Proc. Nati. Acad. Sci. USA Vol. 74, No. 12, pp. 5397-5401, December 1977
Biochemistry
Distribution of rat liver albumin mRNA membrane-bound and free in polyribosomes as determined by molecular hybridization [albumin [3HJcDNA/albumin specific RNA sequences/poly(A)+ RNA vs. poly(A)- RNA]
S. H. YAP, ROGER K. STRAIR, AND DAVID A. SHAFRITZ Departments of Medicine and Cell Biology and The Liver Research Center, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461
Communicated by Alex B. Novikoff, September 19,1977
ABSTRACT Recently, we purified rat liver albumin mRNA and prepared albumin [3HJcDNA Using albumin [3HJcDNA in molecular hybridization experiments, we have now determined the distribution of albumin mRNA sequences in membranebound and free liver polyribosomes prepared by techniques in which there is high recovery of polyribosomes without evidence of degradation. By using molecular hybridization to measure specific mRNA sequence content, numerous problems could be avoided in interpretation of results as obtained by cell-free protein synthesis or immunological methods. Under these conditions, 98% of albumin mRNA sequences in polyribosomes are found in the membrane-bound fraction.
Polyribosomes of animal cells can be separated into two populations, those associated with the endoplasmic reticulum (membrane-bound polyribosomes) and those free in the cytoplasm (free polyribosomes). Evidence from numerous studies (1-11) suggests that secretory proteins are synthesized primarily on membrane-bound polyribosomes, whereas cytosol proteins are synthesized primarily on free polyribosomes. From these studies it has been concluded that mRNAs for secretory proteins are compartmentalized inside the cell. In certain cases, however, these distinctions have been found not to be absolute, because both secretory and intracellular proteins can be synthesized on membrane-bound polyribosomes (12-16). In previous studies, we have examined the synthesis of albumin (a secretory protein) and ferritin (an intracellular protein) in liver membrane-bound and free polyribosomes, as well as their respective RNA extracts. With native liver polyribosomes, 5- to 6-fold differences were found in the percentage synthesis of the complete polypeptide chains for these proteins in these two polyribosomal subpopulations (11, 17). These results agreed with those of other investigators (5, 7, 12) and were consistent with an interpretation that the mRNAs for these proteins are compartmentalized within the liver cell. However, when RNA extracts from the same preparations of membrane-bound and free liver polyribosomes were translated in a reticulocyte cell-free system, little difference was found in the synthesis of either albumin or ferritin (17). These results were not consistent with mRNA segregation and raised the possibility that nontranslated mRNAs for ferritin and albumin were present in membrane-bound and free liver polyribosomes, respectively. They further suggested that the concentration of a given mRNA in a mixed mRNA preparation may not be the sole factor in governing translation of that mRNA. More recently, it has become clear that various factors or conditions may influence in vitro translation of specific eukaryotic mRNAs (18). Therefore, we developed techniques that have allowed us to use molecular hybridization to deterThe 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. §1734 solely to indicate this fact.
mine albumin mRNA sequence content directly in any preparation of RNA without the need for measuring protein synthesis. With this technique, we reexamined the distribution of albumin mRNA in membrane-bound and free liver polyribosomes, using albumin [3H]cDNA prepared from purified albumin mRNA. Our experiments indicate that 98% of polyribosomal albumin mRNA sequences are associated with membrane-bound polyribosomes in the normal rat liver.
MATERIALS AND METHODS Isolation of Membrane-Bound and Free Liver Polyribosomes. Male Sprague-Dawley rats weighing 200-300 g were used throughout and were maintained on standard Purina Chow and water ad libitum until sacrifice. Animals were sacrificed between 0900 and 1000, and both membrane-bound and free liver polyribosomes were prepared according to the procedure of Ramsey and Steele (19). Briefly, livers were perfused with 250 mM sucrose/5 mM MgCl2, homogenized in 250 mM sucrose/10 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (Hepes), pH 7.4/75 mM KCl/5 mM MgCI2/3 mM glutathione, and centrifuged at 131,000 X g for 12 min to separate the larger particulate fraction from free polyribosomes and cell sap. The particulate fraction was resuspended in cell sap containing 250 mM KCI and rehomogenized; membrane-
bound polyribosomes were separated from nuclei and other particulate matter by centrifugation at 1470 X g for 5 min. A 1/h volume of 13% (wt/wt) deoxycholic acid was added to the post-nuclear supernatant fraction to solubilize membranebound polyribosomes. Both membrane-bound and free polyribosomes were purified by centrifugation through discontinuous sucrose gradients as described by Ramsey and Steele (19). With this technique, approximately 95% of total cytoplasmic ribosomal RNA is recovered in membrane-bound and free polyribosomes. Preparation of Cell Sap. One volume of homogenizing buffer was added to pooled livers that had been perfused as described above. After homogenization, the material was centrifuged at 17,000 X g for 10 min in a Sorvall RC5 centrifuge with the HB-4 rotor at 40. Pelleted material was discarded and the supernatant fraction was recentrifuged at 368,000 X g for 95 min in a Beckman 65 rotor at 40. The cell sap was harvested from the upper three-fourths of the supernatant fraction, excluding the lipid layer, and adjusted to 250 mM KCI for use as an RNase inhibitor. Polyribosome Profile Analysis. Sucrose gradient analysis of membrane-bound and free polyribosomes was performed by diluting -10 A20 units of polyribosomes in 400 JAI of homogenizing buffer without sucrose. This material was layered over a 12-ml 10-40% (wt/vol) exponential sucrose gradient Abbreviation: NaDodSO4, sodium dodecyl sulfate.
5397
Biochemistry: Yap etal.
5398
Proc. Natl. Acad; Sci. USA 74 (1977) 100
Z
a
3
.
*'m E > 40 .0 ,~~~%I
is,
20
/ #
2
0,
X N 60 if
El
C
U
80
0
J*/ %
/
0
c),o 1
2 RNA, Mig
3
4
FIG. 2. Hybridization analysis of crude RNA extracted from membrane-bound (0-0) and free (O --- 0) liver polyribosomes. In this experiment, the amounts of RNA were varied from 0.05 to 4Mg.
10
5
0
Bottom ml +--Top FIG. 1. Sucrose gradient analysis [1040% (wt/vol) exponential sucrose gradients] of rat liver membrane-bound (-) and free (- -- -) polyribosomes. After centrifugation at 38,000 rpm in a Beckman SW 41 rotor for 65 min at 2°, gradients were withdrawn from the bottom of each tube and absorbance at 254 nm was monitored with an Altex UV monitor 152.
containing the same buffer and 0.5 mM EDTA. Centrifugation was performed in a Beckman SW 41 rotor at 38,000 rpm for 65
min at 2°. Gradients were withdrawn from the bottom of each tube and absorbance at 254 nm was monitored with an Altex
UV monitor no. 152 (Altex Scientific, Berkeley, CA). Preparation of RNA Extracts and Poly(A)+ RNA from Free and Membrane-Biound Liver Polyribosomes. After separation of free and membrane-bound polyribosomes, RNA extraction was 'performed as previously reported (20). Poly(A)+ RNA was prepared from total RNA extracts of membrane-bound and free
polyribosomes by oligo(dT)-cellulose chromatography (20). RNA was resuspended in 10 mM Tris-Ha, pH 7.4/0.5% sodium dodecyl sulfate (NaDodSO4) and heated to 65° for 5 min. The RNA solution was then cooled rapidly by addition of an equal volume of I M NaCl (4°). This material was placed over a 3-ml packed oligo(dT)-cellulose column. The column was washed extensively with 0.5 M NaCI/10 mM Tris-HCI, pH 7.4/0.5% NaDodSO4 .and then with 0.1 M NaCl/ 10 mM Tris HCI, pH 7.4/0.5% NaDodSO4. Poly(A)+ RNA was eluted from the column with 10 mM Tris-HCI, pH 7.4/0.5% NaDodSO4. The eluted RNA was adjusted to 0.2 M in sodium acetate buffer (pH 5.5) and precipitated overnight by addition of 2 volumes of absolute ethanol at -20°. RNA was pelleted by centrifugation at 12,000 X g for 30 min and washed several times with 70% ethanol.- For hybridization analysis, RNA was dissolved in deionized distilled water and adjusted to appropriate concentrations as indicated in the figures. Preparation of Purified Albumin mRNA and Albumin
[3HlcDNA. Purification of albumin mRNA from total rat liver
polyribosomes was performed as reported (20). The major steps in this procedure include immunoprecipitation of polysomes containing albumin nascent chains, isolation of polyadenylylated RNA from these polyribosomes by phenol extraction and oligo(dT)-cellulose chromatography, and subsequent purification of albumin mRNA by controlled molecular hybridization with albumin-sequence-enriched cDNA-cellulose (20). The isolated albumin mRNA was then transcribed into albumin [3H~cDNA as described (20).
RNAcDNA Hybridization. Analytical RNA-cDNA hybridization was performed at 650 in 5-,ul sealed capillary tubes containing 0.2 M sodium phosphate buffer, pH 6.8/0.5% NaDodSO4 according to the method of Housman et al. (21). The amount of albumin [3H]cDNA used was approximately 400 cpm (specific activity 7.5 X 106 cpm/,gg). The amount of RNA in each assay is indicated in the figures. After 48 hr, samples were diluted 1:400 into 0.1 M sodium acetate buffer, pH 4.5/1 mM ZnSO4/denatured calf thymus DNA (10 Ag/ml). Hybrid formation was monitored by determining the percentage of input [3H]dCTP-labeled cDNA that was insoluble in 10% trichloroacetic acid after digestion with Si nuclease (125 units/ml) at 450 for 30 min. Trichloroacetic acid-insoluble material was collected on nitrocellulose filters. The filters were washed three times with 5% trichloroacetic acid, dissolved in 10 ml of Bray's solution (New England Nuclear), and assayed for radioactivity by liquid scintillation spectroscopy. RESULTS Isolation of Free and Membrane-Bound Polyribosomes. Because our goal was to determine quantitatively the amount of albumin mRNA in membrane-bound and free polyribosomes, it seemed particularly important to utilize isolation procedures that would provide high yields of undegraded polyribosomes. Recently, Ramsey and Steele (19) developed such methods and as shown in Table 1, we obtained excellent yields of polyribosomes with these techniques. Membranebound polyribosomes comprised 75-80% of total hepatic polyribosomes. Sucrose gradient centrifugation established that the polyribosomes were not degraded (Fig. 1). Large- and intermediate-sized polyribosomes were present in both membrane-bound and free polyribosomes, and more than 90% of ribosomal material sedimented in the polyribosome region of the gradient. Albumin mRNA Sequence Content in RNA Extracted from Membrane-Bound and Free Polyribosomes. RNA extracts prepared from membrane-bound and free polyribosomes were utilized for analysis of albumin mRNA sequence content by hybridization to albumin [3H]cDNA. In these experiments, varying amounts of RNA were titrated against a fixed amount of albumin [3H]cDNA. The amount of RNA required to protect 50% of albumin [3H]cDNA measures the concentration of albumin mRNA in that fraction. As shown in Fig. 2, the concehtration of albumin mRNA was approximately 16 times greater in membrane-bound polyribosomal RNA than in free polyribosomal RNA. Although this result would appear to reflect a specific segregation of albumin mRNA into membrane-bound polyribosomes, alternative explanations are possible.
Biochemistry: Yap etal. 80
n'a 20 .E.
P
I
15
IO RNA,
ng
FIG. 3. Hybridization analysis of poly(A)+ RNA prepared from 0) liver polyribosomes. In membrane-bound (0-0) and free (O this experiment, the amounts of RNA were varied from 0.156 to 12.5 ng. ---
In these experiments, albumin mRNA content is expressed function of total RNA in the sample. If the free polyribosomal RNA fraction contained considerably more total RNA relative to mRNA than did the membrane-bound RNA fraction, a high ratio could be obtained without specific segregation of albumin mRNA. Therefore, we prepared poly(A)+ RNA (a fraction enriched for mRNA) from the crude RNA extracts of as a
both membrane-bound and free polyribosomes and analyzed these fractions for albumin mRNA sequence content. As shown in Fig. 3, the ratio of albumin mRNA sequences in poly(A)+ membrane-bound RNA to those in poly(A)+ free polyribosomal RNA was still 16:1. From these results we conclude that albumin mRNA is specifically segregated into membrane-bound polyribosomes. Because the above results as well as those of many other studies of mRNA in eukaryotic cells have been based on the use of poly(A)+ RNA, we thought that it was important to determine exactly what percentage of total polyribosomal albumin mRNA was present in the poly(A)+ RNA. As shown in Fig 4, approximately 3 times as much RNA was required for 50% hybridization with poly(A)- compared to poly(A)+ RNA. Therefore, 75% of albumin mRNA sequences are polyadenylylated [retained on oligo(dT)-cellulose]. This finding suggests that most albumin mRNA in polyribosomes contains a poly(A) sequence. Whether the 25% of albumin mRNA found in the 100 0
80
Z
5399
.~60
300 5 215+5
11.5 0.5 10±2
5.9 3.1
1.2 1.0
4.9 3.1
* Mean + SEM.
poly(A)- RNA fraction represents degradation during the isolation procedure has not been determined. On the basis of yield data in Table 1 and hybridization results in Figs. 2 and 3, it can be calculated that approximately 98% of albumin mRNA in total liver polyribosomes is associated with the membrane-bound fraction (Table 2). From our previous study using purified albumin mRNA and albumin [3H]cDNA (20), we compute that 1 pg of purified albumin mRNA protects 5 cpm of albumin [3H]cDNA. Using this same cDNA probe, we can determine the absolute amount of albumin mRNA in any RNA fraction by directly measuring cpm of albumin [3H]cDNA protected per unit of RNA. In Fig. 2, 0.1 ,ug of membrane-bound polyribosomal RNA protected 68 cpm of albumin [3HJcDNA. Therefore, the content of albumin mRNA in this fractiornis 137 pg per gg RNA. Free polyribosomal RNA protected '/16th this amount of albumin [3HJcDNA for any given amount of RNA (Fig. 2). Therefore, this fraction contains 8.5 pg of albumin mRNA per ,ug of RNA. Because we also determined the total recovery of membrane-bound and free polyribosomes (Table 1), we can calculate the total amount of albumin mRNA in these fractions per g of liver as shown in Table 2. DISCUSSION Previous studies on the distribution of specific mRNAs between membrane-bound and free polyribosomes have been based primarily on measurements of in vitro protein synthesis with identification and quantitation of specific protein products (4, 17, 22, 23). Other studies have utilized immunological procedures to identify polyribosome fractions containing nascent chains of a protein against which a specific antibody has been
N
,00
-
E> .0
20
Table 2. Distribution of albumin mRNA sequences in rat liver membrane-bound and free polyribosomes Liver
polyribosome 132
0
116
1:8 RI NA dilution
0) and FIG. 4. Hybridization analysis of liver poly(A)+ (0 poly(A)- (O 0) polyribosomal RNA. Ten A20 units of total liver polyribosomal RNA was resuspended in 1 ml of 10 mM Tris-HCl, pH 7.4/0.5% NaDodSO4 and heated to 650 for 5 min. The RNA solution was then cooled rapidly and 140 gl of 4 M NaCl (40) was added. This material was passed over a 2-nl packed oligo(dT)-cellulose column. Unadsorbed RNA was washed through the column with 9 ml of 0.5 M NaCl/0.5% NaDodSO4/10 mM Tris-HCl, pH 7.4. Adsorbed RNA was then eluted with 10 ml of 0.5% NaDodSO4/10 mM Tris.HCl, pH 7.4. Adsorbed and unadsorbed RNA fractions were then ethanol precipitated and dissolved in 40 Ml of deionized distilled water. Aliquots (2 pl) of these fractions at various dilutions were hybridized to albumin [3H]cDNA (-400 cpm). -
-
1. Distribution of polyribosomal RNA
fraction Membranebound Free 1 3.6
(ratio) 2. Albumin mRNA sequence content
(ratio) 3. Relative amount of albumin mRNA sequences (line 1 X line 2) (ratio) 4. Relative amount of albumin mRNA sequences as % of albumin mRNA in total polyribosomes 5. Total content of albumin mRNA sequences (ng/g of liver)
16
1
57.6
1
98
2
564
9.5
5400
Biochemistry: Yap etal.
directed (6, 7, 12, 23, 24). Both of these techniques are indirect, and various investigators have reached different conclusions regarding the extent of albumin mRNA segregation (4-12, 17, 22-24). Recently, it has been found that factors other than specific mRNA levels may influence the amount (either relative or absolute) of synthesis of a specific protein in a cell-free system. Such variations may depend on (i) the nature of the cell-free system (18), (ii) the procedure used to quantitate the specific protein product (11, 17), (iii) the concentrations of monovalent and divalent cations (25-27), (iv) the presence of hemin, polyamines, or other activators or inhibitors of protein synthesis (28-35), and (v) the use of purified mixed preparations of mRNA (36-40). With liver mRNA, Tse and Taylor (41) found a 6-fold variation in albumin synthesis in a wheat germ system, depending on K+ AND Mg2+ concentrations. Peterson (42) has also reported differential translational requirements for albumin mRNA compared to total liver mRNA in this cell-free system. In addition, measurement of the relative amount of a specific mRNA by cell-free translation requires functional integrity of the mRNA. Therefore, only a slight modification or degradation of the mRNA may result in differential loss of biological activity. Because most preparations of membrane-bound polyribosomes contain significant amounts of ribonuclease activity (43,44), this may lead to an underestimation of specific mRNA in the cell-free protein synthesis method. Other factors that may influence albumin mRNA distribution include the animal species, the method used for polyribosome isolation, and the nutritional status of the animal (S. H. Yap, R. K. Strair, and D. A. Shafritz, unpublished data). In the present study, we utilized albumin [3H]cDNA to measure albumin mRNA sequence content directly. This assay does not require either in vitro protein synthesis or laborious procedures to identify and quantitate the cell-free product. The results indicate that albumin mRNA is 16-fold enriched in membrane-bound polyribosomes and that this is due to a specific segregation of albumin mRNA. In order to quantitate the amount of albumin mRNA in total liver membrane-bound and free polyribosomes, it was important to use an isolation procedure giving high yields of undegraded polyribosomes. Methods for separation of membrane-bound and free polyribosomes from post-mitochondrial supernatant have been criticized because the yield of membrane-bound polyribosomes may be quite low and significant degradation of membrane-bound polyribosomes may occur (11, 19, 23, 43). Although we could not obtain quantitative results by these earlier techniques, the ratio of albumin mRNA sequences in membrane-bound to free polyribosomes was still -16:1 (unpublished data). In the present study using the isolation technique described by Ramsey and Steele (19), we obtained an excellent recovery of undegraded polyribosomes. Calculations indicate that practically all of the albumin mRNA sequences in rat liver polyribosomes are contained in the membrane-bound fraction. However, free cytoplasmic polyribosomes do contain a low level of albumin mRNA. It is unclear whether this small amount of albumin mRNA in free polyribosomes is due to crosscontamination or represents a fraction of albumin-synthesizing polyribosomes prior to their attachment to endoplasmic reticulum membranes. Because the bulk of albumin mRNA is polyadenylylated and the ratio of albumin mRNA sequences in membrane-bound to those in free polyribosomes is the same for poly(A)+ and total RNA fractions, it would appear that the poly(A) sequence is not a critical feature for association of albumin mRNA with the membrane-bound fraction.
Proc. Natl. Acad. Sci. USA 74 (1977)
The above findings are consistent with a model for synthesis of secretory protein recently proposed by Blobel and coworkers (45). According to this model, protein synthesis with newly synthesized mRNA begins on free polyribosomes (initially as monomers, then dimers, then trimers, etc.). As protein synthesis proceeds, the NH2-terminal sequence of the growing nascent chain becomes exposed through the surface of the 60S subunit. This sequence is hydrophobic for all secretory proteins and serves as a signal for attachment of polyribosomes synthesizing secretory proteins to the rough endoplasmic reticulum. Polypeptide synthesis continues vectorially through the membrane, the NH2-terminal sequence ("signal sequence") is cleaved by an enzyme ("clippase") contained within the rough endoplasmic reticulum, and the final product is extruded into the intracisternal space. Therefore, depending on the rate of entry of secretory mRNAs into the cytoplasm, the rate of initiation and elongation of secretory proteins, and the number of hydrophobio sites available on the rough endoplasmic reticulum, a certain proportion of ribosomes containing mRNA for secretory proteins will be found in the free cytoplasmic fraction. However, once mRNA becomes associated with the membrane-bound fraction, in all probability it is not released again unless all nascent chains are completed or released and protein synthesis reinitiation is inhibited. Recently, several investigators have reported evidence consistent with such a model (46-50), and Strauss et al. (51) have reported that rat albumin synthesized in a wheat germ cell-free system contains an extra hydrophobic NH2-terminal sequence of 24 amino acids. In a previous study, we found that RNA extracted from membrane-bound and free polyribosomes of rabbit liver showed little difference in albumin synthesis when translated in a messenger-dependent reticulocyte cell-free system (17). We suggested that these results were consistent with an interpretation that free polyribosome preparations contained nontranslated albumin mRNA, perhaps as messenger ribonucleoprotein particles. Zahringer et al. (23) suggested another possibility-namely, that the membrane-bound polyribosomes used in the study were dissociated during the isolation procedure and thus contaminated the free polyribosome fraction. Because we had observed a 5- to 6-fold difference in albumin synthesis between membrane-bound and free polyribosomes in the native or intact liver system, this interpretation was unlikely. Based on our present study, the simplest explanation for differences in albumin mRNA content as determined by protein synthesis and by molecular hybridization analysis is unequal translation of albumin mRNA in the various fractions. Although there is no evidence that such variations occur in vivo, the present study clearly demonstrates that 98% of albumin mRNA sequences in rat liver polyribosomes are contained in the membrane-bound fraction. This research was supported in part by National Institutes of Health Grants AM-17609 and AM-17702, Cell and Molecular Biology Training Grant 5-TO1 GM 02209, Medical Scientist Program Grant 5T32-GM 7288, theNetherlands Organization for Advancement of Pure Research (Zuiver Wettenschappelijk Orderzock) and Niels Stensen Stichting to S.H.Y., a National Institutes of Health Research Career Development Award to D.A.S., and the Irma T. Hirschl Charitable Trust of New York. 1. Siekevitz, P. & Palade, G. (1960) J. Biophys. Biochem. Cytol. 7, 619-630. 2. Birbeck, M. S. C. & Mercer, E. H. (1961), Nature 189, 558560. 3. Peters, T., Jr. (1962) J. Biol. Chem. 237, 1186-1189. 4. Hicks, S. J., Drysdale, J. W. & Munro, H. N. (1968) Science 164, 584-585.
Biochemistry: Yap etal. 5. Takagi, M. & Ogata, K. (1968) Biochem. Biophys. Res. Commun. 33,55-60. 6. Redman, C. M. (1969) 1. Biol. Chem. 244, 4308-4315; 7. Takagi, M., Tanaka, T. & Ogata, K. (1970) Biochim. Biophys. Acta 217, 148-158. 8. Andrews, T. M. & Tata, J. R. (1971) Biochem. J. 121, 683694. 9. Uenoyama, K. & Ono, T. (1972) Biochem. Biophys. Res. Commun. 49,713-719. 10. Rolleston, F. S. (1974) Sub-Cell Biochem. 3, 91-117. 11. Shafritz, D. A. (1974) J. Biol. Chem. 249, 81-88. 12. Ikehara, Y. & Pitot, H. C. (1973) J. Cell Biol. 59, 28-44. 13. Ragnotti, G., Lawford, G. R. & Campbell, P. N. (1969) Biochem. J. 112, 139-147. 14. Kadenbach, B. (1970) Eur. J. Biochem. 12,392-398. 15. Sakamoto, T. & Higashi, T. (1973) J. Biochem. (Tokyo) 73, 1083-1088. 16. Bergeron, J. J. M., Berridge, M. V. & Evans, W. H. (1975) Biochim. Biophys. Acta 407,325-337. 17. Shafritz, D. A. (1974) J. Biol. Chem. 249,89-93. 18. Shafritz, D. A. (1977) "Molecular Mechanisms of Protein Biosynthesis," in Molecular Biology Series, eds. Weissback, H. & Pestka, S. (Academic Press, New York), pp. 555-601. 19. Ramsey, J. C. & Steele, W. J. (1976) Biochemistry 15, 17041712. 20. Strair, R. K., Yap, S. H. & Shafritz, D. A. (1977) Proc. Natl. Acad. Sci. USA 74, 4346-4350. 21. Housman, D., Skoultchi, A., Forget, B. G. & Benz, E. J. (1974) Ann. N.Y. Acad. Sci. 241, 280-289. 22. Shore, G. C. & Tata, J. R. (1977) J. Cell Biol. 72, 726-743. 23. Zahringer, J., Baliga, B. S., Drake, R. L. & Munro, H. N. (1977) Biochim. Biophys. Acta 474,234-244. 24. Takagi, M. & Ogata, K. (1971) Biochem. Biophys. Res. Commun. 42, 125-131. 25. Rourke, A. W. & Heywood, S. M. (1972) Biochemistry 11, 2061-2066. 26. Mathews, M. B., Pragnell, I. M., Osborn, M. & Arnstein, H. R. V. (1972) Biochim. Biophys. Acta 287, 113-123. 27. Roberts, B. E. & Paterson, B. M. (1973) Proc. Natl. Acad. Sci. USA
70,2330-2334. 28. Zucker, W. V. & Schulman, H. M. (1968) Proc. Natl. Acad. Sci. USA 59, 582-589.
Proc. Natl. Acad. Sci. USA 74 (1977)
5401
29. Adamson, S. D., Yau, P. M. P., Herbert E. & Zucker W. V. (1972) J. Mol. Biol. 63,247-264. 30. Gross, M. & Rabinowits, M. (1972) Proc. Natl. Acad. Sci. USA 69,1565-1568. 31. Legon, S., Jackson, R. J. & Hunt, T. (1973) Nature New Biol. 241, 150-152. 32. Beuzard, Y. & London, I. M. (1974) Proc. Natl. Acad. Scd. USA 71,2863-2866. 33. Atkins, J. F., Lewis, J. B., Anderson, C. W. & Gesteland, R. R. (1975) J. Biol. Chem. 250, 5688-5695. 34. Ehrenfeld, E. & Hunt, T. (1971) Proc. Natl. Acad. Sci. USA 68, 1075-1078. 35. Heywood, S. M., Kennedy, D. S. & Bester, A. J. (1974) Proc. Natl. Acad. Sct. USA 71,2428-2431. 36. Lingrel, J. B., Lockhard, R. E., Jones, R. F., Burr, H. E. & Holder, J. W. (1971) Ser. Haematol. 4,37-69. 37. Nienhuis, A. W., Canfield, P. H. & Anderson, W. F. (1973) J. Clin. Invest. 52,1735-1745. 38. McKeehan, W. L. (1974) J. Biol. Chem. 249,6517-6526. 39. Palmiter, R. D. (1974) J. Biol. Chem. 249,6779-6787. 40. Sonenshein, G. E. & Brawerman, G. (1976) Biochemistry 15, 5501-5506. 41. Tse, T. P. H. & Taylor, J. M. (1977) J. Biol. Chem. 252, 12721278. 42. Peterson, J. A. (1976) Nucleic Acids Res. 3, 1427-1436. 43. Blobel, G. & Potter, V. R. (1966) Proc. Natl. Acads Sci. USA 55, 1283-1288. 44. Roth, J. S. (1967) Methods Cancer Res. 3, 153-155. 45. Blobel, G. & Dobberstein, B. (1975) J. Cell Biol. 67,835-851. 46. Milstein, C., Brownlee, C. G., Harrison, T. M. & Mathews, M. B. (i972) Nature New Biol. 239, 117-120. 47. Burstein, Y. & Schechter, I. (1977) Proc. Natl. Acad. Sci. USA 74,716-720. 48. Kemper, B., Habener, J. F., Ernst, M. D., Potts, J. T. & Rich, A. (1976) Biochemistry 15, 15-19. 49. Wirth, D. F., Katz, F., Small, B. & Lodish, F. (1977) Cell 10, 253-263. 50. Grubman, M. J., Weinstein, J. A. & Shafritz, D. A. (1977) J. Ce!! Biol. 74, 43-57. 51. Strauss, A. W., Donohue, A. M., Bennett, C. D., Rodkey, J. A. & Alberts, A. W. (1977) Proc. Natl. Acad. Sci. USA 74, 13581362.
Biochemistry
Distribution of rat liver albumin mRNA membrane-bound and free in polyribosomes as determined by molecular hybridization [albumin [3HJcDNA/albumin specific RNA sequences/poly(A)+ RNA vs. poly(A)- RNA]
S. H. YAP, ROGER K. STRAIR, AND DAVID A. SHAFRITZ Departments of Medicine and Cell Biology and The Liver Research Center, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461
Communicated by Alex B. Novikoff, September 19,1977
ABSTRACT Recently, we purified rat liver albumin mRNA and prepared albumin [3HJcDNA Using albumin [3HJcDNA in molecular hybridization experiments, we have now determined the distribution of albumin mRNA sequences in membranebound and free liver polyribosomes prepared by techniques in which there is high recovery of polyribosomes without evidence of degradation. By using molecular hybridization to measure specific mRNA sequence content, numerous problems could be avoided in interpretation of results as obtained by cell-free protein synthesis or immunological methods. Under these conditions, 98% of albumin mRNA sequences in polyribosomes are found in the membrane-bound fraction.
Polyribosomes of animal cells can be separated into two populations, those associated with the endoplasmic reticulum (membrane-bound polyribosomes) and those free in the cytoplasm (free polyribosomes). Evidence from numerous studies (1-11) suggests that secretory proteins are synthesized primarily on membrane-bound polyribosomes, whereas cytosol proteins are synthesized primarily on free polyribosomes. From these studies it has been concluded that mRNAs for secretory proteins are compartmentalized inside the cell. In certain cases, however, these distinctions have been found not to be absolute, because both secretory and intracellular proteins can be synthesized on membrane-bound polyribosomes (12-16). In previous studies, we have examined the synthesis of albumin (a secretory protein) and ferritin (an intracellular protein) in liver membrane-bound and free polyribosomes, as well as their respective RNA extracts. With native liver polyribosomes, 5- to 6-fold differences were found in the percentage synthesis of the complete polypeptide chains for these proteins in these two polyribosomal subpopulations (11, 17). These results agreed with those of other investigators (5, 7, 12) and were consistent with an interpretation that the mRNAs for these proteins are compartmentalized within the liver cell. However, when RNA extracts from the same preparations of membrane-bound and free liver polyribosomes were translated in a reticulocyte cell-free system, little difference was found in the synthesis of either albumin or ferritin (17). These results were not consistent with mRNA segregation and raised the possibility that nontranslated mRNAs for ferritin and albumin were present in membrane-bound and free liver polyribosomes, respectively. They further suggested that the concentration of a given mRNA in a mixed mRNA preparation may not be the sole factor in governing translation of that mRNA. More recently, it has become clear that various factors or conditions may influence in vitro translation of specific eukaryotic mRNAs (18). Therefore, we developed techniques that have allowed us to use molecular hybridization to deterThe 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. §1734 solely to indicate this fact.
mine albumin mRNA sequence content directly in any preparation of RNA without the need for measuring protein synthesis. With this technique, we reexamined the distribution of albumin mRNA in membrane-bound and free liver polyribosomes, using albumin [3H]cDNA prepared from purified albumin mRNA. Our experiments indicate that 98% of polyribosomal albumin mRNA sequences are associated with membrane-bound polyribosomes in the normal rat liver.
MATERIALS AND METHODS Isolation of Membrane-Bound and Free Liver Polyribosomes. Male Sprague-Dawley rats weighing 200-300 g were used throughout and were maintained on standard Purina Chow and water ad libitum until sacrifice. Animals were sacrificed between 0900 and 1000, and both membrane-bound and free liver polyribosomes were prepared according to the procedure of Ramsey and Steele (19). Briefly, livers were perfused with 250 mM sucrose/5 mM MgCl2, homogenized in 250 mM sucrose/10 mM N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (Hepes), pH 7.4/75 mM KCl/5 mM MgCI2/3 mM glutathione, and centrifuged at 131,000 X g for 12 min to separate the larger particulate fraction from free polyribosomes and cell sap. The particulate fraction was resuspended in cell sap containing 250 mM KCI and rehomogenized; membrane-
bound polyribosomes were separated from nuclei and other particulate matter by centrifugation at 1470 X g for 5 min. A 1/h volume of 13% (wt/wt) deoxycholic acid was added to the post-nuclear supernatant fraction to solubilize membranebound polyribosomes. Both membrane-bound and free polyribosomes were purified by centrifugation through discontinuous sucrose gradients as described by Ramsey and Steele (19). With this technique, approximately 95% of total cytoplasmic ribosomal RNA is recovered in membrane-bound and free polyribosomes. Preparation of Cell Sap. One volume of homogenizing buffer was added to pooled livers that had been perfused as described above. After homogenization, the material was centrifuged at 17,000 X g for 10 min in a Sorvall RC5 centrifuge with the HB-4 rotor at 40. Pelleted material was discarded and the supernatant fraction was recentrifuged at 368,000 X g for 95 min in a Beckman 65 rotor at 40. The cell sap was harvested from the upper three-fourths of the supernatant fraction, excluding the lipid layer, and adjusted to 250 mM KCI for use as an RNase inhibitor. Polyribosome Profile Analysis. Sucrose gradient analysis of membrane-bound and free polyribosomes was performed by diluting -10 A20 units of polyribosomes in 400 JAI of homogenizing buffer without sucrose. This material was layered over a 12-ml 10-40% (wt/vol) exponential sucrose gradient Abbreviation: NaDodSO4, sodium dodecyl sulfate.
5397
Biochemistry: Yap etal.
5398
Proc. Natl. Acad; Sci. USA 74 (1977) 100
Z
a
3
.
*'m E > 40 .0 ,~~~%I
is,
20
/ #
2
0,
X N 60 if
El
C
U
80
0
J*/ %
/
0
c),o 1
2 RNA, Mig
3
4
FIG. 2. Hybridization analysis of crude RNA extracted from membrane-bound (0-0) and free (O --- 0) liver polyribosomes. In this experiment, the amounts of RNA were varied from 0.05 to 4Mg.
10
5
0
Bottom ml +--Top FIG. 1. Sucrose gradient analysis [1040% (wt/vol) exponential sucrose gradients] of rat liver membrane-bound (-) and free (- -- -) polyribosomes. After centrifugation at 38,000 rpm in a Beckman SW 41 rotor for 65 min at 2°, gradients were withdrawn from the bottom of each tube and absorbance at 254 nm was monitored with an Altex UV monitor 152.
containing the same buffer and 0.5 mM EDTA. Centrifugation was performed in a Beckman SW 41 rotor at 38,000 rpm for 65
min at 2°. Gradients were withdrawn from the bottom of each tube and absorbance at 254 nm was monitored with an Altex
UV monitor no. 152 (Altex Scientific, Berkeley, CA). Preparation of RNA Extracts and Poly(A)+ RNA from Free and Membrane-Biound Liver Polyribosomes. After separation of free and membrane-bound polyribosomes, RNA extraction was 'performed as previously reported (20). Poly(A)+ RNA was prepared from total RNA extracts of membrane-bound and free
polyribosomes by oligo(dT)-cellulose chromatography (20). RNA was resuspended in 10 mM Tris-Ha, pH 7.4/0.5% sodium dodecyl sulfate (NaDodSO4) and heated to 65° for 5 min. The RNA solution was then cooled rapidly by addition of an equal volume of I M NaCl (4°). This material was placed over a 3-ml packed oligo(dT)-cellulose column. The column was washed extensively with 0.5 M NaCI/10 mM Tris-HCI, pH 7.4/0.5% NaDodSO4 .and then with 0.1 M NaCl/ 10 mM Tris HCI, pH 7.4/0.5% NaDodSO4. Poly(A)+ RNA was eluted from the column with 10 mM Tris-HCI, pH 7.4/0.5% NaDodSO4. The eluted RNA was adjusted to 0.2 M in sodium acetate buffer (pH 5.5) and precipitated overnight by addition of 2 volumes of absolute ethanol at -20°. RNA was pelleted by centrifugation at 12,000 X g for 30 min and washed several times with 70% ethanol.- For hybridization analysis, RNA was dissolved in deionized distilled water and adjusted to appropriate concentrations as indicated in the figures. Preparation of Purified Albumin mRNA and Albumin
[3HlcDNA. Purification of albumin mRNA from total rat liver
polyribosomes was performed as reported (20). The major steps in this procedure include immunoprecipitation of polysomes containing albumin nascent chains, isolation of polyadenylylated RNA from these polyribosomes by phenol extraction and oligo(dT)-cellulose chromatography, and subsequent purification of albumin mRNA by controlled molecular hybridization with albumin-sequence-enriched cDNA-cellulose (20). The isolated albumin mRNA was then transcribed into albumin [3H~cDNA as described (20).
RNAcDNA Hybridization. Analytical RNA-cDNA hybridization was performed at 650 in 5-,ul sealed capillary tubes containing 0.2 M sodium phosphate buffer, pH 6.8/0.5% NaDodSO4 according to the method of Housman et al. (21). The amount of albumin [3H]cDNA used was approximately 400 cpm (specific activity 7.5 X 106 cpm/,gg). The amount of RNA in each assay is indicated in the figures. After 48 hr, samples were diluted 1:400 into 0.1 M sodium acetate buffer, pH 4.5/1 mM ZnSO4/denatured calf thymus DNA (10 Ag/ml). Hybrid formation was monitored by determining the percentage of input [3H]dCTP-labeled cDNA that was insoluble in 10% trichloroacetic acid after digestion with Si nuclease (125 units/ml) at 450 for 30 min. Trichloroacetic acid-insoluble material was collected on nitrocellulose filters. The filters were washed three times with 5% trichloroacetic acid, dissolved in 10 ml of Bray's solution (New England Nuclear), and assayed for radioactivity by liquid scintillation spectroscopy. RESULTS Isolation of Free and Membrane-Bound Polyribosomes. Because our goal was to determine quantitatively the amount of albumin mRNA in membrane-bound and free polyribosomes, it seemed particularly important to utilize isolation procedures that would provide high yields of undegraded polyribosomes. Recently, Ramsey and Steele (19) developed such methods and as shown in Table 1, we obtained excellent yields of polyribosomes with these techniques. Membranebound polyribosomes comprised 75-80% of total hepatic polyribosomes. Sucrose gradient centrifugation established that the polyribosomes were not degraded (Fig. 1). Large- and intermediate-sized polyribosomes were present in both membrane-bound and free polyribosomes, and more than 90% of ribosomal material sedimented in the polyribosome region of the gradient. Albumin mRNA Sequence Content in RNA Extracted from Membrane-Bound and Free Polyribosomes. RNA extracts prepared from membrane-bound and free polyribosomes were utilized for analysis of albumin mRNA sequence content by hybridization to albumin [3H]cDNA. In these experiments, varying amounts of RNA were titrated against a fixed amount of albumin [3H]cDNA. The amount of RNA required to protect 50% of albumin [3H]cDNA measures the concentration of albumin mRNA in that fraction. As shown in Fig. 2, the concehtration of albumin mRNA was approximately 16 times greater in membrane-bound polyribosomal RNA than in free polyribosomal RNA. Although this result would appear to reflect a specific segregation of albumin mRNA into membrane-bound polyribosomes, alternative explanations are possible.
Biochemistry: Yap etal. 80
n'a 20 .E.
P
I
15
IO RNA,
ng
FIG. 3. Hybridization analysis of poly(A)+ RNA prepared from 0) liver polyribosomes. In membrane-bound (0-0) and free (O this experiment, the amounts of RNA were varied from 0.156 to 12.5 ng. ---
In these experiments, albumin mRNA content is expressed function of total RNA in the sample. If the free polyribosomal RNA fraction contained considerably more total RNA relative to mRNA than did the membrane-bound RNA fraction, a high ratio could be obtained without specific segregation of albumin mRNA. Therefore, we prepared poly(A)+ RNA (a fraction enriched for mRNA) from the crude RNA extracts of as a
both membrane-bound and free polyribosomes and analyzed these fractions for albumin mRNA sequence content. As shown in Fig. 3, the ratio of albumin mRNA sequences in poly(A)+ membrane-bound RNA to those in poly(A)+ free polyribosomal RNA was still 16:1. From these results we conclude that albumin mRNA is specifically segregated into membrane-bound polyribosomes. Because the above results as well as those of many other studies of mRNA in eukaryotic cells have been based on the use of poly(A)+ RNA, we thought that it was important to determine exactly what percentage of total polyribosomal albumin mRNA was present in the poly(A)+ RNA. As shown in Fig 4, approximately 3 times as much RNA was required for 50% hybridization with poly(A)- compared to poly(A)+ RNA. Therefore, 75% of albumin mRNA sequences are polyadenylylated [retained on oligo(dT)-cellulose]. This finding suggests that most albumin mRNA in polyribosomes contains a poly(A) sequence. Whether the 25% of albumin mRNA found in the 100 0
80
Z
5399
.~60
300 5 215+5
11.5 0.5 10±2
5.9 3.1
1.2 1.0
4.9 3.1
* Mean + SEM.
poly(A)- RNA fraction represents degradation during the isolation procedure has not been determined. On the basis of yield data in Table 1 and hybridization results in Figs. 2 and 3, it can be calculated that approximately 98% of albumin mRNA in total liver polyribosomes is associated with the membrane-bound fraction (Table 2). From our previous study using purified albumin mRNA and albumin [3H]cDNA (20), we compute that 1 pg of purified albumin mRNA protects 5 cpm of albumin [3H]cDNA. Using this same cDNA probe, we can determine the absolute amount of albumin mRNA in any RNA fraction by directly measuring cpm of albumin [3H]cDNA protected per unit of RNA. In Fig. 2, 0.1 ,ug of membrane-bound polyribosomal RNA protected 68 cpm of albumin [3HJcDNA. Therefore, the content of albumin mRNA in this fractiornis 137 pg per gg RNA. Free polyribosomal RNA protected '/16th this amount of albumin [3HJcDNA for any given amount of RNA (Fig. 2). Therefore, this fraction contains 8.5 pg of albumin mRNA per ,ug of RNA. Because we also determined the total recovery of membrane-bound and free polyribosomes (Table 1), we can calculate the total amount of albumin mRNA in these fractions per g of liver as shown in Table 2. DISCUSSION Previous studies on the distribution of specific mRNAs between membrane-bound and free polyribosomes have been based primarily on measurements of in vitro protein synthesis with identification and quantitation of specific protein products (4, 17, 22, 23). Other studies have utilized immunological procedures to identify polyribosome fractions containing nascent chains of a protein against which a specific antibody has been
N
,00
-
E> .0
20
Table 2. Distribution of albumin mRNA sequences in rat liver membrane-bound and free polyribosomes Liver
polyribosome 132
0
116
1:8 RI NA dilution
0) and FIG. 4. Hybridization analysis of liver poly(A)+ (0 poly(A)- (O 0) polyribosomal RNA. Ten A20 units of total liver polyribosomal RNA was resuspended in 1 ml of 10 mM Tris-HCl, pH 7.4/0.5% NaDodSO4 and heated to 650 for 5 min. The RNA solution was then cooled rapidly and 140 gl of 4 M NaCl (40) was added. This material was passed over a 2-nl packed oligo(dT)-cellulose column. Unadsorbed RNA was washed through the column with 9 ml of 0.5 M NaCl/0.5% NaDodSO4/10 mM Tris-HCl, pH 7.4. Adsorbed RNA was then eluted with 10 ml of 0.5% NaDodSO4/10 mM Tris.HCl, pH 7.4. Adsorbed and unadsorbed RNA fractions were then ethanol precipitated and dissolved in 40 Ml of deionized distilled water. Aliquots (2 pl) of these fractions at various dilutions were hybridized to albumin [3H]cDNA (-400 cpm). -
-
1. Distribution of polyribosomal RNA
fraction Membranebound Free 1 3.6
(ratio) 2. Albumin mRNA sequence content
(ratio) 3. Relative amount of albumin mRNA sequences (line 1 X line 2) (ratio) 4. Relative amount of albumin mRNA sequences as % of albumin mRNA in total polyribosomes 5. Total content of albumin mRNA sequences (ng/g of liver)
16
1
57.6
1
98
2
564
9.5
5400
Biochemistry: Yap etal.
directed (6, 7, 12, 23, 24). Both of these techniques are indirect, and various investigators have reached different conclusions regarding the extent of albumin mRNA segregation (4-12, 17, 22-24). Recently, it has been found that factors other than specific mRNA levels may influence the amount (either relative or absolute) of synthesis of a specific protein in a cell-free system. Such variations may depend on (i) the nature of the cell-free system (18), (ii) the procedure used to quantitate the specific protein product (11, 17), (iii) the concentrations of monovalent and divalent cations (25-27), (iv) the presence of hemin, polyamines, or other activators or inhibitors of protein synthesis (28-35), and (v) the use of purified mixed preparations of mRNA (36-40). With liver mRNA, Tse and Taylor (41) found a 6-fold variation in albumin synthesis in a wheat germ system, depending on K+ AND Mg2+ concentrations. Peterson (42) has also reported differential translational requirements for albumin mRNA compared to total liver mRNA in this cell-free system. In addition, measurement of the relative amount of a specific mRNA by cell-free translation requires functional integrity of the mRNA. Therefore, only a slight modification or degradation of the mRNA may result in differential loss of biological activity. Because most preparations of membrane-bound polyribosomes contain significant amounts of ribonuclease activity (43,44), this may lead to an underestimation of specific mRNA in the cell-free protein synthesis method. Other factors that may influence albumin mRNA distribution include the animal species, the method used for polyribosome isolation, and the nutritional status of the animal (S. H. Yap, R. K. Strair, and D. A. Shafritz, unpublished data). In the present study, we utilized albumin [3H]cDNA to measure albumin mRNA sequence content directly. This assay does not require either in vitro protein synthesis or laborious procedures to identify and quantitate the cell-free product. The results indicate that albumin mRNA is 16-fold enriched in membrane-bound polyribosomes and that this is due to a specific segregation of albumin mRNA. In order to quantitate the amount of albumin mRNA in total liver membrane-bound and free polyribosomes, it was important to use an isolation procedure giving high yields of undegraded polyribosomes. Methods for separation of membrane-bound and free polyribosomes from post-mitochondrial supernatant have been criticized because the yield of membrane-bound polyribosomes may be quite low and significant degradation of membrane-bound polyribosomes may occur (11, 19, 23, 43). Although we could not obtain quantitative results by these earlier techniques, the ratio of albumin mRNA sequences in membrane-bound to free polyribosomes was still -16:1 (unpublished data). In the present study using the isolation technique described by Ramsey and Steele (19), we obtained an excellent recovery of undegraded polyribosomes. Calculations indicate that practically all of the albumin mRNA sequences in rat liver polyribosomes are contained in the membrane-bound fraction. However, free cytoplasmic polyribosomes do contain a low level of albumin mRNA. It is unclear whether this small amount of albumin mRNA in free polyribosomes is due to crosscontamination or represents a fraction of albumin-synthesizing polyribosomes prior to their attachment to endoplasmic reticulum membranes. Because the bulk of albumin mRNA is polyadenylylated and the ratio of albumin mRNA sequences in membrane-bound to those in free polyribosomes is the same for poly(A)+ and total RNA fractions, it would appear that the poly(A) sequence is not a critical feature for association of albumin mRNA with the membrane-bound fraction.
Proc. Natl. Acad. Sci. USA 74 (1977)
The above findings are consistent with a model for synthesis of secretory protein recently proposed by Blobel and coworkers (45). According to this model, protein synthesis with newly synthesized mRNA begins on free polyribosomes (initially as monomers, then dimers, then trimers, etc.). As protein synthesis proceeds, the NH2-terminal sequence of the growing nascent chain becomes exposed through the surface of the 60S subunit. This sequence is hydrophobic for all secretory proteins and serves as a signal for attachment of polyribosomes synthesizing secretory proteins to the rough endoplasmic reticulum. Polypeptide synthesis continues vectorially through the membrane, the NH2-terminal sequence ("signal sequence") is cleaved by an enzyme ("clippase") contained within the rough endoplasmic reticulum, and the final product is extruded into the intracisternal space. Therefore, depending on the rate of entry of secretory mRNAs into the cytoplasm, the rate of initiation and elongation of secretory proteins, and the number of hydrophobio sites available on the rough endoplasmic reticulum, a certain proportion of ribosomes containing mRNA for secretory proteins will be found in the free cytoplasmic fraction. However, once mRNA becomes associated with the membrane-bound fraction, in all probability it is not released again unless all nascent chains are completed or released and protein synthesis reinitiation is inhibited. Recently, several investigators have reported evidence consistent with such a model (46-50), and Strauss et al. (51) have reported that rat albumin synthesized in a wheat germ cell-free system contains an extra hydrophobic NH2-terminal sequence of 24 amino acids. In a previous study, we found that RNA extracted from membrane-bound and free polyribosomes of rabbit liver showed little difference in albumin synthesis when translated in a messenger-dependent reticulocyte cell-free system (17). We suggested that these results were consistent with an interpretation that free polyribosome preparations contained nontranslated albumin mRNA, perhaps as messenger ribonucleoprotein particles. Zahringer et al. (23) suggested another possibility-namely, that the membrane-bound polyribosomes used in the study were dissociated during the isolation procedure and thus contaminated the free polyribosome fraction. Because we had observed a 5- to 6-fold difference in albumin synthesis between membrane-bound and free polyribosomes in the native or intact liver system, this interpretation was unlikely. Based on our present study, the simplest explanation for differences in albumin mRNA content as determined by protein synthesis and by molecular hybridization analysis is unequal translation of albumin mRNA in the various fractions. Although there is no evidence that such variations occur in vivo, the present study clearly demonstrates that 98% of albumin mRNA sequences in rat liver polyribosomes are contained in the membrane-bound fraction. This research was supported in part by National Institutes of Health Grants AM-17609 and AM-17702, Cell and Molecular Biology Training Grant 5-TO1 GM 02209, Medical Scientist Program Grant 5T32-GM 7288, theNetherlands Organization for Advancement of Pure Research (Zuiver Wettenschappelijk Orderzock) and Niels Stensen Stichting to S.H.Y., a National Institutes of Health Research Career Development Award to D.A.S., and the Irma T. Hirschl Charitable Trust of New York. 1. Siekevitz, P. & Palade, G. (1960) J. Biophys. Biochem. Cytol. 7, 619-630. 2. Birbeck, M. S. C. & Mercer, E. H. (1961), Nature 189, 558560. 3. Peters, T., Jr. (1962) J. Biol. Chem. 237, 1186-1189. 4. Hicks, S. J., Drysdale, J. W. & Munro, H. N. (1968) Science 164, 584-585.
Biochemistry: Yap etal. 5. Takagi, M. & Ogata, K. (1968) Biochem. Biophys. Res. Commun. 33,55-60. 6. Redman, C. M. (1969) 1. Biol. Chem. 244, 4308-4315; 7. Takagi, M., Tanaka, T. & Ogata, K. (1970) Biochim. Biophys. Acta 217, 148-158. 8. Andrews, T. M. & Tata, J. R. (1971) Biochem. J. 121, 683694. 9. Uenoyama, K. & Ono, T. (1972) Biochem. Biophys. Res. Commun. 49,713-719. 10. Rolleston, F. S. (1974) Sub-Cell Biochem. 3, 91-117. 11. Shafritz, D. A. (1974) J. Biol. Chem. 249, 81-88. 12. Ikehara, Y. & Pitot, H. C. (1973) J. Cell Biol. 59, 28-44. 13. Ragnotti, G., Lawford, G. R. & Campbell, P. N. (1969) Biochem. J. 112, 139-147. 14. Kadenbach, B. (1970) Eur. J. Biochem. 12,392-398. 15. Sakamoto, T. & Higashi, T. (1973) J. Biochem. (Tokyo) 73, 1083-1088. 16. Bergeron, J. J. M., Berridge, M. V. & Evans, W. H. (1975) Biochim. Biophys. Acta 407,325-337. 17. Shafritz, D. A. (1974) J. Biol. Chem. 249,89-93. 18. Shafritz, D. A. (1977) "Molecular Mechanisms of Protein Biosynthesis," in Molecular Biology Series, eds. Weissback, H. & Pestka, S. (Academic Press, New York), pp. 555-601. 19. Ramsey, J. C. & Steele, W. J. (1976) Biochemistry 15, 17041712. 20. Strair, R. K., Yap, S. H. & Shafritz, D. A. (1977) Proc. Natl. Acad. Sci. USA 74, 4346-4350. 21. Housman, D., Skoultchi, A., Forget, B. G. & Benz, E. J. (1974) Ann. N.Y. Acad. Sci. 241, 280-289. 22. Shore, G. C. & Tata, J. R. (1977) J. Cell Biol. 72, 726-743. 23. Zahringer, J., Baliga, B. S., Drake, R. L. & Munro, H. N. (1977) Biochim. Biophys. Acta 474,234-244. 24. Takagi, M. & Ogata, K. (1971) Biochem. Biophys. Res. Commun. 42, 125-131. 25. Rourke, A. W. & Heywood, S. M. (1972) Biochemistry 11, 2061-2066. 26. Mathews, M. B., Pragnell, I. M., Osborn, M. & Arnstein, H. R. V. (1972) Biochim. Biophys. Acta 287, 113-123. 27. Roberts, B. E. & Paterson, B. M. (1973) Proc. Natl. Acad. Sci. USA
70,2330-2334. 28. Zucker, W. V. & Schulman, H. M. (1968) Proc. Natl. Acad. Sci. USA 59, 582-589.
Proc. Natl. Acad. Sci. USA 74 (1977)
5401
29. Adamson, S. D., Yau, P. M. P., Herbert E. & Zucker W. V. (1972) J. Mol. Biol. 63,247-264. 30. Gross, M. & Rabinowits, M. (1972) Proc. Natl. Acad. Sci. USA 69,1565-1568. 31. Legon, S., Jackson, R. J. & Hunt, T. (1973) Nature New Biol. 241, 150-152. 32. Beuzard, Y. & London, I. M. (1974) Proc. Natl. Acad. Scd. USA 71,2863-2866. 33. Atkins, J. F., Lewis, J. B., Anderson, C. W. & Gesteland, R. R. (1975) J. Biol. Chem. 250, 5688-5695. 34. Ehrenfeld, E. & Hunt, T. (1971) Proc. Natl. Acad. Sci. USA 68, 1075-1078. 35. Heywood, S. M., Kennedy, D. S. & Bester, A. J. (1974) Proc. Natl. Acad. Sct. USA 71,2428-2431. 36. Lingrel, J. B., Lockhard, R. E., Jones, R. F., Burr, H. E. & Holder, J. W. (1971) Ser. Haematol. 4,37-69. 37. Nienhuis, A. W., Canfield, P. H. & Anderson, W. F. (1973) J. Clin. Invest. 52,1735-1745. 38. McKeehan, W. L. (1974) J. Biol. Chem. 249,6517-6526. 39. Palmiter, R. D. (1974) J. Biol. Chem. 249,6779-6787. 40. Sonenshein, G. E. & Brawerman, G. (1976) Biochemistry 15, 5501-5506. 41. Tse, T. P. H. & Taylor, J. M. (1977) J. Biol. Chem. 252, 12721278. 42. Peterson, J. A. (1976) Nucleic Acids Res. 3, 1427-1436. 43. Blobel, G. & Potter, V. R. (1966) Proc. Natl. Acads Sci. USA 55, 1283-1288. 44. Roth, J. S. (1967) Methods Cancer Res. 3, 153-155. 45. Blobel, G. & Dobberstein, B. (1975) J. Cell Biol. 67,835-851. 46. Milstein, C., Brownlee, C. G., Harrison, T. M. & Mathews, M. B. (i972) Nature New Biol. 239, 117-120. 47. Burstein, Y. & Schechter, I. (1977) Proc. Natl. Acad. Sci. USA 74,716-720. 48. Kemper, B., Habener, J. F., Ernst, M. D., Potts, J. T. & Rich, A. (1976) Biochemistry 15, 15-19. 49. Wirth, D. F., Katz, F., Small, B. & Lodish, F. (1977) Cell 10, 253-263. 50. Grubman, M. J., Weinstein, J. A. & Shafritz, D. A. (1977) J. Ce!! Biol. 74, 43-57. 51. Strauss, A. W., Donohue, A. M., Bennett, C. D., Rodkey, J. A. & Alberts, A. W. (1977) Proc. Natl. Acad. Sci. USA 74, 13581362.
