Printed itz Swrdm Copyright 0 1975 by Academic Press, Inc. All rights of reproduction in any form rrscraed
Experimental
TURNOVER
Cell Research 91 (1975) 101-106
OF RIBOSOMAL
RNA
IN MOUSE
FIBROBLASTS
(3T3) IN CULTURE G. M. KOLODNY Radiology
Research
Laboratory,
Massachusetts
General
Hospital,
Boston,
Mass.
02114,
USA
SUMMARY Growing and confluent cultures of mouse fibroblasts were labeled with 3H-uridine and chased with an excess of nonradioactive uridine to investigate the turnover of ribosomal RNA. Growing cultures did not turn over their 18s and 28s i?bosomal RNA; however, confluent cultures did show ribosomal RNA (rRNA) turnover. If the cells were labeled while growing, and chased when confluent, 18s RNA displayed a two-component decay curve, while 28s RNA showed only single-component decay, similar in lifetime to the first component of the 18s RNA decay curve. If the cells were labeled while confluent, both the 18s and 28s RNA showed singlecomponent decay curves with a lifetime possibly only slightly longer than the lifetime of the first component of the 18s RNA and the single component of the 28s RNA of the cultures labeled while growing.
Investigations [lo, 11, 171have been conducted on the lifetime of RNA in rat liver in the intact animal. Recently several studies [3, 20, 231 have examined turnover of RNA in cultured mammalian cells. These latter studies have found a two-component decay curve for messenger RNA (mRNA) in growing cells in culture. It seemedof interest also to examine the kinetics of decay of ribosomal RNA (rRNA) in cultured cells and to compare the mode of decay of rRNA with that derived for mRNA in the above in vitro studies. It has been shown that growing cell cultures do not turn over their rRNA. When cultures become confluent, however, rRNA begins to turn over [9, 251. In the following experiments on cells from a mouse fibroblast line (3T3), the turnover of ribosomal RNA in confluent cultures has been measured during extended chase periods of up to 3 weeks.
METHODS
AND MATERIALS
3T3 cells were cultured in 60 mm plastic dishes (Falcon no. 3002) and at confluence each 60 mm dish contained about lo6 cells. The cultures were maintained at 37°C in an atmosphere of 5 % CO, in Dulbecco medium (GIBCo, Grand Island, N.Y.) with 10 % added calf serum, 75 units/ml penicillin and 50 pg/ml streptomycin. The medium was changed 3 times/week. The cultures were checked periodically for mycoplasma and bacterial contamination [14]. Subconfluent cultures were labeled with 2 pCi/ml SH-uridine (30 Ci/mM, New England Nuclear). After labeling for 24-48 h, the radioactive media were removed and the cultures washed 3 times with Earle’s saline. Fresh whole media containing unlabeled uridine 100 times the concentration in the labeling media was added. At intervals over the next 2 weeks duplicate culture labeled and chased as above were separately combined with 10’ unlabeled carrier cells and phenol extracted [12]. The RNA was precipitated twice with ethanol and separated on sucrose gradients [12]. Fractions comprising the optical density peaks for 18s and 28 S RNA were separately pooled and the optical activity at 260 nm determined on a Gilford spectrophotometer. The radioactivity of 0.5 ml of each fraction was then counted in a Packard liquid scintillation counter after mixing with 20 ml of Bray’s scintillant [2]. The radioactivity was then expressed as dpm/unit of optical activity to correct for loss of RNA during purification. Exptl
Cell
Res 91 (1975)
102 10 6
G. M. Kolodny RESULTS
.._
‘~coN+cE, , , ,,; 2
4
6
6
10
12
14
5 t 2
II
-
.
.
4 CONFLiJENCE
’ 2
’ 4
I I
’ 6
’ 6
’ 10
’
12
’
14
’
16
’
16
’
1. Abscissa: time (days); or&tare: (a) dpm x 10-3/OD,,0; (/I) dpm x 1O-3. (a) Subconfluent cells were labeled 24 h with SHuridine (2 ,&i/ml) and then chased while growing and during confluence. Radioactivity within 18s and 28s rRNA during chase is plotted against days of chase. Radioactivity is unchanged during culture growth but declines after confluency is attained; (b) cultures prepared as in (a) labeled also with 14Cthymidine (0.6 &i/ml) and chased as in (a). 14Cradioactivity not changed during chase period indicating no significant change in cell number. Fig.
Hemocytometer cell counts and radioactive counting of cultures labeled with 14C-thymidine were performed on duplicate cultures during SH-uridine labeling and chase ueriods to correct for possible cell loss during the chase periods. Growing cultures labeled with SH-uridine as above were also labeled with 0.6 ,&i/ml 14C-thymidine (50 mCi/mM, New England Nuclear). Trichloroacetic acid (TCA) precipitable r4C radioactivity was determined at intervals during culture growth and confluence while the cultures were being chased with an excess of nonradioactive uridine, as above [15]. The duplicate cultures were washed with serumless media and then dissolved in 2 ml of 0.25 N NaOH. An equal volume of 25 % TCA was added and after 1 h at 4°C the precipitate was filtered on Millinore. The filters were dried at 60°C for 1 h and counted in toluene-based liquid scintillation fluid containing 4 g/l PPO and 50 mg/l POPOP. Subconfluent cultures were labeled with 5 &i/ml aH-uridine for 48-72 h. When confluencv was attained the cultures were chased, and at intervals the radioactivity in 18 S and 28s ribosomal RNA were determined- as above in duplicate culture dishes. Confluent cultures were also similarly labeled for 2448 h, then chased, and the radioactivity in rRNA determined as above. Exptl
Cell
Res 91 (1975)
Subconfluent cultures were radioactively Iabeled with 3H-uridine and then chased with a large excess of unlabeled uridine while still subconfluent and growing. The chase was continued during the next 2 weeks while the cultures grew to confluence and then were maintained in a non-growing confluent state. At intervals during the chase period the radioactivity was determined in 18s and 28s RNA isolated from sucrose gradients of RNA extracted with phenol from at least 3 replicate cultures. Fig. 1 is a plot of radioactivity vs time in 18S and 28 S during the chase period. The radioactivity remains constant during culture growth until confluence is attained, after which radioactivity begins to decline. The lack of turnover of rRNA during growth, and presence of turnover at confluence demonstrated in these experiments in mouse fibroblasts is consistent with the findings of others [9, 251using chick fibroblasts, that rRNA does not turn over during growth but does turn over when the cells become confluent. Hemocytometer cell counts at confluence during the chaseperiod showed no significant decline in cell number and there were no mitotic cells seen after confluence was attained. 14C-Thymidine radioactivity administered and chased on the same schedule to duplicate 3H-uridine labeled cultures is plotted in fig. 1b, and shows no significant decline during growth or confluence. These findings would indicate that the turnover of 3H-uridine radioactivity was due to RNA turnover rather than cell loss. The kinetics of degradation of 18s and 28s ribosomal RNA in confluent cultures was measured by incubating growing subconfluent cell cultures with 3H-uridine and chasing the cells with an excess of non-
Turnover of ribosomal RNA
105,
a
103
1
5xlO3r .
I
I
I
2
4
L 6
I
I
I
I
I
I
a
10
12
14
16
16
I
d
104
5x103
-
l I 2
1 4
I 6
I 8
I 10
I 12
I 14
I 16
I 16
2. Abscissa: time (days); ordinate:(a, c, d) dpm ODzao; (b) dpm x 10d3. Subconfluent cultures were labeled with SH-uridine (5 ,&X/ml). 48-72 h later when confluent the radioactivity was chased. Radioactivity in 18s (a) and 28s (b) RNA from duplicate cultures taken at intervals during the chase period was averaged and plotted. Confluent cultures were radioactively labeled as in (a) and (6) for 24 h and then chased. Radioactivity in 18s (c) and 28.53(d) RNA from duplicate cultures taken at intervals during the chase period was averaged and is plotted. Fig.
labeled precursor after confluence had been reached. The radioactivity in 18s and 28s RNA was measured at intervals during a 3week chase period and is plotted in fig. 2. In fig. 2a the radioactivity in labeled 18s RNA is plotted on a logarithmic axis and shows a decay curve which appears to have at least two components. The first component shows a half-life of about 2.8 days and the second component appears to have a halflife of about 7 days. The decay curve for 3H-uridine labeled 28s RNA (fig. 2b) on the other hand, appears to be linear, indicating a single component. The half life of 28 S RNA of about 2.8 days is close to the half life of the first component of the 18s decay curve.
The variation between 18S and 28S RNA decay curves seen in the case of RNA labeled while growing, suggested that ribosomal RNAs might possesssome degree of heterogeneity in regard to susceptibility to degradation. Since drastic changes occur in RNA metabolic activity [7, 9, 16, 251at confluence it seemed worthwhile also to investigate the decay of radioactivity in rRNA species in cultures labeled while confluent, rather than labeled while in logarithmic growth, as in preceding experiments. Confluent cultures were radioactively labeled and chased for 3 weeks in a manner similar to the subconfluent cultures described above. The radioactivity in 18S and 28 S Exptl
Cell
Res 91 (1975)
104
G. M. Kolodny
RNA during the chase period is plotted in fig. 2 c and d. Radioactivity in both 18 S (fig. 2c) and 28s RNA (fig. 2d) from cultures labeled while confluent with 3H-uridine show single-component decay curves during a 3 week chase period. For both 18s and 28s RNA the half-life appears to be about 3.5 days. DISCUSSION Several experiments have shown that it is difficult to dilute intracellular uridine pools during short chase periods [6, 241. Murphy & Attardi [18] have shown, however, that radioactive uridine can be chased from the uridine pools if the chase period is sufficiently long. In our experiments very long chase periods, even longer than those of Murphy & Attardi, have been employed. In the above experiments we have confirmed that 3T3 cells, like chick fibroblasts [9, 251 which have been exposed to much shorter labeling and chase periods, show no rRNA turnover while growing, but do show rRNA turnover at confluence. The half life of unstable rRNA in other systems has been calculated to follow first order kinetics and to be between 12 and 121 h [5, 9, 10, 11, 17, 251. In our experiments except for the case of the 18 S RNA prelabeled in growing cells, which will be further discussed, we also found single order kinetics with a half life of unstable ribosomal RNA of 2.8-3.5 days. This half-life is intermediate in the range of half-lives found by others. The discrepancies in half-lives that have been determined could possibly be explained by differences in cell types and their generation times or inaccuracies due to the much shorter labeling and chase periods employed in these other experiments. It is interesting that Murphy & Attardi [lS] found a half life of 3 days for mRNA labeled for 24 h in growing HeLa cells which is close to the half-life of rRNA Exptl
Cell Res 91 (1975)
of our results in cells labeled for 24-48 h. Other investigators have, however, found shorter half-lives for mRNA after labeling for shorter periods. The 18s RNA prelabeled in growing cells shows second-order kinetics in its decay. The first component is similar in lifetime to that of 28s RNA prelabeled in growing cells and not too dissimilar to the life times of both 18s and 28s RNA prelabeled in confluent cultures. The second component appears to have a half-life of about 7 days. The second component i.n the 18 S decay curve in cells labeled while growing could represent a contaminating species of RNA that sediments with 18s RNA on a sucrose gradient or a form of 18s RNA that is more resistant to degradation. Nair & Knight [19] have shown that intermediates in the degradation of 28s RNA can sediment with 18s RNA. However, it is unlikely that the second component in the decay curve of 18 S RNA, labeled while growing, represents a degradation product of 28s RNA, because it does not appear in the decay curve for 18s RNA labeled while confluent. It is difficult to explain the difference in the decay curves for 18 S RNA labeled while growing and labeled while confluent. It has been shown [16] that during confluence there is a change in the ratio of membrane bound to free ribosomes. Perhaps such differences could account for the differences in stability of 18 S RNA depending upon the location of the prelabeled RNA. Several investigators [3, 20, 231 have shown a two-component decay curve for poly A associated mRNA in exponentially growing cells that do not show contact inhibition. In the light of our results with 18s RNA it would be of great interest to compare the kinetics of turnover of mRNA in contactinhibited cells prelabeled while growing and prelabeled while confluent.
Turnover of ribosomal RNA Since completion of our studies, a report [l] has appeared which demonstrated differences in the rates of turnover of 18 S and 28s rRNA in resting 3T3 cells. Confluent cells were labeled with radioactive precursors and the cells were chased for 10 days. They found, as we did with cells labeled when confluent, single-component decay curves for 18 S and 28s rRNA. Their data did not include cultures labeled while growing and chased when confluent as in our experiments. Another possible hypothesis to explain the difference in the decay curve for 18s RNA prelabeled in growing cells and prelabeled in confluent cells may be provided by a recent hypothesis [13] concerning gene regulation. We have suggested that at least some RNA transcription may be regulated by the presence of primer RNAs. These primer RNAs are segments from the normal breakdown of functional cytoplasmic or heterodisperse nuclear RNA. If a segment of RNA from the 5’ end of a degraded RNA molecule is preserved, then it can hybridize with the initial portion of the gene from which it was originally transcribed and thereby act as a primer to resynthesize and replace the entire degraded RNA molecule. Such a scheme might explain steady state turnover of ribosomal RNA in confluent cells. rRNA is synthesized from a 45s precursor molecule [8]. It is not certain whether the 18s rRNA or the 28s rRNA is at the 5’ end of the 45 S precursor. Although some studies have suggested that the 28s rRNA is near the 5’ end of the 45 S precursor [4, 261, other studies [19, 21, 221 indicate that the 18 S rRNA is at the 5’ end of the precursor. If the 18s portion of the molecule is at the 5’ end of the precursor while the 28 S protein is at or near the 3’ end of the precursor, a conserved 5’ primer segment might be expected to be within the 18s molecule. In the decay of 18s RNA labeled while
105
growing, the 5’ end is conserved and therefore shows a longer half life than the remainder of the molecule when turnover begins at confluence. If however, an 18s RNA molecule is labeled while confluent, the 5’ end being conserved will not be labeled. Therefore only the nonconserved labeled portion of the molecule will turn over and only a one component decay curve will be seen. The 28s RNA, being at the 3’ end of the 45s precursor, would show only a single component decay curve since none of this molecule is conserved to be used in the synthesis of new RNA molecules. This would be the case whether it was derived from cells labeled while growing or labeled while confluent. It is of interest that an endonucleolytic enzyme has been described which attacks 28s RNA at multiple sites. The same enzyme has a more restricted attack on 18 S RNA to yield a species more resistant to endonucleolytic digestion [27]. Perhaps in vivo that more resistant portion of the 18 S molecule is conserved to initiate new ribosomal RNA transcription. This hypothesis will obviously require further experimental data to confirm or deny its validity. 1 thank Mrs llana Nitzan for expert technical assistance. This research was aided by grant E651 from the ACS and contract AT (11-l) 3335 from the US AEC.
REFERENCES 1. Abelson. H T. Johnson. L F. Penman. S & Green, H, Cell’1 (1974) 161. ’ ’ 2. Bray, G A, Anal biochem 1 (1960) 279. 3. Che&ers, %’ P & Sheinin, k, Biochim biophys acta 204 (1970) 449. 4. ::5i, Y C & Busch, H, 3 biol them 245 (1970) 5. 6. i: 9. 10.
Cocucci, S M & Sussman, M, J cell biol45 (1970) 399. Cooper, H L, J biol them 243 (1968) 31. - Ibid 244 (1969) 5590. Darnell, J E, Bact rev 32 (1968) 262. Emerson, C P, Nature new biol 232 (1971) 101. Hadjiolov, A A, Biochim biophys acta 119 (1966) 547. Exptl
Cell Res 91 (1975)
106
G. M. Kolodny
11. Hirsch, C A & Hiatt, H H, J biol them 241 (1966) 5936. 12. Kolodny, G M, Exptl cell res 65 (1971) 313. 13. - J cell biol 59 (1973) 175a. Kolodny, G M, in Extracellular matrix influences on gene expression (ed H D Slavkin). Academic Press, New York (1975). In press. 14. Kolodny, G M, Culp, L A & Rosenthal, L J, Exptl cell res 73 (1972) 65. 15. Kolodny, G M & Gross, P R, Exptl cell res 57 (1969) 423. 16. Levine, E M, Becker, Y, Boone, C W &Eagle, H, Proc natl acad sci US 53 (1965) 350. 17. Loeb, J N, Howell, R R & Tomkins, G M, Science 149 (1965) 1093. 18. Murphy, W ‘& Attardi, G, Proc natl acad sci US 70 (1973) 115. 19. Nair, C N & Knight, E, J cell biol 50 (1971) 787.
Exptl
Cell Res 91 (1975)
20. Perry, R P. Unpublished results. 21. Reeder, R H & Grown, D D, 3 mol biol 51 (1970) 361. 22. Siev, M, Weinberg, R & Penman, S, J cell biol 41 (1969) 510. 23. Singer, R H & Penman, S, J mol biol 78 (1973) 321. 24. Warner, J R, Soiero, R, Birnboim, H C, Girard, M & DarnelI, J E, J mol biol 19 (1966) 349. 25. Weber, M H, Nature new biol 235 (1972) 58. 26. Wellauer, P K & Dawid, I B, Proc natl acad sci US 70 (1973) 2827. 27. Winicov, I & Perry, M P, J cell biol 59 (1973) 366a.
Received June 11, 1974 Revised version received August 19, 1974
Experimental
TURNOVER
Cell Research 91 (1975) 101-106
OF RIBOSOMAL
RNA
IN MOUSE
FIBROBLASTS
(3T3) IN CULTURE G. M. KOLODNY Radiology
Research
Laboratory,
Massachusetts
General
Hospital,
Boston,
Mass.
02114,
USA
SUMMARY Growing and confluent cultures of mouse fibroblasts were labeled with 3H-uridine and chased with an excess of nonradioactive uridine to investigate the turnover of ribosomal RNA. Growing cultures did not turn over their 18s and 28s i?bosomal RNA; however, confluent cultures did show ribosomal RNA (rRNA) turnover. If the cells were labeled while growing, and chased when confluent, 18s RNA displayed a two-component decay curve, while 28s RNA showed only single-component decay, similar in lifetime to the first component of the 18s RNA decay curve. If the cells were labeled while confluent, both the 18s and 28s RNA showed singlecomponent decay curves with a lifetime possibly only slightly longer than the lifetime of the first component of the 18s RNA and the single component of the 28s RNA of the cultures labeled while growing.
Investigations [lo, 11, 171have been conducted on the lifetime of RNA in rat liver in the intact animal. Recently several studies [3, 20, 231 have examined turnover of RNA in cultured mammalian cells. These latter studies have found a two-component decay curve for messenger RNA (mRNA) in growing cells in culture. It seemedof interest also to examine the kinetics of decay of ribosomal RNA (rRNA) in cultured cells and to compare the mode of decay of rRNA with that derived for mRNA in the above in vitro studies. It has been shown that growing cell cultures do not turn over their rRNA. When cultures become confluent, however, rRNA begins to turn over [9, 251. In the following experiments on cells from a mouse fibroblast line (3T3), the turnover of ribosomal RNA in confluent cultures has been measured during extended chase periods of up to 3 weeks.
METHODS
AND MATERIALS
3T3 cells were cultured in 60 mm plastic dishes (Falcon no. 3002) and at confluence each 60 mm dish contained about lo6 cells. The cultures were maintained at 37°C in an atmosphere of 5 % CO, in Dulbecco medium (GIBCo, Grand Island, N.Y.) with 10 % added calf serum, 75 units/ml penicillin and 50 pg/ml streptomycin. The medium was changed 3 times/week. The cultures were checked periodically for mycoplasma and bacterial contamination [14]. Subconfluent cultures were labeled with 2 pCi/ml SH-uridine (30 Ci/mM, New England Nuclear). After labeling for 24-48 h, the radioactive media were removed and the cultures washed 3 times with Earle’s saline. Fresh whole media containing unlabeled uridine 100 times the concentration in the labeling media was added. At intervals over the next 2 weeks duplicate culture labeled and chased as above were separately combined with 10’ unlabeled carrier cells and phenol extracted [12]. The RNA was precipitated twice with ethanol and separated on sucrose gradients [12]. Fractions comprising the optical density peaks for 18s and 28 S RNA were separately pooled and the optical activity at 260 nm determined on a Gilford spectrophotometer. The radioactivity of 0.5 ml of each fraction was then counted in a Packard liquid scintillation counter after mixing with 20 ml of Bray’s scintillant [2]. The radioactivity was then expressed as dpm/unit of optical activity to correct for loss of RNA during purification. Exptl
Cell
Res 91 (1975)
102 10 6
G. M. Kolodny RESULTS
.._
‘~coN+cE, , , ,,; 2
4
6
6
10
12
14
5 t 2
II
-
.
.
4 CONFLiJENCE
’ 2
’ 4
I I
’ 6
’ 6
’ 10
’
12
’
14
’
16
’
16
’
1. Abscissa: time (days); or&tare: (a) dpm x 10-3/OD,,0; (/I) dpm x 1O-3. (a) Subconfluent cells were labeled 24 h with SHuridine (2 ,&i/ml) and then chased while growing and during confluence. Radioactivity within 18s and 28s rRNA during chase is plotted against days of chase. Radioactivity is unchanged during culture growth but declines after confluency is attained; (b) cultures prepared as in (a) labeled also with 14Cthymidine (0.6 &i/ml) and chased as in (a). 14Cradioactivity not changed during chase period indicating no significant change in cell number. Fig.
Hemocytometer cell counts and radioactive counting of cultures labeled with 14C-thymidine were performed on duplicate cultures during SH-uridine labeling and chase ueriods to correct for possible cell loss during the chase periods. Growing cultures labeled with SH-uridine as above were also labeled with 0.6 ,&i/ml 14C-thymidine (50 mCi/mM, New England Nuclear). Trichloroacetic acid (TCA) precipitable r4C radioactivity was determined at intervals during culture growth and confluence while the cultures were being chased with an excess of nonradioactive uridine, as above [15]. The duplicate cultures were washed with serumless media and then dissolved in 2 ml of 0.25 N NaOH. An equal volume of 25 % TCA was added and after 1 h at 4°C the precipitate was filtered on Millinore. The filters were dried at 60°C for 1 h and counted in toluene-based liquid scintillation fluid containing 4 g/l PPO and 50 mg/l POPOP. Subconfluent cultures were labeled with 5 &i/ml aH-uridine for 48-72 h. When confluencv was attained the cultures were chased, and at intervals the radioactivity in 18 S and 28s ribosomal RNA were determined- as above in duplicate culture dishes. Confluent cultures were also similarly labeled for 2448 h, then chased, and the radioactivity in rRNA determined as above. Exptl
Cell
Res 91 (1975)
Subconfluent cultures were radioactively Iabeled with 3H-uridine and then chased with a large excess of unlabeled uridine while still subconfluent and growing. The chase was continued during the next 2 weeks while the cultures grew to confluence and then were maintained in a non-growing confluent state. At intervals during the chase period the radioactivity was determined in 18s and 28s RNA isolated from sucrose gradients of RNA extracted with phenol from at least 3 replicate cultures. Fig. 1 is a plot of radioactivity vs time in 18S and 28 S during the chase period. The radioactivity remains constant during culture growth until confluence is attained, after which radioactivity begins to decline. The lack of turnover of rRNA during growth, and presence of turnover at confluence demonstrated in these experiments in mouse fibroblasts is consistent with the findings of others [9, 251using chick fibroblasts, that rRNA does not turn over during growth but does turn over when the cells become confluent. Hemocytometer cell counts at confluence during the chaseperiod showed no significant decline in cell number and there were no mitotic cells seen after confluence was attained. 14C-Thymidine radioactivity administered and chased on the same schedule to duplicate 3H-uridine labeled cultures is plotted in fig. 1b, and shows no significant decline during growth or confluence. These findings would indicate that the turnover of 3H-uridine radioactivity was due to RNA turnover rather than cell loss. The kinetics of degradation of 18s and 28s ribosomal RNA in confluent cultures was measured by incubating growing subconfluent cell cultures with 3H-uridine and chasing the cells with an excess of non-
Turnover of ribosomal RNA
105,
a
103
1
5xlO3r .
I
I
I
2
4
L 6
I
I
I
I
I
I
a
10
12
14
16
16
I
d
104
5x103
-
l I 2
1 4
I 6
I 8
I 10
I 12
I 14
I 16
I 16
2. Abscissa: time (days); ordinate:(a, c, d) dpm ODzao; (b) dpm x 10d3. Subconfluent cultures were labeled with SH-uridine (5 ,&X/ml). 48-72 h later when confluent the radioactivity was chased. Radioactivity in 18s (a) and 28s (b) RNA from duplicate cultures taken at intervals during the chase period was averaged and plotted. Confluent cultures were radioactively labeled as in (a) and (6) for 24 h and then chased. Radioactivity in 18s (c) and 28.53(d) RNA from duplicate cultures taken at intervals during the chase period was averaged and is plotted. Fig.
labeled precursor after confluence had been reached. The radioactivity in 18s and 28s RNA was measured at intervals during a 3week chase period and is plotted in fig. 2. In fig. 2a the radioactivity in labeled 18s RNA is plotted on a logarithmic axis and shows a decay curve which appears to have at least two components. The first component shows a half-life of about 2.8 days and the second component appears to have a halflife of about 7 days. The decay curve for 3H-uridine labeled 28s RNA (fig. 2b) on the other hand, appears to be linear, indicating a single component. The half life of 28 S RNA of about 2.8 days is close to the half life of the first component of the 18s decay curve.
The variation between 18S and 28S RNA decay curves seen in the case of RNA labeled while growing, suggested that ribosomal RNAs might possesssome degree of heterogeneity in regard to susceptibility to degradation. Since drastic changes occur in RNA metabolic activity [7, 9, 16, 251at confluence it seemed worthwhile also to investigate the decay of radioactivity in rRNA species in cultures labeled while confluent, rather than labeled while in logarithmic growth, as in preceding experiments. Confluent cultures were radioactively labeled and chased for 3 weeks in a manner similar to the subconfluent cultures described above. The radioactivity in 18S and 28 S Exptl
Cell
Res 91 (1975)
104
G. M. Kolodny
RNA during the chase period is plotted in fig. 2 c and d. Radioactivity in both 18 S (fig. 2c) and 28s RNA (fig. 2d) from cultures labeled while confluent with 3H-uridine show single-component decay curves during a 3 week chase period. For both 18s and 28s RNA the half-life appears to be about 3.5 days. DISCUSSION Several experiments have shown that it is difficult to dilute intracellular uridine pools during short chase periods [6, 241. Murphy & Attardi [18] have shown, however, that radioactive uridine can be chased from the uridine pools if the chase period is sufficiently long. In our experiments very long chase periods, even longer than those of Murphy & Attardi, have been employed. In the above experiments we have confirmed that 3T3 cells, like chick fibroblasts [9, 251 which have been exposed to much shorter labeling and chase periods, show no rRNA turnover while growing, but do show rRNA turnover at confluence. The half life of unstable rRNA in other systems has been calculated to follow first order kinetics and to be between 12 and 121 h [5, 9, 10, 11, 17, 251. In our experiments except for the case of the 18 S RNA prelabeled in growing cells, which will be further discussed, we also found single order kinetics with a half life of unstable ribosomal RNA of 2.8-3.5 days. This half-life is intermediate in the range of half-lives found by others. The discrepancies in half-lives that have been determined could possibly be explained by differences in cell types and their generation times or inaccuracies due to the much shorter labeling and chase periods employed in these other experiments. It is interesting that Murphy & Attardi [lS] found a half life of 3 days for mRNA labeled for 24 h in growing HeLa cells which is close to the half-life of rRNA Exptl
Cell Res 91 (1975)
of our results in cells labeled for 24-48 h. Other investigators have, however, found shorter half-lives for mRNA after labeling for shorter periods. The 18s RNA prelabeled in growing cells shows second-order kinetics in its decay. The first component is similar in lifetime to that of 28s RNA prelabeled in growing cells and not too dissimilar to the life times of both 18s and 28s RNA prelabeled in confluent cultures. The second component appears to have a half-life of about 7 days. The second component i.n the 18 S decay curve in cells labeled while growing could represent a contaminating species of RNA that sediments with 18s RNA on a sucrose gradient or a form of 18s RNA that is more resistant to degradation. Nair & Knight [19] have shown that intermediates in the degradation of 28s RNA can sediment with 18s RNA. However, it is unlikely that the second component in the decay curve of 18 S RNA, labeled while growing, represents a degradation product of 28s RNA, because it does not appear in the decay curve for 18s RNA labeled while confluent. It is difficult to explain the difference in the decay curves for 18 S RNA labeled while growing and labeled while confluent. It has been shown [16] that during confluence there is a change in the ratio of membrane bound to free ribosomes. Perhaps such differences could account for the differences in stability of 18 S RNA depending upon the location of the prelabeled RNA. Several investigators [3, 20, 231 have shown a two-component decay curve for poly A associated mRNA in exponentially growing cells that do not show contact inhibition. In the light of our results with 18s RNA it would be of great interest to compare the kinetics of turnover of mRNA in contactinhibited cells prelabeled while growing and prelabeled while confluent.
Turnover of ribosomal RNA Since completion of our studies, a report [l] has appeared which demonstrated differences in the rates of turnover of 18 S and 28s rRNA in resting 3T3 cells. Confluent cells were labeled with radioactive precursors and the cells were chased for 10 days. They found, as we did with cells labeled when confluent, single-component decay curves for 18 S and 28s rRNA. Their data did not include cultures labeled while growing and chased when confluent as in our experiments. Another possible hypothesis to explain the difference in the decay curve for 18s RNA prelabeled in growing cells and prelabeled in confluent cells may be provided by a recent hypothesis [13] concerning gene regulation. We have suggested that at least some RNA transcription may be regulated by the presence of primer RNAs. These primer RNAs are segments from the normal breakdown of functional cytoplasmic or heterodisperse nuclear RNA. If a segment of RNA from the 5’ end of a degraded RNA molecule is preserved, then it can hybridize with the initial portion of the gene from which it was originally transcribed and thereby act as a primer to resynthesize and replace the entire degraded RNA molecule. Such a scheme might explain steady state turnover of ribosomal RNA in confluent cells. rRNA is synthesized from a 45s precursor molecule [8]. It is not certain whether the 18s rRNA or the 28s rRNA is at the 5’ end of the 45 S precursor. Although some studies have suggested that the 28s rRNA is near the 5’ end of the 45 S precursor [4, 261, other studies [19, 21, 221 indicate that the 18 S rRNA is at the 5’ end of the precursor. If the 18s portion of the molecule is at the 5’ end of the precursor while the 28 S protein is at or near the 3’ end of the precursor, a conserved 5’ primer segment might be expected to be within the 18s molecule. In the decay of 18s RNA labeled while
105
growing, the 5’ end is conserved and therefore shows a longer half life than the remainder of the molecule when turnover begins at confluence. If however, an 18s RNA molecule is labeled while confluent, the 5’ end being conserved will not be labeled. Therefore only the nonconserved labeled portion of the molecule will turn over and only a one component decay curve will be seen. The 28s RNA, being at the 3’ end of the 45s precursor, would show only a single component decay curve since none of this molecule is conserved to be used in the synthesis of new RNA molecules. This would be the case whether it was derived from cells labeled while growing or labeled while confluent. It is of interest that an endonucleolytic enzyme has been described which attacks 28s RNA at multiple sites. The same enzyme has a more restricted attack on 18 S RNA to yield a species more resistant to endonucleolytic digestion [27]. Perhaps in vivo that more resistant portion of the 18 S molecule is conserved to initiate new ribosomal RNA transcription. This hypothesis will obviously require further experimental data to confirm or deny its validity. 1 thank Mrs llana Nitzan for expert technical assistance. This research was aided by grant E651 from the ACS and contract AT (11-l) 3335 from the US AEC.
REFERENCES 1. Abelson. H T. Johnson. L F. Penman. S & Green, H, Cell’1 (1974) 161. ’ ’ 2. Bray, G A, Anal biochem 1 (1960) 279. 3. Che&ers, %’ P & Sheinin, k, Biochim biophys acta 204 (1970) 449. 4. ::5i, Y C & Busch, H, 3 biol them 245 (1970) 5. 6. i: 9. 10.
Cocucci, S M & Sussman, M, J cell biol45 (1970) 399. Cooper, H L, J biol them 243 (1968) 31. - Ibid 244 (1969) 5590. Darnell, J E, Bact rev 32 (1968) 262. Emerson, C P, Nature new biol 232 (1971) 101. Hadjiolov, A A, Biochim biophys acta 119 (1966) 547. Exptl
Cell Res 91 (1975)
106
G. M. Kolodny
11. Hirsch, C A & Hiatt, H H, J biol them 241 (1966) 5936. 12. Kolodny, G M, Exptl cell res 65 (1971) 313. 13. - J cell biol 59 (1973) 175a. Kolodny, G M, in Extracellular matrix influences on gene expression (ed H D Slavkin). Academic Press, New York (1975). In press. 14. Kolodny, G M, Culp, L A & Rosenthal, L J, Exptl cell res 73 (1972) 65. 15. Kolodny, G M & Gross, P R, Exptl cell res 57 (1969) 423. 16. Levine, E M, Becker, Y, Boone, C W &Eagle, H, Proc natl acad sci US 53 (1965) 350. 17. Loeb, J N, Howell, R R & Tomkins, G M, Science 149 (1965) 1093. 18. Murphy, W ‘& Attardi, G, Proc natl acad sci US 70 (1973) 115. 19. Nair, C N & Knight, E, J cell biol 50 (1971) 787.
Exptl
Cell Res 91 (1975)
20. Perry, R P. Unpublished results. 21. Reeder, R H & Grown, D D, 3 mol biol 51 (1970) 361. 22. Siev, M, Weinberg, R & Penman, S, J cell biol 41 (1969) 510. 23. Singer, R H & Penman, S, J mol biol 78 (1973) 321. 24. Warner, J R, Soiero, R, Birnboim, H C, Girard, M & DarnelI, J E, J mol biol 19 (1966) 349. 25. Weber, M H, Nature new biol 235 (1972) 58. 26. Wellauer, P K & Dawid, I B, Proc natl acad sci US 70 (1973) 2827. 27. Winicov, I & Perry, M P, J cell biol 59 (1973) 366a.
Received June 11, 1974 Revised version received August 19, 1974
