Vol. 253, No. 22, Issue of November 25, pp. 82134220, 1978 Prmted in U.S. A.

Phospholipid Composition and Membrane Function in Phosphatidylserine Decarboxylase Mutants of Escherichia cob*

(Received for publication, October 11, 1977)

Edward Hawrott and Eugene P. Kennedy

From the Department of Biological Chemistry, Harvard Medical School, Boston, Massachusetts 02115

Temperature-sensitive conditional lethal mutants in phosphatidylserine decarboxylase (psd) accumulate large amounts of phosphatidylserine under nonpermis- sive conditions (42°C) prior to cell death. In addition, the ratio of cardiolipin to phosphatidylglycerol is in- creased. At an intermediate temperature (37”C), high levels of phosphatidylserine can be maintained with little effect on cell growth or viability. Under these conditions, both the rate of induction and the function of the lactose transport system are normal. At 42°C addition of Mg2+ or Ca’+ to mutant cultures produces a partial phenotypic suppression. Growth is prolonged and the filaments normally present at 42°C do not form.

Upon transfer to the nonpermiSsive temperature, there is a considerable lag before accumulation of phos- phatidylserine begins and the growth rate is affected. Based on the kinetics of heat inactivation of phospha- tidylserine decarboxylase activity in extracts, in intact nongrowing cells, and in growing cells, it appears that the enzyme newly synthesized at 42°C is more ther- molabile in uiuo than enzyme molecules previously in- serted into the membrane at the lower temperature. Thus, the older, stable enzymatic activity must be di- luted during growth before physiological effects are observed.

Our understanding of membrane biogenesis and function has been greatly enriched by studies of Escherichia coli mutants defective in membrane lipid synthesis (for recent reviews, see Refs. 1 and 2). The isolation of fatty acid auxo- trophs has permitted extensive manipulation of the membrane fatty acids (3, 4). These auxotrophs have been widely used to examine the relationship between the physical properties of membrane lipids and membrane-localized physiological proc- esses (2). A similar study of genetically produced alterations in the polar groups of E. coli membrane phospholipids is just now beginning (5-8). Mutants are now available with blocks in the formation of the major phospholipid in E. coli, phos- phatidylethanolamine (5-9).

Phosphatidylserine decarboxylase, a membrane-bound en- zyme, catalyzes the decarboxylation of phosphatidylserine to form phosphatidylethanolamine in the terminal step of the biosynthetic pathway for E. coli (10). We have previously

* This research was supported by Grants GM19822 and GM22057 from the National Institute of General Medical Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

+ The work described here formed part of a dissertation submitted to the Faculty of Arts and Sciences of Harvard University in partial fulfillment of the requirements for the Ph.D. degree. Present address, Department of Neurobiology, Harvard Medical School, Boston, Mass. 02115.

described-the isolation and mapping of conditional lethal psd mutants having temperature-sensitive phosphatidylserine de- carboxylase activity (6, 8).

In this study, we examine in detail the alteration in phos- pholipid composition produced by the psd defect. In addition, the effects of phosphatidylserine accumulation on growth, viability, and cell morphology have been studied. A partial characterization of the thermolability of one of the psd mu- tants (psd2) suggests an unusual form of enzyme temperature sensitivity. Although the enzyme is indeed thermolabile, the mutant phenotype is similar to that seen in mutants in which the synthesis of a given enzyme is temperature-sensitive. Finally, the effects of phosphatidylserine accumulation on the induction of a membrane transport protein (Zac y gene prod- uct) and on the function of several transport systems are described.

EXPERIMENTAL PROCEDURES

Materials-Culture media were obtained from Difco, Detroit. L-

[“HlProline, L-[‘4C]glutamine, and ‘lP, were supplied by New England Nuclear, Boston. Methyl-a-D-[‘%]glucose (a-methyl glucoside) was obtained from Amersham/Searle, Chicago. P-D-[,‘H]Galactosyl-l- thio-D-P-galactoside (thiodigalactoside) was prepared by the oxida- tion of unlabeled thiodigalactoside with galactose oxidase, followed by reduction with tritiated borohydride, according to the general procedures described by Kennedy et al. (11). Phosphatidyl- [l-‘%]serine was prepared enzymatically as previously described (10). Triton X-100 (octylphenoxy polyethoxyethanol) was a product of Rohm and Haas, Philadelphia.

Media and Strains-Bacterial cultures were grown in a low phos- phate medium, Medium 56.LP (12) containing 0.3 mM phosphate when efficient labeling of cells with ‘lP1 was required. For other experiments, cells were grown in L broth (13). Carbon sources and supplements were added as indicated.

Bacterial strains used in this study were all derivatives of E. coli K-12 and are listed in Table I. In general, all experiments compared temperature-sensitive psd mutants with isogenic wild type (psd+) transductant strains. The turbidity of bacterial cultures was measured with a Coleman Jr. spectrophotometer at 650 nm.

Analysis of Phospholipids-Phospholipids were extracted from bacterial cultures uniformly labeled with “P) by means of the Ames modification (14) of the Bligh and Dyer procedure (15). The phos- pholipid classes were then separated by thin layer chromatography as described previously (6). Medical x-ray fiim (Kodak NSMT and RP/R-54) was used for radioautography of the chromatograms.

Transport Assays-The release of o-nitrophenol from o-nitro- phenyl-P-n-galactoside was used as one measure of lactose transport function since transport of this /?-galactoside across the membrane is the rate-limiting step in the release of o-nitrophenol via the action of intracellular P-galactosidase (16, 17). As an index of the nonspecific leakiness of membranes, hydrolysis of o-nitrophenyl-P-n-galactoside was measured in cells incubated in the presence of excess thiodigalac- toside, a substrate that inhibits the transport of o-nitrophenyl-/-D- galactoside mediated by the lac permease.

Portions (0.5 ml) of the cultures were taken directly from the growth flasks and added to 1.5 ml of medium containing chloram- phenicol (100 pg/ml) and o-nitrophenyl-,&n-galactoside (2 mM). In control experiments, thiodigalactoside (IO mM) was included in the assay mix. After a lo-min incubation at 37”C, 3.5 ml of 1 M KzCO~

8213

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

8214 Phosphatidylserine Decarboxylase Mutants of E. coli

TABLE I Strains of Escherichia coli K-12

Relevant characteristics Strain Source

ES4 EH150

EH150-RI

EH160

EHlGO-Rl

EH170

EH170-RA3

M2508 EH250 EH251 w3110 EH450 EH451, EH471 EH470

purA psd2 derivative (temperature-sensi-

tive phosphatidylserine decarbox- ylase) of ES4

Spontaneous revertant of EH150 that grows normally at 42OC

psd3 derivative (temperature-sensi- tive phosphatidylserine decarbox- ylase) of ES4

Spontaneous revertant of EH160 that grows normally at 42’C

psd4 derivative (temperature-sensi- tive phosphatidylserine decarbox- ylase) of ES4

Spontaneous revertant of EH170 that grows normally at 42’C

Hfr, metB, melA, rel melA’, psd2 transductant of M2508 melA’, psd’ transductant of M2508 Prototroph psd2 transductant of W3110 psd+ transductants of W3110 psd4 transductant of W3110

CGSC” Refs. 6 and 8

See text

Ref. 8

See text

Ref. 8

See text

CGSC Ref. 8 Ref. 8 D. Fraenkel Ref. 8 Ref. 8 Ref. 8

n E. coli Genetic Stock Center, Yale University, New Haven, Conn.

was added and the cells were removed by centrifugation. The for- mation of o-nitrophenol was measured in a Gilford spectrophotometer at 420 nm.

The rates of induction of lactose transport and P-galactosidase activity were determined as described by Nunn and Cronan (17). After a 17-min induction period with thioisopropyl-/s-galactoside, portions (0.5 ml) of the cultures were transferred to 0.5 ml of medium containing chloramphenicol and o-nitrophenyl-P-o-galactoside as de- scribed above. Controls with added 10 IXM thiodigalactoside were also included. For assay of P-galactosidase activity, cells were first dis- rupted by addition of sodium dodecyl sulfate and CHCI,l (18). After incubation with substrate for 17 min at 37”C, 1.5 ml of 1 M KzCO.~ was added and the extent of o-nitrophenyl-,&n-galactoside hydrolysis was determined as described above.

In addition, the function of the lactose system, as well as that of three other transport systems, was measured by the uptake and accumulation of radioactive substrates. Cells (about lo”), suspended in 2.1 ml, were incubated at 37°C for 5 min before the substrate was added in a volume of 0.3 ml. Samples (0.5 ml) were withdrawn, filtered on 25.mm nitrocellulose falters (0.45 pm, Matheson-Higgins or Milli- pore Corp.), and quickly washed twice with 5 ml of Medium 56-LP at 25°C. Controls to determine the level of nonspecific uptake were carried out by the addition of high levels of unlabeled substrate to the system. Radioactivity trapped on the filters was determined by liquid scintillation counting.

Other Procedures-Phosphatidylserine decarboxylase activity was assayed as previously described (6, 19). Mutant psd extracts were assayed in the presence of glycerol which partially stabilizes the mutant activities (6, 8). Extracts were prepared by sonic irradiation with either an MSE-100 ultrasonic disintegrator or a Branson Sonifier. Protein was determined by the method of Lowry et al. (20).

RESULTS

Alterations in Phospholipid Composition in psd Mutant Strains-As previously reported (6), the most dramatic change in the phospholipid composition of psd mutant strains is the accumulation of high levels of phosphatidylserine at nonpermissive temperatures. As indicated in Table II, after 4 h growth at 42”C, all three temperature-sensitive mutants (psd2, psd3, and psd4) accumulate phosphatidylserine to lev- els at least 50-fold higher than normal when compared to revertant strains. The accumulation of phosphatidylserine, in general, corresponds to the decrease of phosphatidylethanol- amine as expected from the pathway of biosynthesis proposed for this organism (10). The psd4 strain, EH170, accumulates an intermediate level of phosphatidylserine at 30°C suggesting

that phosphatidylserine decarboxylase in this strain is rate- limiting for phosphatidylethanolamine synthesis even at this temperature. The phenotypic expression of thepsd4 mutation at 30°C is apparently quite sensitive to the genetic back- ground. Different levels of phosphatidylserine accumulation at 3O”C, varying from 2 to 20 mol R of total lipid, have been observed in different strains into which thepsd4 mutation has been transduced.

In addition to the accumulation of phosphatidylserine by the psd mutants, there is also an increase in the ratio of cardiolipin to phosphatidylglycerol as compared to revertant and wild type strains. For strains EH150 psd2 and EH160 psd3 grown at 42”C, this ratio is increased by a factor of 2 or 3 over and above the normal 2-fold increase seen upon transfer of control cultures from 30°C to 42°C (Table II and see Fig. 4). Strain EH170psd4, with a higher level of phosphatidylser- ine accumulation at 42”C, shows a correspondingly greater increase (6- to 7-fold) in the ratio of cardiolipin to phosphati- dylglycerol in a number of experiments.

Growth and Viability of a Temperature-sensitive psd2

Strain-A culture of mutant strain, EH250 psd2, shifted to 42°C from the permissive temperature of 3O”C, continues to

TABLE II

Phospholipid composition ofpsd mutant and revertant strains at 30°C and 42°C

Cellular phospholipids were uniformly labeled by growth of bac- terial cultures for several generations at 30°C in Medium 56-LP containing ,12Pa (specific activity of 11 Ci/mol). At a cell density of 7 x lO’/ml, the cultures were divided. Half of each culture was shifted to 42°C for 4 h, while growth of the other half continued at 30°C. Phospholipids were extracted and analyzed by thin layer chromatog- raphy. After location of the phospholipids by radioautography, the appropriate regions of the chromatograms were cut out and the radioactivity was determined by liquid scintillation counting. The cultures at 30°C were growing exponentially at the time of the phospholipid extraction. Except for strain EH170psd4, all the cul- tures incubated at 42°C had entered the late log phase of growth after 4 h. Strain EH170 psd4 had ceased growth at the time of lipid extraction. All the chromatograms contained some radioactive ma- terial remaining at the origin. The unidentified material amounted to less than 1% of the total lipid-soluble radioactivity present in cultures grown at 30°C. However, in mutant cultures grown at 42”C, this material on occasion represented as much as 10% of the total chlo- roform-extractable radioactivity. The presence of this labeled mate- rial was not considered in the calculation of the results presented. Revertant derivatives of the mutant strains were spontaneous revert- ants selected by growth at 42°C on agar plates.

Distribution of phospholipid classes

Strain and growth temper- Phospha- Phospha- Phospha-

ature tidyl- tidyl- ethanol- tidyl-

serine amine glycerol

EH150 psd2 30°C 42°C

EH150-Rl 30°C 42°C

EH160 psd3 30°C 42°C

EHlGO-RI 30°C 42°C

EH170 psd4 30°C 42’C

EH170-RA3 30°C 42’C

B of total

1 82 25 59

to.5 84 to.5 84

1 85 21 64

<0.5 83 co.5 83

20 67 48 31

<0.5 86 <I 85

Cardio- lipin

14 3 8 8

14 12

11 3 8 7

13 12

10 7

11 9

2 4

4 5

3 14

3 6

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

Phosphatidylserine Decarboxylase Mutants of E. coli 8215

Id IB

I 2 3 4 5 6 7 8 9 IO II ” I n n 1 ’ ’ ’ ’ ’ ’ ’ 21

HOURS AFTER SHIFT TO 42’

FIG. 1. Growth and viability of mutant strain EH250psd2 at 42°C. Mutant, EH250psd2, and isogenic control, EH251psd’, cultures were grown at 30°C in Medium 56-LP containing 1% glycerol and supple- mented with methionine. At a cell density of 2.5 x 107/ml, the cultures were transferred to an incubator at 42°C. Panel A, bacterial growth at 42°C was measured as the increase in turbidity at 650 nm. Panel B, viable cell count was determined after serial dilution in Medium E (21) by plating diluted samples on plates of supplemented Medium E. Colony formers were scored after incubation at 30°C for 36 to 48 h. The relative viability was normalized by dividing the viable count by the culture turbidity (A&. A, EH251 psd+; l , EH250psd2.

grow at a normal rate for approximately three generations (Fig. 1). The number of generations of growth observed after shift to 42°C varied with the medium. In general, richer media allowed longer growth at 42’C. As Fig. 1A indicates, after about 5 h of growth at 42”C, growth declined in rate and eventually came to a complete halt. Macromolecular synthesis in mutant cells at 42“C, as measured by net synthesis of DNA, RNA, protein, and phospholipid follows the same general time course as the increase in culture turbidity (data not shown). Upon continued incubation of cultures at 42’C, a sharp drop in viability occurs (Fig. 1B).

FIG. 2. Light microscopy of mutant EH170 psd4 cells at 42’C. A culture of strain EH170 psd4 was grown at 42°C in L broth into stationary phase. A sample was air-dried, heat-fixed, and stained by the Gram procedure. The length of a single bacterium is about 2 pm.

As reported previously (22), cell division is also affected in psd mutant cells under conditions of phosphatidylserine ac- cumulation such as in the experiment of Fig. 1. Chains of unseparated mutant cells form upon growth at 42°C (Fig. 2). Separations in these filaments are often irregularly spaced along the length of the chain. The chain formation of mutant psd cultures may explain a small part, but certainly not all, of the observed decrease in viable cell count seen in Fig. 1B where the viability fell by a factor of lo4 after 21 h. In some cases, clumping of cells into aggregates also occurs. This is especially true of the mutant psd transductants of strain W3110. The reason for this aggregation has not been explored further.

Time Course of Effects on Phospholipid Synthesis-Since the temperature-sensitive psd mutant, EH250, continues to grow in liquid culture for a considerable length of time at 42“C (Fig. l), it was of obvious importance to determine the factors governing the relatively long delay in lethal effect. It would be especially useful to know the time at which signifi- cant levels of phosphatidylserine first appear in mutant cells after a shift to 42”C, since the physiological effects of a defect in phosphatidylserine decarboxylase would not be expected to appear prior to this accumulation and might possibly appear only some time later.

The synthesis of phosphatidylethanolamine and phospha- tidylserine was followed in EH250 psd2 and EH251 psd” grown as described in the experiment of Fig. 1 but labeled with “Pi from the time of shift to 42°C. As indicated in Fig. 3A, the earliest accumulation of phosphatidylserine occurred only after 3 h of growth at 42% Correspondingly, incorpora- tion of “Pi into phosphatidylethanolamine was normal over the initial 3 h. With further incubation at 42’C, synthesis of phosphatidylethanolamine was curtailed (Fig. 3B), whereas phosphatidylserine continued to accumulate. After 3 h at 42”C, the total synthesis of phospholipid per ml of culture was markedly reduced in rate (Fig. 3C). Both phospholipid (Fig. 3C) and protein (data not shown) synthesis were affected somewhat prior to the effect on culture turbidity (Fig. 1A). The apparent delayed response of the turbidity of the culture

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

8216 Phosphatidylserine Decarboxylase Mutants of E. coli

t 1 FIG. 3. Synthesis of phosphatidylserine and phosphatidylethanol-

amine in strains EH250 psd2 and EH251 psd’ at 42’C. Bacterial cultures were grown at 30°C in Medium 56.LP in parallel to those described in Fig. 1. At the time of shift to 42”C, “lP, was added at a final specific activity of 11 Ci/mol. At the appropriate times, samples (0.8 ml) of the cultures were removed and the phospholipids were extracted. One-fourth of this extract was used for the analysis of

may be related to the formation of bacterial filaments in the mutant culture.

Fig. 4 summarizes the data on “2Pi incorporation into all four phospholipid classes upon transfer to 42°C. At the end of the experiment, the phospholipids were uniformly labeled with “2PI and the final phospholipid composition was deter- mined. After a 9-h incubation (Fig. 4), the composition (in mole per cent) for strain EH250 pad2 was: phosphatidyletha- nolamine, 37%; phosphatidylserine, 34%; phosphatidylglyc- erol, 12%; and cardiolipin, 17%.

Heat Inactivation of Mutant Phosphatidylserine Decar- boxylase-In order to gain further information on the lag in physiological effect observed in cells of mutant psd2 upon shift to 42°C (Fig. l), it was necessary to determine the extent of inactivation of phosphatidylserine decarboxylase in vivo at nonpermissive temperatures. Mutant psd2 cells, in the ab- sence of growth, are not killed to any greater extent than wild type cells even after a 22-h exposure to a temperature of 42°C (data not shown). This is in marked contrast to the killing observed during growth as seen in Fig. 1.

Cells of mutant psd2 were incubated at 42°C for various periods in the presence of chloramphenicol, added to prevent growth. When phosphatidylserine decarboxylase was then assayed in extracts of these cells, it was found that the activity had decreased only at a very slow rate (Fig. 5). The half-life for inactivation of the enzyme in intact cells was on the order of 12 h. This rate is much too slow to account for the time course of the observed physiological effects. Furthermore, the mutant psd2 decarboxylase bound to the membrane in cell- free extracts is resistant to heat inactivation, like the enzyme from the originally isolated “silent” mutantpsdl (6). Although the mutant psd2 phosphatidylserine decarboxylase is present at an apparent reduced specific activity, in crude extracts it is no more sensitive to heat inactivation than wild type enzyme as long as no detergent is added (Fig. 6). However, when the heat treatment is carried out in the presence of 0.1% Triton X-100, the mutant psd2 activity is rapidly inactivated with a half-life on the order of 25 s while the half-life for inactivation of wild type enzyme is about 80 min. The detergent effect in extracts, together with the results from enzyme inactivation

phospholipids by thin layer chromatography. Panels A and B, the amount of “‘Pi incorporated into phosphatidylserine and phosphati- dylethanolamine was determined as described in Table II. Significant accumulation of phosphatidylserine at the first time point could have been easily detected with the analytical procedures that were used. Panel C, synthesis of total phospholipid. A, EH251 psd’; 0, EH250 psd2.

PG CL 40,."""'., , . . . I I 8. . ,140

d I23456789

g40

8 20

PE

60

L-A-=d~ - .I I.‘.‘.“..1 EH251

123456789 I23456769 HOURS HOURS

FIG. 4. Incorporation of ‘“P, into the individual phospholipid com- ponents upon transfer of EH250 psd2 cells to 42°C. The bacterial phospholipids were labeled with J2Pb, extracted, and analyzed as described in Fig. 3. The radioactivity in each phospholipid is presented as the percentage of the “PI incorporated into total phospholipid at each given time point. The values for cardiolipin have been corrected for its content of 2 mol of phosphate/mol. PG, phosphatidylglycerol; CL, cardiolipin; PS, phosphatidylserine; PE, phosphatidylethanola- mine. 0, EH251 psd+; A, EH250 psd2.

in whole cells, suggest that the mutant psd2 phosphatidylser- ine decarboxylase is resistant to heat inactivation when mem- brane-bound. Once extracted from the membrane, however, it is markedly thermolabile.

Mutant (psd2) levels of activity are similar whether or not extracts are previously incubated in Triton at 28°C (data not shown). This is true for the glycerol-containing assay system used in the heat-inactivation studies. Therefore, the psd2 mutation does not produce Triton sensitivity per se ( i.e. at all temperatures).

The physiological lag in accumulation of phosphatidylserine can now be explained by the following hypothesis. It would appear that mutant psd2 phosphatidylserine decarboxylase synthesized at the permissive temperature in vivo is incorpo-

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

Phosphatidylserine Decarboxylase Mutants of E. coli

1234567 HOURS OF PRIOR INCUBATION

EH450 psd2 ‘ -

0 - 30-I 8 I

I

s 20- jw50 psd2 t TRITON X-100 I I I / 1 I

‘O ’ 30 I

60 90 120 MINUTES OF PRIOR INCUBATION AT 43”

OF CELLS AT 43”

FIG. 5 (left). Heat inactivation of phosphatidylserine decarboxyl- ase in whole cells of EH450 psd2 and EH45 1 psd+. Bacterial cultures were grown at 3O’C to a cell density of 5 x 10H/ml. Cells were harvested and resuspended at lO”/ml in Medium 56%LP containing chloramphenicol (115 pg/ml). The cell suspension was placed in an incubator at 43°C and at the appropriate times samples (4 ml) were removed and the cells disrupted by sonic irradiation. Phosphatidyl- serine decarboxylase activity was assayed at 28°C in the presence of 20%1 glycerol (6, 8). The specific activities of the unheated samples were: 27.1 nmol of CO2 released/min/mg of protein for EH451 psd’ and 10.7 for EH450 p&2. 0, strain EH451 ps&; A, strain EH450

psd2.

FIG. 6 (center). Heat inactivation of phosphatidylserine decarbox- ylase in sonicates of EH450 psd2 and EH451 psd+. Bacterial cultures grown at 30°C were harvested and the cells resuspended at a density of lO”‘/ml. Extracts were prepared by sonic disruption. The irrevers- ible heat inactivation of phosphatidylserine decarboxylase was deter- mined by incubating these preparations (about 1.5 mg of protein/ml)

rated into the membrane and thus remains nearly completely functional during subsequent incubation at elevated temper- atures. On the other hand, the decarboxylase synthesized during incubation at 42°C is presumably rapidly inactivated at some point prior to effective insertion into the membrane. In this way, the overall phenotype should resemble that seen with mutants containing temperature-sensitive defects in en- zyme synthesis (23). The appearance of physiological effects is thus dependent on dilution through growth of the stable enzyme to a point where the enzyme function becomes rate- limiting. This hypothesis was tested by determining the rate of decline of phosphatidylserine decarboxylase specific activ- ity (activity/mg of cell protein) as measured in extracts at the low temperature, in growing cells after transfer to the nonper- missive temperature. It was observed that the decline of specific activity was much more rapid in growing cells than in nongrowing cells (compare Fig. 7 and Fig. 5). Since the mem- brane-bound mutant decarboxylase is fairly resistant to heat treatment of whole cells (Fig. 5), it would appear that the rapid decline of specific activity in growing cells (Fig. 7) is primarily due to a differential heat sensitivity of the newly synthesized mutant phosphatidylserine decarboxylase or to some other effect of growth.

Function of Transport Systems--In order to ascertain the effect of accumulated phosphatidylserine upon membrane function, the activities of several transport systems were ex- amined. For these studies, strain EH470 psd4 was grown at 37”C, a temperature intermediate between the permissive and nonpermissive conditions. Whereas at 3O”C, strain EH470 psd4 contains phosphatidylserine at a level of about 2% of the total phospholipid, at 37°C phosphatidylserine is accumulated

:,;I , , , I 2 3

HOURS OF GROWTH AT 43”

8217

at 43°C for the appropriate time intervals and with the indicated additions. Samples were then quickly placed in an ice bath and the initial rate of phosphatidylserine decarboxylase activity was assayed at 28’C in the presence of 20% glycerol. The specific activities for the unheated samples were: 2 1.4 nmol of COr released/min/mg of protein for EH451 psd+ and 9.0 for NH450 psd2. A, EH451 psd+; 0, EH450 psd2; 0, EH451 psd+ + 0.1% Triton X-100; - - -, EH450psd2 + O.l’%, Triton X-100.

FIG. 7 (r@t). Specific activity of phosphatidylserine decarboxyl- ase in EH450 psd2 and EH451 psd’ cells grown at 43°C. Bacterial cultures grown at 30°C were shifted to 43’C at a cell density of about 5 x lO’/ml. At the times indicated, samples of the growing cultures were removed and the cells were disrupted by sonic irradiation after harvesting by centrifugation. Initial rates of phosphatidylserine de- carboxylase activity were measured at 28°C in the presence of a protective level of glycerol. The initial specific activity of the EH450

psd2 extract was approximately 20% of that in the wild type extract. 0, EH451 psd+; A, EH450 psd2.

to a steady state level of from 13 to 26% of the total phospho- lipid (data not shown). However, in terms of plating efficiency, viability, and growth in liquid culture, 37°C can be considered a permissive temperature for strain EH470psd4. The elevated levels of phosphatidylserine have no, or little, effect on the growth rate in liquid culture, nor on plating efficiency on broth plates at 37°C. In comparison, strain EH450 psd2 will not survive plating at 37°C. The viability of strain EH470 psd4 as measured by colony-forming ability at 30°C after growth at 37°C in liquid culture is usually 70 to 90% of control values. This slight reduction in apparent viability may how- ever be due to the increased filamentation of strain EH470 psd4 under these conditions. It was hoped that by studying membrane function in strain EH470 psd4 at 37”C, the non- specific effects of the psd mutations resulting in the death of cells would be minimized, while the specific effects of phos- phatidylserine accumulation would be accessible to analysis.

Mutant psd4 cells grown at 37°C and containing phospha- tidylserine at a level of 13% of the total phospholipid were tested for function of four different transport systems, as shown in Fig. 8. Clearly, this level of phosphatidylserine had no effect on the intracellular accumulation of thiodigalacto- side, a substrate of the lactose transport system (16). In contrast, the rate and extent of accumulation of L-proline, L-

glutamine, and a-methyl glucoside were all somewhat de- creased in the mutant psd4 cells with the largest effects being observed with L-glutamine and a-methyl glucoside.

As another index of transport function in mutant cells with altered membrane phospholipid composition, the so-called “downhill” transport of another substrate of the lactose sys- tem, o-nitrophenyl-/3-o-galactoside was examined. In the ex-

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

8218 Phosphatidylserine Decarboxylase Mutants of E. coli

6

Y. I

MINUTES 5 IO 15

FIG. 8. Transport activity in whole cells of EH470psd4 and EH471 psd+ grown at 37°C. Bacterial cultures grown in parallel at 30°C were diluted and grown for 7 h in minimal medium (described in Table III) containing 1 mM thioisopropyl-P-galactoside. At a density of about 5 x lO’/ml, cells were harvested by centrifugation and resuspended in Medium 56-LP containing 0.5% glycerol and chloramphenicol (50 pg/ml). A parallel culture was used for a determination of phospha- tidylserine content as described in Table II. Uptake and accumulation of radioactive substrates were determined by Millipore fitration (see “Experimental Procedures”). Substrates were prepared at the follow- ing concentrations and specific activities: 20 mM ,L-[“H]galactosyl-l- thio-D-P-galactoside (thiodigalactoside) (150 cpm/nmol), 1 mM methyl(a-o-[‘%]gluco)pyranoside (n-methyl glucoside) (1500 cpm/nmol), 0.2 mM L-[“Hlproline (12,ooO cpm/nmol), and 0.2 mM L- [‘%]glutamine (35,000 cpm/nmol). Viability of the mutant culture was approximately 80% of the wild type level. 0, EH470 p&44; A, EH471 psd’.

periment described in Table III, mutant psd4 cells attained a phosphatidylserine content of 24% of total phospholipid. Thus, even with nearly a quarter of the phospholipid as phosphati- dylserine, no adverse effect on o-nitrophenyl$-o-galactoside transport was seen. In addition, control experiments with added thiodigalactoside indicated that non-carrier-mediated entry of o-nitrophenyl$o-galactoside and extracellular leak- age of ,B-galactosidase had not increased in the mutant cells.

Induction of the lac Transport System-Although the ac- cumulation of intermediate levels of phosphatidylserine has little effect on the lac transport function per se, the possibility remained that phosphatidylserine might interfere with proper insertion of carrier proteins and other membrane proteins into the membrane. To examine this possibility, the rate of induc- tion of the Zac transport system was determined in mutant psd4 cells at 37°C under conditions where the phosphatidyl- serine content was 19% of total phospholipid. As indicated in Table IV, the rate of induction of transport function in psd4 cells, after normalization to the amount of induced P-galac- tosidase activity, was identical to that of wild type cells.

Phenotypic Suppression-The addition of cations such as Mg’+ or Ca2+ to the growth medium produces a partial phe- notypic suppression of the temperature sensitivity of the psd mutants. In the case of the psd4 mutation, Ca”’ appears to stabilize the enzyme in uiuo. This is suggested by the results of an experiment wherepsd4 cells, after transfer to 37”C, were grown in minimal medium in the presence or absence of 10 mM CaC12. In the control cells, grown without Ca2+, phospha- tidylserine amounted to 31% of the cellular phospholipid after 4’/L h of growth. At the same time, the mutant psd4 cells grown in 10 XTIM CaC12 had a phosphatidylserine content of

only 6%. Moreover, the growth rate was unaffected by this level of Ca”. The fact that Ca” reduces the level of accu- mulated phosphatidylserine in psd4 cells suggests that a higher level of decarboxylase activity is present in these cells as compared to the controls. It is likely, therefore, that Ca’)+ acts in this case by protecting the mutant psd4 decarboxylase from heat inactivation. These findings, together with the previously discussed dependence of phenotypic expression of the psd4 mutation on genetic background, suggest that the psd4 mutant decarboxylase is highly sensitive to the mem- brane environment.

On the other hand, phenotypic suppression may result from a stabilization of the phosphatidylserine-containing mem- brane. Strain EH450 psd2, when transferred to 42°C in rich medium containing 0.8 mM Mg’+, ceases growth after about five doublings. At this point, the phosphatidylserine content of the mutant culture was 48 mol 70. MgS04 was added at a final concentration of 20 mM to a parallel culture upon transfer to 42°C and growth continued for several generations more than in the untreated culture. The final level of phosphati- dylserine attained in the Mg’+-treated culture was increased to 76 mol %, while phosphatidylethanolamine was reduced to only 5% of the total phospholipid. The addition of Mg” under these conditions brings about the greatest accumulation of phosphatidylserine so far observed. It is as yet unclear whether the two different modes of phenotypic suppression are due to differences in sensitivity of the two mutants psd2 and psd4 or rather due to differences in the effects of Ca’+ versus Mg”.

In addition to their effect on increasing growth yield, both Mg”+ and Ca2+ also serve to suppress the filamentation that is coincident with phosphatidylserine accumulation (Fig. 2).

TABLE III Transport of o-nitrophenyl-P-D-galactoside

Bacterial cultures were grown in Medium 56.LP containing 0.5% glycerol, 0.05% casamino acids, and thiamin (1 pg/ml). To induce the lactose operon, thioisopropyl-/3-galactoside was added to the growth medium at a concentration of 1 mM. After overnight growth at 3O”C, the cultures were diluted and grown at 37’C for 4% h. Samples were taken directly from the growing cultures and the transport-limited hydrolysis of o-nitrophenyl-,&D-galactoside was determined in the presence or absence of excess thiodigalactoside. Viability of the mu- tant culture, in which fdaments were present, was 80% of the control culture.

Strain Culture turbid- Transport of o-ni- Control with

ity trophenyl+n- added thiodiga- galactoside lactoside

EH471 psd+ EH470 asrl4

A‘X

0.078 0.070

nmol/mm/mgprotem 603 179 651 109

TABLE IV Znduction of the lactose transport system

Cultures were grown at 30°C in Medium 56.LP supplemented as described in Table III except that no thioisopropyl-P-galactoside was added. The cultures were diluted and subsequently grown at 37°C for 3% h. At a cell density of about 2 x lO”/ml, thioisopropyl-P-galactoside was added at a final concentration of 1 rnM in order to induce the lactose operon. The rate of induction of P-galactosidase and y per- mease was then determined. In this experiment, mutant viability was 72% of the control culture with filaments again being present in the mutant culture.

Strain bidity phenyl-/J-n- ealactoside

sidaszyactlv- P.galacto. sidase

EH45 1 psd’ EH470 asd4

A,,,,~ nmol/min/mgprotein 0.050 222 950 0.042 172 723

0.23 0.24

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

Phosphatidylserine Decarboxylase Mutants of E. coli 8219

Few filaments are seen in mutant psd cultures grown at nonpermissive temperatures in the presence of Mg2’ or Ca”+ despite the continued accumulation of phosphatidylserine.

DISCUSSION

The cytoplasmic membrane of E. coli is the site of a large number of processes and systems essential to the life of the cell. These include the highly organized electron transport system, the energy transduction and conservation systems associated with it, a large number of specific transport sys- tems, and numerous biosynthetic processes. A priori, it might therefore have been anticipated that the lipid composition of the membrane must be rather precisely and specifically reg- ulated if the cell is to continue to live and grow. The present studies, however, reveal the rather surprising fact that the lipid composition may vary quite widely without preventing continued growth.

In the mutants studied, phosphatidylserine replaced phos- phatidylethanolamine during growth at the nonpermissive temperatures. Ordinarily, phosphatidylserine is present only in traces in E. coli since it is decarboxylated almost as rapidly as it is formed. Phosphatidylserine is strikingly different from phosphatidylethanolamine, in that it possesses an additional net negative charge at neutral pH. One of the mutants (psd4) can be grown at 37°C for indefinite periods. Under these conditions, phosphatidylserine constitutes about 13 to 26% of the total phospholipid. Obviously, the presence of these levels of phosphatidylserine is not incompatible with the vital func- tions of the membrane. More strikingly, a psd2 mutant grow- ing in medium containing 20 mM MgS04 did not cease growth until the content of phosphatidylserine reached 76 mol %, while the content of phosphatidylethanolamine had been re- duced to only 5%.

The sequestration of accumulated phosphatidylserine in some structure of the cell other than the cytoplasmic mem- brane might offer an explanation for the resistance of mem- brane function to its presence. However, the accumulated phosphatidylserine, like the phosphatidylethanolamine that it replaces, is almost equally distributed between the inner, cytoplasmic membrane, and the outer membrane.’

The conclusion that the functions of the membrane of E. coli are surprisingly resistant to variations in the composition of membrane lipid is borne out by studies of the Zac transport system, and certain other transport systems. Neither the induction nor the function of the Zac system is affected by the presence of 19 to 24 mol % of phosphatidylserine in the membrane lipids.

Systems for the uptake of a-methyl glucoside, L-proline, and L-glutamine were somewhat reduced in activity in cells which had accumulated phosphatidylserine, but it is not easy to determine whether this reduction is a direct or an indirect consequence of the presence of phosphatidylserine in the membrane. The experiments cited above in which growth continued at relatively high levels of phosphatidylserine sup- port the argument that transport systems in general, although perhaps somewhat reduced in activity, nevertheless function at a rate permitting continued growth.

In the available mutants, the accumulation of phosphati- dylserine begins only after a considerable lag. This observation is consistent with the proposal that phosphatidylserine decar- boxylase activity is in relative excess in cells as compared to phosphatidylserine synthetase activity; a conclusion that is supported by the fact that phosphatidylserine is present only in traces during the growth of wild type cells. Thus, most of the phosphatidylserine decarboxylase activity would have to

I E. Hawrot and E. P. Kennedy, manuscript in preparation.

be lost before any effect on phosphatidylethanolamine syn- thesis would be observed.

Phosphatidylserine decarboxylase appears to be greatly sta- bilized against thermal inactivation when it is localized in the membrane, whether in intact, resting cells (Fig. 5) or in membrane-containing cell-free extracts (Fig. 6). In contrast, there is a rapid drop of specific activity (activity/mg of cell protein) in growing cells at 43°C (Fig. 7). This finding argues that it is the newly synthesized enzyme which is acutely thermolabile. I f all of the newly synthesized decarboxylase were immediately inactivated, the time needed for the specific activity of the enzyme to fall to one-half of its original value (85 min in the experiment of Fig. 7) would be equal to the doubling time for the culture, which was in fact about 60 min. This finding suggests that some of the newly synthesized decarboxylase escapes inactivation and is inserted into the membrane.

Under nonpermissive conditions, the sum of phosphatidyl- serine and phosphatidylethanolamine remains constant in the psd mutants. Thus, the accumulation of phosphatidylserine does not affect the relative rate of synthesis in the phospha- tidylglycerol-cardiolipin branch of the biosynthetic pathway for the lipids of E. coli (10). The increase in the ratio of cardiolipin to phosphatidylglycerol, typical of thepsd mutants at 42”C, may be a result ‘of a general response to growth- limiting conditions as has been previously observed (24).

The specific site of the lethal action of phosphatidylserine in the psd mutants remains unknown. In general, during phosphatidylserine accumulation, the mutant cell membrane appears to remain functionally intact and growth slows grad- ually before finally stopping. The amount of phosphatidylser- ine accumulated at the time of growth cessation is somewhat variable but is in the range of 35 to 50 mol %. The fact that higher levels of phosphatidylserine accumulation are obtained in Mg”+-treated cultures suggests that Mg”+ ions may be shielding the negatively charged phosphatidylserine and thus permitting further synthesis and incorporation of this anionic phospholipid. Under these conditions of increased phospha- tidylserine accumulation, growth of mutant cultures is, in fact, prolonged. The partial phenotypic suppression of the growth effect by cations suggests that ultimately it is not the lack of a critical amount of phosphatidylethanolamine that is lethal to the mutant cell, but instead the accumulation of negatively charged phosphatidylserine is itself responsible for the toxic effects.

REFERENCES

1. SiIbert, D. F. (1975) Annu. Rev. Biochem. 44, 315-340 2. Cronan, J. E., Jr., and Gelmann. E. P. (1975) Bacterial. Rev. 39.

232&6 3. Silbert, D. F., Ruth, F., and Vagelos, P. R. (1968) J. Bacterial.

95, 1658-1665 4. Silbert, D. F., Cronan, J. E., Jr., Beacham, I. R., and Harder, M.

E. (1974) Fed. Proc. 33, 1725-1732 5. Ohta, A., Okonogi, K., Shibuya, I., and Maruo, B. (1974) J. Germ.

Appl. Microbial. 20, 21-32 6. Hawrot, E., and Kennedy, E. P. (1975) Proc. N&l. Acad. Sci. U.

S. A. 72, 1112-1116 7. Raetz, C. R. H. (1976) J. Biol. Chem. 251, 3242-3249 8. Hawrot, E., and Kennedy, E. P. (1976) Mol. & Gen. Genet. 148,

271-279 9. Ohta, A., Shibuya, I., Maruo, B., Ishinaga, M., and Kito, M. (1974)

Biochim. Biophys. Acta 348,449-454 LO. Kanfer, J., and Kennedy, E. P. (1964) J. Biol. Chem. 239,

1720-1726 11. Kennedy, E. P., Rumley, M. K., and Armstrong, J. B. (1974) J.

Biol. Chem. 249, 33-37 12. Cronan, J. E., Jr., Birge, C. H., and Vagelos, P. R. (1969) J.

Bacterial. 100, 601-604 13. Lennox, E. S. (1955) Virology 1, 190-206

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

8220 Phosphatidylserine Decarboxylase Mutants of E. coli

14. Ames, G. F. (1968) J. Bacterial. 95, 833-843 15. Bligh, E. G., and Dyer, W. J. (1959) Can. J. Biochem. Physiol.

37,911-917 16. Kennedy, E. I’. (1970) in The Lactose Operon (Beckwith, J., and

Zipser, D., eds) pp. 49-92, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York

17. Nunn, W. D., and Cronan, J. E., Jr. (1974) J. Biol. Chem. 249, 724-731

18. Miller, J. H. (1972) Experiments in Molecular Genetics pp. 352-355, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York

19. Dowhan, W., Wickner, W. T., and Kennedy, E. P. (1974) J. Biol. Chem. 249,3079-3084

20. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275

21. Vogel, H. J., and Bonner, D. M. (1956) J. Biol. Chem. 218,97-106 22. Hawrot, E., and Kennedy, E. P. (1975) Fed. Proc. 34, 668 23. Oeschger, M. P., and Berlyn, M. K. B. (1975) Proc. N&l. Acad.

Sci. U. S. A. 72, 911-915 24. Cronan, J. E., Jr., and Vagelos, P. R. (1972) Biochim. Biophys.

Acta 265, 25-60

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from

E Hawrot and E P Kennedydecarboxylase mutants of Escherichia coli.

Phospholipid composition and membrane function in phosphatidylserine

1978, 253:8213-8220.J. Biol. Chem. 

  http://www.jbc.org/content/253/22/8213Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/253/22/8213.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on January 27, 2020http://w

ww

.jbc.org/D

ownloaded from