Regulation of Transcription : Induction of the proU B Transport Gene Is Dependent on Accumulation of Intracellular Potassium Osmotic etaine LESLEY SUTHERLAND ,1 JOHN CAIRNEY ,1 MICHAEL J. ELMORE ,2 IAN R. BOOTH ,2 AND CHRISTOPHER F. HIGGINS ' * Molecular Genetics Laboratory , Department of Biochemistry , University of Dundee , Dundee DDJ 4HN ,2 and Department of Microbiology , Marischal College , University of Aberdeen , Aberdeen AB9 lAS ,1 Scotland Received 17 March 1986/Accepted 4 August 1986 The proU locus , which encodes a high-affinity betaine transport system , and the kdp operon , which encodes a potassium transport system , are the principal osmoresponsive genes in Escherichia coli and Salmonella typhimurium .
The kdp operon is known to be induced in response to changes in cell turgor .
We have investigated the control of proU expression and shown that it differs from that of kdp in a number of fundamental ways .
Rather than responding to changes in turgor , proU expression is principally determined by the intracellular accumulation of potassium ions .
Potassium and betaine were shown to play distinct osmoprotective roles .
Potassium serves as the principal osmoprotectant and is accumulated in response to low-level osmotic-stress to restore turgor .
As external osmolarity is increased to a level at which the corresponding increase in internal potassium concentrations is potentially deleterious to enzyme function , betaine ( when available ) is accumulated in preference to potassium .
The different mechanisms ofproU and kdp regulation reflect the different physiological roles of these two osmoprotectants .
Bacterial cells are frequently subjected to fluctuations in the osmolarity of the medium in which they are growing .
Most organisms respond to osmotic-stress in a similar manner , by accumulating high intracellular concentrations of compatible solutes ( 30 , 31 ) .
These solutes not only balance external osmolytes such that cell turgor is maintained but must also be compatible with the function of intracellular enzymes .
In Escherichia coli and Salmonella typhimurium the accumulation of K + ions plays a primary role in maintaining the osmotic balance of the cell ( 6 ; W. Epstein , FEMS Microbiol .
An ionic balance is normally maintained by the intracellular synthesis of equivalent amounts of glutamate ( 41 ) .
Potassium uptake is mediated by two independent transport systems , a low-affinity system , Trk , and a high-affinity system , Kdp ( 36 ) .
The Kdp system contributes to the accumulation of K + whenever the Trk system is not sufficiently active to maintain turgor ; the expression of the kdp operon is specifically induced under such conditions ( 26 ) .
It is well established that in the absence of exogenous solutes such as proline or betaine the intracellular concentration of K + ions is essentially proportional to the osmolarity of the external medium ( 16 ) .
However , in complex media the situation is less simple .
Certain organic-compounds ( compatible solutes ) can serve as osmoprotectants and , if accumulated by the cell , will significantly enhance growth universally rates at high-osmolarity .
The most adopted compatible solutes are proline and betaine ( N,N,N-trimethylglycine ) .
Betaine is accumulated during osmotic-stress by such diverse organisms as halotolerant plant spe-cies ( 44 ) , marine animals ( 4 ) , halophilic bacteria ( 22 ) , and cyanobacteria ( 35 ) .
Both proline and betaine can serve as osmoprotectants for the gram-negative bacteria E. coli and S. typhimurium .
Thus , increased synthesis of proline ( 13 ) or betaine ( 27 , 40 ) or the uptake of these compounds when supplied exogenously ( 6 , 7 , 11 , 14 , 29 , 34 ) can markedly enhance growth-rates at high-osmolarity .
There are two distinct transport systems for betaine in S. typhimurium ( 6 , 7 ) .
The proP gene encodes a low-affinity betaine uptake system which also transports proline ( 6 ) , whereas the proU gene , originally defined as a proline transport system ( 14 , 15 ) , encodes a high-affinity transport system with betaine as its principal substrate ( 7 ) .
ProU is a periplasmic binding protein-dependent transport system ( unpublished results ) , and expression of the proU gene is induced under conditions of osmotic-stress ( 7 , 15 , 17 ) .
The proP and proU genes have also been identified in E. coli and probably serve analogous roles ( 17 , 18 ) .
Osmotic stress affects the expression of only a very limited number of bacterial genes ( 7 , 12 ) .
The relative expression of the porins OmpC and OmpF is affected by the osmolarity of the medium ( 19 , 20 ) .
However , many other factors also influence porin expression ( 42 ) , and the precise role of osmolarity is obscure and may be indirect .
Expression ofphoA and malB is also affected by medium osmolarity ( 43 ) , but again these affects are not major and may well be an indirect effect on the cellular concentration of their relevant inducers ( 5 ) .
In E. coli the enzymes involved in the conversion of choline to betaine are only synthesized in osmotically stressed cells ( 40 ) .
The genes encoding these enzymes are located near the lac operon and are absent in S. typhimurium and many E. coli strains ( 40 ) .
It seems that their likely regulation is at the level of transcription , but this has not yet been demonstrated .
Unlike of the any above-mentioned , osmotically influenced genes , kdp and proU are fully repressed at low-osmolarity , and expression is induced over 100-fold by osmotic-stress .
Thus , kdp and proU are the only two known genes which are directly induced at the transcriptional level in response to osmotic-stress .
The kdp and proU genes encode high-affinity potassium and betaine transport systems , respectively .
The relative roles of K + and betaine in restoring turgor and facilitating growth at high-osmolarity are presently obscure .
In this paper we show that these two osmoprotectants play distinct physiological roles .
It has been demonstrated that the kdp gene is primarily regulated in response to changes in turgor pressure ( 26 ) .
Initial studies showed that , like kdp expression , proU expression is also induced by osmotic-stress ( 7 , 15 , 17 ) .
We show here that there are a number of fundamental differences in the responses of kdp and proU to osmotic-stress .
These differences are reconciled by the demonstration that induction of proU expression is not directly controlled by changes in cell turgor , but is apparently dependent upon the intracellular concentration of potassium ions .
Thus , when a cell is osmotically stressed the consequent changes in cell turgor activate K + transport , leading to an increase in intracellular K + concentrations which induces proU expression .
The different mechanisms of induction of pro Uand kdp are shown to reflect distinct physiological roles for betaine and potassium as osmoprotectants .
MATERIALS AND METHODS Bacteeial strains and growth media .
All strains are derivatives of S. typhimurium LT2 or of E. coli K-12 .
Their genotypes and sources are detailed in Table 1 .
Transductions in S. typhimurium were carried out with a high-transducing derivative of phage P22 int4 as described by Roth ( 37 ) .
E. coli transductions were with phage P1 vir ( 32 ) .
After transduction of any transposon or Mu derivative , the correct location of the element and the presence of only a single copy of that element in the transductant were checked by marker rescue .
Cells were grown aerobically at 37 °C unless otherwise stated .
For strain constructions cells were grown in LB-medium ( 37 ) .
For all other experiments the growth medium used was LOM ( low-osmolarity medium [ 6 ] ) , which contains 0.4 mM MgSO4 , 6 , uM ( NH4 ) 2SO4 .
FeSO4 , 20 mM ( NH4 ) 2HP04 , 10 mM bistrispropane , 5 mM NaCl , 10 mM glucose , and 1 mM KCl .
For certain experiments the potassium concentration was varied by adding extra KCl as appropriate .
For potassium limitation KCl was omitted ; flame photometry showed the potassium concentration in this `` potassium-free '' medium to be < 8 1xM .
Isolation of a stable proU-lacZ fusion in S. typhimurium .
Although we have previously constructed pro U-lacZ operon fusions with the Mu derivatives Mu dl and Mu dl-8 ( 7 ) , a mini-Mu derivative ( Mu dJ , originally called Mu dI1734 ; [ 10 ] ) has recently been described which lacks all transposition functions and is therefore completely stable .
A Mu dJ-mediated proU-lacZ fusion was isolated by transducing LT2 to Kanr with a P22 lysate of TT10288 .
TT10288 contains a his : : Mu dJ insertion and , in addition , a very closely linked Mu dl helper phage that provides the genes encoding transposition functions and that can be copackaged with the Mu dJ in P22 transducing particles .
About 104 Kanr colonies , each of which contained an independent Mu dJ insertion into the chromosome , were pooled , washed twice in minimal-medium , and plated on LOM-agar plates containing 0.3 M NaCl and 40 , ug of 5-bromo-4-chloro-3-indolyl-3-D-galactoside per ml .
Deep blue ( Lac ' ) colonies were picked and screened on LOM-agar plates containing 5-bromo-4-chloro-3-indolyl - , - D-galactoside but lacking the added NaCl .
Any colonies which expressed 3-galactosidase on the plates containing 0.3 M NaCl but remained white ( Lac - ) on low-salt plates were purified and characterized further .
On the basis of previous results ( 7 ) we anticipated that most of these isolates would contain proU-lacZ-fusions .
One such derivative , CH1301 , was shown to contain a single Mu dJ insertion , and the insertion was shown to be in proU by transductional mapping and marker rescue as described previously ( 7 ) .
Expression of , B-galactosidase from this fusion was regulated in a manner identical to that reported previously for Mu dl-8-mediated pro U-lacZ-fusions ( 7 ) .
proU-lacZ-fusions in E. coli .
A mini-Mu-mediated proU-lacZ operon fusion was selected from a collection of random Mu dI1681 insertions .
These random mini-Mu insertions were constructed by infection of MC4100 ( 9 ) with a temper-ature-induced lysate of strain POI1681 , which contains both the mini-Mu and a Mu helper phage ( 10 ) .
pro'U - lacZ-fusions were selected from this pool as described above for the S. typhimurium fusions , except that MacConkey mediumlactose was used as an indicator rather than 5-bromo-4-chloro-3-indolyl-p-D-galactoside .
As for S. typhimurium , all osmotically induced lacZ-fusions selected in this manner mapped to a single locus which was shown to be proU by cotransduction with a srl : : TnJO insertion .
kdp-lacZ-fusions were not selected by this procedure since MacConke medium-lactose plates contain sufficient K + to fully repress kdp expression .
As discussed below , proU-lacZ-fusions isolated in E. coli were found to be regulated in a matiner identical to that of the more completely characterized pro UlacZ-fusions in S. typhimurium .
The E. coli kdpAS : : Mu dl ( Apr lac ) fusion ( 26 ) was stabilized by selecting temperature-resistant colonies at 42 °C .
These temperature-resistant de-rivatives were screened for the absence of Mu dl transposition when transduced into a Mu-free strain .
One such stabilized kdp-lacZ fusion transduced into MJF1 ( 38 ) ; was the resultant strain , MJF261 , shown contain single was to a Mu dl insertion , and the insertion was shown to be in kdp , by marker rescue and genetic mapping .
This stabilized fusion was used for all experiments with kdp and is referred to as kdpAS : : Mu dlX .
Growth of cells under potassium limitation .
Cells grown overnight in LOM were washed twice in LOM lacking KCl and resuspended in the same K + - free medium at a 1:50 dilution .
Cells stopped growing at an optical density at 600 nm of about 0.3 ( see Fig. 3 ) .
Since growth could be rapidly restored to normal rates by the addition of 1 mM KCl , we concluded that growth was indeed limited by potassium .
The two osmolytes used routinely were 0.3 M NaCl and 0.44 M sucrose .
These are isoosmotic concentrations of the two solutes .
Cells were osmotically stressed in one of two ways .
Steady-state adaptation to osmotic-stress was achieved by growing cells overnight in the presence of the appropriate osmolyte , subculturing the cell into the same medium at a 1:50 dilution , And growing to the midexponential phase before assaying .
Alternatively , a rapid osmotic shock was achieved by growing cells to midexponential-growth in LOM medium before adding sucrose to 0.44 M or NaCl to 0.3 M. Pulse-labeling .
Samples ( 1.0 ml ) of cells grown as indicated below were pulsed for 2 min with [ 35S ] methionine ( 800 Ci mmol-1 ; 5 , uCi ) followed by a 2-min chase with unlabeled methionine ( 100 , ug ml - ' ) .
The cells were then rapidly sedimented by centrifugation , resuspended in 50 , u1 of Laemnli sample buffer ( 25 ) , boiled for 2 min , and separated on a 9 % sodium dodecyl sulphate-polyacrylamide gel ( 2 , 25 ) .
Labeled proteins were visualized by autoradiography after the gel was dried .
, - Galactosidase was assayed by the procedure of Miller ( 32 ) .
Cells were permeabilized by the sodium dodecyl sulfate-chloroform procedure .
All assays were performed in triplicate and repeated on at least three independent occasions .
Measurement of intracellular KV .
Cells were grown to the midexponential phase in LOM containing NaCl , KCl , and betaine as appropriate .
Cells from 1 ml of culture were rapidly harvested by centrifugation for 30 s in an Eppendorf a Cells of S. typhimurium CH1301 ( proU-lacZ ) , E. coli MJF261 ( kdp-lacZ ) , and E. coli CH1366 ( proU-lacZ ) to the midexponential phase in were grown LOM , with osmolytes and K + added as indicated , and assayed for,-galactosidase activity .
Bacterial strains Genotype Source/Construction Strain S. typhimurium CH272 oppA305 : : Mu dl galE503 Abio-561 putP202 : Mu dl gal-503 Abio-561 AoppBC250 tppB84 : Mu dl-8 proUl 702 : Mu dl-8 proUl 705 : : Mu dJ ( Kanr lac ) Wild type hisD9953 : : Mu dJa his-9944 : : Mu dl ( 23 ) ( 8 ) CH322 CH776 CH946 CH1301 LT2 TT10288 ( 24 ) ( 7 ) This paper B. N. Ames J. Roth E. coli CH1366 CH1407 EMG2 MC4100 MJF1 MJF261 MJF266 PB13 P011681 This paper This paper MC4100 proUl706 : : Mu d11681 ( Kanr ) MC4100 proUl706 : : Mu dI1681 ( Kan9 kdp : : TnJO CGSCb ( 9 ) ( 38 ) This paper W. Epstein M. Jones-Mortimer ( 10 ) Wild type araD139 rpsL thi A ( argF-lac ) U169 relA araFJ39 araE thi gltSo A ( his-gnd ) A ( argF-lac ) U169 rpsL flaB deo MJF1 kdpA5 : : Mu dIX MJF1 kdp : : TnlO cysE recA srl : : TnlO Mu d11681 ( Kan9 ara : : ( Mu cts ) 3 A ( proAB-argF-lac ) XIII rpsL a Originally called Mu d11734 ( 10 ) .
b E. coli Genetic Stock Center , Yale University , New Haven , Conn. .
Osmolyte K + proU-lacZ pro U-lacZ kdp : : TnlO 100 , uM 23 24 350 F.M 24 22 1 mM 23 24 10 mM 23 23 50 mM 55 26 0.3 M NaCI 100 , uM 1,570 64 0.3 M NaCI 350 , uM 1,606 475 0.3 M NaCl 1 mM 1,293 830 0.3 M NaCl 10 mM 1,060 866 0.3 M NaCI 50 mM 1,201 923 a Cells of E. coli strains CH1366 ( proU-lacZ ) and CH1407 ( proU-lacZ kdp : : TnlO ) were grown to the midexponential phase in LOM , with osmolytes and K + added as indicated , and 3-galactosidase activity was assayed .
centrifuge , all of the supernatant was removed , and the pellet was suspended in 1 ml of deionized water .
The suspension was boiled for I min to release k ' from the cell , and the K + concentration was determined by flame photometry .
Each data point is an average of at least three separate determinations that did not vary by more than - + -10 % .
Results obtained with this inexpensive and rapid procedure were very similar to those obtained with 42K ( see below ) .
Cell volumes were determined by the distribution of 3H20 and [ 14C ] sucrose by the method of Stock et al. ( 39 ) as described previously ( 1 ) .
Values obtained for cell volumes were consistent with previously reported results ( 34 , 39 ) .
Increasing the osmolarity of LOM by adding 0.3 M NaCl caused a decrease in cell volume of approximately 35 % ( from 3.0 to 1.95 , ul mg - ' ) .
Measurement of relative intracellular K + and betaine concentrations .
Cells were grown to saturation in LOM containing 0.3 M NaCl , harvested by centrifugation , and suspended at 10 mg of protein per ml in LOM containing chloramphen-icol ( 50 , ug ml - ' ) and NaCl as appropriate .
42KHCO3 ( 150 mCi mmol-1 ) and [ 14C ] betaine ( 10 mCi mmol-1 ) were added to a final concentration of 1 mM , and the cells were incubated at 20 °C for 90 min .
Preliminary experiments showed this to be sufficient time for equilibration of the internal and external pools .
Cells were then collected by filtration through a 0.45-p .
m membrane filter ( Millipore Corp. ) and washed in the same medium in which they had been incubated but lacking the radioisotopes , and the 42K ' and [ 14C ] betaine accumulated by the cells were determined by scintillation counting .
Due to the very short half-life of 42K and the overlap of the scintillation windows , the 14C was counted after allowing sufficient time for 42K decay .
RESULTS Because direct assays for proU and kdp function are difficult to perform , expression of these genes was assessed by assaying 0-galactosidase expression from isolates with kdp-lacZ or proU-lacZ operon fusions .
Assays were performed during steady-state growth in the appropriate me-dium , rather than after sudden osmotic shock , unless otherwise indicated .
The proU gene and the ProU transport system are considerably better characterized in S. typhimurium than they are in E. coli .
However , strains with kdp-lacZ-fusions have never been isolated in S. typhimurium ; despite several attempts with various selections , we have so far been unabl to isolate strains with such fusions in this species .
This suggests there may be some differences in the Kdp system between these two species .
However , any difference is likely to be minor , for example , in the affinity of the transport system for its substrate , rather than a major difference in the actual mechanism of kdp regulation .
It is known that S. typhimurium possesses a Kdp transport system , the genes have been cloned by virtue of their sequence homology to the E. coli kdp genes ( W. Epstein , personal communication ) , and , like in E. coli , the Kdp proteins of S. typhimurium are induced by potassium starvation ( 3 ; unpublished results ) .
We have therefore studied the regulation of kdp expression in E. coli MJF261 and the regulation of proU expression primarily in S. typhimurium CH1301 .
However , to exclude the possibility that the differences in regulation we observe between proU and kdp ( see below ) are due to a species difference , and to show that these differences are not a consequence of any species difference in Kdp , also we isolated a proU-lacZ fusion in E. coli ( CH1366 ) and repeated some experiments with this species .
As described below , no significant difference in the regulation of proU expression was seen between E. coli and S. typhimurium .
Potassium limitation induces kdp but not proU .
In minimalsalts medium , K + accumulation provides the principal means of maintaining cell turgor ( 16 ) .
When cells are grown under potassium limitation , potassium can not be accumulated to sufficient intracellular concentrations to maintain cell turgor , and any turgor-sensitive gene would be expected be induced .
Strains containing either to a proU-lacZ or a kdp-lacZ fusion were grown under potassium limitation ( < 8 TABLE 3 .
Effects of a kdp mutation on proU expressiona Medium additives 1-Galactosidase activity ( U ) TAB3LE 2 .
Differences in expression of proU and kdpa Medium additives 1-Galactosidase activity ( U ) E. coli kdp-lac 878 11 1,359 2 Osmolyte K + S. typhimurium proU-lac E. coli proU-lac < 8 , uM 50 mM 5 9 23 55 364 468 1 mM 50 mM 0.3 M NaCI 0.3 M NaCl 1,293 1,201 , uM ) , and P-galactosidase activity was assayed ( Table 2 ) .
Our results with kdp-lacZ-fusions were in full agreement with those of Laimins et al. ( 26 ) ; expression of kdp at low-osmolarity was strongly induced by potassium limitation ( < 8 , uM K + ) .
However , under identical growth-conditions no induction of proU expression was observed in either E. coli or S. typhimurium .
This indicates an important difference from the regulation of kdp and , in addition , implies that the expression of proU can not be simply in response to changes in cell turgor .
Excess potassium represses kdp but not proU .
When sufficient potassium is present in the growth medium it is accumulated by the cell to restore turgor .
It is well established that intracellular potassium concentrations increase in more or less direct proportion to external osmolarity , both to a Cells were grown under potassium limitation ( 50 , uM ) or excess ( 1 mM ) in medium containing 0.3 M NaCl , and , B-galactosidase activity was assayed during midexponential-growth .
Leucine ( 1 mM ) was included in the growth medium for CH776 to induce expression of tppB ( 23 ) .
oppA ( 23 ) and putP ( 8 ) are both constitutively expressed under these growth-conditions maintain an osmotic balance and to restore turgor ( 16 ; see below ) .
Thus , during steady-state growth at high-osmolarity in the presence of excess potassium ( 50 mM ) , sufficient K + can be accumulated to partially or completely restore turgor , and expression of turgor-sensitive genes would be expected to be repressed .
Consistent with this , and with the view that kdp is turgor regulated , we found that kdp was not expressed during steady-state growth in medium of high-osmolarity as long as sufficient K + ( 50 mM ) was present ( Table 2 ) .
However , at lower K + concentrations ( 1 mM ) , turgor could not be maintained in these Kdp-strains , and kdp expression was induced .
In marked contrast to kdp , expression of proU was maintained at a fully induced level during steady-state growth at high-osmolarity , even in the presence of 50 mM K + ( Table 2 ) .
Again , there was no obvious difference in the regulation of proU-lacZ-fusions in E. coli and S. typhi-murium .
These results illustrate another clear difference between kdp and proU and again imply that induction of proU expression can not be a direct response to changes in cell turgor .
Effect of kdp mutations on proU expression .
Because the pro U-lacZ strains are Kdp + , whereas the kdp-lacZ strain we used is phenotypically Kdp - , it is important to ascertain that the differences in regulation between proU and kdp are not simply a result of the presence or absence of the Kdp transport system .
Thus , a kdp : : TnlO insertion was introduced into an E. coli proU-lacZ fusion strain , and the expression of proU was compared with that of its Kdp + parent ( Table 3 ) .
The differences in the regulation of proU and kdp can not be attributed to the presence or absence of a functional Kdp system .
As expected ( 16 , 26 ) , growth of the Kdp-derivative was limited by potassium even at extracellular concentrations as high as 350 , uM ( data not shown ) .
This is because the remaining K + transport system , Trk , has a low affinity for K + ( K , = 1 mM ) and is unable to accumulate K + at sufficient rates to maintain optimal intracellular concentrations ( see below ) .
In contrast , the Kdp + derivative grew normally at both 100 and 350 , uM K + .
proU expression was not induced by K + limitation , even when such limitation was made more severe by the introduction of a kdp mutation ( Table 3 ) .
Similarly , expression of proU was not repressed by excess K + ( 50 mM ) in either the Kdp-or the Kdp + derivative .
The only effect of introducing a kdp lesion was that proU expression at high-osmolarity was reduced when extracellular K + concentrations were low enough ( 100 or 350 j , M ) to be limiting for K + accumulation .
This is in contrast to the Kdp + derivative , where K + does not become limiting at these extracellular concentrations ( see below ) and where proU expression remains relatively constant even at low extracellular K + concentrations .
Thus , when the cell can not accumulate sufficient K + to restore turgor and maintain optimum growth-rates , osmotic-stress is increased , yet proU expression is actually reduced .
Significantly , these results are completely opposite to those found for kdp , which is induced rather than repressed as K + becomes limiting ( Table 2 ) , and confirm that proU and kdp are induced by very different signals .
These observations suggest that intracellular potassium concentrations may be important in regulating proU expression .
proU is induced under conditions where intracellular potassium concentrations would be expected to be high and is repressed when intracellular potassium concentrations would be expected to be low .
To confirm these interpretations , we measured intracellular potassium concentrations under the various growth-conditions ( Table 4 ) .
In Kdp + strains , intracellular K + concentrations were found not to vary significantly with various extracellular concentrations between 100 puM and 50 mM but to increase with increasing medium osmolarity .
This is in excellent agreement with the data of Epstein and Schultz ( 16 ) .
In contrast , in a Kdpstrain intracellular potassium levels are reduced at extracellular levels below the of Trk and 350 potassium K , ( 100 , uM ) .
Again , these data are consistent with published results ( 26 ) ; in addition , the reduced intracellular K + concentrations correlate well with reduced growth-rates under the same conditions ( data not shown ) .
In all cases , the measured intracellular concentrations of K + ions , altered by varyin the extracellular K + concentrations or medium osmolarity or by introducing kdp mutations , are compatible with the view that induction of proU is dependent upon the accumulation of intracellular potassium .
Expression of proU depends upon intracellular potassium .
The data above show that kdp and proU expression is regulated differently and demonstrate that proU can not respond directly to changes in turgor .
Other than turgor , the most direct physiological response to osmotic-stress is the accumulation of K + ions within the cell .
In the absence of osmoprotectants such as betaine , intracellular K + concentrations increase in proportion to extracellular osmolarity ( 16 ; this paper ) .
It therefore seemed possible that proU expression might be determined by the intracellular concentration of K + ions .
This hypothesis is strongly supported by the data presented above , which show that proU expression is induced whenever intracellular K + concentrations are high and is repressed whenever concentrations are low .
( i ) Iso-osmotic concentrations of different osmolytes which increase intracellular K + to similar extents ( 16 ) have similar effects on the induction of proU .
( ii ) At high-osmolarity the intracellular K + concentration remains high , even when sufficient extracellular K + is supplied to permit restoration of turgor ( 16 ) .
proU expression is fully induced under these conditions even during steady-state growth .
( iii ) Potassium limitation , which reduces intracellular K + concentrations , does not induce proU expression even though the cell is under turgor stress .
( iv ) In a Kdp-strain , proU expression at high-osmolarity is considerably reduced when the external K + concentrations are low .
These K + concentrations are below the K , for Trk ( 36 ) , and the ability of the Kdp-strain to accumulate sufficient intracellular K + becomes limiting .
To test this hypothesis , strain CH946 ( proU-lacZ ) was grown under K + limitation ( 50 , uM ) in medium containing 0.3 M NaCl .
After 3 h of growth , expression of proU was reduced to less than 50 % of the level in cells grown with 1 mM K + .
Expression of several control fusions of lacZ to other genes ( opp , tpp , and putP ) was essentially identical whatever the extracellular K + concentration ( Table 5 ) .
These results imply that K + ions play a specific role in proU expression .
To define more precisely the effects of K + ions on proU expression , cells from an overnight culture of strain CH1301 were washed in LOM lacking K + ( < 8 , uM K + ) , and resuspended at a 1:50 dilution in the same medium .
Growth was followed by monitoring the optical density of the culture at 600 nm ( Fig. 1 ) .
At an optical density of about 0.3 , growth ceased due to potassium starvation ; growth could be restored to normal by the addition of potassium to the medium .
At various stages during-growth ( arrows in Fig. 1 ) , cells were taken and osmotically shocked by adding sucrose to a final concentration of 0.44 M. IfproU induction were mediated by an increase in internal K + concentrations , then an osmotic shock under K + limitation would not be expected to induce expression .
However , if K + ( 1 mM ) were added at the same time as the osmotic shock is given , this will be accumulated and normal induction would be anticipated ; this is precisely what was observed ( Table 6 ) .
Expression of proU was not induced by osmotic shock unless K + ions were provided .
Essentially identical results were obtained with a pro U-lacZ fusion in E. coli ( data not shown ) .
When this experiment was repeated with NaCl as the osmolyte , significant induction of proU was seen even when the osmotic shock was given in TABLE 4 .
0.3 M 0.3 M 0.3 M 0.3 M 0.3 M NaCl NaCl NaCl NaCl NaCI 100 , uM 350 , uM 159 271 E. coli CH1407 E. coli CH1407 0.3 M NaCl 0.3 M NaCl 1 mM 332 395 E. coli CH1407 E. coli CH1407 0.3 M NaCl 0.3 M NaCl 10 mM 50 mM 416 E. coli CH1407 0.3 M NaCl BL TA E 5 .
K for p + requirement roU inductiona P-Galactosidase activity ( U ) + K + ¬ Strain Fusion K + CH946 proU-lacZ oppA-lacZ tppB-lacZ putP-lacZ 436-766-248 70 181-716-281 65 CH272 CH776 CH322 TA BL E 6 .
K + requirement for pr oU inductiona P-Galactosidase activity ( U ) Strain Fusion Inducer Induction point + K + - K + CH1301 proU-lacZ 0.44 M Sucrose 190 39 1 CH1301 proU-lacZ 0.44 M Sucrose 204 27 2 CH1301 proU-lacZ 0.44 M Sucrose 143 34 3 CH776 tppB-lacZ 1 mM Leucine 276 218 3 EMG2 lacZ 4 mg of IPTGb per ml and 5 mM cAMP 1,416 1,434 3 a Cells were grown to potassium starvation , and the inducer was added , with or without 1 mM K + , at the points on the growth curve indicated in Fig. 3 .
13-Galactosidase activity was assayed 30 min after the inducer was added .
For the lacZ experiment with EMG2 , glycerol rather than glucose was used as carbon source .
b IPTG , Isopropyl - , - D-thiogalactopyranoside .
' p 0i O1 1 2 5 7 6 3 4 Time ( hours ) FIG. 1 .
Growth of cells under potassium limitation .
An overnight culture of S. typhimurium CH1301 ( proU-lacZ ) grown in LOM containing 1 mM KCl was washed twice in LOM lacking K + ( < 8 , uM K + ) and resuspended at a 1:50 dilution in the same medium .
Growth was followed by monitoring the optical density of the culture at 600 nm ( O.D. 6m ) .
After about 5 h , growth ceased due to K + limitation ; growth could be rapidly restored by adding 1 mM KCl .
The arrows indicate the points at which samples were taken for osmotic shock ( see text and Table 6 ) .
Effect of different osmolytes on kdp expression .
Cells of E. coli MJF261 ( kdp-lacZ ) were grown overnight in LOM containing the indicated concentrations of KCI and added osmolytes as appropriate , and , B-galactosidase activity was assayed .
Symbols : x , no added osmolyte ; E , grown in 0.44 M sucrose ; 0 , grown in 0.3 M NaCl the absence of added K + .
However , this was found to be an artifact , since 0.3 M NaCl ( Analar grade ) contains approximately 50 , uM K + as a contaminant , sufficient to permit proU induction in a Kdp + strain .
Because cell growth was reduced and eventually halted by potassium limitation during these experiments , it was important to establish that protein synthesis was identical whether or not potassium was added at the same time as the cells were given the osmotic shock .
Thus , cells were grown and osmotically shocked under conditions identical to those described above , but instead of assaying , B-galactosidase we pulse-labeled cells with [ 35S ] methionine .
The labeled proteins were separated by sodium dodecyl sulfate-polyacryl-amide gel electrophoresis ( Fig. 2 ) .
Quite clearly , protein synthesis is unaffected by K + starvation and is essentially identical whether or not K + is supplied at the same time as the cells are shocked .
Presumably , the internal K + concentration is maintained at a sufficient level for normal ribosome function but at insufficient levels to restore turgor and permit growth .
As a further control , we wished to show that lacZ expression from promoters other than pro U could be induced under conditions of potassium starvation .
Unlike proU-lacZ-fusions , isopropyl-p-D-thiogalactopyranoside induction of chromosomally encoded lacZ and the induction of a tppB-lacZ fusion by leucine were little affected by potassium starvation ( Table 6 ) .
Thus , neither protein synthesis nor the specific synthesis of P-galactosidase was significantly reduced by K + starvation .
It must therefore be concluded that proU expression requires the accumulation of intracellular K + ions .
Different osmolytes have different effects on kdp induction .
The expression of kdp is induced by sudden osmotic shock with any osmolyte ( with the exception of glycerol , which is freely permeable through the membrane and hence does not cause a change in turgor ) ( 26 ) .
In apparent contradiction , it has been reported ( 17 ) that during steady-state growth at a Cells of strains CH946 ( proU-lacZ ) and MJF261 ( kdp-lacZ ) were grown in LOM containing K + , proline-betaine , and NaCI as indicated .
lB-Galactosidase activity was assayed at midexponential-growth .
b ND , Not determined .
high-osmolarity kdp is induced by ionic solutes but not by nonionic solutes at equivalent osmolarity .
To resolve this problem we examined the effects of different osmolytes on kdp expression in more detail .
Cells were assayed for kdp expression after growth in medium containing various amounts of K + .
In the absence of any added osmolyte , expression of kdp depended upon the extracellular K + concentration ; kdp was expressed in low K + and repressed by high K + ( Fig. 3 ) .
This is in complete agreement with the data of Laimins et al. ( 26 ) and implies that at K + concentrations below about 5 mM , potassium uptake via Trk is insufficient to fully restore turgor and hence kdp is induced .
Significantly , the K , of Trk , the only remaining K + transport system in this Kdp-strain , is around 1 mM ( 36 ) .
When the same experiment was carried out for cells grown with 0.44 M sucrose or 0.3 M choline-chloride added as an osmolyte , the response to various levels of K + was unaltered .
Thus , even though the increased osmolarity increases the intracellular K + concentration , it seems that the rate of K + uptake required to maintain this intracellular concentration is unaffected .
In contrast , NaCl at an equivalent osmolarity ( 0.3 M ) shifted the response curve such that higher concentrations of K + were required to fully repress kdp expression .
Thus , both ionic and nonionic solutes induce kdp expression at the steady state , although solutes containing Na + ions shift the K + response curve slightly .
This shift in the K + response curve accounts for the failure of Gowrishankar to observe induction of kdp by nonionic solutes ( 17 ) ; he measured the induction of kdp at just a single K + concentration which happened to be that concentration at which the differences between ionic and nonionic solutes are most marked .
The most probable explanation for the effect of Na + ions on the K + response curve is that , unlike sucrose and choline-chloride , Na + ions either decrease the activity of Trk ( kdp is mutated in these strains ) and hence reduce the rate of K + uptake at any given extracellular K + concentration or , alternatively , that Na + ions increase the rate of K + efflux .
Betaine and proline have different effects on kdp and proU expression .
It is well known that betaine or proline can enhance the growth-rate of cells at high-osmolarity ( 6 , 7 , 29 , 34 ) .
These solutes can play two distinct roles in osmoprotection ; they contribute to the restoration of cell turgor , and they are able to protect enzymes against the adverse effect of high ionic strength ( 21 , 28 , 33 , 45 ) .
During growth at high-osmolarity in 1 mM K + , which in a kdp strain limits turgor and induces kdp expression ( Table 2 ) ( 26 ) , betaine markedly enhances growth-rate ( Fig. 4 ) .
However , even when strains are grown at high-osmolarity in excess K + ( 50 mM ) , which is believed to restore turgor and fully represses kdp expression , growth-rates are still significantly enhanced by betaine .
It is , of course , possible that turgor is only partially restored by K + accumulation , even in the presence of excess extracellular K + , and that betaine results in full restoration of turgor and thereby enhances growth .
However , this explanation seems improbable because it has been demonstrated that at these osmolarities sufficient K + can be accumulated to restore turgor ( 16 ) .
More probably , the enhancement of growth-rates by betaine in excess potassium is not due to restoration of turgor but to protection of enzyme function from high intracellular ionic strength .
Interestingly , at low K + concentrations ( 100 p , M ) betaine does not enhance growth , implying that at these concentrations K + itself is growth limiting .
The effects of proline and betaine on proU and kdp expression were examined ( Table 7 ) .
Cells were assayed during steady-state growth in LOM ( 1 mM K + ) at high-osmolarity ( 0.3 M NaCl ) , with 1 mM betaine or 1 mM proline or both added as appropriate .
As we have reported previously ( 7 ) , expression of proU in S. typhimurium is considerably reduced by added betaine or proline .
This effect is not simply an effect of growth-rate on pro U expression ; altering the growth-rate in other ways ( i.e. , amino-acid starvation ) did not influence proU expression .
Interestingly , betaine caused a greater reduction in proU expression than did proline .
The significance of this observation is discussed below .
Similar results were obtained for pro U-lacZ-fusions in E. coli ( data not presented ) .
In contrast , the addition of betaine or proline to the growth medium had little or no effect on the expression of kdp .
In fact , betaine actually increased kdp expression slightly .
The absence of any major effect of betaine on kdp expression is in agreement with the growth data presented above , which imply that betaine does not play a role in restoring cell turgor under these conditions .
These data provide yet another distinction between the regulation of proU and kdp expression .
If , as implied by our data above , proU expression were dependent upon intracellular potassium concentrations , we would predict that betaine and proline , which reduce proU expression , also reduce the intracellular K + pools .
Table 4 shows that this is indeed the case .
However , despite their rather different effects on proU expression , proline and betaine actually reduce the intracellular K + pools to similar extents .
The implications of this apparently inconsistent result are discussed below .
It should be noted that Gowrishankar reported that the addition of proline to the growth medium does not affect proU expression ( 17 ) , in apparent contradiction to the results reported here .
However , his experiments were carried out at 0.4 M NaCl , at which osmolarity proline uptake is effectively abolished in a proU strain ( 6 , 7 ) .
In addition , we show below that there is a difference between growth at 0.3 and 0.4 M NaCl in terms of the relative roles of K + and betaine in restoring cell turgor .
Distinct roles for K + and betaine at different osmolarities .
Because expression of proU is only fully induced at relatively high extracellular osmolarities , once a significant intracellular K + concentration has already been achieved , it seemed possible that K + and betaine might play rather distinct roles in osmoprotection .
We therefore measured the relative amounts of K + and betaine accumulated by cells grown at different osmolarities .
As the osmolarity was increased the relative role of betaine as the intracellular osmolyte became more important ( Fig. 5 ) .
In media containing 50 mM NaCl , betaine accounted for less than 10 % of the total pool of the two osmolytes , whereas at 0.8 M NaCl betaine accounted for more than 80 % of this pool .
DISCUSSION It is now well established that the transcription of two bacterial genes , kdp and proU , is determined primarily by osmotic-stress ( 7 , 15 , 17 , 26 ) .
However , in neither case is the molecular basis of this regulation understood .
The elegant studies of Epstein have led to the conclusion that expression of the kdp operon is induced in response to a lowering of cell turgor and that induction of kdp is independent of either intracellular or extracellular K + concentrations ( 26 ) .
This system provides a reference point for the investigation of the expression of other genes under osmotic control .
The finding that expression of proU , like that of kdp , is induced in cells under osmotic-stress ( 7 , 15 , 17 ) implied that this gene might also be sensitive to changes in turgor .
However , the data presented here illustrate a number of important differences between the regulation of proU and kdp expression .
( i ) Expression of kdp during-growth at low-osmolarity is induced by potassium starvation , whereas proU remains repressed .
( ii ) Expression of kdp during-growth at high-osmolarity is repressed by added potassium , whereas expression of proU remains unaffected .
( iii ) Expression of kdp in cells grown at high-osmolarity is unaffected by the addition of betaine or proline , whereas expression ofproU is significantly reduced .
These data quite clearly demonstrate a difference in the osmotic regulation of the two genes .
It is also clear from these observations that changes in turgor can not directly influence proU expression .
For example , proU is not expressed under potassium limitation , which reduces turgor .
Similarly , proU is expressed during steady-state growth at high-osmolarity , even in the presence of excess potassium which results in the substantial , if not complete , restoration of turgor .
We have now obtained strong evidence that proU expression is induced in response to an increase in intracellular K concentrations .
We have manipulated the intracellular po-tassium concentrations in a number of ways , by altering medium osmolarity , by limiting potassium availability , and by introducing potassium transport mutants .
The changes in intracellular K + concentrations resulting from these procedures have been measured and are in excellent agreement with previously published results ( 16 , 26 ; Epstein , in press ) .
We find a strong correlation between proU expression and intracellular potassium : ( i ) increases in external osmolarity , which result in a corresponding increase in intracellular K + , induce proU .
( ii ) Potassium starvation , which leads to turgor stress but does not increase intracellular K + , does not induce proU .
( iii ) At high-osmolarity , limitation of K + ( either by starvation or by introducing potassium transport lesions ) decreases proU expression , despite the fact that this limitation increases turgor stress .
It should of course be remembered that the induction curve for proU expression is sigmiodal ( 7 , 17 ) , implying cooperativity , and that consequently a precise linear relationship between K + concentrations and proU expression would not be anticipated .
To more critically demonstrate a requirement for potassium , we have shown that sudden osmotic shock fails to induce proU when potassium is limiting for growth , even though protein synthesis and the induction of other genes are unaffected by K + starvation .
Thus , the accumulation of intracellular K + seems to be both required and sufficient for the induction of proU .
The only circumstances in which the correlation between K + concentrations and proU expression breaks down is when betaine and proline are provided in the growth medium .
However , the lack of correlation here can readily be explained by roles for betaine and proline other than in turgor regulation ( see below ) .
How might intracellular potassium effect the induction of proU ?
One possibility is that the interaction of RNA polymerase with the proU promoter is directly influenced by ionic strength .
It is well known that ionic strength can alter promoter function in-vitro , and increases in ionic strength of the magnitude which occurs under severe osmotic-stress might induce structural transitions in DNA .
More probably a cytoplasmic regulatory protein is involved in modulating proU expression .
This protein would undergo a conformational change in response to increased concentrations of intracellular potassium and activate proU transcription .
Either a positive or a negative regulatory protein might be involved .
Such a conformational change could be achieved in one of two ways .
The simplest hypothesis is that K + binds to a specific site on the protein .
This site would have a binding constant for K + of around 150 mM , the concentration which results in half-maximal expression of proU ( 7 , 17 ) .
Alternatively , the conformational changes in the putative regulatory protein could simply be the result of sensitivity of the protein structure to fluctuations in ionic strength .
Such a model permits an explanation for the finding that betaine causes a greater reduction in the expression of proU than does proline , whereas both have similar effects on intracellular K + concentrations .
It is well known that betaine is a better osmoprotectant than proline ( 6 ) and that it is better able to protect proteins from ionic or other denatur-ation ( 30 , 31 , 33 , 45 ) .
The different effects of betaine and proline on proU expression may therefore reflect their relative efficiencies at protecting the putative regulatory protein from conformational changes at high internal ionic strength .
We have shown that , under the conditions used here , betaine is able to enhance growth-rates even under conditions where sufficient potassium is present to restore turgor and that therefore betaine must be providing another function such as the protection of proteins against denatur-ation .
It is interesting to note that the induction profile for proU expression is sigmoidal with respect to osmolarity ( 7 ) .
This is compatible with a model in which the putative proU regulatory protein is a dimer ( or larger oligomer ) and that renaturation-denaturation by K + involves a monomer-dimer transition .
Betaine has been shown in-vitro to facilitate the multimerization of proteins under otherwise dissociating conditions ( 21 ) .
Although unlikely , it is of course possible that K + ions act indirectly .
For example , it may be that the glutamate , which is synthesized in response to K + accumulation to restore the ionic balance ( 41 ) , provides the actual signal for proU induction .
Alternatively , the signal may simply be an effect of ionic strength within the cell , to which K + makes the major contribution .
Since the intracellular concentration of Na + is maintained at a low level compared with that of K + ( 16 ) , it is not possible to distinguish in-vivo whether K + functions in a specific manner or is simply the primary contributing ion to an overall increase in intracellular ionic strength .
We are currently testing this possibility in an in-vitro system .
However , even if the effects of K + are indirect , the accumulation of K + still remains the primary event that leads to the induction of proU expression .
The differential regulation of proU and kdp expression is entirely consistent with the different roles played by betaine and potassium in osmoregulation .
We have found that at low osmolarities potassium is the principal osmotic species , whereas at higher osmolarities betaine ( when available ) becomes more important .
Because K + , unlike betaine , is almost always available to cells it is clearly the most appropriate primary osmoprotectant and , when cells are subjected to mild osmotic-stress , is accumulated to restore turgor .
However , at high external osmolarities the correspondingly higher intracellular K + concentrations are potentially deleterious to enzyme function and can reduce growth-rates .
Thus , if it is available , betaine is taken up in preference to K + .
The accumulated betaine contributes to the restoration of turgor , reducing the need to accumulate potentially deleterious levels of K + .
In addition , betaine is known to protect proteins from ionic denaturation ( 4 , 45 ) and may therefore also serve to protect intracellular enzymes from inhibition by the K + ions which are accumulated .
Thus , the cell possesses an integrated response to achieve optimum growth at different osmolarities .
This scheme is best achieved by inducing the expression of the potassium uptake system ( kdp ) in response to changes in turgor and of the betaine uptake system ( proU ) in response to intracellular K + concentrations .
Protein synthesis during potassium starvation .
Cells of S. typhimurium CH1301 were grown in potassium-free medium as described in the legend to Fig. 1 .
At the appropriate points on the growth curve ( Fig. 1 ) , samples were osmotically shocked by adding sucrose to 0.44 M ; 1 mM K ' was added together with the sucrose where indicated .
At 30 min after the osmotic shock cells were pulse-labeled with [ 35S ] methionine , and the proteins were separated by electrophoresis on a 9 % 1 sodium dodecyl sulfate-polyacrylamide gel .
1400O 1200 F @i 1000 v0 ) u 800 F da 600 ) ) t.l a 4VVrL 200 I , 6 2 4 8 10 50 TABLE 7 .
Effects of proline and betaine on proU and kdp expressiona Medium additives P-Galactosidase activity ( U ) kdp-proU-lacZ lacZ KM Betaineproline Osmolyte ( MM ) 1 1 1 1 1 1,406 1,359 NDb ND 1,514 7 523-155-419-160 0.3 M 0.3 M 0.3 M 0.3 M NaCI 1 mM Betaine 1 mM Proline NaCl NaCl NaCl 1 mM Betaine plus 1 mM proline 5 984 406 1,108 108 0.3 M NaCl 0.3 M NaCl 1 mM Betaine plus 1 mM proline 5 129-506-109-141 0.3 M NaCl 0.3 M NaCl 10 10 1 mM Betaine plus 1 mM proline 1-4 * 12 .
Stimulation of growth by betaine and proline .
An over-night culture of E. coli MJF261 ( kdp-lacZ ) was washed and suspended at a 1:50 dilution in LOM containing 0.3 M NaCl and the indicated amounts of KCI ; 1 mM betaine and 1 mM proline were included in the medium ( A , * , E , V ) omitted ( A , O , V ) , or 0 , as appropriate .
Growth monitored by measuring the optical denwas sity at 600 nm ( O.D. 6w ) .
Potassium was at the following concentrations : 50 mM ( A , A ) , 10 mM ( 0 , 0 ) , 1 mM ( - , 10 ) , 100 , uM ( V , V ) , < 8 , uM ( x , + ) .
Relative accumulation of K + and betaine .
Cells were induce washed , and grown at high-osmolarity ( 0.3 M NaCl ) to proU , equilibrated in media containing the indicated concentrations of NaCl , and the intracellular K + and betaine concentrations were determined as described in Materials and Methods .
The results are plotted as a percentage of the total pool of K + plus betaine .
Symbols : vJ , betaine ; , K + .
We are grateful Epstein stimulating suggestions as to the possible role of ionic strength in regulating proU expression .
He also generously provided the kdp-lacZ fusion strain .
We also thank Rob Reed for assistance with the measurements of intracellular K + concentrations and Evert Bakker for many helpful discussions .
This work was supported by a research grant from the Medical Research Council to C.F.H. and I.R.B. C.F.H. is a Lister Institute Research Fellow .
Ahmed , S. , and I. R. Booth .
The use of valinomycin , nigericin and tetrachlorosalicylanilide in the control of the protonmotive force .
Resolution of bacterial proteins by polyacrylamide gel electrophoresis on slabs .
Chem proteins of Salmonella typhimurium and Escherichia coli .
Bowlus , R. D. , and G. N. Somero .
Solute compatibility with enzyme function and structure : rationales for the selection of osmotic agents and end products of anerobic metabolism in marine invertebrates .
Bukau , B. , M. Ehrmann , and W. Boos .
Osmoregulation of the maltose regulon in Escherichia coli .
Cairney , J. , I. R , Booth , and C. F. Higgins .
Salmonella typhimurium proP gene encodes a transport system for the osmoprotectant betaine .
Cairney , J. , I. R. Booth , and C. F. Higgins .
Osmoregulation of gene expression in Salmonella typhimurium : proU encodes an osmotically induced betaine transport system .
Cairney , J. , C. F. Higgins , and I. R. Booth .
Proline uptake through the major transport system of Salmonella typhimurium is coupled to sodium ions .
Transposition and fusion of the lac genes to selected promoters in E. coli using bacteriophage lambda and Mu .
Castilho , B. A. , P. Olfson , and M. J. Casadaban .
Plasmid insertion mutagenesis and lac gene fusion with mini-Mu bacteriophage transposons .
Christian , J. H. B. 1955 .
The influence of nutrition on the water relations of Salmonella oranienburg .
Clark , D. , and J. Parker .
Proteins induced by high osmotic pressure in Escherichia coli .
Proline over-production results in enhanced osmotolerance in Salmonella typhimurium .
A third L-proline permease in Salmonella typhimurium which fuinctions in media of elevated osmotic strength .
Dunlap , V. J. , and L. N. Csonka .
Osmotic regulation of L-proline transport in Salmonella typhimurium .
Epstein , W. , and S. G. Schultz .
Cation transport in Escherichia coli .
V. Regulation of cation content .
Identification of osmoresponsive genes in Escherichia coli : evidence for participation of potassium and proline transport systems in osmoregulation .
Grothe , S. , R. L. Krogsrud , D. J. McClellan , J. C. Milner , and J. M. Wood .
Proline transport and osmotic-stress response in Escherichia coli .
Hall , M. N. , and T. J. Silhavy .
The ompB locus and the regulation of the major outer membrane porin proteins of Escherichia coli K12 .
Hall , M. N. , and T. J. Silhavy .
Genetic analysis of the ompB locus of Escherichia coli K12 .
Hand , S. C. , and G. N. Somero .
Urea and methylamine effects on rabbit muscle phosphofructokinase .
Imhoff , J. F. , and F. Rodriguez-Valera .
Betaine is the main compatible solute of halophilic eubacteria .
Jamieson , D. J. , and C. F. Higgins .
Anaerobic and leucine-dependent expression of a peptide transport gene in Salmonella typhimurium .
Jamieson , D. J. , and C. F. Higgins .
Two genetically distinct pathways for transcriptional regulation of anaerobic gene expression in Salmonella typhimurium .
Cleavage of structural proteins during the assembly of the head of bacteriophage T4 .
Laimins , L. A. , D. B. Rhoads , and W. Epstein .
Osmotic control of kdp operon expression in Escherichia coli .
Landfald , B. , and A. R. Strom .
Choline-glycine betaine pathway confers a high level of osmotic tolerance in Escherichia coli .
LeRudulier , D. , T. Bernard , G. Goas , and J. Hamelin .
Osmoregulation in Klebsiella pneumoniae : enhancement of anaerobic-growth and nitrogen fixation under stress by proline betaine , a-butyrobetaine and other related compounds .
LeRudulier , D. , and L. Bouillard .
Glycine betaine , an osmotic effector in Klebsiella pneumoniae and other members of the Enterobacteriaceae .
LeRudulier , D. , A. R. Strom , A. M. Dandekar , L. T. Smith , and R. C. Valentine .
Molecular biology of osmoregulation .
LeRudulier , D. , and R. C. Valentine .
Genetic engineering in agriculture : osmoregulation .
Experiments in molecular genetics .
Cold Spring Harbor Laboratory , Cold Spring Harbor , N.Y. 33 .
Paleg , L. G. , G. R. Stewart , and R. Starr .
The effects of compatible solutes on proteins .
Perroud , B. , and D. LeRudulier .
Glycine betaine transport in Escherichia coli : osmotic modulation .
Reed , R. H. , J. A. Chudek , R. Foster , and W. D. P. Stewart .
Osmotic adjustment in cyanobacteria from hypersaline environments .
Rhoads , D. B. , F. B. Waters , and transport in Escherichia coli .
Genetic techniques in studies metabolism .
Rowland , G. C. , P. M. Gifford , and I. R. Booth .
phs locus of Escherichia coli , a mutation causing lesions in metabolism , is an rpoA allele .
Stock , J. B. , B. Rauch , and S. Roseman .
Periplasmic space in Salmonella typhimurium and Escherichia coli .
Styrvold , D. B. , P. Falkenberg , B. Landfald , M. W. Eshoo , T. Bjornsen , and A. R. Strom .
Selection , mapping and characterization of osmoregulatory mutants of Escherichia coli blocked in the choline-glycine betaine pathway .
Tempest , D. W. , J. L. Meers , and C. M. Brown .
Influence of environmnent on the content and composition of microbial free amino-acid pools .
van Alphen , W. , and B. Lugtenberg .
Influence of osmolarity of the growth medium on the outer membrane protein pattern of Escherichia coli .
Villarejo , M. , J. L. Davis , and S. Granett .
Osmoregulation of alkaline phosphatase synthesis in Escherichia coli K-12 .
Wyn Jones , R. G. , and R. Storey .
In L. G. Paleg and D. Aspinall ( ed .
) , The physiology and biochemistry of drought-resistance in plants .
Academic Press , Inc. , Sydney , Australia .
Yancey , P. H. , M. E. Clark , S. C. Hand , R. Bowlus , and G. N. Somero .
Living with water stress : evolution of osmolyte systems .