distribution of mrna, polysome size, and the relative amounts of ribosomes in polysomes. The relative amount of ribosomes in polysomes was calculated as described in the legend to Fig. 1C. The control values are from the same data set as shown in Fig. 1C. and extent of mrna binding, in contrast to m7gtp which decreased the fraction of mrna in polysomes. For example, 8 min after mrna addition, ATA only reduced the fraction of mrna in polysomes to 56% as compared to 65% for the control. m7gtp reduced the fraction of mrna in polysomes to 45%. Both drugs resulted in reductions in polysome size. However, ATA reduced polysome size to a much greater extent than did m7gtp. m7gtp did not significantly affect polysome size, as compared to the control, until after 4 min when polysome size stopped increasing with m7gtp, while continuing to increase with the control (compare Figs. 3A and 3B). These two inhibitors are both thought to inhibit mrna binding to 43 S initiation complexes. ATA also inhibits the formation of ternary complexes (Vazquez et al., 1982). m7gtp behaves as if it were competing with mrna for some factor (Chu and Rhoads, 1980; Suzuki, 1978). Neither inhibitor is thought to effect the elongation rate (Pestka, 1977; Suzuki, 1978). Although they both inhibit mrna binding, it is clear that they have different effects on polysome formation. These results suggest that there are sites in the protein synthetic machinery which can exert differential effects on polysome size and the fraction of mrna in polysomes. Thus, it is likely FG. 4. Entry of unlabeled globin mrna into polysomes. 20 pg/ml globin mrna was added to a nuclease-treated reticulocyte lysate; aliquots were removed at the indicated times and processed as described under Experimental Procedures. The 0-min time point was taken as rapidly as possible after the addition of the globin mrna. that these same parameters can be regulated independently of each other in the cell. Differences between Histone and Globin mrna-globin mrna, which is actually a combination of a- and &globin mrnas is often used as a model mrna. Most of the work described here used histone mrna preparations. A major reason for this is that in vivo labeled mrna preparations are less likely to contain damaged mrnas which will yield artifactual results in ribosome binding studies than in vitro labeled preparations. Histone mrnas are similar in size to globin mrnas; they both sediment at about 9 S in sucrose gradients. However, histone mrna differs from globin mrna in not having a poly(a) tail (Adesnik and Darnell, 1972). Thus, histone mrna preparations are not as pure as globin mrna since they cannot be isolated using oligo(dt)- cellulose. The histone mrna preparations are quite pure on the basis of radioactivity, since there is little or no ribosomal RNA synthesis and a very high rate of histone mrna synthesis during the period in which the mrna was labeled (Brandhorst, 1984). The results obtained with histone mrnas appear not to be limited to histone mrna but also are seen with globin mrna. Fig. 4 shows a series of polysome profiles at various time intervals after the addition of globin mrna. Polysome size continues to increase for at least 16 min after the addition of globin mrna. The major difference we observed between the histone and globin mrna is the presence of higher levels of half-mers in polysome profiles with the histone mrnas. These are thought to be polysomes with an additional 40 S ribosomal subunit that has bound but that still lacks bound 60 S subunit (Hoerz and McCarty, 1971). This may be caused by a lower affinity of 60 S subunits for histone mrna-43 S initiation complexes. Another possibility is that the AUG site on the histone mrna less is accessible to the 40 S subunit so that there is a delay between the time a 40 S subunit binds the mrna and the time it reaches the AUG site and is capable of binding the 60 S subunit. Another difference is that histone mrna may have a higher binding constant for ribosomes than does globin mrna. The largest polysomes we have observed using tracings for globin were 7 mers (polysomes with 7 ribosomes), whereas we observed 9 mers with histone mrna. DSCUSSON The kinetic observations made regarding polysome loading can be summarized as follows. 1) nitiation by the first ribo-

5 some on an mrna is very rapid; 2) the subsequent increase in polysome size is slow relative to the transit time occurs but at a rate sufficient to yield full size polysomes by 16 min after mrna addition; 3) reinitiation of terminated ribosomes probably occurs at a rate greater than the initiation of new ribosomes on mrna (primary initiation events); 4) the fraction of mrna bound to ribosomes reaches a maximum very early and then decreases during the time that polysome size is increasing. Regulation of Polysome Size different sized polysomes. Using a computer model to consider steric interactions, it was concluded that ribosome transit At the most fundamental level, polysome size will be deter- times were unaffected by the degree of polysome loading mined by the ratio of initiation frequency to termination (Bergmann and Lodish, 1979). However, their analysis does frequency. Termination is generally thought not to be rate- not consider the many other parameters which could affect limiting; thus the time a ribosome will be bound to an mrna elongation rate in the cell or in the cell-free system. We have will be determined by the ribosome transit time (Bergmann commonly observed that the rate of incorporation of labeled and Lodish, 1979). After one ribosome transit time, polysomes amino acid in a cell-free translation system is linear from an should be fully loaded. For polysome size to increase over a early time after the addition of mrna. The relatively long period of time, the ratiof initiation frequency to transit time period of time required to fully load polysomes suggests that must increase. There are several parameters which could act the rate of protein synthesis should continue to increase to increase the initiation rate during the period of time during the period of time that the number of ribosomes in polysome size is increasing. f ribosomes which terminate polysomes is increasing. However, this is not observed. The from a given mrna were to reinitiate more efficiently on simplest explanation for this is that the elongation rate is that mrna, then occasional initiation of new ribosomes faster early in the experiment when there are mainly small would result in polysomes which gradually increased in size. polysomes and decreases as polysomes become fully loaded. The experimental data with the two initiation inhibitors Such a decrease in elongation rate could be caused by paramsuggests that reinitiation events and primary initiation events eters such as the depletion, in the immediate environment of may be regulated differently. Thus, it is possible to have the polysome, of rate-limiting substrates such as charged different rate constants for ribosome binding for the two trna or GTP or more complex steric interactions. different types of initiation events. The gradual increase in n addition, intact reticulocytes recovering from NaF inhipolysome size at the same time that the total amount of bition of initiation show linear incorporation of labeled amino mrna bound to ribosomes is decreasing suggests that poly- acid within 3 min of NaF removal (Conconi et al., 1966). soma1 mrna is able to compete for ribosomes more effec- However, the amount and size of polysomes continued to tively than free mrna. This could be mediated by a cooper- increase for at least 60 min. The measured average ribosome ative effect in which the binding of the first ribosome increases transit time at 3 min, when 25% of the ribosomes was in the affinity of the mrna for subsequent ribosomes. This polysomes, was 18 s. At 51 min, when 51% of the ribosomes cooperative effect might be caused by the initiating ribosome was in polysomes, the transit time had increased to 42 s. disrupting mrna secondary structure so that subsequent These measurements were made at an incubation temperature ribosomes could initiate more efficiently. Such situations have of 37 "C (Conconi et al., 1966). These experiments support been found in bacterial systems (Lodish, 1976). An additional parameter which will promote reinitiation is the hypothesis that larger polysomes have slower elongation rates. However, this work must be interpreted with caution the distribution of ribosomes in the lysate and the actual since the NaF may have had other side effects, and exact ribosome concentration in the immediate environment of the 5' end of the mrna. t has been calculated that a ribosome which terminates at the 3' end of the mrna is closer to the 5' end of the mrna than are other ribosomes in the lysate (Howard et al., 1970). However, all of these factors do not explain the fact that free mrna binds to the first ribosome very rapidly. f there is a 1-min transit time and at 16 min the average polysome contains 4 ribosomes, then the initiation rate must be at least 4/min. Complex ad hoc models can be constructed which would allow rapid binding of the first ribosomes, slow binding of subsequent ribosomes, and still allow polysomal mrnas to outcompete free mrna for ribosomes. Such models are not very informative or useful, however. Another model, which can explain the slow formation of large polysomes and is consistent with the initial rapid binding of ribosomes to mrna, is one in which the elongation rates are variable for different sized polysomes. n this model the elongation rate for small polysomes is faster than for large polysomes. Thus, as a polysome increases in size, its elongation rate decreases, allowing more time for ribosomes to bind. This process would continue until other parameters became limiting. For example, if the transit time for a polysome with 1 or 2 ribosomes was 30 s and if mrnas bound ribosomes on an average of about once every 15 s, then polysomes with 2 ribosomes would rapidly form. This is what we actually observe. n Fig. 1 it can be seen that polysomes with 2 ribosomes are formed by 1 min after mrna addition. f the occasional binding of a third ribosome resulted in a decrease in the elongation rate, this would favor the formation of polysomes with 3 ribosomes and allow the occasional formation of poly- somes with 4 ribosomes. This process would continue until fully loaded polysomes were formed. Elongation rates are generally assumed to be the same for values for polysome size were not given. Nevertheless, the simplest explanation for these results is that large polysomes have slower elongation rates. A decrease in the elongation rate in larger polysomes does not exclude the possibility that the parameters discussed earlier, which affect the relative initiation rate, play a role in the slow formation of large polysomes. The concurrent increase in polysome size and decrease in the amount of mrna in polysomes suggests that polysomal mrna has a higher intrinsic affinity for ribosomes than does free mrna. However, decreased elongation rates for larger polysomes, under conditions where ribosomes are limiting, could yield similar results. t is likely that both reductions in the elongation rate and increases in the relative initiation rates may be responsible for mediating the slow increase in polysome size. Additional experimental results will be required to determine the role of changes in initiation and elongation rates during polysome loading. Direct measurements of elongation rates for different sized polysomes and of initiation and reinitiation rates should be possible. n summary, we have presented data which suggests that the process of polysome formation is more complex than previous stochastic models would indicate. Polysome size and the fraction of mrna in polysomes appear to be regulated at least partially independently of each other. n addition, differences in initiation rates and in ribosome elongation rates

6 11506 Regulation of Polysome Size in different sized polysomes may play a role in determining Farmer, S. R., Ben-Ze'ev, A., Benecke, B.-J., and Penman, S. (1978) polysome loading rates. t is possible that all of these param- Cell 16, Godefroy-Colburn, T., and Thach, R. S. (1981) J. Biol Chem. 256, eters are modulated in cells during different forms of trans lational regulation. Hershey, J. W. B. (1982) in Protein Biosynthesis in Eucaryotes (Perez- Bercaff, R., ed) pp , Plenum Publishing Corp., New York Acknowledgments-We thank Gisela Kramer and Roger Duncan Hoerz, W., and McCarty, K. S. (1971) Biochirn. Biophys. Acta 228, for helpful discussions and comments on the manuscript. We also Howard, G. A., Adamson, S. D., and Herbert, E. (1970) J. Biol. Chern. thank Lewis Patterson for his assistance in the preparation of this 245, manuscript. Humphreys, T. (1971) Deu. Biol. 26, Jackson, R. J. (1982) in Protein Biosynthesis in Eucaryotes (Perez- REFERENCES Bercaff, R., ed) pp , Plenum Publishing Corp., New York Jackson, R. J., and Hunt, T. (1983) Methods Enzyrnol. 96, Jagus, R., Anderson, W. F., and Safer, B. (1981) Prog. Nucleic Acid Res. Mol. Biol. 25, Laskey, R. A., Mills, A. D., Gordon, J. B., and Partington, G.A. (1977) Cell 11, Lodish, H. F. (1976) Annu. Rev. Biochem. 45, Petka, S. (1977) in Molecular Mechanisms of Protein Biosynthesis (Weissbach, H., and Pestka, S., eds) pp , Academic Press, Orlando, FL Suzuki, H. (1978) J. Biochern. (Tokyo) 84, Thomas, G., Martin-Perez, J., Siegmann, M., and Otto, A. M. (1982) Cell 30, Vazquez, D., Zaera, E., Doh, H., and Jimenez, A. (1982) in Protein Adesnik, M., and Darnell, J. E. (1972) J. Mol. Biol. 67, Asselbergs, F. A. M., Peters, W., van Venrooij, W. J., and Bloemendal, H. (1978) Eur. J. Biochem. 88, Benecke, B. J., Ben-Ze'ev, A., and Penman, S. (1978) Cell 14, Bergmann, J. E., and Lodish, H. F. (1979) J. Biol. Chem. 254, Brandhorst, B. P. (1984) in Deueloprnental Biology: A Comprehensiue Synthesis (Browder, L., ed) pp , Plenum Publishing Corp., New York Calzone, F. J., Angerer, R. C., and Gorovsky, M.A. (1983) J. Biol. Chem. 258, Chu, L.-Y., and Rhoads, R. E. (1980) Biochemistry 19, Conconi, F. M., Bank, A,, and Marks, P. A. (1966) J. Mol. Biol. 19, Duncan, R., and McConkey, E. H. (1982) Eur. J. Bwchem. 123, Biosvnthesis in Eucarvotes (Perez-Bercaff. R.. ed) DD " Pleium Publishing Civ., New York Walden. W. E.. Godefrev-Colburn.. T... and Thach. R. E. (1981).. J. Biol. Chem. 256, Winkler, M. M., Nelson, E. M., Lashbrook, C., and Hershey, J. W. B. (1985) Deu. Biol. 107,