Sep 8, 1980 - short poly(A) sequences were sensitive to digestion with micro- coccal nuclease, suggesting that they were not associated with pro- tein.
Proc. Natl Acad. Sci. USA Vol. 78, No. 1, pp. 83-87, January 1981 Biochemistry
Cytoplasmic polyadenylate processing events accompany the transfer of mRNA from the free mRNP particles to the polysomes in Physarum (poly(A)-protein complex dissociation/poly(A) degradation/microinjection/translational regulation)
DAVID S. ADAMS*, DANIEL NOONAN, AND WILLIAM R. JEFFERYt Department of Zoology, University of Texas, Austin, Texas 78712
Communicated by Philip Siekevitz, September 8, 1980
ABSTRACT The relationship between the mRNA in the polysomes and the free cytoplasmic messenger ribonucleoprotein of Physarum polycephalum was studied by microinjection techniques. Labeled free cytoplasmic ribonucleoprotein, prepared from donor plasmodia, was microinjected into unlabeled host plasmodia, and its fate was followed in the host ribonucleoprotein particles. Approximately one-half of the poly(A)-containing RNA [poly(A)+RNA] that originated from the microinjected particles was incorporated into the host polysomes by normal translational processes within 1 hr. Very short poly(A) sequences (" 15 nucleotide residues) were found in these poly(A)+RNA molecules. These short poly(A) sequences were sensitive to digestion with micrococcal nuclease, suggesting that they were not associated with protein. Because the poly(A)+RNA molecules of the microinjected free cytoplasmic mRNP had originally contained poly(A) sequences 50-65 nucleotides long and were associated with protein, extensive poly(A) degradation and poly(A)-protein complex dissociation must have occurred during their incorporation into the polysomes or during their translation. These results demonstrate a precursor-product relationship between free cytoplasmic mRNP and polysomal mRNA and suggest that the incorporation process in Physarum is accompanied by structural modifications in the poly(A) region of mRNA. They also imply that the polysome is a site for disruption of the poly(A)-protein complex and poly(A) degradation.
Eukaryotic messenger RNA (mRNA) occurs in two types of cytoplasmic particles (1). Usually, most of it is found in the polvsomes, and the remainder is associated with specific proteins in a class of lighter particles known as the free cvtoplasmic messenger ribonucleoprotein (mRNP). Very little is known about the biological role of the free cvtoplasmic mRNP, even though in some cells it contains 30-40% of the total mRNA (2-7). Pulsechase labeling experiments and kinetic studies have been unable to establish a precursor-product relationship between the free cytoplasmic mRNP and the polysomal mRNA (4, 7-9). Recent studies have shown that many of the mRNA species in the free cytoplasmic mRNP of actively growing cells are inefficiently translated and mav essentially represent inactive forms of mRNA (9-13). Structural differences between mRNA molecules and their associated proteins in the free cytoplasmic mRNP and in the polysomes may suggest ways by which certain messages are selected for translation. We have reported the existence of structural differences between the poly(A) region of mRNA molecules located in the polysomes and that in the free cytoplasmic mRNP of the acellular slime mold Physarutn polycephalurn (5). Polysomal poly(A)+mRNA contains very short adenylate tracts that are free of associated protein, but free cvtoplasmic mRNP contains much longer polv(A) sequences organized in a polv(A)-protein complex. These findings suggested that strucThe publication costs ofthis article were defrayed in part b1 page charge payment. This article must therefore be herehv marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.
tural modifications in the polv(A) region of mRNA might accompany its transfer to the polysomes (5). In the study reported here, we used microinjection techniques to provide direct evidence for the existence of these modifications and of a precursor-product relationship between the free cytoplasmic mRNP and the polysomal mRNA. Cytoplasmic processing events involving poly(A)-protein complex dissociation and extensive poly(A) degradation appear to accompany the incorporation of mRNA into the polysomes. MATERIALS AND METHODS Preparation of Labeled mRNP. Physaruin polycephalm (Carolina strain) was cultured in the microplasmodial form, using the axenic medium and culture conditions described (14, 15). The cultures were labeled by the addition ofsterile aqueous [2,8-3H]adenosine (48 Ci/mmol; 1 Ci = 3.7 X 101" becquerels), [5,6-3H]uridine (41 Ci/mmol), or both (ICN) directly to the medium to final concentrations of 20 ,uCi/ml. Labeled polysomal and postpolysomal RNP fractions were deposited from postmitochondrial supernatants by differential centrifugation as described (5) and stored at liquid N2 temperature. The nonpolysomal RNP fraction contained the free cytoplasmic mRNP, most of the 80S monoribosomes, and the ribosomal subunits but no detectable polysomes (5). Microinjection of Labeled mRNP. Healthy, 48-hr-old macroplasmodia prepared by conventional methods (16) from microplasmodial cultures were used as hosts. Micropipets were pulled to a diameter of 10-30 gum from custom glass tubing. The tubing was pretreated with chromic acid to remove leached alkalis and nuclease activity, washed with 20 mM EDTA to prevent coagulation of the plasmodial protoplasm (17), and finally rinsed in distilled water. The labeled nonpolysomal RNP suspensions were centrifuged at 5000 rpm in a Beckman JA-14 rotor before injection to remove the aggregates that sometimes formed during N2 storage. About 10-20 ,ul of concentrated nonpolysomal RNP suspension was microinjected into the protoplasmic veins at 5-10 sites in each host plasmodium (=z2 ,ul of suspension per site in ==0.2-,ul portions). The microinjections were separated by 1-min intervals, which were required for the clearing of the injected materials through the vein. This gradual introduction of the RNP suspension into the plasmodial vein prevented the formation of exudates at the wound and the recurrent tendency of the plasmodia to wall off and destroy protoplasm in the vicinity of the injection site. The total injection process required -5-10 min. The microinjected plasmodia were incubated at 25°C. Abbreviation: mRNP, messenger ribonucleoprotein. Present address: The Rockefeller University, New York, NY 10021. t To whom reprint requests should be addressed.
*
83
Biochemistry: Adams et al.
84
Proc. Nad Acad. Sci. USA 78 (1981)
Fractionation and Assay of Microinjected mRNP. Media developed for the isolation of mRNP from Physarum macroplasmodia (18) were used for fractionation of the microinjected macroplasmodia. The polysomal and postpolysomal RNP fractions were obtained by differential centrifugation as described for microplasmodia (5). All other procedures have been discussed (5, 15, 18, 19).
RESULTS Incorporation of Microinjected mRNP into the Polysomes. Physarum macroplasmodia proved well suited for the microinjection and tracing of concentrated RNP suspensions. The microinjected material moved rapidly through the veins and was distributed through the plasmodium by cytoplasmic streaming. About 70% of the acid-insoluble radioactivity was recovered in host plasmodia after 1 hr of incubation. A considerable portion (74%) of this acid-insoluble material, most of which probably represents ribosomal RNA, was found in the RNP fractions. When the subcellular distribution of labeled poly(A)+RNA was examined, the polysomal fraction was found to contain about 4 times the radioactivity present in the RNP particles (Table 1). To test the possibility that new transcription causes the appearance of labeled poly(A)+RNA in the polysomes, the microinjection experiments were repeated using host plasmodia that had been preincubated with actinomycin D at 300 ,ug/ml, a concentration known to severely depress RNA synthesis in Physarum (20). The proportion of labeled poly(A)+RNA that appeared in the polysomal fraction was not reduced in actinomycin D-treated plasmodia. Thus, the presence of labeled poly(A)+RNA in the polysomes cannot be accounted for by endogenous RNA synthesis. To determine whether cytoplasmic polyadenylation, a process resistant to actinomycin D (21-23), was responsible for the appearance of labeled poly(A)+RNA, the polysomal fraction of host plasmodia incubated for 1 hr after the microinjection of labeled nonpolysomal RNP was extracted with phenol, and the distribution oflabeled adenosine residues within the RNA molecules was determined by pancreatic RNase digestion. If the labeling were caused by cytoplasmic polyadenylation, it would occur exclusively within the poly(A) sequence and, as a result, show resistance to pancreatic RNase. Instead, about 96% of the poly(A)+RNA radioactivity was sensitive to pancreatic RNase Table 1. Distribution of microinjected poly(A)+RNP between polysomal and postpolysomal RNP fractions of the host plasmodia Actinomycin Dpretreated Normal plasmodia plasmodia Poly(A)- MicroPoly(A)+ MicroRNA, injected RNA, injected cpm activity, % Fraction cpm activity, %
Microinjected nonpolysomal RNP Polysomal RNP Postpolysomal RNP Total recovery in RNP
1800 580
100.0 32.2
2100 740
100.0 35.2
150
8.3
230
10.9
730
40.5
970
46.1
In actinomycin D-pretreated cultures, the drug was added to a final concentration of 300 pg/ml 60 min before microinjection and was retained in the culture medium for the duration of the subsequent incubation period. Poly(A)+RNA radioactivity was measured by filtration of deproteinized fractions through poly(U)-impregnated filters (5).
1.5 o15
'-4
4
X
1.0 0.5 D
C -4~~~~~
4
F 1.5 0.
6
4
2
6
4
2
Distance sedimented, cm
FIG. 1. Sedimentation of endogenous and microinjected poly(A)+RNP from the polysomal and postpolysomal fractions of Physarum plasmodia. (A) Endogenous polysomal poly(A)+RNP. (B) Endogenous nonpolysomal poly(A)+RNP. (C) Microinjected poly(A)+RNP in the polysomal fraction. (D) Microinjected poly(A)+RNP in the nonpolysomal fraction. The polysomal and nonpolysomal fractions were prepared from plasmodia labeled in vivo for 2 hr with [3H]-adenosine (A andB), or unlabeled plasmodia microinjected with [3Hluridine- and (3H]adenosine-labeled nonpolysomal RNP (C and D). Centrifugation was for 90 min (A and C) or for 5 hr (B and D) at 35,000 rpm through 10-50% linear sucrose gradients in buffer A (50 mM Tris HCl, pH 7.6/ 100 mM NaCl/100 mM MgCl2) in a Beckman SW 41 rotor at 4°C. Light lines represent absorbance tracings at 254 nm. Heavy lines (e) represent poly(A)+RNA radioactivity in each fraction as determined by poly(U) filtration after deproteinization (5). Position of the 80S marker was determined by centrifugation of Ehrlich ascites cell monoribosomes on parallel gradients.
and, therefore, must have been in the nonpoly(A) portions of the molecules. This result is inconsistent with the possibility that the appearance of labeled poly(A)+RNA in the polysomal fraction is due to cytoplasmic polyadenylation. The sedimentation profiles of microinjected and endogenous poly(A)+RNP derived from the polysomal and postpolysomal fractions ofthe host plasmodia are shown in Fig. 1. The microinjected poly(A)+RNP recovered from the host polysomal and postpolysomal fractions had sedimentation profiles that were similar to the endogenous poly(A)+RNP of these subcellular fractions. The adenosine- and uridine-labeled poly(A)+RNP in the polysomal fraction, as would be expected for poly(A)+RNA of polysomal origin, was converted to 80S particles by mild RNase digestion and to 20-60S RNP after treatment with EDTA (Fig. 2). Although these results suggest that the microinjected poly(A)+RNP is actually incorporated into the polysomes, further experiments using specific translational inhibitors were conducted to determine whether the incorporation process occurred by normal initiation steps (Fig. 3). Microinjection of labeled nonpolysomal RNP into host plasmodia pretreated with NaF, an inhibitor of the initiation of protein synthesis (24), resulted in the virtual absence oflabeled poly(A)+RNP in the polysomes (Fig. 3A); rather, it accumulated primarily in particles that sedimented at -40 S. In contrast, labeled poly(A)+RNP appeared to accumulate primarily in 80S particles when microinjected into plasmodia pretreated with cyclohex-
Proc. NatL Acad. Sci. USA 78 (1981)
Biochemistry: Adams et A A.
A E
4
80S
-
g2
60S 40S
0
6
4
2 6 Distance sedimented, cm
4
2
FIG. 2. Effects of mild (1 ,ug/ml) pancreatic RNase, treatment (A) and 40 mM EDTA treatment (B) on sedimentation rate of microinjected adenosine- and uridine-labeled poly(A)+RNP from polysomal fraction of unlabeled host plasmodia. Polysomes dissolved in buffer A were treated for 15 min on ice before layering on the gradients. Centrifugation was for 90 min (A) or for 5 hr (B). The gradients used in B contained 20 mM EDTA. Centrifugation and other conditions were similar to those given in the legend for Fig. 1. Positions of the 80S, 60S, and 40S markers were determined by centrifugation ofEhrlich ascites cell monosomes, and EDTA-treated polysomes on parallel gradients.
About 4% of the endogenous polysomal poly(A)+RNA was resistant to RNase, whereas the proportion of resistant sequence in the free cytoplasmic mRNP was greater, usually about 10%. This difference presumably reflects the existence of longer poly(A) sequences in the free cytoplasmic mRNP. The proportion of RNase-resistant sequence was 9.6% in the microinjected poly(A)+RNP particles from the free cytoplasmic mRNP of the host plasmodia but only 4.6% in the molecules that were incorporated into polysomes. Thus, a substantial loss of poly(A) could accompany the incorporation of microinjected mRNP into the polysomes. This possibility was confirmed by direct measurement ofthe poly(A) sizes by polyacrylamide gel electrophoresis. The electrophoretic mobilities of poly(A) sequences prepared from endogenous polysomal RNA, the microinjected free cytoplasmic mRNP, and the microinjected mRNP that entered the host polysomes are shown in Fig. 4. To determine the number of poly(A) sequences in each size class, the electrophoretic mobilities were converted into length-frequency distributions (Fig. 4 B, D, and F). As shown previously (5), the average size Poly (A) size, nucleotides 160 96 220 101
A
imide ( Fig. 3B), an inhibitor of polypeptide translocation that has little effect on initiation (25, 26). These results suggest that the microinjected poly(A)+RNA was incorporated into the polysomes via 80S initiation complexes. It is concluded that the labeled poly(A)+RNA molecules found in the polysomal fraction after microinjection are incorporated into the polysomes by normal translational processes. Poly(A) Degradation Accompanies Incorporation of Microinjected mRNP into Polysomes. It has been reported that poly(A) sequences from the polysomal RNA of Physarum are much shorter than those from the free cytoplasmic mRNP (5). This suggests that the poly(A) sequence might be subject to degradation during the incorporation of mRNA into polysomes. To determine whether poly(A) is hydrolyzed as the microinjected mRNP enters the polysomes, we measured the percentage of RNase-resistant radioactivity in poly(A)+RNA molecules from the polysomal and postpolysomal fractions of host plasmodia.
85
28
160 96 220 101
116
54
28 54
16
B
tN~~~~s .
4 Il l l
r
T
10
D~~~~
20 .5
1
.0.
poly(A)~ seune ~~~ inmconetdnnoysmlpl()RP(
6 2 Distance sedimented, cm
FIG. 3. Effects of NaF (A) and cycloheximide (B) on subcellular distribution of microinjected poly(A)+RNP. (A) Plasmodia were pretreated with 20 mM NaF or cycloheximide 20 ,ug/ml for 2 hr prior to microinjection and during the 1-hr incubation period. Centrifugation was for 90 min. Other details were similar to those given in the legend for Fig. 1. Light lines represent absorbance tracings (254 nm) of Physarum polysomes isolated from control plasmodia and are included for marker reference only.
n
ai2 (9 w d
2b
Fity,and right-ohand
(otane nd number average sizes mobilis aers
filractionof RNA aeraged from bydvdn in) haiactivity oa waeachindbyply tance of migration along the gel). In each panel, the left-hand dashed line represents the average mobility or sequence number ofthe larger poly(A) class (from free mRNP) and the right-hand one represents the same features for the smaller (polysomal) poly(A) class. Poly(A) sizes (upper abcissa) were determined from the electrophoretic mobility of poly(A) standards run on parallel gels. A mobility of 1.0 represents a migration distance of 4.8 + 0.4 cm through the gel toward the anode.
86
Proc. Nad Acad. Sci. USA 78 (1981)
Biochemistry: Adams et al.
Table 2. Micrococcal nuclease sensitivity of the poly(A) moiety of Physarum mRNP fractions after microinjection of labeled free cytoplasmic mRNP into plasmodia Radioactivity retained on filter, cpm Micrococcal Micrococcal nuclease nuclease resistance, Fraction Undigested digest % Endogenous nonpolysomal RNP 836 637 76.2 Postinjection polysomal 960 128 RNP 13.3
Postinjection nonpolysomal RNP
347
304
87.6
of the nonpolysomal poly(A) was =60 nucleotides, which is =4 times the length of the endogenous polysomal poly(A) (Fig. 4 B and D). Short poly(A) sequences, averaging only =15 nucleotides, were also observed in the microinjected poly(A)+RNA molecules recovered from the host polysomes (Fig. 4F). Although a small proportion of longer poly(A) sequences were also observed in this fraction, the short poly(A) sequences were by far the dominant components. The longer poly(A) sequences observed in this fraction may represent poly(A)+RNA that has recently entered the polysome and thus has poly(A) sequences in intermediate stages of degradation. Because the short poly(A) sequences were virtually absent from the microinjected nonpolysomal mRNP, their appearance in the polysomes must be a consequence of poly(A) degradation. These results suggest that the incorporation of mRNA into the polysomes in Physarum is accompanied by extensive poly(A) degradation. Poly(A)-Protein Complex Dissociation Accompanies Incorporation of Microinjected mRNP into Polysomes. Because the endogenous poly(A)+RNA of Physarum polysomes does not contain a poly(A)-protein complex (5), further microinjection experiments were conducted to determine whether the poly(A)associated proteins of free cytoplasmic mRNP are removed during its transfer to the polysomes. Micrococcal nuclease sensitivity was used to assay for the existence of a poly(A).protein com-
plex. Deproteinized poly(A) and the short poly(A) sequences in the polysomes of Physarum are destroyed by this enzyme, whereas the long poly(A) sequences in poly(A)-protein complexes show 70-100% resistance (5, 19). We found that labeled poly(A)-containing components from the polysomes of microinjected plasmodia, unlike those from the postpolysomal fraction, were almost entirely hydrolyzed by the enzyme (Table 2). This result suggests that the poly(A)-protein complex of free cytoplasmic mRNP is dissociated during the formation of translational complexes or during subsequent translation, which implies that the polysome is a site for the disruption of the poly(A)-protein complex and for poly(A) degradation. DISCUSSION
The results described in this paper represent a direct demonstration of a precursor-product relationship between the mRNA in the free cytoplasmic mRNP and that in the polysomes. The procedure used was to follow the fate oflabeled free cytoplasmic mRNP microinjected into unlabeled Physarum plasmodia. Previous attempts to establish such a relationship, primarily pulsechase labeling experiments, used drugs with possible side effects (7, 13, 27) and were largely unsuccessful [only a small pro-
portion, if any, of the free cytoplasmic mRNP could be designated as polysomal precursors (4, 7-9, 13)]. After correcting for the loss of radioactivity due to poly(A) degradation, we estimate that =50% of the microinjected poly(A)+RNA molecules enter the polysomes during the 1-hr incubation period following microinjection. This estimate is a minimum value because (i) some poly(A) sequences might have been completely lost during the poly(A) degradation process and thus would not be accounted for in our assay of polysomal material and (ii) the proportion of poly(A)+RNA in the polysomes was not measured for incubation periods longer than 1 hr. It is not surprising that a large portion of the microinjected mRNP is incorporated into the polysomes in such a short incubation period if it is recalled that a relatively small mass of mRNP is introduced into a vast excess of potential translational machinery in the host plasmodium. Thus, we believe that a large proportion of the free cytoplasmic mRNP is potentially capable of translation when saturated with competent ribosomes and other translational factors. This hypothesis is supported by recent studies that show that most ofthe free cytoplasmic mRNP of mammalian cells can eventually be driven into translational complexes by cycloheximide treatment (13, 27). Our results also suggest that the polysome is the site of cytoplasmic processing events involving the poly(A) region of Physarum mRNA. The existence of the two cytoplasmic processing events described in this report-poly(A)-protein complex dissociation and poly(A) degradation-could have been predicted from previous results on the structure and behavior of poly(A)+RNA in Physarum (5, 15). However, although the previous results were consistent with the occurrence of the two processing events, it remained to be shown that the poly(A)+RNA in free cytoplasmic mRNP can serve as a precursor to the polysomal mRNA. We have shown this by finding that the poly(A)+RNA from a substantial portion of microinjected nonpolysomal RNP particles is incorporated into the host polysomes and that this process is accompanied by poly(A).protein complex dissociation and extensive poly(A) degradation. The poly(A) degradation process we observed during this transfer should not be confused with the age-dependent poly(A) shortening phenomenon described previously (28, 29). That process removes no more than 15 adenylate residues, resulting in a poly(A) steady-state size of 45-50 nucleotides, and is insensitive to cycloheximide. The poly(A) degradation process described in this paper removes at least 45-50 residues and is cycloheximide sensitive (15). It remains to be determined whether the cytoplasmic processing events involved in mRNA maturation in Physarum also occur in other eukaryotes. There is some question as to whether they do; there are many reports ofthe presence of poly(A)-protein complexes in mRNA obtained from the polysomes (30-36). In most of these studies, however, the polysomes may not have
been entirely separated from the largest free cytoplasmic mRNP, which sometimes overlaps them in sedimentation range (37). In Physarum, the relatively narrow sedimentation range of the free cytoplasmic mRNP (5) may permit a more effective separation between these two classes ofsubcellular particles. There are several positive indications that a poly(A) degradation process such as the one we have described in Physarum may exist in other eukaryotes. In fertilized sea urchin eggs, the poly(A) sequences of maternal mRNA are extensively degraded at the same time that increased levels of polysomes appear in the cytoplasm (38). There are a number of reports of existence of two forms of mRNA that code for a single protein, one lacking and one containing poly(A) (39-42). The presence of a short adenylate tract at the 3' terminus of the so-called poly(A)-RNA cannot be excluded in any of these cases. Protamine mRNA from
Biochemistry:
Adams et al.
trout testes contains both poly(A)+ and poly(A)- forms (42). The poly(A)+ form is found only in the free cytoplasmic mRNP, whereas the poly(A)- form and poly(A)+protamine mRNA coexist in the polysomes (43). Similar to the situation in Physarum, it has been suggested that the poly(A)- form of protamine mRNA is derived from the poly(A)+ form during the process of translation (44). Recent studies with mammalian cells (45) suggest that the poly(A) sequence can be selectively cleaved from mRNA in vitro and that this process also occurs in vivo and may be responsible for the formation of the poly(A)-lacking messenger RNA sequences. At present, the relationship between the poly(A)-associated processing events described in this paper and the mRNA translational cycle is unknown. Because the poly(A) degradation probably occurs after the mRNA has entered the polysome, we feel that it is probably not involved in mrRNA translational selection. A more likely possibility is that this process is related to mRNA stability. We thank Ms. Pricilla Kemp for technical assistance. This work was supported by Grants PCM-77-24767 from the National Science Foundation and GM-24119 from the National Institutes-of Health. 1. Preobrazhensky, A. & Spirin, A. (1978) Prog. Nucleic Acid Res. Mol. Biol. 21, 1-37. 2. Lindberg, U. & Persson, T. (1972)J. Mol. Biol. 86, 451-468. 3. Jelinek, W., Adesnick, M., Salditt, M., Sheiness, D., Wall, R., Molloy, G., Phillipson, L. &.Darnell, J. (1973)J. Mol. Biol. 75, 515-532. 4. MacLeod, M. (1975) Biochemistry 14,4011-4018. 5. Adams, D., Noonan, D, .& Jeffery, W. (1980) Biochemistry 19, 1965-1970. 6. Levy, W. & Rizzino, A. (1977) Exp. Cell Res. 106, 377-380. 7. Spohr, G., Granboulan, N., Morel, C. & Scherrer, K. (1970) Eur. J. Biochem. 17, 296-318. 8. Dworkin, M. & Infante, A. (1976) Dev. Biol. 53, 73-90. 9. Enger, M. & Hanners, J. (1978) Biochim. Biophys. Acta 521, 606-618. 10. Liautard, J. & Egly, J. (1980) Nucleic Acids Res. 8, 1793-1804. 11. McMullen, M., Shaw, P. & Martin, T. (1979)J. Mol. Biol. 132, 679-694. 12. Rudensey, L. & Infante, A. (1979) Biochemistry 18, 3056-3063. 13. Georghegan, T., Cereghini, S. & Brawerman, G. (1979) Proc. Natl. Acad. Sci. USA 76, 5587-5591.
Proc. NatL Acad. Sci. USA 78 (1981)
87
14. Daniel, J. & Baldwin, H. (1964) Methods Cell Physiol. 1, 9-41. 15. Adams, D. & Jeffery, W (1978) Biochemistry 17, 4519-4524. 16. Mittermayer, C., Braun, R. & Rusch, H. (1965) Exp. Cell Res. 38, 33-41. 17. Ridgway, E. & Durham, A. (1976)J. Cell Biol. 69, 223-226. 18. Adams, D., Noonan, D. & Jeffery, W (1980) Anal. Biochem. 103, 408-412. 19. Adams, D., Noonan, D. & Jeffery, W. (1980) FEBS Lett. 114, 115-118. 20. Sauer, H., Babcock, K. & Rusch, H. (1969) Biochim. Biophys. Acta 195,410-421. 21. Diez, J; &. Brawerman, G. (1974) Proc. Natl. Acad. Sci. USA 71, 4091-4095. 22. Brawerman, G. & Diez, J. (1975) Cell 5, 271-280. 23. Sawicki, S., Jelinek, W & Darnell, J. (1977) J. Mol. Biol. 113, 219-235. 24. Hoerz, W. & McCarty, K. (1969) Proc. Natl. Acad. Sci. USA 63, 1206-1213. 25. Obrig, T., Irvin, J., Culp, W. & Hardesty, B. (1971) Eur. J. Biochem. 21, 31-41. 26. Siegel, M. &. Sisler, H. (1964) Biochim. Biophys. Acta 87, 70-82. 27. Lee, G. & Engelhardt, D. '(1979)J. Mol. Biol. 129, 221-233. 28. -Brawerman, G. (1973) Mol. Biol. Rep. 1, 7-13. 29. Sheiness, D. & Darnell, J. (1973) Nature (London) New Biol. 241, 265-268. 30. Kwan, S.-W. & Brawerman, G. (1972) Proc. Natl. Acad. Sci. USA 69, 3247-3250. 31. Blobel, G. (1973) Proc. Natl. Acad. Sci. USA 70,- 924-928. 32. Jeffery, W. & Brawerman, G. (1974).Biochemistry 13,4633-4637. 33. Barrieux, A., Ingraham, H., David, D. & Rosenfeld, M. (1975) Biochemistry 14, 1815-1821. 34. Kish, V. & Pederson, T. (1976)J. Biol. Chem. 251, 5888-5894. 35. Schwartz, H. & Darnell, J. (1976)J. Mol. Biol. 104, 833-851. 36. Jeffery, W. (1977)J. Biol. Chem. 252, 3525-3532. 37. Knochel, W. & Tiedemann, H. (1975) Biochim. Biophys. Acta 269, 104-117. 38. Dolecki, G., Duncan, R. & Humphreys, T. (1977) Cell 11, 339-344. 39. Cann, A., Grambino, R., Banks, J. & Bank, A. (1974)J. Biol. Chem. 249, 7536-7540. 40. Montagnier, L., Collandre, H., DeMaeyer-Guignard, J. & De41. 42. 43. 44. 45.
Maeyer, E. (1974) Biochem. Biophys. Res. Commun. 59, 1031-1038. Houdebine, L. (1976) FEBS Lett. 66, 110-113. Gedamu, L. & Dixon, G. (1976)J. Biol. Chem. 251, 1446-1454. Iatrou, K. & Dixon, G. (1977) Cell 10, 433-441. Gedamu, L., Iatrou, K. & Dixon, G. (1977) Cell 10, 441-451. Bergmann, I. & Brawerman, G. (1980)J. Mol. Biol. 139, 439-454.
