Escolar Documentos
Profissional Documentos
Cultura Documentos
USA
Vol. 94, pp. 7263–7268, July 1997
Biochemistry
Communicated by Charles S. Peskin, New York University, Hartsdale, NY, April 24, 1997 (received for review August 26, 1996)
ABSTRACT Bacterial chemotaxis is widely studied because Phospho–CheY interacts with switch proteins in the flagellar
of its accessibility and because it incorporates processes that are motors to augment CW rotation, CCW being the default state in
important in a number of sensory systems: signal transduction, the absence of Phospho–CheY. CheZ assists in dissipating CW
excitation, adaptation, and a change in behavior, all in response signals by enhancing the dephosphorylation of CheY. Tar con-
to stimuli. Quantitative data on the change in behavior are trols the flux of phosphate through this circuit by forming a stable
available for this system, and the major biochemical steps in the ternary complex (TaryCheWyCheA) that modulates CheA au-
signal transductionyprocessing pathway have been identified. tophosphorylation in response to changes in ligand occupancy or
We have incorporated recent biochemical data into a mathemat- methylation state. CheA phosphorylates more slowly when the
ical model that can reproduce many of the major features of the receptor is occupied than when it is not.
intracellular response, including the change in the level of Changes in MCP methylation state are responsible for sensory
chemotactic proteins to step and ramp stimuli such as those used adaptation. Tar has four residues that are reversibly methylated
in experimental protocols. The interaction of the chemotactic by a methyltransferase, CheR, and demethylated by a methyles-
proteins with the motor is not modeled, but we can estimate the terase, CheB. CheR activity is unregulated, whereas CheB, like
degree of cooperativity needed to produce the observed gain CheY, is activated by phosphorylation via CheA. Thus, receptor
under the assumption that the chemotactic proteins interact methylation level is regulated by feedback signals from the
directly with the motor proteins. signaling complex, which can probably shift between two confor-
mational states having different rates of CheA autophosphory-
lation. Attractant binding and demethylation shift the equilibrium
Chemotaxis (or more accurately, chemokinesis), the process by
toward a low CheA activity state; attractant release and methyl-
which a cell alters its speed or frequency of turning in response
ation shift the equilibrium toward a high CheA activity state. A
to an extracellular chemical signal, has been most thoroughly receptor complex that is bound to attractant but not highly
studied in the peritrichous bacterium Escherichia coli. E. coli methylated can be thought of as a ‘‘sequestered’’ state (5) which
exhibits sophisticated responses to many beneficial or harmful is unable to autophosphorylate at a significant rate.
chemicals. In isotropic environments the cell swims about in a E. coli can sense and adapt to ligand concentrations that
random walk produced by alternating episodes of counterclock- range over five orders of magnitude (5). In addition, the
wise (CCW) and clockwise (CW) flagellar rotation. CCW rota- machinery for detection and transduction is exquisitely sensi-
tion pushes the cell forward in a fairly straight ‘‘run,’’ CW rotation tive to chemical stimuli. The cell can respond to exponential
triggers a random ‘‘tumble’’ that reorients the cell. The durations increases (‘‘ramps’’) in attractant levels that correspond to
of both runs and tumbles are exponentially distributed, with rates of change in fractional occupancy of only 0.1% per
means of 1.0 s and 0.1 s, respectively (1). In a chemoeffector second. Consequently, the cell can respond even if only a small
gradient, the cell carries out chemotactic migration by extending fraction of its receptors have changed occupancy state during
runs that happen to carry it in favorable directions. Using specific a typical sampling period.
chemoreceptors to monitor its chemical environment, E. coli These experimental observations raise several questions. First,
perceives spatial gradients as temporal changes in attractant or how does the cell achieve the extreme sensitivity that is observed?
repellent concentration. The cell in effect compares its environ- Second, why are there multiple methylation states of the recep-
ment during the past second with the previous 3–4 s and responds tor? In this paper we describe a mathematical model based on
accordingly. Attractant increases and repellent decreases tran- known kinetic properties of Tar and the phosphorelay signaling
siently raise the probability of CCW rotation, or ‘‘bias,’’ and then components that accounts for this exquisite sensitivity. In the
a sensory adaptation process returns the bias to baseline, enabling model, excitation results from the reduction in the autophos-
the cell to detect and respond to further concentration changes. phorylation rate of CheA when Tar is bound to a ligand, and
The response to a small step change in chemoeffector concen- adaptation arises from methylation of the receptor. The disparity
tration in a spatially uniform environment occurs over a 2- to 4-s in the time scales of these processes produces a ‘‘derivative-’’ or
time span (2). Saturating changes in chemoeffector concentration ‘‘temporal-sensing’’ mechanism with respect to the ligand con-
can increase the response time to several minutes (3). centration. The model makes essential use of the multiple meth-
Many bacterial chemoreceptors belong to a family of trans- ylation states to achieve adaptation, and can be used to derive
membrane methyl-accepting chemotaxis proteins (MCPs) (re- quantitative relations between certain key rates and the behav-
viewed in ref. 4). Among the best-studied MCPs is Tar, the E. ioral responses to experimental protocols, and to predict the
coli receptor for the attractant aspartate. Tar has a periplasmic cooperativity needed to achieve the observed gain.
binding domain and a cytoplasmic signaling domain that
communicates with the flagellar motors via a phosphorelay A Qualitative Description of the Response
sequence involving the CheA, CheY, and CheZ proteins (see to Different Stimuli
Fig. 1). CheA, a histidine kinase, first autophosphorylates and We assume that Tar is the only receptor type, that the Tar–
then transfers its phosphoryl group to CheY. CheA–CheW complex does not dissociate, and that Tar, CheA,
and CheW are found only in this complex. We also assume that
The publication costs of this article were defrayed in part by page charge methylation of the multiple sites occurs in a specified order (6, 7).
payment. This article must therefore be hereby marked ‘‘advertisement’’ in
accordance with 18 U.S.C. §1734 solely to indicate this fact.
Abbreviations: CCW, counterclockwise; CW, clockwise; MCP, meth-
© 1997 by The National Academy of Sciences 0027-8424y97y947263-6$2.00y0 yl-accepting chemotaxis protein.
PNAS is available online at http:yywww.pnas.org. ‡To whom reprint requests should be addressed.
7263
7264 Biochemistry: Spiro et al. Proc. Natl. Acad. Sci. USA 94 (1997)
FIG. 4. Response of the system depicted in Fig. 2 to various chemoattractant (aspartate) stimulus protocols: a ramp of rate 0.015 s21 (Left), a
step at t 5 10 s from zero concentration to a concentration (0.11 mM) that is 11% of the Kd for ligand binding (Center), and a step at t 5 100 s
from zero concentration to a concentration (1 mM) that is 1,000 times the Kd for ligand binding (Right). Time series in the top row of figures are
for Che Yp (solid line), CheBp (short dashes), and aspartate (long dashes). The concentrations and rates used are listed and compared to measured
values in Tables 2 and 3. The maximum change in [Che Yp] from baseline for both the ramp and small step responses is 9%. Time series in the
bottom row are the corresponding bias responses, using a Hill coefficient of 11 for binding of both Che Yp and the competitive inhibitor Che Y
to the motor. (Compare the left trace to figure 4A in ref. 12; the center trace to figure 4 in ref. 2, figure 2 in ref. 14, and figure 8 in ref. 13; and
the right trace to figure 8 in ref.3.
S D
respectively, which bind to the motor, M represents motor
db d ln z unbound to either CheY species, M(Yp)j represents motor in
g5 2 y~1 2 y! 12 5 v~1 1 w!. [7]
dy d ln p complex with Yp , and M(Y)k represents motor in complex with
Y. We assume that the binding reactions Eqs. 8 and 9 equili-
Here v [ 2y(1 2 y)dbydy is the gain due only to cooperative brate rapidly, and study only the steady-state quantities
interaction of CheY with the motor, and w [ 2dlnzydlnp is,
for small steps at least, the fractional change in CheZ activity M 1
m; 5 , [10]
per fractional change in receptor occupancy (10). We have M 0 1 1 A j y 1 A k~ 1 2 y ! k
j
assumed here that CheZ does not interact directly with the
motor (19), and acts only by dephosphorylating CheY. This M ~ Y p! j Ajyj
mj ; 5 , [11]
expression makes it clear that the two possible sources of M0 1 1 A j y j 1 A k~ 1 2 y ! k
cooperativity act multiplicatively.
To determine the gain achievable, consider first the factor M~Y!k A k~ 1 2 y ! k
mk ; 5 . [12]
v. We assume that CheY species interact with the motor by M0 1 1 A j y j 1 A k~ 1 2 y ! k
binding to the flagellar switch, and we allow for the possibility
that both CheYp and unphosphorylated CheY may bind (19). Here M0 is the total concentration of motor, and Aj and Ak are
Interaction with the motor may involve cooperativity, either the ratios of the association and dissociation constants for
CheYp and unphosphorylated CheY, respectively.
Table 2. Conserved quantities used in the model (see ref. 3) We assume that the unbound motor M exists in CCW mode
Species T R B Y Z with probability one (19), and that the Mj and Mk states have
probabilities fj and fk, respectively, of being in CCW mode.
Concentration, mM 8 0.3 1.7 20 40
Then the bias is given by
Biochemistry: Spiro et al. Proc. Natl. Acad. Sci. USA 94 (1997) 7267
Table 3. Rates used in the model, and corresponding values from the literature
Reaction Rate constant Value Literature value Ref.
T 2R 3 T 3 1 R k 1c 0.17 s21 0.17 s21 12
T 3R 3 T 4 1 R k 2c 0.1k 1c .0.02k 1c* 13
LT 2R 3 LT 3 1 R k 3c 30k 1c 15k 1c 2 30k 1c† 13
LT 3R 3 LT 4 1 R k 4c 30k 2c 15k 2c 2 30k 2c† 13
T n 1 R º T nR k 1byk 1a, . . . , k 4byk 4a 1.7 mM 1.7 mM 12
T3 1 Bp 3 T2 1 Bp k 21 4 3 10 5 M21zs21 3 3 10 4 M21zs21 11
T4 1 Bp 3 T3 1 Bp k 22 3 3 10 4 M21zs21 .3.0k 21‡ 13
LT 3 1 B p 3 LT 2 1 B p k 23 k 21 k 21§ 13
LT 4 1 B p 3 LT 3 1 B p k 24 k 22 k 22§ 13
L 1 T 3 LT k5, k 6, k 7 7 3 10 7 M21zs21 7 3 10 7 M21zs21 3¶
LT 3 L 1 T k 25, k 26, k 27 70 s21 70 s21 3¶
T 2 3 T 2p k8 15 s21 17 s21 14
T 3 3 T 3p k9 3k 8
T 4 3 T 4p k 10 3.2k 8
LT 2 3 LT 2p k 11 0
LT 3 3 LT 3p k 12 1.1k 8
LT 4 3 LT 4p k 13 0.72k 10 k 10 15
B 1 T np 3 B p 1 T n kb 8 3 10 5 M21zs21 8 3 10 5 M21zs21 11i
Y 1 T np 3 Y p 1 T n ky 3 3 10 7 M21zs21 3 3 10 7 M21zs21 14
Bp 3 B k 2b 0.35 s21 0.35 s21 14
Yp 1 Z 3 Y 1 Z k 2y 5 3 10 5 M21zs21 5 3 10 5 M21zs21 14
*Methylation rates of different methylation sites vary by a factor of up to 50.
†Ligand binding increases methylation rates of different methylation sites by a factor of 15–30.
‡Demethylation rates of different methylation sites vary by a factor of up to 3.
§Ligand binding has little effect on demethylation rate.
£Estimated from figure 10 in ref. 3.
i
Estimated from figure 3 in ref. 11.
1 1 f j A j y j 1 f k A k~ 1 2 y ! k g
b 5 m 1 f jm j 1 f km k 5 .[13] w$4 2 1. [18]
1 1 A j y j 1 A k~ 1 2 y ! k n
One can show that the optimal values of fj and fk are 0 and 1, Thus, without cooperative effects at the switch, we must have
respectively (10), and using these v is given by w $ 10, and for moderate cooperativity at the switch, for
example n 5 6 (20), we must have w $ 0.8. A gain of 6 (17)
v ~ y# ! 5
1
4
Sj ~ 1 2 y# ! 1 ky#
ak
1 1 ak
. D [14]
requires w $ 23 and w $ 3 in these respective cases.
Finally, we note that an additional cooperative step probably
occurs in the interactions between flagella, since the bias of an
where ak 5 Ak(1 2 y)k and y is the baseline CheYp concen- individual flagellum in the absence of stimulation [.0.64 (2)]
tration. Because this expression is monotonically increasing in is less than that of a swimming cell [.0.9, assuming mean run
ak, v(y# ) is maximized with respect to ak when ak is as large as and tumble durations of 1.1 s and 0.14 s, respectively (21)]. We
possible, indicating that both CheY species should have strong can estimate what the threshold number of flagella might be
for this cooperative interaction if we adopt the ‘‘voting hy-
binding affinities for the flagellar switch. In the limit as ak 3
pothesis’’ (21, 22), whereby the biases of the individual flagella
`, Eq. 14 becomes
are identical and independent, and the probability that the
1 flagella will form a bundle is one when the number of flagella
v ~ y# ! 5 ~ j ~ 1 2 y# ! 1 ky# ! . [15] turning CCW equals or exceeds a threshold u and zero
4 otherwise. Then the bias B of the cell is given by
Finally, maximizing Eq. 15 with respect to y# (subject to y# [ [0, 1])
OS D
N
shows that the maximum possible gain at the flagellar switch is N j
B5 b ~ 1 2 b ! N2j, [19]
j
j5u
1
v ~ y# ! 5 n, n ; max~ j, k ! . [16] N
4 (not B 5 ( j5u b j(1 2 b) N2j, as claimed by Weis and Koshland
(22)], where N is the total number of flagella and b is the bias
Thus, a gain of 2.7 requires a Hill coefficient of at least 11 if of an individual flagellum. For b 5 0.64 and either n 5 6 or
the only cooperativity is in the interaction of CheY species with n 5 8 (23), we find that B . 0.9 when u 5 Ny2, and thus a
the switch. In Fig. 4 D–F we show the bias resulting from the simple majority rules.
CheYp response in Fig. 4 A–C using a Hill coefficient of 11 for
Discussion
both CheY species (j 5 k 5 11).
The full expression for the gain now becomes In our model of aspartate signal transduction via Tar, excitation
is the result of the reduced autophosphorylation rate of the
1 ligand-bound state of the receptor, which reduces the level of
g# n~1 1 w!, [17]
4 CheAp and CheYp, thereby reducing the tumbling rate. Adapta-
tion results from the enhanced rate of methylation of bound states
and so the fractional change in CheZ activity relative to the and the fact that methylation increases the rate of autophosphor-
fractional change in receptor occupancy is ylation, which returns CheAp and CheYp to their prestimulus
7268 Biochemistry: Spiro et al. Proc. Natl. Acad. Sci. USA 94 (1997)
levels. The results we present demonstrate that the model can one another, a possibility raised by the existence of a ‘nose spot’
reproduce the experimentally observed responses to both step of elevated receptor density (26).
increases and slow ramps using experimentally determined values We have shown that the effects of cooperativity at different
for most of the parameters. The disparity between the time scale locations in the signal transduction pathway are likely to be
of excitation, which is fast, and that of adaptation, which is slow, multiplicative, and thus the total system gain may be the result
implies that the transduction system can function as a ‘‘derivative of moderate cooperativity occurring at two or more of the
sensor’’ with respect to the ligand concentration: the DC com- above-mentioned locations. For example, a Hill coefficient of
ponent of a signal is ultimately ignored if it is not too large. This 6 at the flagellar switch (20) and a moderate degree of CheZ
provides a bacterium with a temporal sensing mechanism without modulation are sufficient to produce the desired gain.
the need for any type of memory beyond that embodied in the Although the model analyzed herein is specific to signal
disparity in time scales between excitation and adaptation. transduction in bacterial chemotaxis, the structure of the
In ref. 10 we show that, as is seen experimentally (16), the network in Fig. 2 is very similar to those of other signal
magnitude of the response to a slow ramp is an increasing transduction processes, such as those modeled in refs. 27–29.
function of ramp rate, and so we may understand the ramp This suggests that the type of analysis done here will have
response as the result of a difference between the rate at which applicability to other systems. A more complete discussion of
receptors enter the sequestered state via increases in receptor aspects of adaptation not treated here, including an evaluation
occupancy, and the rate at which they exit via transitions between of general models of adaptation, is given in ref. 10.
methylation states. Steeper ramps result in larger differences
between the entrance and exit rates, and thus in larger responses. This work was supported in part by National Institutes of Health
The threshold ramp response (16) occurs when these rates are Grant GM29123 to H.G.O. and National Institutes of Health Grant
equal. In response to a ramp, the deviation from baseline of the GM19559 to J.S.P.
level of sequestered receptor is an approximately linear function
of ramp rate. If the downstream steps in the transduction pathway 1. Berg, H. C. (1990) Cold Spring Harbor Symp. Quant. Biol. 55,
539–545.
operate in a linear range, then the bias response will in turn be 2. Block, S. M., Segall, J. E. & Berg, H. C. (1982) Cell 31, 215–226.
approximately linear in the ramp rate, consistent with the finding 3. Stock, J. B. (1994) in Regulation of Cellular Signal Transduction
of Block et al. (16). Pathways by Desensitization and Amplification, eds. Sibley, D. R.
Three methylation states were necessary to accurately repro- & Houslay, M. D. (Wiley, New York), pp. 3–24.
duce the responses to both step and ramp stimuli, because these 4. Stock, J. B. & Surette, M. G. (1996) Escherichia coli and Salmo-
responses place competing restrictions on the effective methyl- nella: Cellular and Molecular Biology (Am. Soc. Microbiol.,
ation and demethylation rates. For step stimuli given at a baseline Washington, DC).
attractant concentration of zero (17, 18), adaptation is dominated 5. Bourret, R. B., Borkovich, K. A. & Simon, M. I. (1991) Annu.
by the fast transitions between the two lowest methylation states, Rev. Biochem. 60, 401–441.
which provide a sufficiently fast adaptation response. At the 6. Engström, P. & Hazelbauer, G. L. (1980) Cell 20, 165–171.
7. Springer, M. S., Zanolari, B. & Pierzchala, P. A. (1982) J. Biol.
higher attractant concentrations of ramp experiments (16, 18),
Chem. 257, 6861–6866.
receptors are on average more highly methylated, and so the 8. Boyd, A. & Simon, M. I. (1980) J. Bacteriol. 143, 809–815.
slower transitions between the two highest methylation states are 9. Bray, D., Bourret, R. B. & Simon, M. I. (1993) Mol. Biol. Cell 4,
more prominent, providing a significant ramp response. 469–482.
The sensitivity, or gain, of the signal transduction system is 10. Spiro, P. A. (1997) Ph.D. dissertation (Univ. of Utah, Salt Lake
found experimentally to be quite high, but the source of this City).
sensitivity is unknown. We have shown that the sensitivity ob- 11. Lupas, A. & Stock, J. B. (1989) J. Biol. Chem. 264, 17337–17342.
served in response to ramp stimuli (16, 18) is consistent with the 12. Simms, S. A., Stock, A. M. & Stock, J. B. (1987) J. Biol. Chem.
moderate estimate of 6 for the maximum gain of the system (17) 262, 8537–8543.
obtained from step experiments, but that cooperativity equivalent 13. Terwilliger, T. C., Wang, J. Y. & Koshland, D. E. (1986) J. Biol.
Chem. 261, 10814–10820.
to a Hill coefficient on the order of 11 is necessary to produce the
14. Bray, D. & Bourret, R. B. (1995) Mol. Biol. Cell 6, 1367–1380.
desired gain. There are several potential sources of cooperativity. 15. Borkovich, K. A., Alex, L. A. & Simon, M. I. (1992) Proc. Natl.
Binding of CheYp to the flagellar switch is thought to be Acad. Sci. USA 89, 6756–6760.
cooperative (20), and we have shown that competitive inhibition 16. Block, S. M., Segall, J. E. & Berg, H. C. (1983) J. Bacteriol. 154,
by unphosphorylated CheY can enhance the sensitivity. Because 312–323.
the motor contains 26 or 27 subunits of switch protein FliF (19), 17. Khan, S., Castellano, F., Spudich, J., McCray, J., Goody, R., Reid,
it is conceivable that high cooperativity occurs either in binding G. & Trentham, D. (1993) Biophys. J. 65, 2368–2382.
to the switch or else in interactions among switch subunits bound 18. Segall, J. E., Block, S. M. & Berg, H. C. (1986) Proc. Natl. Acad.
individually to CheYp. CheZ phosphatase activity may also be Sci. USA 83, 8987–8991.
modulated in a manner that exhibits cooperativity. The system 19. Macnab, R. M. (1995) in Two-Component Signal Transduction,
eds. Hoch, J. A. & Silhavy, T. J. (Am. Soc. Microbiol., Washing-
gain could be significant if CheZ activity were positively corre-
ton, DC), pp. 181–199.
lated with the level of sequestered CheA, because small fractional 20. Kuo, S. C. & Koshland, D. E. (1989) J. Bacteriol. 171, 6279–6287.
changes in receptor occupancy can correspond to large fractional 21. Ishihara, A., Segall, J. E., Block, S. M. & Berg, H. C. (1983) J.
changes in sequestered CheA. As an example, the cells possess a Bacteriol. 155, 228–237.
short form of CheA (CheAs) which may bind several molecules 22. Weis, R. M. & Koshland, D. E. (1990) J. Bacteriol. 172, 1099–
of CheZ (24). If it does so when the associated transducer is 1105.
unbound to attractant or highly methylated, and releases these 23. Stewart, R. C. & Dahlquist, F. W. (1987) Chem. Rev. 87, 997–
molecules into the cytoplasm upon attractant binding, the effec- 1025.
tive concentration of CheZ will be raised following the latter 24. Amsler, C. D. & Matsumura, P. (1995) in Two-Component Signal
event. Alternatively, the CheAs–CheZ complex may amplify Transduction, eds. Hoch, J. A. & Silhavy, T. J. (Am. Soc. Micro-
CheZ phosphatase activity [Wang (1996) cited in ref. 25], which biol., Washington, DC), pp. 89–103.
25. Blat, Y. & Eisenbach, M. (1996) J. Biol. Chem. 271, 1232–1236.
could produce high gain if the complex were to form upon 26. Parkinson, J. S. & Blair, D. F. (1993) Science 259, 1701–1702.
attractant binding. Polymerization of CheZ in the presence of 27. Katz, B. & Thesleff, S. (1957) J. Physiol. (London) 138, 63–80.
CheYp is another potential mechanism for signal amplification, 28. Knox, B. E., Devreotes, P. N., Goldbeter, A. & Segel, L. A.
though it is more likely that this reaction instead enhances the (1986) Proc. Natl. Acad. Sci. USA 83, 2345–2349.
adaptation response (25). A further source of gain might be found 29. De Young, G. & Keizer, J. (1992) Proc. Natl. Acad. Sci. USA 89,
in cooperative interactions among receptors in close proximity to 9895–9899.
letters to nature
possibility is that the key properties of biochemical networks are
robust; that is, they are relatively insensitive to the precise values of
Robustness in simple biochemical parameters. Here we explore the issue of robustness of
one of the simplest and best-known signal transduction networks: a
biochemical networks biochemical network responsible for bacterial chemotaxis. Bacteria
such as Escherichia coli are able to sense (temporal) gradients of
N. Barkai & S. Leibler chemical ligands in their vicinity2. The movement of a swimming
Departments of Physics and Molecular Biology, Princeton University, Princeton, bacterium is composed of a series of ‘smooth runs’, interrupted by
New Jersey 08544, USA events of ‘tumbling’, in which a new direction for the next run is
.........................................................................................................................
chosen randomly. By modifying the tumbling frequency, a bac-
Cells use complex networks of interacting molecular components terium is able to direct its motion either towards attractants or away
to transfer and process information. These ‘‘computational from repellents. A well established feature of chemoxis is its property
devices of living cells’’1 are responsible for many important of adaptation10,13–16: the steady-state tumbling frequency in a
cellular processes, including cell-cycle regulation and signal homogeneous ligand environment is insensitive to the value of
transduction. Here we address the issue of the sensitivity of the ligand concentration. This property allows bacteria to maintain
networks to variations in their biochemical parameters. We their sensitivity to chemical gradients over a wide range of attractant
propose a mechanism for robust adaptation in simple signal or repellent concentrations.
transduction networks. We show that this mechanism applies in The different proteins that are involved in chemotactic response
particular to bacterial chemotaxis2–7. This is demonstrated within have been characterized in great detail, and much is known about
a quantitative model which explains, in a unified way, many the interactions between them (Fig. 1a). In particular, the receptors
aspects of chemotaxis, including proper responses to chemical that sense chemotactic ligands are reversibly methylated. Biochem-
gradients8–12. The adaptation property10,13–16 is a consequence of ical data indicate that methylation is responsible for the adaptation
the network’s connectivity and does not require the ‘fine-tuning’ property: changes in methylation of the receptor can compensate
of parameters. We argue that the key properties of biochemical for the effect of ligand on tumbling frequency. Theoretical models
networks should be robust in order to ensure their proper proposed in the past assumed that the biochemical parameters are
functioning. fine-tuned to preserve the same steady-state behaviour at different
Cellular biochemical networks are highly interconnected: a per- ligand concentrations17,18. We present an alternative picture in
turbation in reaction rates or molecular concentrations may affect which adaptation is a robust property of the chemotaxis network
numerous cellular processes. The complexity of biochemical net- and does not rely on the fine-tuning of parameters.
works raises the question of the stability of their functioning. One We have analysed a simple two-state model of the chemotaxis
possibility is that to achieve an appropriate function, the reaction network closely related to the one proposed previously2,19. The two-
rate constants and the enzymatic concentrations of a network need state model assumes that the receptor complex has two functional
to be chosen in a very precise manner, and any deviation from the states: active and inactive. The active receptor complex shows a
‘fine-tuned’ values will ruin the network’s performance. Another kinase activity: it phosphorylates the response regulator molecules,
adaptation, p, and the adaptation time, t, were measured. The assay was Cytokines are secreted proteins that regulate important cellular
repeated for various reference model systems, with different values of bio- responses such as proliferation and differentiation1. Key events in
chemical parameters and of am, and different variants of the model. The cytokine signal transduction are well defined: cytokines induce
robustness of adaptation (Fig. 3) is independent of these choices. receptor aggregation, leading to activation of members of the JAK
Chemotactic drift velocity. The behaviour of a model system in the presence family of cytoplasmic tyrosine kinases. In turn, members of
of a linear gradient of attractant, =l, was simulated. The movement of the the STAT family of transcription factors are phosphorylated,
system was assumed to be composed of a series of smooth runs at a constant dimerize and increase the transcription of genes with STAT
velocity of 20 mm s 2 1 , interrupted by tumbling events. The tumbling frequency recognition sites in their promoters1–4. Less is known of how
was taken to be a sigmoidal function of the system activity (Hill coefficient, cytokine signal transduction is switched off. We have cloned a
q ¼ 2. Different values of q lead to the same qualitative picture; the sensitivity complementary DNA encoding a protein SOCS-1, containing an
increases with q). The trajectories were also subject to a rotation diffusion, with SH2-domain, by its ability to inhibit the macrophage differen-
D ¼ 0:125 rad2 s 2 1 (ref. 9). Attractant concentration was increasing along the x tiation of M1 cells in response to interleukin-6. Expression of
direction, (with l ¼ 1 mM at x ¼ 0). The chemotactic drift velocity was SOCS-1 inhibited both interleukin-6-induced receptor phos-
estimated by measuring the average x position of a hundred identical simulated phorylation and STAT activation. We have also cloned two rela-
systems as a function of time. tives of SOCS-1, named SOCS-2 and SOCS-3, which together with
the previously described CIS (ref. 5) form a new family of
Received 31 December 1996; accepted 17 April 1997. proteins. Transcription of all four SOCS genes is increased rapidly
1. Bray, D. Protein molecules as computational elements in living cells. Nature 376, 307–312 (1995). in response to interleukin-6, in vitro and in vivo, suggesting they
2. Stock, J. B. & Surette, M. in E. coli and S. typhimurium: Cellular and Molecular Biology (ed. Neidhardt,
F. C.) 1103–1129 (American Soceity of Microbiology, Washington DC, 1996).
may act in a classic negative feedback loop to regulate cytokine
3. Parkinson, J. S. Signal transduction schemes of bacteria. Cell 73, 857–871 (1993). signal transduction.
4. Hazelbauer, G. L., Berg, H. C. & Matsumura, P. M. Bacterial motility and signal transduction. Cell 73, To identify cDNAs encoding proteins capable of suppressing
15–22 (1993).
5. Bourret, R. B., Borkovich, K. A. & Simon, M. I. Signal transduction pathways involving protein cytokine signal transduction, we used an expression cloning
phosphorylation in prokaryotes. Annu. Rev. Biochem. 60, 401–441 (1991). approach. The strategy used the murine monocytic leukaemic M1
6. Adler, J. Chemotaxis in bacteria. Annu. Rev. Biochem. 44, 341–356 (1975).
7. Bray, D., Bourret, R. B. & Simon, M. I. Computer simulation of the phosphorylation cascade
cell line that differentiates into mature macrophages and ceases
controlling bacterial chemotaxis. Mol. Biol. Cell 4, 469–482 (1993). proliferation in response to various cytokines, including interleu-
8. Adler, J. Chemotaxis in bacteria. Science 153, 708–716 (1996). kin-6 (IL-6), and in response to the steroid, dexamethasone6,7.
9. Berg, H. C. & Brown, D. A. Chemotaxis in E. coli analysed by three-dimensional tracking. Nature 239,
500–504 (1972). Parental M1 cells were infected with the RUFneo retrovirus, into
10. Macnab, R. M. & Koshland, D. E. The gradient-sensing mechanism in bacterial chemotaxis. Proc. Natl which a library of cDNAs from the factor-dependent haemopoietic
Acad. Sci. USA 69, 2509–2512 (1972).
11. Block, S. M., Segall, J. E. & Berg, H. C. Impulse responses in bacterial chemotaxis. Cell 31, 215–226
cell line FDC-P1 had been inserted8. Retrovirally infected M1 cells
(1982). that were unresponsive to IL-6 were selected in semi-solid agar
12. Koshland, D. E. A response regulator model in a simple sensory system. Science 196, 1055 (1977). culture by their ability to generate compact colonies in the presence
13. Berg, H. C. & Tedesco, P. M. Transient response to chemotactic stimuli in E. coli. Proc. Natl Acad. Sci.
USA 72, 3235–3239 (1975). of IL-6 and geneticin. One stable IL-6-unresponsive clone, 4A2, was
14. Springer, M. S., Goy, M. F. & Adler, J. Protein methylation in behavioural control mechanism and in obtained after examining 104 infected cells (Fig. 1). A 1.4 kilobase
signal transduction. Nature 280, 279–284 (1979).
15. Koshland, D. E., Goldbeter, A. & Stock, J. B. Amplification and adaptation in regulatory and sensory
pair (kbp) cDNA insert, which we have named suppressor of
systems. Science 217, 220–225 (1982). cytokine signalling-1, or SOCS-1, was recovered by polymerase
16. Khan, S., Spudich, J. L., McCray, J. A. & Tentham, D. R. Chemotactic signal integration in bacteria. chain reaction (PCR) from the retrovirus that had integrated into
Proc. Natl Acad. Sci. USA 92, 9757–9761 (1995).
17. Segel, L. A., Goldbeter, A., Devreotes, P. N. & Knox, B. E. A mechanism for exact sensory adaptation genomic DNA of 4A2 cells. The SOCS-1 PCR product was used to
based on receptor modification. J. Theor. Biol. 120, 151–179 (1986).
18. Hauri, D. C. & Ross, J. A model of excitation and adaptation in bacterial chemotaxis. Biophys. J. 68,
708–722 (1995).
19. Asakura, S. & Honda, H. Two-state model for bacterial chemoreceptor proteins. J. Mol. Biol. 176, 349–
367 (1984).
20. Spudich, J. L. & Koshland, D. E. Non-genetic individuality: chance in the single cell. Nature 262, 467–
471 (1976).
21. Kleene, S. J., Hobson, A. C. & Adler, J. Attractants and repellents influence methylation and
demethylation of methyl-accepting proteins in an extract of E. coli. Proc. Natl Acad. Sci. USA 76,
6309–6313 (1979).