JouRNAL OF BACTERIOLOGY, July 1979, p. 185-194 0021-9193/79/07-0185/10$02.00/0
Vol. 139, No. 1
Levels of Major Proteins of Escherichia coli During Growth at Different Temperatures SHERRIE L. HERENDEEN, RUTH A. VANBOGELEN, AND FREDERICK C. NEIDHARDT* Department of Microbiology, The University of Michigan, Ann Arbor, Michigan 48109 Received for publication 26 March 1979
The adaptation of Escherichia coli B/r to temperature was studied by measuring the levels of 133 proteins (comprising 70% of the cell's protein mass) during balanced growth in rich medium at seven temperatures from 13.5 to 460C. The growth rate of this strain in either rich or minimal medium varies as a simple function of temperature with an Arrhenius constant of approximately 13,500 cal (ca. 56,500 J) per mol from 23 to 370C, the so-called normal range; above and below this range the growth rate decreases sharply. Analysis of the detailed results indicates that (i) metabolic coordination within the normal (Arrhenius) range is largely achieved by modulation of enzyme activity rather than amount; (ii) the restricted growth that occurs outside this range is accompanied by marked changes in the levels of most of these proteins; (iii) a few proteins are thermometer-like in varying simply with temperature over the whole temperature range irrespective of the influence of temperature on cell growth; and (iv) the temperature response of half of the proteins can be predicted from current information on their metabolic role or from their variation in level in different media at 370C. In general, individual bacteria can grow over been extensive, largely because of the previous a range of approximately 40 (Celsius) degrees inadequacy of methods to measure the levels of (23). For Escherichia coli, a typical mesophile, many enzymes simultaneously. Work has mostly balanced growth can be sustained from 10 to been directed at the sub- and supranormal temalmost 490C (depending at the upper extreme peratures to learn what proteins might be limon the nutrients present in the medium). In the iting growth. An initiation factor in protein synmiddle of this temperature range-from approx- thesis has been implicated as the vulnerable imately 20 to 37°C-the rate of growth of E. coli element both at high (16) and low (2) temperavaries with temperature as though cellular ture. Other work has shown that growth at high growth, no matter what the medium, were a temperature in minimal medium can be resimple chemical process with a temperature stricted by inactivation of a (methionine) biocharacteristic (j) of 12,000 to 14,000 cal (ca. synthetic enzyme (20, 21). In some sense the 50,230 to 58,600 J) per mol (8). Increasing the isolation of mutants with temperature-sensitive temperature above 400C, or decreasing it below enzymes has provided a ready source of possible 200C, leads to progressively slower growth until analogs of the wild-type cell at high and at low temperature (e.g., ref. 7). No systematic study finally growth ceases altogether at 9 or 490C. It is not known how E. coli manages its met- has been made of the levels of individual proabolic affairs so as to maximize growth rate teins as a function of temperature within or between 20 and 370C, nor is it understood what beyond the normal range, but Schaechter et al. prevents this optimization outside this range. A (22) showed that cell volume, mass, RNA, DNA, prevalent and reasonable view is that one or and the number of nuclei per cell in Salmonella more reactions become rate limiting above 370C typhimurium were nearly constant for a given (16) and below 200C (2, 3, 8, 13) as a result of medium at 37 and 250C. With the advent of methods for resolving total the inability of the cell to compensate for thermally (hot or cold) induced changes in confor- cell protein (14, 15) and accurately measuring mation of proteins. The "nornal" or "Arrhen- their levels (e.g., ref. 17), comprehensive studies ius" range in this view is therefore the range are now possible. We recently reported unexover which the cell is successful in adjusting pected changes in transient rates of synthesis of reaction rates that get out of line: adjustment individual proteins after quite modest shifts in that might entail changing the amount of the growth temperature (9). Here we present the first picture of how the levels of the major proproteins involved, their activity, or both. Experimental exploration of this view has not teins of E. coli vary during steady-state growth 185
186
HERENDEEN, VANBOGELEN, AND NEIDHARDT
J. BACTwERIOL.
at different temperatures within and beyond the nornal, Arrhenius range.
MATERIALS AND METHODS Bacterial strain. The E. coli B/r derivative NC3 (11) was used in all experiments. Media. All media used were based on the defined MOPS medium (11) and were made by supplementing the MOPS medium with 0.4% glucose (wt/vol), amino acids (minus leucine; 0.12 mM valine and 0.08 mM isoleucine), five vitamins, and four bases in concentrations given previously (26). At 13.5, 15, 42, and 46°C twice the normal concentration of MOPS was necessary to obtain steady-state growth at the desired cell densities. Bacterial growth. Cells were grown aerobically on a rotary shaker at seven temperatures. The temperature was controlled to within +0.10C with a thermistor probe. All cultures were started at an optical density at 420 nm of 0.01 and grown to an optical density of 1.0, approximately 108 cells per ml. In all cases growth was monitored with flasks growing in paallel to the flask containing radioactive isotope. Radioactive labeling. Steady-state cultures of strain NC3 were grown in rich medium containing ['4C]leucine (356 mCi/mmol; 40 pCi/ml) at each of seven chosen temperatures (13.5, 15, 23, 30, 37, 42, and 46°C). A reference culture was prepared at 37°C in the same medium but with [3H]leucine (502 mCi/mmol; 63 yCi/ml). CelLs were harvested, and portions of cells from the reference culture were mixed with cells of each of the seven experimental cultures. Determination of steady-state levels. Extracts were prepared in the manner described by Blumenthal et al. (1). Portions (20 pl) of extracts were processed by the O'Farrell equilibrium and nonequilibrium gel systems (14, 15). The amounts of 14C and 3H isotopes in the individual protein spots and in the unresolved extracts were measurd, after sample oxidation, in the manner described by Lemaux et al. (9). The '4C:3H value for each protein, divided by the '4C:3H ratio of the unresolved mixture, is the level of that protein in the experimental culture relative to the reference culture.
RESULTS The growth rate of E. coli NC3 (or its close derivative strain NC81) was measured in a series of steady-state cultures at different temperatures in rich and in minimal media. The log of the specific growth rate constant (k) is plotted as a function of the inverse of absolute temperature in Fig. 1. This function appears linear between 21 and 37°C, and has approximately the same ,u value (13,000 to 14,000 cal; ca. 54,400 to 58,600 J) in rich and in minimal medium. On the basis of this information, three temperatures within the linear range were chosen for study: 23, 30, and 37°C. Outside this range four temperatures were chosen to provide celLs with various levels of restricted growth (defined as percentage of the value predicted at each tempera-
45'a 33.
.: 1.0
j
0.5
4
25'~~~~~q
3
21l
3.023 19. Gn
00.4
0.3 LI)
0.21
0.1 1I 3.1 3.2
3.3
3.4
5' 3.5
1000/ T(OK)
FIG. 1. Growth rate of E. coli Bir as a fuinction of temperature. Cultures were grown to steady state at each temperature, and the rate of growth was measured. The logarithm of the growth rate constant, k (h-I), is plotted on the ordinate against the inverse of the absolute temperature (-K) on the abscissa. Individual data points are marked with the corresponding degrees Celsius. (0) Strain NC3 in glucose-rich medium; (0) strain NC81 (identical to strain NC3 except for lacI lacP37 lacP5 thi) in glucose minimal medium.
ture by extrapolating the linear Arrhenius relationship): 13.50C, 56% normal rate; 15°C, 63% normal rate; 420C, 72% normal rate; and 46°C, 28% normal rate. The steady-state levels of individual proteins resolved by the O'Farrell technique were measured at each of the seven temperatures relative to their levels in reference cells growing at 370C. The results are presented in Table 1, which also contains, where available, the identification of the individual proteins, their metabolic class, and their chemical abundance in the cell. We have prepared several additional displays of these data to aid in analysis. In Fig. 2 we have grouped the proteins according to the magnitude of variation in their level within the normal range of temperature, 23 to 370C. This histogram reveals that the amounts of most of the 111 measured proteins change very little throughout the normal Arrhenius temperature range: only 2 change more than 2.5fold, and they change only 4-fold, and 83 of the 111 change less than 1.6-fold. No transcription or translation protein changed more than 1.4fold, and most changed 1.2-fold or less. A similar histogram has been prepared for the 133 proteins for which we have data over the entire range of 13.5 to 460C. Figure 3 displays
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
VOL. 139, 1979
TABLE 1. Steady-state levels of individual proteins Protein
Muxiura)
Protein Identificationb)
A13.0
Metabolic
Regulation Groupc)
Weight fraction of total protein in glucose rich aedium at 37°a a' (ca/eu)
Level at each temperature relative to level at 37 C 13.5 C
15°C
23
L7
Ic
3.05
2.21
-
-
0.91
0.24
0.79
L12
Ic
9.96
0.25
B18.4
Ia4
1. 68*
B18. 7
Ib
1.44*
B20.9
IIbl
7.17*
3.40 1.28 0.95
A165 813.0
B35. 1 840. 7 B46. 7
RNP, a ATPase, a
850.3 B56. 5
Al
groE
861
lb
0.85e
0.63
Ic
5.27
0.89
Ia4
5.30
Ic
9.66
1.15 1.19
Ic
16.47
0.94
1.16 0.73
1.05 0.78 0.69
3. 01*
S1
Ic
26.27
B66
Ic
14.09
C15. 3
Ic
0.68
C22.7
Ic
2.04*
C30. 7
EF-Ts
Ic
2.12
C31.6
EF-Ts
Ic
5.65
0.66 0.63 0.58 0.80 0.84
Ib
2.39
0.42
Ib
5.17
1.48
C44. 2
IIal
3.12*
0.84
C44.6
IIal
4.01
0.44 2.06
C34. 3 C40.
3f)
C56
1. 68*
-
C58. 5
_
2.61*
C60. 7
Ic
4. 18*
C62. 5
Ic
C62. 7
Ic
4.45* 2.18*
C70
Ic
3.27*
C78
Ic
4. 29*
D31.5
Ilbl
1. 36
D32. 5
IIal
1.24
D40.7
Ia3
1.59 0.57 0.49 0.79 0.73 3.22
D44.5
2.71*
D46
2. 55*
D47
2.82*
D47.5
2. 38^
D49.2
1. 31*
0.63 0.53 5.60 2.30 0.89 1.49 1.63 1.12
Ic
1.48
0.88
Ic
4.99
Ic
28.22
2.16 1.02 0.72 0.75
D58. 5
LysS
D74
D84
EF-G
D87. 5 D94
lb
PheS,
D99 D100 D157
LeuS
FRP,0
Ic
1.05
2.45* 2.26
0.62
0.89
0.55 0.79
0.88 0.79
1.01 0.73 0.84 0.49 1.25
1.02 0.68 1.08 0.73 1.18
Ic
84.92
E48. 7
2.69*
_
E58
ArgS
Ic
1.37
E77. 5
GlyS
Ic
2.74
IIb2
4.69
Ic
2.11
IIa2
0.86
0.73 0.49 0.36 0.78 0.64
1.11
0.56
0. 85*
E140
-
F32. 3
Ia3
0.48
P32. 5
IIa2
2.06
1.28 2.47 1.43 1.49 0.73 2.03 1.27
F38
112
2.94
1.85
F39
lb
2.74*
-
6.40*
2.81 0.76
F14.7
Ia4
F24.5
Ib
F24.6
-
Ic
16.20
1.20*
1.41* 4.02
(0.82) (0.82) (1.07)
46°C 0.50 25.03
0.68
0. 37
1.02
1.49
1.11 0.87 0.86
1.17
1.35
1. 38
2.16 0.93
1.03
1.18 1.28
0.96 0.84
0.76
1.65
1.07
0.94 0.97
1.00 0.78 0.55 0.93 1.00 1.02
0.59
0.59 3.21
0.64 2.01
0.63
0.75
0.57 4.03 1.45 0.75 1.03 0.97 1.02 0.85 2.16 1.08 1.03
0.68 1.53
1.32 0.65 0.75 0.99 0.85 0.99 1.82 1.10 1.08
1.03
0.67 0.87 0.76 0.77
1.14
(0.46)
(0.89) (1.19)
0.82 0.95 0.97 0.85 1.42
0.51 0.87
1.24 1.02
1.09
0. 52 0.46 3.06 5.65
1.11 0.55 0.60
0.81 1.90 0.65 0.66 0.29
1.10 0.55 2.73 0.63 0.69 0.56
1.11 0.93
1.34
1.98 1.06 4.35
1.27
1.46
2.06
1.02
1.21
0.93
1.18 1.08 1.08 0.90 0.57 0.94 0.79 1.08 0.81 0.90 0.90
3.06 1.02 1. 50 0.93 0.49 0.24 0.49 0.65
1.02
(1.04)
1.04
(1.09)
1.45 1.12 1.09
(0.97) (1.35) (0.99)
1.08 1.41 0.94
(1.24)
1.09 0.88 0.87 0.94 1.17 1.02 1.32
(0.84)
1.02
1.05 1.69 0.62 7.22
0.85 1.32 0.97 0.83 1.05
0.97
0.74
1.02 1.00
1.12 0.83 1.94
1.02
0.47
1.30
1.19 1.15 0.79
0.82
0.59
1.42
1.00 0.85 1.13 0.88 1.10
1.11
0.61
1.26
1.19
0.70
1.01 0.51
1.25*
(0.59)
42°C 0.82 0.77
1.32
1.04
0.63
(1.16)
300C 1.01 0.89 0.74 1.30
0.62
7.43
_
(1.17) (0.68)
-
Ic
Ib
F42 .2
1.14 0.97 0.63
1.79 1.29 1.20
E24. 8
F30.2
0.61
0.89
E38.5
F26
1.29 1.31
2.05
0.55 1.08 1.41 0.89 1.12
E133
(0.54) (0.99)
0.89
0.50
VailS
0.56 0.99
1.88 0.73
1. 08*
E106
(0.95)
1.05
0.64*
E79
0.69
2.05*
Ib
EF-Tu
-
0.63
Ic
hIal
E42
(1.10)
1.60
Ia3
023
E22.8
(0.96)
-
1.92 0.66 0.68 0.62 0.87 1.26 1.08 0.85
Ic
B65
-
C)
1.27 1.16
0.58 1.36 0.60 0.60
0.50
0.82 0.81 1.24
0. 51
0.93
1.06
1.05 1.49
0.93
1.80
2.63
0.84 0.59
0.93
(0.96)
0.91
0.96
0.59
0.71
(0.87)
0.52
0.79
1.17
0.98 0.71
1.68
1.19 1.16 1.22 0.85
0.82
0.49
0.77 0.69 1.05 1.52
0.92
0.91
l.09
0.66
1.00
0.78
0.58
1.11
1. 37 1.15
2.42 2.40 1.01 0.46 9.03 2.75 1. 16 1.46 0.91
-
-
1.29
0.99
1.15 1.00 0.59 1.61 1.15 1.51 1.52 0.74
1.01
(1.09)
(1.24)
1.20
1.05 1.11 0.82 0.82 1.35 1.30
1.22 0.82
1.23 0.92
0.98 0.65 0.60 1.41
(1.17)
1.07 0.97 2.04
1.28 1.08 1.16 1.01
0.55
0.85 2.47
187
188
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT TABLE 1-Continued Weight
fraction of total protein
Protein
Protein Identifi-
Numbera) cationb)
l4etabolic
in
Regulation
medium
Groupc)
'
at
each temperature relative
at
13.5 C
150C
1.44
1.51
238Ce) 1.43
42°C
46°C
1.11
1.23
1.43
0.79
0.58
0.75
1.15 1.22 1.27
0.78
1.03
0.63
1.23
1.39
1.87
1.22
0.85
0.46
3.53
1.74
1.33
1.28
1.51
0.95
0.70
0.81
F50.3
Ia3
0.24
2.64
1.95
1.15
0.87
F56
Ia2
2.00*
0.68
1.14
0.98
0.95
GluS, 8
F56.2
5.03
Ic
at 37 C
300C
Ic
F48.1
level
1.04
Ib
F43.9
to
37°d
(pg/mg)
10.54*
Ic
F43.8
Level
glucose rich
0.63
(0.87)
0.94
0.86
0.65
2.84
1.86
-
1.69
0.81
0.61
0.98
Ic
1.64l
1.48
1.15
1.25
F60.3
Ia3
0.23
2.16
1.70
1.00
0.93
1.45
2.76
F63.4
Ic
0.56
0.66
0.99
0.78
1.23
0.90
0.97
F63.5
-
1.24
0.57
1.04
0.80
0.77
0.88
0.97
F56.5 F58.5
AspS
(1.13)
F63.8
Ia4
1.78
1.29
0.74
0.83
0.89
1.18
0.73
F64.5
IIal
2.91
0.82
0.94
0.89
0.97
0.89
0.70
Ia3
0.19
5.35
4.06
2.15
1.13
IIa2
0.79
2.20
1.58
1.28
F82.5 F84
(2.20)
1.17
1.59
1.25
1.01
1.84 4.66 0.57
Ia4
0.73
0.28
0.35
0.46
(2.41)
0.51
0.94
F88
IIa3
1.13
2.79
5.71
4.18
(1.34)
1.02
0.88
F99
Ic
24.12
0.69
0.58
0.62
0.71
0.98
0.47
Ic
2.79
1.08
1.23
1.20
1.08
0.78
0.83
F178
Ib
0.63
1.90
2.42
1.58
1.20
-
G25.3
Ib
2.01
1.14
0.80
0.78
1.04
1.27
1.54
G27.2
Ial
0.45*
3.19
3.10
2.30
1.52
0.91
0.30
G29.6 G32.8
-
0.50*
23.90
1.77
1.20
0.94
1.44
0.84
-
0.75
0.65
1.41
0.96
(0.58)
0.66
0.38
0.56
Ic
1.12
1.16
0.77
0.78
(0.75)
0.82
1.07
0.69
-
1.55a
1.79
0.49
0.77
1.08
0.95
2.92
Ib
2.07
5.95
1.40
1.16
1.02
1.01
0.89
F84.1
F107
G36
IleS
PheS,
a
G41
G41.2
(1.26)
(2.31)
0.53
G41.3
Ic
5.49
0.28
0.39
0.47
0.70
0.93
0.25
G41.4
IIa2
0.41
2.30
0.93
1.53
1.34
0.85
2.87
G43.2 G43.8 G43.9 G44 G50.5 G51 G54.7 G57
IIa2
1.01
0.96
1.88
1.75
1.43
1.52
IIa2
1.99
1.42
0.99
1.28
1.19
1.12
2.15
G61
ATPase, B
GlnS
1.08
Ia4
3.54*
0.55
0.42
0.49
0.73
1.00
0.91
Ib
2. 69*
0.57
0.54
0.72
0.86
1.13
1.16
IIbl
6.49
0.94
0.72
0.75
1.09
1.18
1.20
Ia4
6.00
0.99
1.19
1.27
1.21
0.99
1.73
Ic
3.62
1.37
0.89
0.86
0.95
1.02
0.91
Ic
0.54
0.59
0.86
1.12
Ic
1.49
0.93
0.82
0.83
.38
1.68
0.82
0.72
1.66
3.21
3.47
(0.95)
1.08
0.97
0.72
0.85
0.94
0.83
0.78
0.64
1.75
0.95
0.77
1.54
G62.8 G74 G76
Ia4 Ia2
Ia3
-
1.83
1.77
1.17
1.24
1.07
1.24
G78 G93 G97
Ic Ia4
0.88
0.69
0.74
1.26
1.13
1.04
0.97
1.16
0.78
0.80
0.91
1.00
1.13
1.17
Ia3
0.60
2.42
2.68
1.46
(0.89)
0.90
1.12
G117 G127 H52.7 H54.6
Ic
1.79
1.65
1.34
(1.14)
1.09
1.04
1.03
0.99
0.99
1.32
0.36
I14.2
Ib
Ribosomal Pro.
2.34
-
1.08
-
1.01
(2.99)
0.89
1.04
0.98
1.41
Ic
4.07
0.31
0.41
0.62
Ia2
0.53
1.08
2.10
1.89
1.38
1.08
1.52
1.08
0.99
0.49
Ic
-
0.78
-
-
(0.38)
Pro.
Ic
-
0.75
-
-
1.05
0.98
0.55
I14.7
Ribosomal Pro.
Ic
-
0.63
-
-
1.03
0.93
0.43
116.4
Ribosomal Pro. Ribosomal Pro.
Ic
-
0.69
-
-
0.96
1.07
0.54
Ic
-
0.74
-
-
0.99
1.00
0.47
I19.5 I21.1
-
-
2.06
-
-
1.33
1.28
1.02
-
-
0.58
-
-
o.82
0.97
0.50
I21.3
-
3.81*
-
1.06
0.69
0.64
-
-
1.41 0.98
-
I21.4
-
-
0.90
0.75
1.30
814.4 Ribosomal
I17.2
I23.0
I26.0 I26.3 I33.5 I35.2
Ribosomal Pro. Ribosomal Pro. Ribosomal Pro.
Ic
-
0.66
-
-
0.99
1.02
0.49
Ic
-
0.65
-
-
0.98
0.97
0.45 0.48 1.28
0.76
-
-
1.01
-
8.63* 8.58*
0.96
1.50
-
-
1.00
1.04
Ib
-
1.15
-
-
0.95
1.17
1.25
847.2
Ic
0.93
-
-
0.98
1.08
0.85
I49.6
Ic
0.70
-
-
1.12
0.91
0.41
I58.4
-
0.13
-
-
0.90
0.05
0.05
I63.5
Ic
5.60* 1.24* 2.34* 1.71*
5.82
-
-
1.64
0.77
0.42
Ic
a Proteins are numbered as described in Pedersen et al. (17). Letter prefixes refer to the position of the protein in the axis parallel to the isoelectric focusing dimension (A-I indicates progression from acid to base);
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
VOL. 139, 1979
189
TABLE 1-Continued numbers refer to the apparent molecular weight of the protein as determined by the distance moved in the sodium dodecyl sulfate electrophoresis dimension (56.5 indicates molecular weight of 56,500). b Abbreviations used: S-aminoacyl-tRNA synthetase; L7, L12, Si-identified ribosomal proteins; Ribosomal Pro.-proteins known to be ribosomal proteins, but not identified further; RNP, a-RNA polymerase, a subunit; RNP, /?-RNA polymerase, ,B subunit; EF-Ts, EF-Tu, EF-G-protein synthesis elongation factors Ts, Tu, and G; A, groE-the A protein of Subramanian et al. (25), also identified as the product of the groE gene by Drahos and Hendrix (personal communication). 'The proteins have been grouped according to the way in which their relative levels change in cells grown in different media. The numerical values are published in Pedersen et al. (17), and the behavior of each class is shown in Fig. 6. d Most of these values were measured by determining the total radioactivity in individual spots on a gel made from cells grown on uniformly labeled ['4C]glucose, and are taken from Table 1 of Pedersen et al. (17). The values with an asterisk were measured in the same manner, but with cells grown on ['4C]leucine; these values are influenced by any differences in the leucine content of the individual proteins relative to the average for E. coli protein, and therefore should be regarded as approximate. 'The values in parentheses were measured in a separate experiment (strain NC81) in which both the experimental culture (230C) and the reference culture (37°C) were grown in glucose minimal medium. IProtein spot C40.3 has recently been recognized to contain, in most gels, two individual proteins. The data shown reflect the behavior of the dominant protein of this pair, but must be regarded as approximate. 40 All Proteins( 11 1) 23tC to 37c
,o 0
0
n E
3
45'
z
Transcriptional and Translational Proteins(18)
230Cto37*C
10°
1
2
3
4
5
6
Maximum Level of Protein/Minimum Level of Protein
FIG. 2. Distribution of E. coliproteins with respect to magnitude of variation in level within the normal temperature range, 23 to 37°C. For each protein its
maximum level between 23 and 37°C was divided by its minimum level in this range. Proteins were then grouped according to the magnitude of their variation: 1.0-1.2, 1.2-1.4, 1.4-1.6. The upper panel shows all 111 proteins for which data are available; the lower panel shows 18 transcriptional and translational proteins. All data were taken from Table 1.
what is evident to the unaided eye from the autoradiograms of the O'Farrell gels; the relative abundance of individual proteins at 460C is easily distinguished from that at 370C. Of 133 proteins, 83 vary more than 2-fold, 18 vary more than 5-fold, and 9 vary more than 10-fold. The two proteins with the greatest change in level,
G29.6 (30-fold) and A165 (100-fold), are virtually constant throughout the normal temperature range, and are nearly invisible in autoradiograms of gels made from cells grown at these temperatures. Their behavior is presented in Fig. 4. Three proteins, D74, G44, and G27.2, have levels that vary as linear functions of temperature over the entire range of 13.5 to 460C (Fig. 5), despite the multiphasic effect of temperature upon cell growth over this temperature span (Fig. 1). These three "thermometer" proteins, plus 10 (B40.7, D49.2, E24.8, E42, F42.2, F74.5, G61, G127, I35.2, and I47.2) that show less than a 1.5-fold change at any temperature, are the major exceptions to the general rule that proteins vary little in the normal range (23 to 370C) and vary significantly above and below this range. Many of the proteins measured in this study were included in our recent report on the amounts of individual proteins during balanced growth at 370C in five different media (17). In Fig. 6 we present all of the proteins for which both "metabolic regulation" and temperature response data are available. The left side of each panel shows the variation of proteins as a function of growth rate in different media at 370C, and the right side of each panel shows the variation of these same proteins as a function of temperature in rich medium. Of the 133 proteins in the current study, 13 were not measured in our studies of metabolic regulation, and 10 exhibited individual and unique patterns of behavior in the five media and are therefore not displayed in Fig. 6. Of those remaining, 50 belong to three classes of metabolic regulation (Ia4, Ib, and Ic) that display a heterogeneous response to temperature; there is no way of predicting how a protein belonging to these three metabolic
190
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT ,,
All Proteins(133) 13.50C to 46°C
40-
30-
i
20-
.& 104--
0 0. 0
m M
0
a F Th I I/ I
E
l1
.0
5
10
15
20
25
a'
I
f
100
30 I
z
I
10
,Transcriptional and Translational
au -'MI
1015
--I
Proteins(25) 1 3.5°C to 460C
.1a 105 30 100 25 20 15 10 Maximum Level of Protein/Minimum Level of Protein FIG. 3. Distribution of E. coli proteins with respect to magnitude of variation in level over the temperature range 13.5 to 460C. For each protein its maximum level between 13.5 and 46C was divided by its minimum level in this range. Proteins were then grouped according to the magnitude of their variation: 1.0-1.5,1.5-2.0, 2.0-2.5, etc. The upper panel shows all 133 proteins for which data are available; the lower panel shows 25 transcriptional and translational proteins. AU data were taken from Table 1. 1
5
classes will vary in level in response to temperature. For the 60 remaining proteins, one can predict how they will respond to temperatures by knowing either that they belong to metabolic regulation classes Ia2, Ia3, fIal, IIa2, and IIbl, or that they are a transcriptional or translational protein of class Ic. Because this group of 60 proteins includes only 11 of the ribosomal proteins and 10 of the aminoacyl-tRNA synthetases, a reasonable guess is that 115 proteins can now have their temperature response predicted by their metabolic response and function. Translational and transcriptional proteins are reduced in level at both high and low temperatures (Fig. 6). From Table 1, it can be calculated that these proteins (together with the as yet unmeasured remai g ribosomal proteins) constitute 38% of the cell's total protein mass at 37°C, but only 22% at 460C. The summed decrease in these proteins is in fact equal to the increase in just three proteins (B56.5, B66.0, and F24.5) that together constitute 4% of the cell's protein at 370C, but 20% at 460C. It is the changes in these three proteins, in the transcriptional and transational proteins, and in protein A165 (Fig. 4) that produce the different appearance of autoradiograms of 460C cells and 370C cells.
Ribosomal protein L7 (protein A13.0) is an acetylated form of L12 (protein B13.0). The sum of these proteins varies coordinately with the other ribosomal proteins, both at different temperatures (Table 1) and in different media (17). The ratio of the acetylated form to the unacetylated form, however, differs dramatically at low temperature from that at moderate and high temperature (Table 2). As might be expected, there is coordinate control at all temperatures for the measured polypeptide subunits of 30S and 50S ribosomes and of the three enzymes, ATPase, RNA polymerase, and phenylalanyl-tRNA synthetase (Table 1).
DISCUSSION From the data displayed in Fig. 2, it appears that the evolution of protein catalysts in E. coli has led to the ability to grow efficiently between 23 and 37°C with only a small number of proteins having to be adjusted in level. We do not know how close to a value of 14,000 cal (ca. 58,600 J) per mol is the temperature characteristic of each reaction in the cell over this temperature range, but whatever their individual values, the reaction rates can remain coordinated simply by modulations of enzyme activity
VOL. 139, 1979
191
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
fold at 13.50C. Exploring the function of these proteins may help answer the question of what restricts growth at these temperature extremes. A third protein in this group, B56.5, has already been shown to be related to macromolecule synthesis and to be growth essential; it is the product of the groE gene (6; David J. Drahos and Roger W. Hendrix, personal communication). The synthesis of this protein has been shown to be subject to transient, extreme changes in rate following temperature shifts; these transient t(U rates permit rapid attainment (within 10 to 15 6( min) of the new steady-state level of this polypeptide. O'Farrell gels produced from cells grown at 23 and 370C are indistinguishable to the untrained eye. In contrast, the pattern at 460C is quite distinct with respect to the relative abundance of different proteins. The rate of degradation of E. coli protein increases at temperatures above 370C (24), so we have initiated a study to learn whether the pattern at 460C reflects changes in the rates of synthesis of individual proteins, or 45 30 20 25 40 35 15 changes in their rates of degradation. The most Temperature(C) important preliminary result is easily summaFIG. 4. Variation in level of two proteins of E. coli rized: protein degradation is enhanced at 460C as a function of temperature. The level of eachprotein (log scale) relative to itslevel at 37°C is plotted as a compared to 370C, but the levels of proteins at
-
function of temperature. The data were taken from Table 1. (0) Protein G2p.6; (0) protein A165.
(as by allosteric effectors), and changes in enzyme levels are not necessary. Three proteins vary in level in a simple, linear fashion with temperature. This regularity over the range from 13.5 to 460C is not readily explained on the basis of the effect of temperature on promoter activity, on repressor activity, or on metabolic pools, because the absolute rate of growth is quite restricted at the high and low temperatures. A reasonable view is that the levels of these proteins are actively regulated by the cell, and that they are required in amounts that vary simply with temperature. A more precise guess is not now possible. Nine proteins (A165, B56.5, D40.7, F32.3, F82.5, F84.1, G29.6, G41.2, and I63.5) are markedly elevated in level (5- to 25-fold) during growth at either high or low temperature. The impression gained is that these proteins are needed in greatly increased amounts, but whether this is because the proteins are not fully active at these temperatures or because their products are in greater demand cannot be told. Two of these proteins have particularly interesting behavior: A165 is constant and barely detectable at temperatures from 13.5 to 420C, and then abruptly increases 25-fold at 460C, whereas G29.6 is low and nearly constant in level from 460C down to 150C and then abruptly jumps 23-
f
I
.
3.0 -
0) 2.0 H -J
I
.
II
I
30
40
G 27.2
D 74.0
0L)
4r..' *
0)
1.0
0
F---
10
20
50
Temperature (0C) FIG. 5. Variation in level of three "thermometer" proteins of E. coli as a function of temperature. The level of each protein relative to its level at 37°C is plotted as a function of temperature. The data were taken from Table 1. (0) Protein G27.2; (0) protein G44.0; (A) protein D74.0.
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT
192
Group 250 _
Ribosomal Proteins
5.0
_
Terature
Medium
2.0
mparature
T
Medium
20
10
1.0
-----
0.5
0.5
02
02
--
.0
AminoacylktRNA Svnthetases,
Group 1a3
EF-G, Ef-Tu, EF-Ts, RNA Polymerase
Msd.v
50
/ \~~~~~~~~II. ~ ~ ~ ~ ~~~wu
2.0 1.0
A
Medium
T
Tem
nperatureZ C5k(r.0
pwerstrfC
025
025
-, 0
405
1.0
-1 ru 20 2.010 1.5-a
30
40
k(hr1) Temperature(T) A A (hr Temperaoure 5 _ d
so-
~~~~~~0 _50_Memk(hr) 0.5
1.0
Group fIal
3,z
rou la4
2.010
30
40
so
Ib
50 1.
1
Tempereature
20
Temperature(C)
B
Group
T--p-
5 _Medium
1.5
_50Medium
0,2
Tefnperature
0.2
Group lb
Group 1-a2
10
5,0
20~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
10~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
50otein of the same regulatory class in glucose-rich medium each~ 50
Ternperature(C k(hr-iumwere taken from 37°C2.0These data Table1.
at the indicated temperaturesrelativeato Tmeatr(c
k(r5.0D
0 2-
5 0~~~~~~~~~~~~~~~~~~~0
MkdhrnI
Temperaturelt
of grwth rte indifGroupei t 7Crltvet lcs minbma
(r)
D
Tmeauet
(ac(tatemntimalk=d038 Groupm
the same reglatory class i glucose-rich mdium at the inicated temperatresTrelativer t of each proteinof aawr ae rmTbe1
2rCThs0
VOL. 139, 1979
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
TABLE 2. Effect of temperature on acetylation of ribosomalprotein L7/L12 Acetylateda Temp (OC)
(%)
13.5 .................. 30.0 .................. 37.0 .................. 42.0 .................. 46.0 ..................
59.3 28.6 23.4 26.5 28.4
a The levels of L7 and L12 at 13.5, 30, 42, and 460C relative to their level at 37°C (from Table 1) were multiplied by their weight fraction of total protein at 37°C (a' value from Table 1) to find their weight fraction (a') at each of these temperatures. Because L7 and L12 have very nearly identical molecular weight, the percent acetylation of L7/L12 at each temperature was calculated by dividing the a' value (x100) of L7 by the total a' of L7 plus L12.
460C reflect changes in rates of synthesis, not degradation. A gel made from pulse-labeled cells at 460C resembles one from steady-state-labeled cells at 460C, and a gel made from pulse-labeled cells at 370C resembles one from steady-statelabeled cells at that temperature (VanBogelen and Neidhardt, manuscript in preparation). At both extremes of the temperature range of growth, E. coli has lowered levels of ribosomal proteins and other translational and transcriptional proteins (Fig. 6). In this respect, the restricted growth at low and high temperature resembles restricted growth in general, and no diagnostic feature helps us identify what might be limiting the growth rate at these temperature extremes. Minimal medium will not support the growth of E. coli at temperatures above 440C, indicating that inactivation of biosynthetic enzymes determines the maximum temperature of growth under this condition. In the present study, a rich medium (glucose-MOPS supplemented with amino acids, purines and pyrimidines, and vitamins) was deliberately chosen to avoid growth restriction brought about by biosynthetic insufficiency and to require that inactivation of some generally irreplaceable enzyme be responsible for growth restriction. As a brief survey, however, 40 proteins were monitored in cells grown in glucose miniimal medium as 23 and 370C. The results (Table 1) indicate that about 90% of these proteins behave similarly over this span of temperature whether the cells are grown in rich or miniimal medium. The remaining 10% (proteins F84.1, F88, G97, and H54.6) are clearly of interest, and must be suspected of being involved in amphibolic or biosynthetic pathways. The enhanced level of acetylation of ribosomal protein L7/L12 at 13.50C is interesting. Our earlier data (17) indicated that at 370C the
193
level of acetylation varies regularly with growth rate, from 23.4% in glucose-rich medium to 90.4% in acetate minimal medium. This result confirmed those of others, who have shown that acetylation varies with media (5) and growth phase (18, 19). The current study reveals that restricted growth at 460C in rich medium does not result in greatly elevated acetylation, whereas growth at 13.50C does (up to 60% acetylation, Table 2). Nothing now known about the conversion of L12 to L7 by acetylation, or about the function of this reaction, helps us to interpret this finding. Perhaps it is a simple matter of the acetylating enzyme being unusually active at low temperature. Until the function of acetylation is known, however, this would be an unwise conclusion. Protein L7/L12 is believed to participate in the initiation process of protein synthesis (10), and since this initiation may be the rate-limiting step in cell growth at low temperature, a more interesting conclusion about the high acetylation at 13.50C may eventually be possible. Attention should be called to the remarkable behavior of group Ia3 proteins. The eight proteins in this group exhibit highly similar response to growth in different media at 370C; they are the only proteins that have an absolute rate of synthesis that is inversely proportional to growth rate, and their level in rich medium is the lowest of any class. As pointed out earlier (17), proteins with this behavior might be suspected of being subject to passive control. Their promoters are probably weak, and would be saturated with transcription and translation factors at low growth rates where the major proteins involved in protein synthesis are repressed, and would be working below saturation in cells growing rapidly with derepressed ribosomes and other core proteins. The response of group Ia3 proteins to growth restriction at high and low temperature is consistent with this suspicion, for under both conditions where translational and transcriptional proteins are lowered in level, these proteins are elevated. Finally, the homogeneity of the temperature responses of proteins in each of the metabolic regulation groups Ia2, Ia3, Ilal, IIa2, and IIbl and the transcriptional and translational proteins of group Ic may indicate metabolic and physiological similarities of individual proteins in these groups. By this argument the proteins in groups Ia4, Ib, and Ic would seem to be quite heterogeneous. We are currently engaged in a long-range project of identifying specific proteins on these gels. Interpretation of these temperature responses of these heterogeneous groups will undoubtedly be assisted as this work progresses.
194
HERENDEEN, VANBOGELEN, AND NEIDHARDT
ACKNOWLEDGMENTS This work was supported by grant GB26461 from the National Science Foundation and Public Health Service grant GM-17892 from the National Institute of General Medical Science. Martin Smith contributed some of the information on metabolic regulation of proteins, and Philip Bloch provided helpful discusion. LITERATURE CITED 1. Blumenthal, R. M., S. Reeh, and S. Pedersen. 1976. Regulation of transcription factor p and the a subunit of RNA polymerase in Escherichia coi B/r. Proc. Natl. Acad. Sci. U.S.A. 73:2286-2288. 2. Broeze, R. J., C. J. Solomon, and D. H Pope. 1978. Effects of low temperature on in vivo and in vitro protein synthesis in Escherichia coli and Pseudmontu fluorescens. J. Bacteriol. 134:861-874. 3. Das, H. K., and A. Goldstein. 1968. Limited capacity for protein synthess at zero degrees centigade in Escherichia coli. J. Mol. Biol. 31:209-226. 4. Dennis, P. P., and H. Bremer. 1974. Macromolecular composition during steady-state growth of Escherichia coli B/r. J. Bacteriol. 119:270-281. 5. Deuser, E. 1972. Heterogeneity ofribosomal populations in Eacherichia coli cells grown in different media. Mol. Gen. Genet. 119:249-258. 6. Hendrix, RI W., and L Tsui. 1978. Role of the host in virus assembly: cloning of Escherichia coli groE gene and identification of its protein product. Proc. Natl. Acad. Sci. US.A. 75:136-139. 7. Ingraham, J. L 1969. Factors which preclude growth of bacteria at low temperature. Cryobiology 6:188193. 8. Ingraham, J. L., and A. G. Marr. 1963. Control of enzyme biosynthesis at temperature near the minimum for growth of Escherichia coli. Colloq. Int. C. N. R. S. 124:319-328. 9. Lemaux, P. G., S. L Herendeen, P. L Bloch, and F. C. Neidhardt. 1978. Transient rates of synthesis of individual polypeptides in E. coli following temperature ahifts Cell 13:427-434. 10. Lockwood, A. H., and U. Maitra. 1974. The role of ribosomal proteins L7 and L,2 in polypeptide chain initiation in Escherichia coli. J. Biol. Chem. 249:12131218. 11. Neidhardt, F. C., P. L Bloch, S. Pede , and S. Reeh. 1977. Chemical measurement of steady state levels of ten aminoacyltransfer ribonucleic acid synthetase in Ewcherichia coli. J. Bacteriol. 1:378-387. 12. Neidhardt, F. C., P. L Bloch, and D. F. Smith 1974. Culture medium for enterobacteria. J. Bacteriol. 119: 736-747.
J. BACTERIOL.
13. Ng, H., J. L Ingraham, and A. G. Marr.-1962. Damage and derepresson in Esherichia coil rlting from growth at low temperatures. J. Bacteriol. 75:331-339. 14. O'Farrell, P. H. 1975. High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250:40074021. 15. O'Farrell, P. &, H. ML Goodman, and P. H. O'Farrell. 1977. High resolution two dimensional electrophoresis of basic as well as acidic proteins. Cell 12:1133-1142. 16. Patterson, D., and D. Gillespie. 1972. Effect of elevated temperatures on protein synthesis in Eswherichia coli. J. BacterioL 112:1177-1183. 17. Pedersen, S., P. L Bloch, S. Reeh, and F. C. Neidhardt. 1978. Pattern of protein synthesis in E. coli: a catalog of the amount of 140 individual proteins at different growth rates. Cell 14:179-190. 18. R gopal, S., and A. IL Subra_mana. 1974. Alteration in acetylation level of ribosomal protein L12 during growth cycle of Escherichia coil. Proc. Natl. Acad. Sci. U.S.A. 71:2136-2140. 19. Ramagopal S., and A. Subramalian. 1975. Growthdependent regulation in production and utilization of acetylated ribosomal protein L7. J. Mol. Biol. 94:633641. 20. Ron, E. A., and B. D. Davis. 1971. Growth rate of Escherichia coli at elevated temperatures limitation by methionine. J. Bacteriol. 107:391-396. 21. Ron, E. A, and IL Shani. 1971. Growth rate of Escherichia coli at elevated temperature: reversible inhibition of homoserine tranmuccinylase. J. Bacteriol. 107:
397-400. 22. Schaechter, M, 0. Maaloe, and N. 0. Kjeldgaard 1958. Dependency on medium and temperature of cell size and chemical composition during balanced growth of SabnoneUa typhimurium. J. Gen. Microbiol. 19:692606. 23. Stanier, R. Y, E. A. Adelberg and J. gra 1976. The microbial world, p. 306-309. Prentice-Hall, Inc. 24. St. John, A. C., K. Conkln, E. Rosenthal, and A. L Goldbert. 1978. Further evidence for the involvement of charged tRNA and gu ne tetraphosphate in the control of protein degradation in Escherichia coli. J. Biol. Chem. 263:3946-61. 26. Subramanuan, A. RI, C. Haase, and IL Gelse 1976. Isolation and characterization of a growth-cycle-reflecting, high molecular weight protein, assdiated with Eswherichia coli ribosomes. Eur. J. Biochem. 67:691601. 26. Wanner, B. L, RI Kodaira, and F. C. Neidhardt. 1977. Physiological regulation of a decontrolled lac operon. J. Bacteriol. 130:212-222.
Vol. 139, No. 1
Levels of Major Proteins of Escherichia coli During Growth at Different Temperatures SHERRIE L. HERENDEEN, RUTH A. VANBOGELEN, AND FREDERICK C. NEIDHARDT* Department of Microbiology, The University of Michigan, Ann Arbor, Michigan 48109 Received for publication 26 March 1979
The adaptation of Escherichia coli B/r to temperature was studied by measuring the levels of 133 proteins (comprising 70% of the cell's protein mass) during balanced growth in rich medium at seven temperatures from 13.5 to 460C. The growth rate of this strain in either rich or minimal medium varies as a simple function of temperature with an Arrhenius constant of approximately 13,500 cal (ca. 56,500 J) per mol from 23 to 370C, the so-called normal range; above and below this range the growth rate decreases sharply. Analysis of the detailed results indicates that (i) metabolic coordination within the normal (Arrhenius) range is largely achieved by modulation of enzyme activity rather than amount; (ii) the restricted growth that occurs outside this range is accompanied by marked changes in the levels of most of these proteins; (iii) a few proteins are thermometer-like in varying simply with temperature over the whole temperature range irrespective of the influence of temperature on cell growth; and (iv) the temperature response of half of the proteins can be predicted from current information on their metabolic role or from their variation in level in different media at 370C. In general, individual bacteria can grow over been extensive, largely because of the previous a range of approximately 40 (Celsius) degrees inadequacy of methods to measure the levels of (23). For Escherichia coli, a typical mesophile, many enzymes simultaneously. Work has mostly balanced growth can be sustained from 10 to been directed at the sub- and supranormal temalmost 490C (depending at the upper extreme peratures to learn what proteins might be limon the nutrients present in the medium). In the iting growth. An initiation factor in protein synmiddle of this temperature range-from approx- thesis has been implicated as the vulnerable imately 20 to 37°C-the rate of growth of E. coli element both at high (16) and low (2) temperavaries with temperature as though cellular ture. Other work has shown that growth at high growth, no matter what the medium, were a temperature in minimal medium can be resimple chemical process with a temperature stricted by inactivation of a (methionine) biocharacteristic (j) of 12,000 to 14,000 cal (ca. synthetic enzyme (20, 21). In some sense the 50,230 to 58,600 J) per mol (8). Increasing the isolation of mutants with temperature-sensitive temperature above 400C, or decreasing it below enzymes has provided a ready source of possible 200C, leads to progressively slower growth until analogs of the wild-type cell at high and at low temperature (e.g., ref. 7). No systematic study finally growth ceases altogether at 9 or 490C. It is not known how E. coli manages its met- has been made of the levels of individual proabolic affairs so as to maximize growth rate teins as a function of temperature within or between 20 and 370C, nor is it understood what beyond the normal range, but Schaechter et al. prevents this optimization outside this range. A (22) showed that cell volume, mass, RNA, DNA, prevalent and reasonable view is that one or and the number of nuclei per cell in Salmonella more reactions become rate limiting above 370C typhimurium were nearly constant for a given (16) and below 200C (2, 3, 8, 13) as a result of medium at 37 and 250C. With the advent of methods for resolving total the inability of the cell to compensate for thermally (hot or cold) induced changes in confor- cell protein (14, 15) and accurately measuring mation of proteins. The "nornal" or "Arrhen- their levels (e.g., ref. 17), comprehensive studies ius" range in this view is therefore the range are now possible. We recently reported unexover which the cell is successful in adjusting pected changes in transient rates of synthesis of reaction rates that get out of line: adjustment individual proteins after quite modest shifts in that might entail changing the amount of the growth temperature (9). Here we present the first picture of how the levels of the major proproteins involved, their activity, or both. Experimental exploration of this view has not teins of E. coli vary during steady-state growth 185
186
HERENDEEN, VANBOGELEN, AND NEIDHARDT
J. BACTwERIOL.
at different temperatures within and beyond the nornal, Arrhenius range.
MATERIALS AND METHODS Bacterial strain. The E. coli B/r derivative NC3 (11) was used in all experiments. Media. All media used were based on the defined MOPS medium (11) and were made by supplementing the MOPS medium with 0.4% glucose (wt/vol), amino acids (minus leucine; 0.12 mM valine and 0.08 mM isoleucine), five vitamins, and four bases in concentrations given previously (26). At 13.5, 15, 42, and 46°C twice the normal concentration of MOPS was necessary to obtain steady-state growth at the desired cell densities. Bacterial growth. Cells were grown aerobically on a rotary shaker at seven temperatures. The temperature was controlled to within +0.10C with a thermistor probe. All cultures were started at an optical density at 420 nm of 0.01 and grown to an optical density of 1.0, approximately 108 cells per ml. In all cases growth was monitored with flasks growing in paallel to the flask containing radioactive isotope. Radioactive labeling. Steady-state cultures of strain NC3 were grown in rich medium containing ['4C]leucine (356 mCi/mmol; 40 pCi/ml) at each of seven chosen temperatures (13.5, 15, 23, 30, 37, 42, and 46°C). A reference culture was prepared at 37°C in the same medium but with [3H]leucine (502 mCi/mmol; 63 yCi/ml). CelLs were harvested, and portions of cells from the reference culture were mixed with cells of each of the seven experimental cultures. Determination of steady-state levels. Extracts were prepared in the manner described by Blumenthal et al. (1). Portions (20 pl) of extracts were processed by the O'Farrell equilibrium and nonequilibrium gel systems (14, 15). The amounts of 14C and 3H isotopes in the individual protein spots and in the unresolved extracts were measurd, after sample oxidation, in the manner described by Lemaux et al. (9). The '4C:3H value for each protein, divided by the '4C:3H ratio of the unresolved mixture, is the level of that protein in the experimental culture relative to the reference culture.
RESULTS The growth rate of E. coli NC3 (or its close derivative strain NC81) was measured in a series of steady-state cultures at different temperatures in rich and in minimal media. The log of the specific growth rate constant (k) is plotted as a function of the inverse of absolute temperature in Fig. 1. This function appears linear between 21 and 37°C, and has approximately the same ,u value (13,000 to 14,000 cal; ca. 54,400 to 58,600 J) in rich and in minimal medium. On the basis of this information, three temperatures within the linear range were chosen for study: 23, 30, and 37°C. Outside this range four temperatures were chosen to provide celLs with various levels of restricted growth (defined as percentage of the value predicted at each tempera-
45'a 33.
.: 1.0
j
0.5
4
25'~~~~~q
3
21l
3.023 19. Gn
00.4
0.3 LI)
0.21
0.1 1I 3.1 3.2
3.3
3.4
5' 3.5
1000/ T(OK)
FIG. 1. Growth rate of E. coli Bir as a fuinction of temperature. Cultures were grown to steady state at each temperature, and the rate of growth was measured. The logarithm of the growth rate constant, k (h-I), is plotted on the ordinate against the inverse of the absolute temperature (-K) on the abscissa. Individual data points are marked with the corresponding degrees Celsius. (0) Strain NC3 in glucose-rich medium; (0) strain NC81 (identical to strain NC3 except for lacI lacP37 lacP5 thi) in glucose minimal medium.
ture by extrapolating the linear Arrhenius relationship): 13.50C, 56% normal rate; 15°C, 63% normal rate; 420C, 72% normal rate; and 46°C, 28% normal rate. The steady-state levels of individual proteins resolved by the O'Farrell technique were measured at each of the seven temperatures relative to their levels in reference cells growing at 370C. The results are presented in Table 1, which also contains, where available, the identification of the individual proteins, their metabolic class, and their chemical abundance in the cell. We have prepared several additional displays of these data to aid in analysis. In Fig. 2 we have grouped the proteins according to the magnitude of variation in their level within the normal range of temperature, 23 to 370C. This histogram reveals that the amounts of most of the 111 measured proteins change very little throughout the normal Arrhenius temperature range: only 2 change more than 2.5fold, and they change only 4-fold, and 83 of the 111 change less than 1.6-fold. No transcription or translation protein changed more than 1.4fold, and most changed 1.2-fold or less. A similar histogram has been prepared for the 133 proteins for which we have data over the entire range of 13.5 to 460C. Figure 3 displays
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
VOL. 139, 1979
TABLE 1. Steady-state levels of individual proteins Protein
Muxiura)
Protein Identificationb)
A13.0
Metabolic
Regulation Groupc)
Weight fraction of total protein in glucose rich aedium at 37°a a' (ca/eu)
Level at each temperature relative to level at 37 C 13.5 C
15°C
23
L7
Ic
3.05
2.21
-
-
0.91
0.24
0.79
L12
Ic
9.96
0.25
B18.4
Ia4
1. 68*
B18. 7
Ib
1.44*
B20.9
IIbl
7.17*
3.40 1.28 0.95
A165 813.0
B35. 1 840. 7 B46. 7
RNP, a ATPase, a
850.3 B56. 5
Al
groE
861
lb
0.85e
0.63
Ic
5.27
0.89
Ia4
5.30
Ic
9.66
1.15 1.19
Ic
16.47
0.94
1.16 0.73
1.05 0.78 0.69
3. 01*
S1
Ic
26.27
B66
Ic
14.09
C15. 3
Ic
0.68
C22.7
Ic
2.04*
C30. 7
EF-Ts
Ic
2.12
C31.6
EF-Ts
Ic
5.65
0.66 0.63 0.58 0.80 0.84
Ib
2.39
0.42
Ib
5.17
1.48
C44. 2
IIal
3.12*
0.84
C44.6
IIal
4.01
0.44 2.06
C34. 3 C40.
3f)
C56
1. 68*
-
C58. 5
_
2.61*
C60. 7
Ic
4. 18*
C62. 5
Ic
C62. 7
Ic
4.45* 2.18*
C70
Ic
3.27*
C78
Ic
4. 29*
D31.5
Ilbl
1. 36
D32. 5
IIal
1.24
D40.7
Ia3
1.59 0.57 0.49 0.79 0.73 3.22
D44.5
2.71*
D46
2. 55*
D47
2.82*
D47.5
2. 38^
D49.2
1. 31*
0.63 0.53 5.60 2.30 0.89 1.49 1.63 1.12
Ic
1.48
0.88
Ic
4.99
Ic
28.22
2.16 1.02 0.72 0.75
D58. 5
LysS
D74
D84
EF-G
D87. 5 D94
lb
PheS,
D99 D100 D157
LeuS
FRP,0
Ic
1.05
2.45* 2.26
0.62
0.89
0.55 0.79
0.88 0.79
1.01 0.73 0.84 0.49 1.25
1.02 0.68 1.08 0.73 1.18
Ic
84.92
E48. 7
2.69*
_
E58
ArgS
Ic
1.37
E77. 5
GlyS
Ic
2.74
IIb2
4.69
Ic
2.11
IIa2
0.86
0.73 0.49 0.36 0.78 0.64
1.11
0.56
0. 85*
E140
-
F32. 3
Ia3
0.48
P32. 5
IIa2
2.06
1.28 2.47 1.43 1.49 0.73 2.03 1.27
F38
112
2.94
1.85
F39
lb
2.74*
-
6.40*
2.81 0.76
F14.7
Ia4
F24.5
Ib
F24.6
-
Ic
16.20
1.20*
1.41* 4.02
(0.82) (0.82) (1.07)
46°C 0.50 25.03
0.68
0. 37
1.02
1.49
1.11 0.87 0.86
1.17
1.35
1. 38
2.16 0.93
1.03
1.18 1.28
0.96 0.84
0.76
1.65
1.07
0.94 0.97
1.00 0.78 0.55 0.93 1.00 1.02
0.59
0.59 3.21
0.64 2.01
0.63
0.75
0.57 4.03 1.45 0.75 1.03 0.97 1.02 0.85 2.16 1.08 1.03
0.68 1.53
1.32 0.65 0.75 0.99 0.85 0.99 1.82 1.10 1.08
1.03
0.67 0.87 0.76 0.77
1.14
(0.46)
(0.89) (1.19)
0.82 0.95 0.97 0.85 1.42
0.51 0.87
1.24 1.02
1.09
0. 52 0.46 3.06 5.65
1.11 0.55 0.60
0.81 1.90 0.65 0.66 0.29
1.10 0.55 2.73 0.63 0.69 0.56
1.11 0.93
1.34
1.98 1.06 4.35
1.27
1.46
2.06
1.02
1.21
0.93
1.18 1.08 1.08 0.90 0.57 0.94 0.79 1.08 0.81 0.90 0.90
3.06 1.02 1. 50 0.93 0.49 0.24 0.49 0.65
1.02
(1.04)
1.04
(1.09)
1.45 1.12 1.09
(0.97) (1.35) (0.99)
1.08 1.41 0.94
(1.24)
1.09 0.88 0.87 0.94 1.17 1.02 1.32
(0.84)
1.02
1.05 1.69 0.62 7.22
0.85 1.32 0.97 0.83 1.05
0.97
0.74
1.02 1.00
1.12 0.83 1.94
1.02
0.47
1.30
1.19 1.15 0.79
0.82
0.59
1.42
1.00 0.85 1.13 0.88 1.10
1.11
0.61
1.26
1.19
0.70
1.01 0.51
1.25*
(0.59)
42°C 0.82 0.77
1.32
1.04
0.63
(1.16)
300C 1.01 0.89 0.74 1.30
0.62
7.43
_
(1.17) (0.68)
-
Ic
Ib
F42 .2
1.14 0.97 0.63
1.79 1.29 1.20
E24. 8
F30.2
0.61
0.89
E38.5
F26
1.29 1.31
2.05
0.55 1.08 1.41 0.89 1.12
E133
(0.54) (0.99)
0.89
0.50
VailS
0.56 0.99
1.88 0.73
1. 08*
E106
(0.95)
1.05
0.64*
E79
0.69
2.05*
Ib
EF-Tu
-
0.63
Ic
hIal
E42
(1.10)
1.60
Ia3
023
E22.8
(0.96)
-
1.92 0.66 0.68 0.62 0.87 1.26 1.08 0.85
Ic
B65
-
C)
1.27 1.16
0.58 1.36 0.60 0.60
0.50
0.82 0.81 1.24
0. 51
0.93
1.06
1.05 1.49
0.93
1.80
2.63
0.84 0.59
0.93
(0.96)
0.91
0.96
0.59
0.71
(0.87)
0.52
0.79
1.17
0.98 0.71
1.68
1.19 1.16 1.22 0.85
0.82
0.49
0.77 0.69 1.05 1.52
0.92
0.91
l.09
0.66
1.00
0.78
0.58
1.11
1. 37 1.15
2.42 2.40 1.01 0.46 9.03 2.75 1. 16 1.46 0.91
-
-
1.29
0.99
1.15 1.00 0.59 1.61 1.15 1.51 1.52 0.74
1.01
(1.09)
(1.24)
1.20
1.05 1.11 0.82 0.82 1.35 1.30
1.22 0.82
1.23 0.92
0.98 0.65 0.60 1.41
(1.17)
1.07 0.97 2.04
1.28 1.08 1.16 1.01
0.55
0.85 2.47
187
188
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT TABLE 1-Continued Weight
fraction of total protein
Protein
Protein Identifi-
Numbera) cationb)
l4etabolic
in
Regulation
medium
Groupc)
'
at
each temperature relative
at
13.5 C
150C
1.44
1.51
238Ce) 1.43
42°C
46°C
1.11
1.23
1.43
0.79
0.58
0.75
1.15 1.22 1.27
0.78
1.03
0.63
1.23
1.39
1.87
1.22
0.85
0.46
3.53
1.74
1.33
1.28
1.51
0.95
0.70
0.81
F50.3
Ia3
0.24
2.64
1.95
1.15
0.87
F56
Ia2
2.00*
0.68
1.14
0.98
0.95
GluS, 8
F56.2
5.03
Ic
at 37 C
300C
Ic
F48.1
level
1.04
Ib
F43.9
to
37°d
(pg/mg)
10.54*
Ic
F43.8
Level
glucose rich
0.63
(0.87)
0.94
0.86
0.65
2.84
1.86
-
1.69
0.81
0.61
0.98
Ic
1.64l
1.48
1.15
1.25
F60.3
Ia3
0.23
2.16
1.70
1.00
0.93
1.45
2.76
F63.4
Ic
0.56
0.66
0.99
0.78
1.23
0.90
0.97
F63.5
-
1.24
0.57
1.04
0.80
0.77
0.88
0.97
F56.5 F58.5
AspS
(1.13)
F63.8
Ia4
1.78
1.29
0.74
0.83
0.89
1.18
0.73
F64.5
IIal
2.91
0.82
0.94
0.89
0.97
0.89
0.70
Ia3
0.19
5.35
4.06
2.15
1.13
IIa2
0.79
2.20
1.58
1.28
F82.5 F84
(2.20)
1.17
1.59
1.25
1.01
1.84 4.66 0.57
Ia4
0.73
0.28
0.35
0.46
(2.41)
0.51
0.94
F88
IIa3
1.13
2.79
5.71
4.18
(1.34)
1.02
0.88
F99
Ic
24.12
0.69
0.58
0.62
0.71
0.98
0.47
Ic
2.79
1.08
1.23
1.20
1.08
0.78
0.83
F178
Ib
0.63
1.90
2.42
1.58
1.20
-
G25.3
Ib
2.01
1.14
0.80
0.78
1.04
1.27
1.54
G27.2
Ial
0.45*
3.19
3.10
2.30
1.52
0.91
0.30
G29.6 G32.8
-
0.50*
23.90
1.77
1.20
0.94
1.44
0.84
-
0.75
0.65
1.41
0.96
(0.58)
0.66
0.38
0.56
Ic
1.12
1.16
0.77
0.78
(0.75)
0.82
1.07
0.69
-
1.55a
1.79
0.49
0.77
1.08
0.95
2.92
Ib
2.07
5.95
1.40
1.16
1.02
1.01
0.89
F84.1
F107
G36
IleS
PheS,
a
G41
G41.2
(1.26)
(2.31)
0.53
G41.3
Ic
5.49
0.28
0.39
0.47
0.70
0.93
0.25
G41.4
IIa2
0.41
2.30
0.93
1.53
1.34
0.85
2.87
G43.2 G43.8 G43.9 G44 G50.5 G51 G54.7 G57
IIa2
1.01
0.96
1.88
1.75
1.43
1.52
IIa2
1.99
1.42
0.99
1.28
1.19
1.12
2.15
G61
ATPase, B
GlnS
1.08
Ia4
3.54*
0.55
0.42
0.49
0.73
1.00
0.91
Ib
2. 69*
0.57
0.54
0.72
0.86
1.13
1.16
IIbl
6.49
0.94
0.72
0.75
1.09
1.18
1.20
Ia4
6.00
0.99
1.19
1.27
1.21
0.99
1.73
Ic
3.62
1.37
0.89
0.86
0.95
1.02
0.91
Ic
0.54
0.59
0.86
1.12
Ic
1.49
0.93
0.82
0.83
.38
1.68
0.82
0.72
1.66
3.21
3.47
(0.95)
1.08
0.97
0.72
0.85
0.94
0.83
0.78
0.64
1.75
0.95
0.77
1.54
G62.8 G74 G76
Ia4 Ia2
Ia3
-
1.83
1.77
1.17
1.24
1.07
1.24
G78 G93 G97
Ic Ia4
0.88
0.69
0.74
1.26
1.13
1.04
0.97
1.16
0.78
0.80
0.91
1.00
1.13
1.17
Ia3
0.60
2.42
2.68
1.46
(0.89)
0.90
1.12
G117 G127 H52.7 H54.6
Ic
1.79
1.65
1.34
(1.14)
1.09
1.04
1.03
0.99
0.99
1.32
0.36
I14.2
Ib
Ribosomal Pro.
2.34
-
1.08
-
1.01
(2.99)
0.89
1.04
0.98
1.41
Ic
4.07
0.31
0.41
0.62
Ia2
0.53
1.08
2.10
1.89
1.38
1.08
1.52
1.08
0.99
0.49
Ic
-
0.78
-
-
(0.38)
Pro.
Ic
-
0.75
-
-
1.05
0.98
0.55
I14.7
Ribosomal Pro.
Ic
-
0.63
-
-
1.03
0.93
0.43
116.4
Ribosomal Pro. Ribosomal Pro.
Ic
-
0.69
-
-
0.96
1.07
0.54
Ic
-
0.74
-
-
0.99
1.00
0.47
I19.5 I21.1
-
-
2.06
-
-
1.33
1.28
1.02
-
-
0.58
-
-
o.82
0.97
0.50
I21.3
-
3.81*
-
1.06
0.69
0.64
-
-
1.41 0.98
-
I21.4
-
-
0.90
0.75
1.30
814.4 Ribosomal
I17.2
I23.0
I26.0 I26.3 I33.5 I35.2
Ribosomal Pro. Ribosomal Pro. Ribosomal Pro.
Ic
-
0.66
-
-
0.99
1.02
0.49
Ic
-
0.65
-
-
0.98
0.97
0.45 0.48 1.28
0.76
-
-
1.01
-
8.63* 8.58*
0.96
1.50
-
-
1.00
1.04
Ib
-
1.15
-
-
0.95
1.17
1.25
847.2
Ic
0.93
-
-
0.98
1.08
0.85
I49.6
Ic
0.70
-
-
1.12
0.91
0.41
I58.4
-
0.13
-
-
0.90
0.05
0.05
I63.5
Ic
5.60* 1.24* 2.34* 1.71*
5.82
-
-
1.64
0.77
0.42
Ic
a Proteins are numbered as described in Pedersen et al. (17). Letter prefixes refer to the position of the protein in the axis parallel to the isoelectric focusing dimension (A-I indicates progression from acid to base);
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
VOL. 139, 1979
189
TABLE 1-Continued numbers refer to the apparent molecular weight of the protein as determined by the distance moved in the sodium dodecyl sulfate electrophoresis dimension (56.5 indicates molecular weight of 56,500). b Abbreviations used: S-aminoacyl-tRNA synthetase; L7, L12, Si-identified ribosomal proteins; Ribosomal Pro.-proteins known to be ribosomal proteins, but not identified further; RNP, a-RNA polymerase, a subunit; RNP, /?-RNA polymerase, ,B subunit; EF-Ts, EF-Tu, EF-G-protein synthesis elongation factors Ts, Tu, and G; A, groE-the A protein of Subramanian et al. (25), also identified as the product of the groE gene by Drahos and Hendrix (personal communication). 'The proteins have been grouped according to the way in which their relative levels change in cells grown in different media. The numerical values are published in Pedersen et al. (17), and the behavior of each class is shown in Fig. 6. d Most of these values were measured by determining the total radioactivity in individual spots on a gel made from cells grown on uniformly labeled ['4C]glucose, and are taken from Table 1 of Pedersen et al. (17). The values with an asterisk were measured in the same manner, but with cells grown on ['4C]leucine; these values are influenced by any differences in the leucine content of the individual proteins relative to the average for E. coli protein, and therefore should be regarded as approximate. 'The values in parentheses were measured in a separate experiment (strain NC81) in which both the experimental culture (230C) and the reference culture (37°C) were grown in glucose minimal medium. IProtein spot C40.3 has recently been recognized to contain, in most gels, two individual proteins. The data shown reflect the behavior of the dominant protein of this pair, but must be regarded as approximate. 40 All Proteins( 11 1) 23tC to 37c
,o 0
0
n E
3
45'
z
Transcriptional and Translational Proteins(18)
230Cto37*C
10°
1
2
3
4
5
6
Maximum Level of Protein/Minimum Level of Protein
FIG. 2. Distribution of E. coliproteins with respect to magnitude of variation in level within the normal temperature range, 23 to 37°C. For each protein its
maximum level between 23 and 37°C was divided by its minimum level in this range. Proteins were then grouped according to the magnitude of their variation: 1.0-1.2, 1.2-1.4, 1.4-1.6. The upper panel shows all 111 proteins for which data are available; the lower panel shows 18 transcriptional and translational proteins. All data were taken from Table 1.
what is evident to the unaided eye from the autoradiograms of the O'Farrell gels; the relative abundance of individual proteins at 460C is easily distinguished from that at 370C. Of 133 proteins, 83 vary more than 2-fold, 18 vary more than 5-fold, and 9 vary more than 10-fold. The two proteins with the greatest change in level,
G29.6 (30-fold) and A165 (100-fold), are virtually constant throughout the normal temperature range, and are nearly invisible in autoradiograms of gels made from cells grown at these temperatures. Their behavior is presented in Fig. 4. Three proteins, D74, G44, and G27.2, have levels that vary as linear functions of temperature over the entire range of 13.5 to 460C (Fig. 5), despite the multiphasic effect of temperature upon cell growth over this temperature span (Fig. 1). These three "thermometer" proteins, plus 10 (B40.7, D49.2, E24.8, E42, F42.2, F74.5, G61, G127, I35.2, and I47.2) that show less than a 1.5-fold change at any temperature, are the major exceptions to the general rule that proteins vary little in the normal range (23 to 370C) and vary significantly above and below this range. Many of the proteins measured in this study were included in our recent report on the amounts of individual proteins during balanced growth at 370C in five different media (17). In Fig. 6 we present all of the proteins for which both "metabolic regulation" and temperature response data are available. The left side of each panel shows the variation of proteins as a function of growth rate in different media at 370C, and the right side of each panel shows the variation of these same proteins as a function of temperature in rich medium. Of the 133 proteins in the current study, 13 were not measured in our studies of metabolic regulation, and 10 exhibited individual and unique patterns of behavior in the five media and are therefore not displayed in Fig. 6. Of those remaining, 50 belong to three classes of metabolic regulation (Ia4, Ib, and Ic) that display a heterogeneous response to temperature; there is no way of predicting how a protein belonging to these three metabolic
190
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT ,,
All Proteins(133) 13.50C to 46°C
40-
30-
i
20-
.& 104--
0 0. 0
m M
0
a F Th I I/ I
E
l1
.0
5
10
15
20
25
a'
I
f
100
30 I
z
I
10
,Transcriptional and Translational
au -'MI
1015
--I
Proteins(25) 1 3.5°C to 460C
.1a 105 30 100 25 20 15 10 Maximum Level of Protein/Minimum Level of Protein FIG. 3. Distribution of E. coli proteins with respect to magnitude of variation in level over the temperature range 13.5 to 460C. For each protein its maximum level between 13.5 and 46C was divided by its minimum level in this range. Proteins were then grouped according to the magnitude of their variation: 1.0-1.5,1.5-2.0, 2.0-2.5, etc. The upper panel shows all 133 proteins for which data are available; the lower panel shows 25 transcriptional and translational proteins. AU data were taken from Table 1. 1
5
classes will vary in level in response to temperature. For the 60 remaining proteins, one can predict how they will respond to temperatures by knowing either that they belong to metabolic regulation classes Ia2, Ia3, fIal, IIa2, and IIbl, or that they are a transcriptional or translational protein of class Ic. Because this group of 60 proteins includes only 11 of the ribosomal proteins and 10 of the aminoacyl-tRNA synthetases, a reasonable guess is that 115 proteins can now have their temperature response predicted by their metabolic response and function. Translational and transcriptional proteins are reduced in level at both high and low temperatures (Fig. 6). From Table 1, it can be calculated that these proteins (together with the as yet unmeasured remai g ribosomal proteins) constitute 38% of the cell's total protein mass at 37°C, but only 22% at 460C. The summed decrease in these proteins is in fact equal to the increase in just three proteins (B56.5, B66.0, and F24.5) that together constitute 4% of the cell's protein at 370C, but 20% at 460C. It is the changes in these three proteins, in the transcriptional and transational proteins, and in protein A165 (Fig. 4) that produce the different appearance of autoradiograms of 460C cells and 370C cells.
Ribosomal protein L7 (protein A13.0) is an acetylated form of L12 (protein B13.0). The sum of these proteins varies coordinately with the other ribosomal proteins, both at different temperatures (Table 1) and in different media (17). The ratio of the acetylated form to the unacetylated form, however, differs dramatically at low temperature from that at moderate and high temperature (Table 2). As might be expected, there is coordinate control at all temperatures for the measured polypeptide subunits of 30S and 50S ribosomes and of the three enzymes, ATPase, RNA polymerase, and phenylalanyl-tRNA synthetase (Table 1).
DISCUSSION From the data displayed in Fig. 2, it appears that the evolution of protein catalysts in E. coli has led to the ability to grow efficiently between 23 and 37°C with only a small number of proteins having to be adjusted in level. We do not know how close to a value of 14,000 cal (ca. 58,600 J) per mol is the temperature characteristic of each reaction in the cell over this temperature range, but whatever their individual values, the reaction rates can remain coordinated simply by modulations of enzyme activity
VOL. 139, 1979
191
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
fold at 13.50C. Exploring the function of these proteins may help answer the question of what restricts growth at these temperature extremes. A third protein in this group, B56.5, has already been shown to be related to macromolecule synthesis and to be growth essential; it is the product of the groE gene (6; David J. Drahos and Roger W. Hendrix, personal communication). The synthesis of this protein has been shown to be subject to transient, extreme changes in rate following temperature shifts; these transient t(U rates permit rapid attainment (within 10 to 15 6( min) of the new steady-state level of this polypeptide. O'Farrell gels produced from cells grown at 23 and 370C are indistinguishable to the untrained eye. In contrast, the pattern at 460C is quite distinct with respect to the relative abundance of different proteins. The rate of degradation of E. coli protein increases at temperatures above 370C (24), so we have initiated a study to learn whether the pattern at 460C reflects changes in the rates of synthesis of individual proteins, or 45 30 20 25 40 35 15 changes in their rates of degradation. The most Temperature(C) important preliminary result is easily summaFIG. 4. Variation in level of two proteins of E. coli rized: protein degradation is enhanced at 460C as a function of temperature. The level of eachprotein (log scale) relative to itslevel at 37°C is plotted as a compared to 370C, but the levels of proteins at
-
function of temperature. The data were taken from Table 1. (0) Protein G2p.6; (0) protein A165.
(as by allosteric effectors), and changes in enzyme levels are not necessary. Three proteins vary in level in a simple, linear fashion with temperature. This regularity over the range from 13.5 to 460C is not readily explained on the basis of the effect of temperature on promoter activity, on repressor activity, or on metabolic pools, because the absolute rate of growth is quite restricted at the high and low temperatures. A reasonable view is that the levels of these proteins are actively regulated by the cell, and that they are required in amounts that vary simply with temperature. A more precise guess is not now possible. Nine proteins (A165, B56.5, D40.7, F32.3, F82.5, F84.1, G29.6, G41.2, and I63.5) are markedly elevated in level (5- to 25-fold) during growth at either high or low temperature. The impression gained is that these proteins are needed in greatly increased amounts, but whether this is because the proteins are not fully active at these temperatures or because their products are in greater demand cannot be told. Two of these proteins have particularly interesting behavior: A165 is constant and barely detectable at temperatures from 13.5 to 420C, and then abruptly increases 25-fold at 460C, whereas G29.6 is low and nearly constant in level from 460C down to 150C and then abruptly jumps 23-
f
I
.
3.0 -
0) 2.0 H -J
I
.
II
I
30
40
G 27.2
D 74.0
0L)
4r..' *
0)
1.0
0
F---
10
20
50
Temperature (0C) FIG. 5. Variation in level of three "thermometer" proteins of E. coli as a function of temperature. The level of each protein relative to its level at 37°C is plotted as a function of temperature. The data were taken from Table 1. (0) Protein G27.2; (0) protein G44.0; (A) protein D74.0.
J. BACTERIOL.
HERENDEEN, VANBOGELEN, AND NEIDHARDT
192
Group 250 _
Ribosomal Proteins
5.0
_
Terature
Medium
2.0
mparature
T
Medium
20
10
1.0
-----
0.5
0.5
02
02
--
.0
AminoacylktRNA Svnthetases,
Group 1a3
EF-G, Ef-Tu, EF-Ts, RNA Polymerase
Msd.v
50
/ \~~~~~~~~II. ~ ~ ~ ~ ~~~wu
2.0 1.0
A
Medium
T
Tem
nperatureZ C5k(r.0
pwerstrfC
025
025
-, 0
405
1.0
-1 ru 20 2.010 1.5-a
30
40
k(hr1) Temperature(T) A A (hr Temperaoure 5 _ d
so-
~~~~~~0 _50_Memk(hr) 0.5
1.0
Group fIal
3,z
rou la4
2.010
30
40
so
Ib
50 1.
1
Tempereature
20
Temperature(C)
B
Group
T--p-
5 _Medium
1.5
_50Medium
0,2
Tefnperature
0.2
Group lb
Group 1-a2
10
5,0
20~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
10~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
50otein of the same regulatory class in glucose-rich medium each~ 50
Ternperature(C k(hr-iumwere taken from 37°C2.0These data Table1.
at the indicated temperaturesrelativeato Tmeatr(c
k(r5.0D
0 2-
5 0~~~~~~~~~~~~~~~~~~~0
MkdhrnI
Temperaturelt
of grwth rte indifGroupei t 7Crltvet lcs minbma
(r)
D
Tmeauet
(ac(tatemntimalk=d038 Groupm
the same reglatory class i glucose-rich mdium at the inicated temperatresTrelativer t of each proteinof aawr ae rmTbe1
2rCThs0
VOL. 139, 1979
PROTEIN LEVELS AT DIFFERENT TEMPERATURES
TABLE 2. Effect of temperature on acetylation of ribosomalprotein L7/L12 Acetylateda Temp (OC)
(%)
13.5 .................. 30.0 .................. 37.0 .................. 42.0 .................. 46.0 ..................
59.3 28.6 23.4 26.5 28.4
a The levels of L7 and L12 at 13.5, 30, 42, and 460C relative to their level at 37°C (from Table 1) were multiplied by their weight fraction of total protein at 37°C (a' value from Table 1) to find their weight fraction (a') at each of these temperatures. Because L7 and L12 have very nearly identical molecular weight, the percent acetylation of L7/L12 at each temperature was calculated by dividing the a' value (x100) of L7 by the total a' of L7 plus L12.
460C reflect changes in rates of synthesis, not degradation. A gel made from pulse-labeled cells at 460C resembles one from steady-state-labeled cells at 460C, and a gel made from pulse-labeled cells at 370C resembles one from steady-statelabeled cells at that temperature (VanBogelen and Neidhardt, manuscript in preparation). At both extremes of the temperature range of growth, E. coli has lowered levels of ribosomal proteins and other translational and transcriptional proteins (Fig. 6). In this respect, the restricted growth at low and high temperature resembles restricted growth in general, and no diagnostic feature helps us identify what might be limiting the growth rate at these temperature extremes. Minimal medium will not support the growth of E. coli at temperatures above 440C, indicating that inactivation of biosynthetic enzymes determines the maximum temperature of growth under this condition. In the present study, a rich medium (glucose-MOPS supplemented with amino acids, purines and pyrimidines, and vitamins) was deliberately chosen to avoid growth restriction brought about by biosynthetic insufficiency and to require that inactivation of some generally irreplaceable enzyme be responsible for growth restriction. As a brief survey, however, 40 proteins were monitored in cells grown in glucose miniimal medium as 23 and 370C. The results (Table 1) indicate that about 90% of these proteins behave similarly over this span of temperature whether the cells are grown in rich or miniimal medium. The remaining 10% (proteins F84.1, F88, G97, and H54.6) are clearly of interest, and must be suspected of being involved in amphibolic or biosynthetic pathways. The enhanced level of acetylation of ribosomal protein L7/L12 at 13.50C is interesting. Our earlier data (17) indicated that at 370C the
193
level of acetylation varies regularly with growth rate, from 23.4% in glucose-rich medium to 90.4% in acetate minimal medium. This result confirmed those of others, who have shown that acetylation varies with media (5) and growth phase (18, 19). The current study reveals that restricted growth at 460C in rich medium does not result in greatly elevated acetylation, whereas growth at 13.50C does (up to 60% acetylation, Table 2). Nothing now known about the conversion of L12 to L7 by acetylation, or about the function of this reaction, helps us to interpret this finding. Perhaps it is a simple matter of the acetylating enzyme being unusually active at low temperature. Until the function of acetylation is known, however, this would be an unwise conclusion. Protein L7/L12 is believed to participate in the initiation process of protein synthesis (10), and since this initiation may be the rate-limiting step in cell growth at low temperature, a more interesting conclusion about the high acetylation at 13.50C may eventually be possible. Attention should be called to the remarkable behavior of group Ia3 proteins. The eight proteins in this group exhibit highly similar response to growth in different media at 370C; they are the only proteins that have an absolute rate of synthesis that is inversely proportional to growth rate, and their level in rich medium is the lowest of any class. As pointed out earlier (17), proteins with this behavior might be suspected of being subject to passive control. Their promoters are probably weak, and would be saturated with transcription and translation factors at low growth rates where the major proteins involved in protein synthesis are repressed, and would be working below saturation in cells growing rapidly with derepressed ribosomes and other core proteins. The response of group Ia3 proteins to growth restriction at high and low temperature is consistent with this suspicion, for under both conditions where translational and transcriptional proteins are lowered in level, these proteins are elevated. Finally, the homogeneity of the temperature responses of proteins in each of the metabolic regulation groups Ia2, Ia3, Ilal, IIa2, and IIbl and the transcriptional and translational proteins of group Ic may indicate metabolic and physiological similarities of individual proteins in these groups. By this argument the proteins in groups Ia4, Ib, and Ic would seem to be quite heterogeneous. We are currently engaged in a long-range project of identifying specific proteins on these gels. Interpretation of these temperature responses of these heterogeneous groups will undoubtedly be assisted as this work progresses.
194
HERENDEEN, VANBOGELEN, AND NEIDHARDT
ACKNOWLEDGMENTS This work was supported by grant GB26461 from the National Science Foundation and Public Health Service grant GM-17892 from the National Institute of General Medical Science. Martin Smith contributed some of the information on metabolic regulation of proteins, and Philip Bloch provided helpful discusion. LITERATURE CITED 1. Blumenthal, R. M., S. Reeh, and S. Pedersen. 1976. Regulation of transcription factor p and the a subunit of RNA polymerase in Escherichia coi B/r. Proc. Natl. Acad. Sci. U.S.A. 73:2286-2288. 2. Broeze, R. J., C. J. Solomon, and D. H Pope. 1978. Effects of low temperature on in vivo and in vitro protein synthesis in Escherichia coli and Pseudmontu fluorescens. J. Bacteriol. 134:861-874. 3. Das, H. K., and A. Goldstein. 1968. Limited capacity for protein synthess at zero degrees centigade in Escherichia coli. J. Mol. Biol. 31:209-226. 4. Dennis, P. P., and H. Bremer. 1974. Macromolecular composition during steady-state growth of Escherichia coli B/r. J. Bacteriol. 119:270-281. 5. Deuser, E. 1972. Heterogeneity ofribosomal populations in Eacherichia coli cells grown in different media. Mol. Gen. Genet. 119:249-258. 6. Hendrix, RI W., and L Tsui. 1978. Role of the host in virus assembly: cloning of Escherichia coli groE gene and identification of its protein product. Proc. Natl. Acad. Sci. US.A. 75:136-139. 7. Ingraham, J. L 1969. Factors which preclude growth of bacteria at low temperature. Cryobiology 6:188193. 8. Ingraham, J. L., and A. G. Marr. 1963. Control of enzyme biosynthesis at temperature near the minimum for growth of Escherichia coli. Colloq. Int. C. N. R. S. 124:319-328. 9. Lemaux, P. G., S. L Herendeen, P. L Bloch, and F. C. Neidhardt. 1978. Transient rates of synthesis of individual polypeptides in E. coli following temperature ahifts Cell 13:427-434. 10. Lockwood, A. H., and U. Maitra. 1974. The role of ribosomal proteins L7 and L,2 in polypeptide chain initiation in Escherichia coli. J. Biol. Chem. 249:12131218. 11. Neidhardt, F. C., P. L Bloch, S. Pede , and S. Reeh. 1977. Chemical measurement of steady state levels of ten aminoacyltransfer ribonucleic acid synthetase in Ewcherichia coli. J. Bacteriol. 1:378-387. 12. Neidhardt, F. C., P. L Bloch, and D. F. Smith 1974. Culture medium for enterobacteria. J. Bacteriol. 119: 736-747.
J. BACTERIOL.
13. Ng, H., J. L Ingraham, and A. G. Marr.-1962. Damage and derepresson in Esherichia coil rlting from growth at low temperatures. J. Bacteriol. 75:331-339. 14. O'Farrell, P. H. 1975. High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250:40074021. 15. O'Farrell, P. &, H. ML Goodman, and P. H. O'Farrell. 1977. High resolution two dimensional electrophoresis of basic as well as acidic proteins. Cell 12:1133-1142. 16. Patterson, D., and D. Gillespie. 1972. Effect of elevated temperatures on protein synthesis in Eswherichia coli. J. BacterioL 112:1177-1183. 17. Pedersen, S., P. L Bloch, S. Reeh, and F. C. Neidhardt. 1978. Pattern of protein synthesis in E. coli: a catalog of the amount of 140 individual proteins at different growth rates. Cell 14:179-190. 18. R gopal, S., and A. IL Subra_mana. 1974. Alteration in acetylation level of ribosomal protein L12 during growth cycle of Escherichia coil. Proc. Natl. Acad. Sci. U.S.A. 71:2136-2140. 19. Ramagopal S., and A. Subramalian. 1975. Growthdependent regulation in production and utilization of acetylated ribosomal protein L7. J. Mol. Biol. 94:633641. 20. Ron, E. A., and B. D. Davis. 1971. Growth rate of Escherichia coli at elevated temperatures limitation by methionine. J. Bacteriol. 107:391-396. 21. Ron, E. A, and IL Shani. 1971. Growth rate of Escherichia coli at elevated temperature: reversible inhibition of homoserine tranmuccinylase. J. Bacteriol. 107:
397-400. 22. Schaechter, M, 0. Maaloe, and N. 0. Kjeldgaard 1958. Dependency on medium and temperature of cell size and chemical composition during balanced growth of SabnoneUa typhimurium. J. Gen. Microbiol. 19:692606. 23. Stanier, R. Y, E. A. Adelberg and J. gra 1976. The microbial world, p. 306-309. Prentice-Hall, Inc. 24. St. John, A. C., K. Conkln, E. Rosenthal, and A. L Goldbert. 1978. Further evidence for the involvement of charged tRNA and gu ne tetraphosphate in the control of protein degradation in Escherichia coli. J. Biol. Chem. 263:3946-61. 26. Subramanuan, A. RI, C. Haase, and IL Gelse 1976. Isolation and characterization of a growth-cycle-reflecting, high molecular weight protein, assdiated with Eswherichia coli ribosomes. Eur. J. Biochem. 67:691601. 26. Wanner, B. L, RI Kodaira, and F. C. Neidhardt. 1977. Physiological regulation of a decontrolled lac operon. J. Bacteriol. 130:212-222.
