Molec. gen. Genet. 157, 205-214 (1977) © by Springer-Verlag 1977
Exchange of Ribosomal Proteins among the Ribosomes of Escherichia coil W.R. Robertson, S.J. Dowsett*, and S.J.S. Hardy Department of Biology, University of York, Heslington, York, England
Summary. The exchange of ribosomal proteins among ribosomes of E. coli has been measured, using a density label technique. As expected most of the proteins do not exchange appreciably. However a substantial fraction of each of proteins S1, $2, $21, L7/L12, L9, L10, L l l , L26 and L33 is found to exchange, but exchange of S1, $2, LT/L12, L10, L l l and L26 is found to occur in vitro after lysis of the cells, and therefore it is not possible to say whether or not these proteins also exchange in vivo. In contrast $21, L9 and L33 do not exchange after lysis of the cells and we therefore conclude that these proteins exchange in vivo. The maximum level of exchange of $21, L9 and L33 is attained so rapidly that we were unable to show whether or not it was dependent on protein synthesis.
Introduction Almost all the macromolecules which have been identified as components of the ribosomes of E. coli, are stable during exponential growth of the bacteria (Davern and Meselson, 1960; Dennis, 1974). In addition the most recent data indicate that there is probably one molecule of each macromolecule in every mature ribosome, with the single exception of protein L7/L12 (Hardy, 1975), of which there are-probably four molecules per ribosome (Subramanian, 1975). Thus ribosomes are generally considered to be fixed and stable structures. However, it has not yet been clearly demonstrated that ribosomal proteins do not exchange among the ribosomes as they function. The * Present address. Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland For offprints contact." Dr. S. Hardy, Department of Biology,
University of York, Heslington, York, England
demonstration of such exchange would alter our view of the structure and function of the ribosome, as well as our view of the exchanging macromolecules as structural components of it. Evidence that the bulk of the ribosomal protein does not exchange in vivo was obtained in the elegant experiments of Kaempfer et al. (1968), in which an upper limit of 3% exchange of the total protein was set for the 50S subunits. For the 30S subunits an upper limit was not calculated because of the problem of contamination of the particles with supernatant proteins. However, it is clear from the data that exchange of proteins is at most a minor phenomenon for the small subunit too. Nevertheless, exchange of a few specific proteins is not ruled out by these results, and we have carried out a series of experiments to detect and characterize ribosomal protein exchange in vivo. This was done by using heavy isotopes in a method similar to, but more amenable to quantitative analysis than, that of Kaempfer et al. (1968). Although as expected, the great majority of proteins do not exchange appreciably, we have found to our surprise that there is very rapid, but not complete exchange among the ribosomes, of three proteins, $21, L9 and L33, within the bacterial cells. In addition, we have shown that proteins S1, $2, L7/LI2, L10, L11, L26 and the non-ribosomal proteins found bound to the ribosomes after density gradient centrifugation, exchange to a considerable extent (20% or more) among the ribosomes in vitro after lysis of the bacteria. For these proteins we are not able to say whether or not they exchange in vivo.
Materials and Methods Chemicals
Radioactive lysine was obtained from the Radiochemical Centre, Amersham. The compounds labelled with heavy isotopes, ammo-
206 nium chloride (98.5 atom percent lSN), deuterium oxide (99.8%) and D glucose (74.3 atom percent IaC) were obtained from Prochem, a subsidiary of the British Oxygen Co. Ltd. In early experiments, an algal whole hydrolysate, (1 gram in 3 ml D20) containing 90, 90, and 98 atom percent ~3C, lSN and D respectively, made by Merck Sharp and Dohme was used. This was a generous gift from Linda Randall.
Bacteria, Growth Media, and Conditions of Growth E. coli strain SH7, a stringent derivative of Hfr P4X Met was used in ali experiments. The basic growth medium was M-9 salts containing 50 gg/ml methionine, and glucose as a carbon source. The doubling time of the bacteria in this medium was about 55 rain. Four different 'heavy' growth media were used in the experiments reported here: i) 4 mg/ml glycerol, M-9 salts in D20 with heavy algal hydrolysate diluted 1 part in 50. The doubling time of SH7 in this medium was about 70 rain. ii) 4 mg/ml glucose, M-9 salts without ammonium chloride in D20, 0.2 mg/ml of lSN ammonium chloride. Doubling time about 180 min. iii) The same medium as ii) but with 0.7 mg/ml 13C glucose instead of 12C glucose. Doubling time about 200 rain. iv) The same medium as iii) but with D20 replaced by H20. Doubling time about 65 rain. The procedure for growing the bacteria first in heavy medium and then in light medium was as follows. An overnight culture of SH7 in M-9 salts with limiting glucose (0.6 mg/ml) and reduced ammonium chloride (0.2 mg/ml) was diluted 50-fold into heavy medium and incubated with shaking at 37°C. Growth was monitored using a Bausch and Lomb Spectronic 20 at a wavelength of 450 rim. When the absorbance reached 0.4 (2.5 x 108 cells/ml) part of the culture was added to frozen crushed buffer (0.01 M Tris, 0.003 M succinic acid, 0.01 M MgC12) to use for the controls, and the remainder of the bacteria were collected by centrifugation at 3000xg for 5 rain in a sterile tube, in a rotor prewarmed to 37°C. The bacterial pellet was resuspended in twice the original volume of prewarmed light medium and incubated at 37°C with shaking. At various times after the medium change, 3 to 5 ml (i volume) of the culture were added to 0.35 volumes of frozen crushed buffer and the bacteria were collected by centrifugation at 2°C for 2 rain at 10,000 g. For the controls (see below), portions of a fully light culture grown in parallel, were added to the fully heavy bacteria saved during the change of medium, and the mixed bacteria were collected in the same way. Radioactive labelling of the bacterial proteins was carried out by adding 14C or 3H-lysine to either the heavy or the light medium, without any non-radioactive carrier. The quantity of radioactivity added was usually 100 gCi of 3H or 10 gCi of l¢C-lysine in each 3 to 5 ml portion of the culture harvested. For experiments in which short labelling times with 3H lysine were required, a chase of 40 gg/ml of non-radioactive lysine was added after the labelling period. Labelling was always begun at least one generation before the change of medium, or at least fifteen minutes after, so that the radioactivity would only be incorporated into fully heavy or fully light ribosomal subunits, and not into particles which were in the process of synthesis and assembly during the change of medium. In all experiments a separate culture grown exclusively in light medium, and used as an internal standard for losses of radioactivity (see below), was labelled with either 14C or 3H lysine as described above. Preparation of Ribosomal Subunits, Extraction of Proteins, Electrophoresis and Estimation of Radioactivity These steps were carried out as described previously (Hardy, 1975) with only a few minor changes. The procedure is described briefly
W.R. Robertson et al. : Exchange of Ribosomal Proteins below. Bacterial pellets were resuspended in Tris magnesium buffer and lysed by ultrasonic vibration. The entire lysate was dialysed against a low magnesium buffer (usually 6 x 10 -~ M MgC12) and the subunits were separated on 15 30% (w/v) sucrose gradients in the same buffer. The heavy and light ribosomal subunits were pooled separately, carrier subunits were added together with the radioactive subunits used as an internal standard (see below) and the particles were concentrated by alcohol precipitation at - 2 0 ° C overnight. This step tends to cause loss of protein L7/L12 and to a lesser extent L10. The proteins in the precipitate were extracted with acetic acid and the extract was dialysed into the buffer used in the sample gel of the two dimensional elctrophoresis. Reduction with/~-mercaptoethanol was carried out during this dialysis. The proteins were subjected to two-dimensional electrophoresis on polyacrylamide gels and the ribosomal protein spots were cut out and counted. A piece of gel just below and to the anode (first dimension) side of the sample, containing several stained spots was also cut out and counted. These spots together are referred to as non-ribosomal proteins (NRP), and are assumed to be mainly nonspecifically adsorbed protein contaminants of the ribosomal subunits, although they may also contain protein synthesis factors.
Measurement of the Fraction of a Protein Exchanged After growth in heavy medium followed by growth in light medium, each bacterium contains both heavy and light ribosomal subunits, which are separable from each other on sucrose gradients. Since the bacterial proteins are labelled only in one of the two media, and since most of the ribosomal proteins do not exchange (Kaempfer et al., 1968), the great majority of the radioactivity in the ribosomes is confined to either the light or the heavy subunits (Fig. 1). Separation of the heavy and light subunits was of course never complete, so that even in the absence of exchange, some of the radioactivity present in the labelled subunits was found in the unlabelled subunits, because o f overlap of the heavy and light peaks. Exchange is indicated for a particular protein when the distribution of its radioactivity between heavy and light subunits is different from that of a non-exchanging protein, i.e. a larger proportion of the radioactivity is found in the unlabelled subunits and a smaller proportion is found in the labelled subunits. The fraction of the protein which has exchanged, x, can be calculated from R, the ratio of the protein's radioactivity in the labelled subunits to that in the unlabelled subunits, by using the equation x
(R0 --R) (Rf+ 1) (R o --Rf) ( R + 1)
(1)
where R0 is the value of R when there is no exchange and Rf is the value of R when there is full exchange. Ro is obtained from the value of R for a protein which is assumed not to exchange ($4 for the small ribosomal subunit, L3 for the large ribosomal subunit). Full exchange of a protein would be characterised by an even distribution of its radioactivity over all the ribosomes, and therefore Rf is the ratio of the absorbances of the pools of heavy and light subunits. This was obtained from the gradient traces by first drawing a perpendicular to the base line at the position of the division between the pooled heavy and light subunits, then by cutting out the two peaks generated and weighing them. Here we have assumed that the molar extinction coefficient of heavy and light subunits is the same, and that both have the same protein composition. Equation (1) is easily derived by considering the ribosomes as the sum of two populations, a fraction x among which the protein has exchanged completely, and a fraction 1 - x among which the protein has not exchanged at all. It is assumed in the derivation that the proportion of heavy and light ribosomes in each fraction is the same. It is obvious from the form of the equation that the accuracy of the measurement of x is dependent
W.R. Robertsou et al.: Exchange of Ribosomal Proteins
207 in different cells, they could not exchange components until lysis, after which exchange of proteins should be the same as in the experimental lysates. One further elaboration of this control was carried out in some experiments, to eliminate the possibility that the exchange we observed was promoted in vitro (i.e. after lysis), by a factor present only in lysates of cultures which had been switched from heavy to light medium. In this control, radioactive cells (either heavy or light) were added to a nonradioactive culture which had been subjected to the medium change. Thus the hypothetical factor would be present in the lysate of this raixed culture and should promote exchange. Since this kind of control gave the same results as the simpler one described above, we have combined the results in what follows.
on a large difference between R 0 and Rf. The average value of R f / R o in this work has been 0.22. The measurement of the radioactivity in each protein must be carried out so that losses which occur during extraction of proteins from the subunits, or during electrophoresis of the proteins, have no effect on the value of R. This was done in the following way. To the separate pools of 3H (or 14C) heavy and light ribosomal subunits obtained from the sucrose gradient, equal volumes of 14C (or 3H) labelled ribosomal subunits were added, and the mixture was processed as described. The 3H/~aC ratio of each protein was measured. The ratio of the value of 3H/1¢C of a protein in the heavy subunits to the value of 3H/1¢C of that protein in the light subunits, gave R (or its reciprocal), because an identical quantity of each 1¢C (or 3H) protein had been added to each kind of subunit. The ~4C (or 3H) labelled subunits added in this way were thus an internal standard for losses. In some experiments the volumes of these internal standard ribosomes added to the pooled heavy and light subunits were not equal. In these experiments the ratio of the volumes added is used as a correction factor in calculating R. An examination of Tables 1 and 2, where typical data and the calculations from them are shown, should make these points clearer. Obviously exchange of proteins among ribosomes could occur (and did) after lysis of the bacteria, during the lengthy dialysis before heavy and light subunits were separated, and during the centrifugation which separated them. In all but the first experiment, estimates of this exchange were made by mixing the heavy culture, harvested before the change of medium, with a light culture grown at the same time, one of the two being appropriately labelled. Since in these controls the heavy and light ribosomes were initially
Results
In order to show exchange of ribosomal proteins in vivo it is necessary to be able to separate the ribosomes made during one period of time from those made during another. Then if the old ribosomes contain new proteins characterised by a radioactive label, and vice versa, exchange must be occurring. After growing bacteria first in a medium containing heavy isotopes, and then in a normal medium, old heavy ribosomes can be separated from new light ribosomes on the basis of their mass and density. When the ribosomal subunits from a lysate of cells
b A260 50S ).,t
0.4 50S
f
).:
0.3
30S
30S
I' 1 I I
0.2
p 0.1
I'
j
0-2 i
I
!
I I I
l
f
Fig. l a and b. Sucrose density gradients of radioactive bacterial lysates made after growth in heavy medium followed by growth in light medium. Solid line: A26o, dashed line: radioactivity. Sedimentation from right to left. a Bacteria were grown in heavy medium iii and labelled with 3H lysine in light medium. 45 rain after addition of the label they were harvested, lysed, and dialysed. The dialysate was applied to a 17-ml 15-30% sucrose gradient. The gradient was centrifuged for 15 h at 27,000 rpm at 2°C in the Spmco SW27 rotor and analysed with an ISCO gradient fractionator. Fractions were collected and a small sample of each fraction was counted. b As for a except that the heavy medium lacked D20 and that labelling was carried out with 3H lysine in the heavy medium, about I generation before the change to light medium
208
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 1. Typical data and calculation of fraction exchanged for the 30S proteins Protein
S1 $2 $3 $4 $5 $6 $7 $8 $9+Sll S10 S12 S13 S 14 S15+S16 S17 S18 S19 $20 $21 NRP
Light 30S
Heavy 30S
3H
14C
3H
3H
14C
aH
cts/min
cts/min
~4C
cts/min
cts/min
~4C
426 3,454 3,318 14,660 10,578 3,365 1,081 11,840 9,i78 3,799 856 1,805 2,446 8,818 3,084 1,677 6,136 5,008 550 1,139
391 936 744 2,554 1,859 656 813 1,922 2,180 742 183 308 458 1,487 495 332 1,021 821 250 1,049
1.1 3.7 4.5 5.7 5.7 5.1 1.3 6.2 4.2 5.1 4.7 5.9 5.3 5.9 6.2 5.1 6.0 6.1 2.2 1.1
175 1,463 705 2,517 2,371 418 78 2,141 1,394 729 298 400 666 2,012 350 341 1,146 1,286 300 506
184 429 266 1,228 1,103 201 268 969 1,081 396 135 194 275 991 151 194 613 582 150 544
1.0 3.4 2.6 2.0 2.2 2.1 0.3 2.2 1.3 1.8 2.2 2.1 2.4 2.0 2.3 1.8 1.9 2.2 2.0 0.9
R 3H (light)
Fraction exchanged
3H(heavy)
(%)
2.3 2.2 3.4 5.6 5.3 4.9 9.2 5.6 6.5 5.6 4.2 5.7 4.4 5.8 5.4 5.7 6.4 5.5 2.2 2.3
53 57 27 0 3 6 18 0 7 0 14 1 12 2 1 1 6 1 56 51
-
-
The heavy and light 3H lysine labelled 30S subunits from the gradient shown in Figure 1 a were pooled separately. As an internal standard for the recovery of proteins, two volumes of 1"C lysine labelled 30S subunits were added to the light 30S pool, and one volume to the heavy 30S pool, together with carrier 30S subunits. The proteins were extracted, separated and counted. The data with background subtracted (56 cts/min for 3H, 24 cts/min for 14C) is given in columns 2, 3, 5 and 6. R is obtained by dividing the ratio in column 4 by that in column 7 and multiplying by 2, The fraction exchanged is obtained using equation 1, with Ro equal to 5.60 and Rr equal to 1.28. N R P are the non ribosomal proteins associated with the 30S subunits
grown first in heavy medium iii) (See Materials and Methods), and then in light medium with 3H lysine, were separated in a sucrose gradient, the bulk of the radioactivity was found associated with the light subunits (Fig. l a). This indicates that exchange of proteins among ribosomes is at most a minor phenomenon, a finding which agrees with that of Kaempfer et al. (1968). The fractions containing the heavy and light ribosomal subunits were pooled separately, 1~C lysine labelled subunits were added to each pool for internal standards, and the radioactivity in each protein was measured. The results are shown in Tables 1 and 2 together with the calculation of R and the fraction of each protein exchanged. These data and calculations confirm that for the great majority of proteins the fraction exchanged is less than 20%. We define these proteins to be non exchanging. In contrast to the non exchanging proteins a major fraction (>50%) of six ribosomal proteins (S1, $2, $21, L7/L12, L9, L33) and a minor fraction (20-30%) of five other proteins ($3, L10, L11, L25, L32), were observed to exchange (Tables 1 and 2). In addition a major fraction of the non ribosomal proteins also exchanged, as would be expected of non specifically
adsorbed protein contaminants of ribosomes, which would be expected to associate randomly with heavy and light subunits. Indeed for this class of proteins one would expect to get complete exchange. Examination of the data in Tables 1 and 2 shows that, compared with the recovery of the standard protein $4 or L3, the recoveries of other proteins are variable. Recoveries of S1 and L7/L12 are very low (approximately 1%) while those of $7 (uniquely in this experiment) $21, L9 and L10 are low (approximately 10%). Losses of S1, LT/L12 and L10 occur during centrifugation and alcohol precipitation of the ribosomes. Other losses, particularly of S1, $21 and L9 occur during extraction of the proteins from the ribosomes, two-dimensional gel electrophoresis and extraction of the radioactivity from the gels : compare for instance the data of Hardy (1975) and Weber (1972). These losses make it essential to use internal standard ribosomes in estimating radioactivity. Losses which occur before the addition of the internal standard can affect the estimates of exchange and are discussed later. It was possible that the exchange observed in the experiment shown in Figure 1 a and Tables 1 and 2,
209
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 2. Typical data and calculation of fraction exchanged for the 50S proteins Protein
L1 L2 L3 L4 L5 L6 L7/L12 L9 L10 L11 L 13 L14 L15 L16 L17 L18 L19 L21 L22 L23 L24 L25 L26 L27 L28
L29 L30 L32 L33 NRP
Light 50S
Heavy 50S
3H
14 C
3H
3H
1¢ C
cts/min
cts/min
14C
cts/min
ets/min
3H ~4C
6,322 6,747 6,357 6,255 7,179 6,597 88 361 273 1,904 7,862 3,986 5,187 1,446 4,694 5,333 7,406 1,961 8,705 3,944 8,042 6,149 932 5,427 1,552 3,043 2,847 4,203 2,420 324
4,215 2,615 1,562 1,646 2,182 2,055 316 148 1,092 2,441 2,154 1,098 1,449 459 1,345 1,426 1,954 565 2,473 1,108 2,041 1,798 261 1,395 423 805 802 1,184 964 348
1.5 2.6 4.1 3.8 3.3 3.2 0.3 2.4 0.3 0.8 3.7 3.6 3.6 3.2 3.5 3.7 3.8 3.5 3.5 3.6 3.9 3.4 3.6 3.9 3.7 3.8 3.6 3.6 2.5 0.9
798 1,061 680 647 698 1,324 45 664 57 43l 1,266 825 615 i69 598
1,734 1,00I 486 490 652 774 129 196 379 897 924 573 421 115 407 674 733 222 1,044 539 434 673 88 555 112 414 364 282 317 172
0.5 1.1 1.4 1.3 1.1 1.7 0.4 3.4 0.2 0.5 1.4 1.4 1.5 1.5 1.5 1.4 1.5 1.4 1.3 1.4 1.3 2.2 1.4 1.5 1.4 1.5 1.5 2.2 3.6 1.4
923 1,129 300 1,368 776 573 1,487 125 816 153 604 539 632 1,154 236
R 3H(light) 3H(heavy)
Fraction exchanged
6.5 4.9 5.8 5.8 6.1 3.8 1.6 1.4 3.3 3.3 5.3 5.0 4.9 4.3 4.7 5.5 4.9 5.1 5.4 4.9 6.0 3.1 5.0 5.3 5.4 5.2 4.8 3.2 1.4 1.4
-4 6 0 0 - 2 17 65 71 23 24 3 5 6 12 7 2 6 4 3 6 - 1 26 5 3 3 4 7 25 74 75
(%)
As for TabIe 1, but for the 50S subunits from the gradient shown in Figure la. Here R 0 is 5.81 and Rf is 0.94. Spots of proteins LS, L20, L31 and L34 were not seen on the electropherograms under our conditions
had taken place after lysis of the cells during the sonication, dialysis or centrifugation. The controls for such in vitro exchange, in which heavy non-radioactive bacteria were mixed with light 3H lysine labelled bacteria before lysis showed that exchange of S1, $2, LT/L12, L10, L11, and the non-ribosomal proteins, quantitatively similar to that observed in the experiment, also occurred in vitro. In marked contrast, exchange of $21, L9 and L33 was negligible in the controls, which indicated that the exchange observed for these three proteins (Tables 1 and 2) had occurred in vivo, before lysis of the bacteria. Exchange of $3 and L25 was also negligible in the controls of this experiment, but with these proteins the difference between the in vivo and in vitro exchange was not great, and we were not confident that it was significant. Since the in vivo exchange that we observed for $21, L9 and L33 was reproducible, although with
considerable variation, in a number of early experiments, it was obviously important to show that it was not the result of physiological perturbations due to growth in heavy medium, or to the change of medium. When bacteria are taken out of heavy media ii) or iii) and resuspended in light medium, the growth rate is radically increased, and this may give rise to other physiological effects, such as those associated with the similar growth pattern obtained on changing a culture from growth on a poor to growth on a rich medium. One such effect could be the exchange of ribosomal proteins, or the destruction of certain proteins on the old ribosomes and their replacement by proteins made in the new medium, a process which would be indistinguishable from partial exchange in experiments similar to the one outlined above. To eliminate these possibilities we used a heavy medium (medium i v - s e e Methods) in which the growth rate of the bacteria was not very different from that in
210
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 3. Summary of exchange data for 30S subunit proteins
Table 4. Summary of exchange data for 50S subunit proteins
Protein
Significance of difference
Protein
9 11 11 13 13 10 13
-
13
-
13 9 8 10 9 13 10 10 12 13 9 13
-
L1 4_+ 2 L2 3+ 1 L3 (standard) 0 L4 1_+ 1 L5 1_+ 1 L6 14-+ 2 L7+L12 55-+10 L9 61_+ 3 L10 36_+ 8 Lll 32+ 4 L13 I-+ 1 L14 2_+ 1 L15 5_+ 1 LI6 9_+ 3 L17 3_+ 1 L18 0_+ 1 L19 2_+ 1 L21 1_+ 2 L22 0_+ 2 L23 4_+ 2 L24 - 1_+ 1 L25 21_+ 5 L26 39+_11 L27 4-+ 2 L28 5_+ 2 L29 2+ 1 L30 4_+ 2 L32 20_+ 4 L33 57_+ 6 NRP 101_+21
Experimental
Control
Fraction No. of exchanged samples (0/°)
Fraction No. of exchanged samples (%)
S1 67_+23 $2 69+ 5 $3 25+ 6 S4(standard) 0 $5 9_+ 4 $6 4_+ 6 $7 - 2_+ 2 $8 1_+ 1 $9÷Sll 2_+ 1 SI0 9_+ 4 S12 7_+ 3 S13 i3_+ 4 S14 18_+ 7 S15÷S16 0_+ 2 S17 1_+ 2 S18 1_+ 1 S19 2_+ 1 $20 2_+ 2 $21 67-+11 NRP 121_+20
11 16 15 17 16 12 15 16
16 11 12 14 14 16 11 14 15 15 13 17
103+18 61_+ 7 13_+ 8 0 1_+ 2 3_+ 1 2_+ 2 1_+ 1
1-+ 4_+ 1_+ 7_+ 3_+ 2_+ 2_+ 2_+ 2_+ 1_+ 2_+ 87_+
1 5 2 3 5 1 1 1 2 2 6 9
Exchange of Ribosomal Proteins among the Ribosomes of Escherichia coil W.R. Robertson, S.J. Dowsett*, and S.J.S. Hardy Department of Biology, University of York, Heslington, York, England
Summary. The exchange of ribosomal proteins among ribosomes of E. coli has been measured, using a density label technique. As expected most of the proteins do not exchange appreciably. However a substantial fraction of each of proteins S1, $2, $21, L7/L12, L9, L10, L l l , L26 and L33 is found to exchange, but exchange of S1, $2, LT/L12, L10, L l l and L26 is found to occur in vitro after lysis of the cells, and therefore it is not possible to say whether or not these proteins also exchange in vivo. In contrast $21, L9 and L33 do not exchange after lysis of the cells and we therefore conclude that these proteins exchange in vivo. The maximum level of exchange of $21, L9 and L33 is attained so rapidly that we were unable to show whether or not it was dependent on protein synthesis.
Introduction Almost all the macromolecules which have been identified as components of the ribosomes of E. coli, are stable during exponential growth of the bacteria (Davern and Meselson, 1960; Dennis, 1974). In addition the most recent data indicate that there is probably one molecule of each macromolecule in every mature ribosome, with the single exception of protein L7/L12 (Hardy, 1975), of which there are-probably four molecules per ribosome (Subramanian, 1975). Thus ribosomes are generally considered to be fixed and stable structures. However, it has not yet been clearly demonstrated that ribosomal proteins do not exchange among the ribosomes as they function. The * Present address. Faculty of Medicine, Memorial University of Newfoundland, St. John's, Newfoundland For offprints contact." Dr. S. Hardy, Department of Biology,
University of York, Heslington, York, England
demonstration of such exchange would alter our view of the structure and function of the ribosome, as well as our view of the exchanging macromolecules as structural components of it. Evidence that the bulk of the ribosomal protein does not exchange in vivo was obtained in the elegant experiments of Kaempfer et al. (1968), in which an upper limit of 3% exchange of the total protein was set for the 50S subunits. For the 30S subunits an upper limit was not calculated because of the problem of contamination of the particles with supernatant proteins. However, it is clear from the data that exchange of proteins is at most a minor phenomenon for the small subunit too. Nevertheless, exchange of a few specific proteins is not ruled out by these results, and we have carried out a series of experiments to detect and characterize ribosomal protein exchange in vivo. This was done by using heavy isotopes in a method similar to, but more amenable to quantitative analysis than, that of Kaempfer et al. (1968). Although as expected, the great majority of proteins do not exchange appreciably, we have found to our surprise that there is very rapid, but not complete exchange among the ribosomes, of three proteins, $21, L9 and L33, within the bacterial cells. In addition, we have shown that proteins S1, $2, L7/LI2, L10, L11, L26 and the non-ribosomal proteins found bound to the ribosomes after density gradient centrifugation, exchange to a considerable extent (20% or more) among the ribosomes in vitro after lysis of the bacteria. For these proteins we are not able to say whether or not they exchange in vivo.
Materials and Methods Chemicals
Radioactive lysine was obtained from the Radiochemical Centre, Amersham. The compounds labelled with heavy isotopes, ammo-
206 nium chloride (98.5 atom percent lSN), deuterium oxide (99.8%) and D glucose (74.3 atom percent IaC) were obtained from Prochem, a subsidiary of the British Oxygen Co. Ltd. In early experiments, an algal whole hydrolysate, (1 gram in 3 ml D20) containing 90, 90, and 98 atom percent ~3C, lSN and D respectively, made by Merck Sharp and Dohme was used. This was a generous gift from Linda Randall.
Bacteria, Growth Media, and Conditions of Growth E. coli strain SH7, a stringent derivative of Hfr P4X Met was used in ali experiments. The basic growth medium was M-9 salts containing 50 gg/ml methionine, and glucose as a carbon source. The doubling time of the bacteria in this medium was about 55 rain. Four different 'heavy' growth media were used in the experiments reported here: i) 4 mg/ml glycerol, M-9 salts in D20 with heavy algal hydrolysate diluted 1 part in 50. The doubling time of SH7 in this medium was about 70 rain. ii) 4 mg/ml glucose, M-9 salts without ammonium chloride in D20, 0.2 mg/ml of lSN ammonium chloride. Doubling time about 180 min. iii) The same medium as ii) but with 0.7 mg/ml 13C glucose instead of 12C glucose. Doubling time about 200 rain. iv) The same medium as iii) but with D20 replaced by H20. Doubling time about 65 rain. The procedure for growing the bacteria first in heavy medium and then in light medium was as follows. An overnight culture of SH7 in M-9 salts with limiting glucose (0.6 mg/ml) and reduced ammonium chloride (0.2 mg/ml) was diluted 50-fold into heavy medium and incubated with shaking at 37°C. Growth was monitored using a Bausch and Lomb Spectronic 20 at a wavelength of 450 rim. When the absorbance reached 0.4 (2.5 x 108 cells/ml) part of the culture was added to frozen crushed buffer (0.01 M Tris, 0.003 M succinic acid, 0.01 M MgC12) to use for the controls, and the remainder of the bacteria were collected by centrifugation at 3000xg for 5 rain in a sterile tube, in a rotor prewarmed to 37°C. The bacterial pellet was resuspended in twice the original volume of prewarmed light medium and incubated at 37°C with shaking. At various times after the medium change, 3 to 5 ml (i volume) of the culture were added to 0.35 volumes of frozen crushed buffer and the bacteria were collected by centrifugation at 2°C for 2 rain at 10,000 g. For the controls (see below), portions of a fully light culture grown in parallel, were added to the fully heavy bacteria saved during the change of medium, and the mixed bacteria were collected in the same way. Radioactive labelling of the bacterial proteins was carried out by adding 14C or 3H-lysine to either the heavy or the light medium, without any non-radioactive carrier. The quantity of radioactivity added was usually 100 gCi of 3H or 10 gCi of l¢C-lysine in each 3 to 5 ml portion of the culture harvested. For experiments in which short labelling times with 3H lysine were required, a chase of 40 gg/ml of non-radioactive lysine was added after the labelling period. Labelling was always begun at least one generation before the change of medium, or at least fifteen minutes after, so that the radioactivity would only be incorporated into fully heavy or fully light ribosomal subunits, and not into particles which were in the process of synthesis and assembly during the change of medium. In all experiments a separate culture grown exclusively in light medium, and used as an internal standard for losses of radioactivity (see below), was labelled with either 14C or 3H lysine as described above. Preparation of Ribosomal Subunits, Extraction of Proteins, Electrophoresis and Estimation of Radioactivity These steps were carried out as described previously (Hardy, 1975) with only a few minor changes. The procedure is described briefly
W.R. Robertson et al. : Exchange of Ribosomal Proteins below. Bacterial pellets were resuspended in Tris magnesium buffer and lysed by ultrasonic vibration. The entire lysate was dialysed against a low magnesium buffer (usually 6 x 10 -~ M MgC12) and the subunits were separated on 15 30% (w/v) sucrose gradients in the same buffer. The heavy and light ribosomal subunits were pooled separately, carrier subunits were added together with the radioactive subunits used as an internal standard (see below) and the particles were concentrated by alcohol precipitation at - 2 0 ° C overnight. This step tends to cause loss of protein L7/L12 and to a lesser extent L10. The proteins in the precipitate were extracted with acetic acid and the extract was dialysed into the buffer used in the sample gel of the two dimensional elctrophoresis. Reduction with/~-mercaptoethanol was carried out during this dialysis. The proteins were subjected to two-dimensional electrophoresis on polyacrylamide gels and the ribosomal protein spots were cut out and counted. A piece of gel just below and to the anode (first dimension) side of the sample, containing several stained spots was also cut out and counted. These spots together are referred to as non-ribosomal proteins (NRP), and are assumed to be mainly nonspecifically adsorbed protein contaminants of the ribosomal subunits, although they may also contain protein synthesis factors.
Measurement of the Fraction of a Protein Exchanged After growth in heavy medium followed by growth in light medium, each bacterium contains both heavy and light ribosomal subunits, which are separable from each other on sucrose gradients. Since the bacterial proteins are labelled only in one of the two media, and since most of the ribosomal proteins do not exchange (Kaempfer et al., 1968), the great majority of the radioactivity in the ribosomes is confined to either the light or the heavy subunits (Fig. 1). Separation of the heavy and light subunits was of course never complete, so that even in the absence of exchange, some of the radioactivity present in the labelled subunits was found in the unlabelled subunits, because o f overlap of the heavy and light peaks. Exchange is indicated for a particular protein when the distribution of its radioactivity between heavy and light subunits is different from that of a non-exchanging protein, i.e. a larger proportion of the radioactivity is found in the unlabelled subunits and a smaller proportion is found in the labelled subunits. The fraction of the protein which has exchanged, x, can be calculated from R, the ratio of the protein's radioactivity in the labelled subunits to that in the unlabelled subunits, by using the equation x
(R0 --R) (Rf+ 1) (R o --Rf) ( R + 1)
(1)
where R0 is the value of R when there is no exchange and Rf is the value of R when there is full exchange. Ro is obtained from the value of R for a protein which is assumed not to exchange ($4 for the small ribosomal subunit, L3 for the large ribosomal subunit). Full exchange of a protein would be characterised by an even distribution of its radioactivity over all the ribosomes, and therefore Rf is the ratio of the absorbances of the pools of heavy and light subunits. This was obtained from the gradient traces by first drawing a perpendicular to the base line at the position of the division between the pooled heavy and light subunits, then by cutting out the two peaks generated and weighing them. Here we have assumed that the molar extinction coefficient of heavy and light subunits is the same, and that both have the same protein composition. Equation (1) is easily derived by considering the ribosomes as the sum of two populations, a fraction x among which the protein has exchanged completely, and a fraction 1 - x among which the protein has not exchanged at all. It is assumed in the derivation that the proportion of heavy and light ribosomes in each fraction is the same. It is obvious from the form of the equation that the accuracy of the measurement of x is dependent
W.R. Robertsou et al.: Exchange of Ribosomal Proteins
207 in different cells, they could not exchange components until lysis, after which exchange of proteins should be the same as in the experimental lysates. One further elaboration of this control was carried out in some experiments, to eliminate the possibility that the exchange we observed was promoted in vitro (i.e. after lysis), by a factor present only in lysates of cultures which had been switched from heavy to light medium. In this control, radioactive cells (either heavy or light) were added to a nonradioactive culture which had been subjected to the medium change. Thus the hypothetical factor would be present in the lysate of this raixed culture and should promote exchange. Since this kind of control gave the same results as the simpler one described above, we have combined the results in what follows.
on a large difference between R 0 and Rf. The average value of R f / R o in this work has been 0.22. The measurement of the radioactivity in each protein must be carried out so that losses which occur during extraction of proteins from the subunits, or during electrophoresis of the proteins, have no effect on the value of R. This was done in the following way. To the separate pools of 3H (or 14C) heavy and light ribosomal subunits obtained from the sucrose gradient, equal volumes of 14C (or 3H) labelled ribosomal subunits were added, and the mixture was processed as described. The 3H/~aC ratio of each protein was measured. The ratio of the value of 3H/1¢C of a protein in the heavy subunits to the value of 3H/1¢C of that protein in the light subunits, gave R (or its reciprocal), because an identical quantity of each 1¢C (or 3H) protein had been added to each kind of subunit. The ~4C (or 3H) labelled subunits added in this way were thus an internal standard for losses. In some experiments the volumes of these internal standard ribosomes added to the pooled heavy and light subunits were not equal. In these experiments the ratio of the volumes added is used as a correction factor in calculating R. An examination of Tables 1 and 2, where typical data and the calculations from them are shown, should make these points clearer. Obviously exchange of proteins among ribosomes could occur (and did) after lysis of the bacteria, during the lengthy dialysis before heavy and light subunits were separated, and during the centrifugation which separated them. In all but the first experiment, estimates of this exchange were made by mixing the heavy culture, harvested before the change of medium, with a light culture grown at the same time, one of the two being appropriately labelled. Since in these controls the heavy and light ribosomes were initially
Results
In order to show exchange of ribosomal proteins in vivo it is necessary to be able to separate the ribosomes made during one period of time from those made during another. Then if the old ribosomes contain new proteins characterised by a radioactive label, and vice versa, exchange must be occurring. After growing bacteria first in a medium containing heavy isotopes, and then in a normal medium, old heavy ribosomes can be separated from new light ribosomes on the basis of their mass and density. When the ribosomal subunits from a lysate of cells
b A260 50S ).,t
0.4 50S
f
).:
0.3
30S
30S
I' 1 I I
0.2
p 0.1
I'
j
0-2 i
I
!
I I I
l
f
Fig. l a and b. Sucrose density gradients of radioactive bacterial lysates made after growth in heavy medium followed by growth in light medium. Solid line: A26o, dashed line: radioactivity. Sedimentation from right to left. a Bacteria were grown in heavy medium iii and labelled with 3H lysine in light medium. 45 rain after addition of the label they were harvested, lysed, and dialysed. The dialysate was applied to a 17-ml 15-30% sucrose gradient. The gradient was centrifuged for 15 h at 27,000 rpm at 2°C in the Spmco SW27 rotor and analysed with an ISCO gradient fractionator. Fractions were collected and a small sample of each fraction was counted. b As for a except that the heavy medium lacked D20 and that labelling was carried out with 3H lysine in the heavy medium, about I generation before the change to light medium
208
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 1. Typical data and calculation of fraction exchanged for the 30S proteins Protein
S1 $2 $3 $4 $5 $6 $7 $8 $9+Sll S10 S12 S13 S 14 S15+S16 S17 S18 S19 $20 $21 NRP
Light 30S
Heavy 30S
3H
14C
3H
3H
14C
aH
cts/min
cts/min
~4C
cts/min
cts/min
~4C
426 3,454 3,318 14,660 10,578 3,365 1,081 11,840 9,i78 3,799 856 1,805 2,446 8,818 3,084 1,677 6,136 5,008 550 1,139
391 936 744 2,554 1,859 656 813 1,922 2,180 742 183 308 458 1,487 495 332 1,021 821 250 1,049
1.1 3.7 4.5 5.7 5.7 5.1 1.3 6.2 4.2 5.1 4.7 5.9 5.3 5.9 6.2 5.1 6.0 6.1 2.2 1.1
175 1,463 705 2,517 2,371 418 78 2,141 1,394 729 298 400 666 2,012 350 341 1,146 1,286 300 506
184 429 266 1,228 1,103 201 268 969 1,081 396 135 194 275 991 151 194 613 582 150 544
1.0 3.4 2.6 2.0 2.2 2.1 0.3 2.2 1.3 1.8 2.2 2.1 2.4 2.0 2.3 1.8 1.9 2.2 2.0 0.9
R 3H (light)
Fraction exchanged
3H(heavy)
(%)
2.3 2.2 3.4 5.6 5.3 4.9 9.2 5.6 6.5 5.6 4.2 5.7 4.4 5.8 5.4 5.7 6.4 5.5 2.2 2.3
53 57 27 0 3 6 18 0 7 0 14 1 12 2 1 1 6 1 56 51
-
-
The heavy and light 3H lysine labelled 30S subunits from the gradient shown in Figure 1 a were pooled separately. As an internal standard for the recovery of proteins, two volumes of 1"C lysine labelled 30S subunits were added to the light 30S pool, and one volume to the heavy 30S pool, together with carrier 30S subunits. The proteins were extracted, separated and counted. The data with background subtracted (56 cts/min for 3H, 24 cts/min for 14C) is given in columns 2, 3, 5 and 6. R is obtained by dividing the ratio in column 4 by that in column 7 and multiplying by 2, The fraction exchanged is obtained using equation 1, with Ro equal to 5.60 and Rr equal to 1.28. N R P are the non ribosomal proteins associated with the 30S subunits
grown first in heavy medium iii) (See Materials and Methods), and then in light medium with 3H lysine, were separated in a sucrose gradient, the bulk of the radioactivity was found associated with the light subunits (Fig. l a). This indicates that exchange of proteins among ribosomes is at most a minor phenomenon, a finding which agrees with that of Kaempfer et al. (1968). The fractions containing the heavy and light ribosomal subunits were pooled separately, 1~C lysine labelled subunits were added to each pool for internal standards, and the radioactivity in each protein was measured. The results are shown in Tables 1 and 2 together with the calculation of R and the fraction of each protein exchanged. These data and calculations confirm that for the great majority of proteins the fraction exchanged is less than 20%. We define these proteins to be non exchanging. In contrast to the non exchanging proteins a major fraction (>50%) of six ribosomal proteins (S1, $2, $21, L7/L12, L9, L33) and a minor fraction (20-30%) of five other proteins ($3, L10, L11, L25, L32), were observed to exchange (Tables 1 and 2). In addition a major fraction of the non ribosomal proteins also exchanged, as would be expected of non specifically
adsorbed protein contaminants of ribosomes, which would be expected to associate randomly with heavy and light subunits. Indeed for this class of proteins one would expect to get complete exchange. Examination of the data in Tables 1 and 2 shows that, compared with the recovery of the standard protein $4 or L3, the recoveries of other proteins are variable. Recoveries of S1 and L7/L12 are very low (approximately 1%) while those of $7 (uniquely in this experiment) $21, L9 and L10 are low (approximately 10%). Losses of S1, LT/L12 and L10 occur during centrifugation and alcohol precipitation of the ribosomes. Other losses, particularly of S1, $21 and L9 occur during extraction of the proteins from the ribosomes, two-dimensional gel electrophoresis and extraction of the radioactivity from the gels : compare for instance the data of Hardy (1975) and Weber (1972). These losses make it essential to use internal standard ribosomes in estimating radioactivity. Losses which occur before the addition of the internal standard can affect the estimates of exchange and are discussed later. It was possible that the exchange observed in the experiment shown in Figure 1 a and Tables 1 and 2,
209
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 2. Typical data and calculation of fraction exchanged for the 50S proteins Protein
L1 L2 L3 L4 L5 L6 L7/L12 L9 L10 L11 L 13 L14 L15 L16 L17 L18 L19 L21 L22 L23 L24 L25 L26 L27 L28
L29 L30 L32 L33 NRP
Light 50S
Heavy 50S
3H
14 C
3H
3H
1¢ C
cts/min
cts/min
14C
cts/min
ets/min
3H ~4C
6,322 6,747 6,357 6,255 7,179 6,597 88 361 273 1,904 7,862 3,986 5,187 1,446 4,694 5,333 7,406 1,961 8,705 3,944 8,042 6,149 932 5,427 1,552 3,043 2,847 4,203 2,420 324
4,215 2,615 1,562 1,646 2,182 2,055 316 148 1,092 2,441 2,154 1,098 1,449 459 1,345 1,426 1,954 565 2,473 1,108 2,041 1,798 261 1,395 423 805 802 1,184 964 348
1.5 2.6 4.1 3.8 3.3 3.2 0.3 2.4 0.3 0.8 3.7 3.6 3.6 3.2 3.5 3.7 3.8 3.5 3.5 3.6 3.9 3.4 3.6 3.9 3.7 3.8 3.6 3.6 2.5 0.9
798 1,061 680 647 698 1,324 45 664 57 43l 1,266 825 615 i69 598
1,734 1,00I 486 490 652 774 129 196 379 897 924 573 421 115 407 674 733 222 1,044 539 434 673 88 555 112 414 364 282 317 172
0.5 1.1 1.4 1.3 1.1 1.7 0.4 3.4 0.2 0.5 1.4 1.4 1.5 1.5 1.5 1.4 1.5 1.4 1.3 1.4 1.3 2.2 1.4 1.5 1.4 1.5 1.5 2.2 3.6 1.4
923 1,129 300 1,368 776 573 1,487 125 816 153 604 539 632 1,154 236
R 3H(light) 3H(heavy)
Fraction exchanged
6.5 4.9 5.8 5.8 6.1 3.8 1.6 1.4 3.3 3.3 5.3 5.0 4.9 4.3 4.7 5.5 4.9 5.1 5.4 4.9 6.0 3.1 5.0 5.3 5.4 5.2 4.8 3.2 1.4 1.4
-4 6 0 0 - 2 17 65 71 23 24 3 5 6 12 7 2 6 4 3 6 - 1 26 5 3 3 4 7 25 74 75
(%)
As for TabIe 1, but for the 50S subunits from the gradient shown in Figure la. Here R 0 is 5.81 and Rf is 0.94. Spots of proteins LS, L20, L31 and L34 were not seen on the electropherograms under our conditions
had taken place after lysis of the cells during the sonication, dialysis or centrifugation. The controls for such in vitro exchange, in which heavy non-radioactive bacteria were mixed with light 3H lysine labelled bacteria before lysis showed that exchange of S1, $2, LT/L12, L10, L11, and the non-ribosomal proteins, quantitatively similar to that observed in the experiment, also occurred in vitro. In marked contrast, exchange of $21, L9 and L33 was negligible in the controls, which indicated that the exchange observed for these three proteins (Tables 1 and 2) had occurred in vivo, before lysis of the bacteria. Exchange of $3 and L25 was also negligible in the controls of this experiment, but with these proteins the difference between the in vivo and in vitro exchange was not great, and we were not confident that it was significant. Since the in vivo exchange that we observed for $21, L9 and L33 was reproducible, although with
considerable variation, in a number of early experiments, it was obviously important to show that it was not the result of physiological perturbations due to growth in heavy medium, or to the change of medium. When bacteria are taken out of heavy media ii) or iii) and resuspended in light medium, the growth rate is radically increased, and this may give rise to other physiological effects, such as those associated with the similar growth pattern obtained on changing a culture from growth on a poor to growth on a rich medium. One such effect could be the exchange of ribosomal proteins, or the destruction of certain proteins on the old ribosomes and their replacement by proteins made in the new medium, a process which would be indistinguishable from partial exchange in experiments similar to the one outlined above. To eliminate these possibilities we used a heavy medium (medium i v - s e e Methods) in which the growth rate of the bacteria was not very different from that in
210
W.R. Robertson et al. : Exchange of Ribosomal Proteins
Table 3. Summary of exchange data for 30S subunit proteins
Table 4. Summary of exchange data for 50S subunit proteins
Protein
Significance of difference
Protein
9 11 11 13 13 10 13
-
13
-
13 9 8 10 9 13 10 10 12 13 9 13
-
L1 4_+ 2 L2 3+ 1 L3 (standard) 0 L4 1_+ 1 L5 1_+ 1 L6 14-+ 2 L7+L12 55-+10 L9 61_+ 3 L10 36_+ 8 Lll 32+ 4 L13 I-+ 1 L14 2_+ 1 L15 5_+ 1 LI6 9_+ 3 L17 3_+ 1 L18 0_+ 1 L19 2_+ 1 L21 1_+ 2 L22 0_+ 2 L23 4_+ 2 L24 - 1_+ 1 L25 21_+ 5 L26 39+_11 L27 4-+ 2 L28 5_+ 2 L29 2+ 1 L30 4_+ 2 L32 20_+ 4 L33 57_+ 6 NRP 101_+21
Experimental
Control
Fraction No. of exchanged samples (0/°)
Fraction No. of exchanged samples (%)
S1 67_+23 $2 69+ 5 $3 25+ 6 S4(standard) 0 $5 9_+ 4 $6 4_+ 6 $7 - 2_+ 2 $8 1_+ 1 $9÷Sll 2_+ 1 SI0 9_+ 4 S12 7_+ 3 S13 i3_+ 4 S14 18_+ 7 S15÷S16 0_+ 2 S17 1_+ 2 S18 1_+ 1 S19 2_+ 1 $20 2_+ 2 $21 67-+11 NRP 121_+20
11 16 15 17 16 12 15 16
16 11 12 14 14 16 11 14 15 15 13 17
103+18 61_+ 7 13_+ 8 0 1_+ 2 3_+ 1 2_+ 2 1_+ 1
1-+ 4_+ 1_+ 7_+ 3_+ 2_+ 2_+ 2_+ 2_+ 1_+ 2_+ 87_+
1 5 2 3 5 1 1 1 2 2 6 9
