ADAPTATION OF MEMBRANE-BOUND
ENZYMES TO ETHANOL
BENJAMIN FISHER DICKENS, JR.
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHYY
UNIVERSITY OF FLORIDA
To my wife, Bonnie, who stood by me through the years it
took to achieve this goal; to my son, Brian, who has brightened
my life; to my parents, to whom I owe more than I will ever be
able to repay; this dissertation is dedicated to all of you.
I would like to offer special thanks to Dr. L. O. Ingram, who
as my major professor offered me the benefit of his knowledge and
experience. I would also like to express my appreciation to Dr. E.
M. Hoffmann and Dr. C. M. Allen Jr. for serving on my graduate
committee and for their assistance during my graduate career. In
addition, I am very grateful to the faculty of the Department of
Microbiology and Cell Science who were always willing to share
their time and resources.
TABLE OF CONTENTS
ACKNOWLEDGMENTS .......................... iii
KEY TO ABBREVIATIONS ......................... v
ABSTRACT ................................... vi
INTRODUCTION ............................... 1
MATERIAL AND METHODS ...................... 3
Lactose Permease .......................... 3
ATPase .................................. 4
Arrhenius Plots of Lac Permease and ATPase .... 5
Fatty Acid Analysis ......................... 6
Chemicals ................................ 6
RESULTS ....................................... 8
Effect of Ethanol on the Lac Permease ......... 8
Effect of Ethanol on ATPase ................. 11
Arrhenius Plots ............................ 22
Reconstitution of Membrane-bound ATPase ..... 33
Cooperativity of ATPase Inhibition by Sodium ... 36
In Vitro Acclimation of ATPase ............... 46
DISCUSSION ................................. 49
BIBLIOGRAPHY .................................. 54
BIOGRAPHICAL SKETCH .......................... 58
KEY TO ABBREVIATIONS
ATP: Adenosine Triphosphate
ATPase: Adenosine Triphosphatase
Ea: Energy of Activation
lac permease: Lactose Permease
ONPG: o-nitrophenol ,-D-galactopyranoside
S. D.: Standard Deviation
16:0 : Palmitic acid
18:1 : Oleic acid
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fullfilment of the Requirements
for the Degree of Doctor of Philosophy
ADAPTATION OF MEMBRANE-BOUND ENZYMES TO ETHANOL
BENJAMIN FISHER DICKENS JR.
Chairman: Lonnie O. Ingram
Major Department: Microbiology and Cell Science
Previous studies have shown that the proportion of unsaturated
fatty acids increases in Escherichia coli during growth in the presence
of ethanol. These lipid changes presumably represent an adaptive
response to compensate for the direct physical effect of ethanol on
the membrane lipid bilayer. The effects of growth in the presence of
ethanol on two membrane-bound enzymes, the lactose permease and
adenosine triphosphatase, were investigated. All but one of the prop-
erties of these membrane-bound enzymes, the activity of membrane-
bound ATPase with Ca++ as a cofactor, were found to adapt to the
presence of ethanol. The role of new protein synthesis and
the ethanol-induced changes in the bulk lipid composition in the
adaptation of membrane-bound enzymes to ethanol was also investigated.
Results of these investigations led to a model proposing that ethanol's
effect on both membrane-bound enzymes and the adaptation of these
enzymes to ethanol is mediated through the enzymes annular lipids.
Living organisms display a remarkable ability to adapt their
membranes to environmental changes such as temperature, by altering
their lipid composition (Fulco, 1974). These lipid changes are thought
to maintain the proper environment within the membrane for biological
function (Sinensky, 1974). Ethanol has been shown to directly inter-
calate within biological membranes altering membrane organization (Hill
and Bangham, 1975; Lee, 1977; and Seeman, 1972). Growth of E. coli
in the presence of ethanol results in the synthesis of membrane lipids
with an altered fatty acid composition, analogous to the changes observed
following a decrease in growth temperature (Ingram, 1976).
The enzymes of fatty acid synthesis appear to be a major site for
the regulation of lipid composition in E. coli (Cronan, 1975). Temperature
induced changes in the fatty acid composition are mediated in part through
direct effects on these soluble enzymes (Cronan, 1975). Investigation
of the mechanism of the ethanol-induced fatty acid changes indicate
that the site of action is also the soluble fatty acid synthetase enzyme
(Buttke and Ingram, 1978). This has recently been confirmed in our
laboratory in vivo.
Ethanol has been shown to affect the activity of a variety of
membrane-bound enzymes. Membrane-bound ATases in eukaryotic
cells have been frequently studied (Grisham and Barnett, 1972; Mitjavila
et al., 1976; Post et al., 1972; and Sun, 1976). Ethanol generally
inhibits the oubain-sensitive (Na,K)-ATPase but has little effect on
the activity of the Ca -ATPase. Transport systems have also been
studied and are usually inhibited by ethanol in eukaryotic cells (Chang
et al., 1967; Fox et al., 1978; Hoyumpa et al., 1977; Israel et al.,
1968; and Worthington et al., 1978). In E. coli, alcohols of different
chain lengths have opposite effects on the /-galactoside transport
enzyme, the lac permease. Alcohols of chain lengths greater than four
stimulate lac permease activity (Sullivan et al., 1974) while alcohols
of shorter chain lengths are inhibitory (Fried and Novick, 1973). A
similar differential effect of short and long chain alcohols has been
reported on fatty acid composition (Ingram, 1976).
The fatty acid changes induced by ethanol in E. coli have been
proposed as part of an adaptive response, compensating for the direct
physical effects of ethanol (Ingram, 1976). The similarities between
the lipid changes induced by ethanol to those resulting from a decrease
in growth temperature suggest that both may be involved in the mainte-
nance of the proper lipid environment for membrane function. In this
paper, we have investigated the significance of ethanol-induced
changes for membrane function in E. coli using two enzymes, ATPase
and lac permease. The effects of ethanol on these enzymes were
compared between cells grown in the presence and in the absence of
MATERIALS AND METHODS
Escherichia coli K12 strain Kl-221 (lac i ), generously donated by
Dr. R. P. Boyce (University of Florida, Gainesville, Florida, 32611),
was used to examine the effect of ethanol on the lac permease. Cultures
(250mL) were grown in mineral salts medium M63 (Miller, 1972) supple-
mented with succinic acid (5g/L), L-threonine (40mg/L), L-tyrosine
(20mg/L), L-proline (30mg/L), L-histidine (22mg/L), uracil (40mg/L)
and thiamine (20mg/L) in oversized culture tubes with continuous
aeration or in 2-L flask with continuous shaking at either 30 OC or 37 C.
Lac permease activity was assayed as described by Fried and Novick
(1973). Cells were harvested in exponential phase by centrifugation
and resuspended in inhibitor buffer. Portions of this cell suspension
were removed and supplemented with o-nitrophenol/- D-galactopyranoside
(ONPG). The reaction was stopped by the addition of calcium bicarbonate
and the o-nitrophenol concentration determined spectrophotometrically
at 420nm. Formaldehyde treated controls were used to correct for cryptic
transport (Kepes, 1971).
In contrast to the above assay, the in vivo permease activity was
determined in actively growing cells by supplementing ImL samples of
growing cultures with ONPG (1.85mM) and continuing to incubate under
growth conditions for 3-7 minutes. Pernease activity was measured as
o-nitrophenol produced using a formaldehyde control. Total
9-galactosidase activity was measured by the procedure of Putman
and Koch (1975). Cell numbers were estimated by turbimetric
Escherichia coli KI2 strain CSH-2, obtained from the Cold Spring
Harbor Laboratory (Cold Spring Harbor, N. Y.), and a derivative of
this strain blocked in fatty acid degradation (Buttke and Ingram, 1978),
strain TB-4 (fad E), were used to examine the effects of ethanol on
ATPase. Cultures (250mL) were grown in glucose-supplemented
Luria broth (Luria and Delbruck, 1943) in 2-L flask with continuous
agitation at either 30 C or 37 C. Cells were harvested in exponential
phase (2 x 108 cells/mL) by centrifugation and resuspended in a
minimum of buffer, lysed using a French pressure cell and the membranes
prepared as described by Futai et al. (1974). ATPase activity was
assayed as described by Evans (1969) with corrections for spontaneous
hydrolysis of ATP. The reaction was terminated by the addition of
trichloroacetic acid. Liberated inorganic phosphate was determined
colorimetrically by the method of Rathburn and Betlach (1969). Protein
was determined by the method of Lowry et al. (1951). The specific
activity of the ATPase preparations was 37.6 f5.6 micromoles ATP
hydrolysed/(mg protein x hour) at 300C.
Reconstitution experiments were performed with ATPase as described
by Rosen and Adler (1975). ATPase was solubilized using low ionic
strength buffers. This procedure solubilizes the FI ATPase with
bound annular lipids (Peter and Ahlers, 1975). Stripped membranes
were prepared by repeated extraction with the solubilizing buffer and
hydrolysed less than 0.20 micromoles of ATP/(mg protein x hour) at
300C. Reconstituted membranes used for Arrhenius plots hydrolysed
about 24 micromoles of ATP/(mg protein x hour) at 300C.
Arrhenius Plots of Lac Permease and ATPase
During the determination of permease activity, freshly prepared
suspensions of cells were held in inhibitor buffer at 0C. Incubations
were performed using individual water baths for a total of 10 to 12
temperature points per plot. During the determination of membrane-
bound ATPase activity, fresh membrane preparations were held at 00C.
During the determination of soluble ATPase activity, freshly solubilized
enzyme was held at room temperature. For ATPase, fifteen temperatures
were assayed concurrently in duplicate using a linear temperature
gradient block. Slopes were determined by least square regression
analysis. The intercepts were determined mathematically by setting
the equations describing each line (above and below the break) equal
at log activity and solving for temperature. Coefficients of correlation
were determined for each line and ranged from 0.985 to 0.998 for the
lac permease and 0.990 to 0.999 for ATPase activity. Omission of the
two points nearest the break had little effect on these slopes.
The temperature gradient block used for Arrhenius plots of ATPase
consists of an aluminum block 22.5 x 4 x 4 inches. An array of inch
holes (15 x 3) was centered in this block (2.5 inches deep) Pieces
(1.5 inches) were cut from each end, routed to form two water channels
and fitted with inlet and outlet connections. These were bolted back
on the block using a rubber gasket as a seal. A Haake E52 thermostated
circulator was connected to one end (hot water) and a Neslab RTE-3
circulator was connected to the other end (cold water/ethylene glycol).
The block was encased in 1.0 inch of styrofoam with corresponding
holes, mounted in a wooden frame and bolted on an Eberbach recipro-
cating shaker (Scientific Products, Orlando, Florida). The temperature
ranges in the block were established by adjusting the circulator tempera-
tures and the block formed a stable gradient within two hours. The
temperature gradient as measured in each well using a NBS standard
thermometer did not deviate from linearity with a coefficient of
correlation greater than 0.9999. Wells accommodate 13 x 50 mm tubes.
Approximately 0.5 mL of water was added to each well to ensure good
thermal contact between the block and the sample tube.
Fatty Acid Analysis
Lipids were extracted from cells which had been inactivated with
trichloroacetic acid (5%) as described by Kanfer and Kennedy (1963).
Fatty acids were transesterified and analysed by gas chromatography
(Ingram, 1976). Compositions are reported as percentage of peak area.
All bicchemicals used in this study were obtained from Sigma
Chemical Company (St. Louis, Missouri). During the course of this
study Cantley et al. (1977) found that the ATP from Sigma used in this
report contained a potent inhibitor of (Na,K)-ATPase. This inhibitor,
an inorganic vanadium compound, does not effect the Ca -ATPase
activity (Cantley et al., 1977) or bacterial ATPase. This was confirmed
in our laboratory using E. coli ATPase with the vanadium free ATP which
was also purchased from Sigma.
Effect of Ethanol on the Lac Permease
Ethanol causes a dose dependent inhibition of the lac permease
(figure 1). Cryptic transport (diffusion and other transport systems)
is slightly stimulated by ethanol. These effects are reversible upon
removal of ethanol by washing. A variety of other moderately hydro-
phobic compounds have been shown to induce changes in E. coli lipid
composition alalogous to ethanol (Ingram, 1977). Of these, acetone
(20g/L), dimethyl sulfoxide (74g/L), methanol (32g/L) and dioxane
(20g/L) were tested for their effect on the lac permease. In all cases,
these agents inhibited the permease activity, Sullivan et al. (1974)
have previously shown that long chain alcohols such as hexanol
stimulate permease activity. This stimulation by long chain alcohols
was confirmed in strain K1-221. Growth of E. coli in the presence of
these long chain alcohols results in changes in fatty acid composition
opposite to those caused by ethanol (Ingram, 1976). Chloroform (0.5g/L)
and amyl acetate (0.3g/L) induce changes in lipid composition like
hexanol (Ingram, 1977) and these agents also stimulate permease activity,
The initial inhibition of permease activity by ethanol is relieved
during subsequent growth in the presence of ethanol. Both the permease
and# -galactosidase from cells grown in the presence of ethanol exhibit
Figure 1. Acute effect of ethanol on P-galactoside
uptake. Lac permease mediated transport, ; and cryptic
entry, O .
1 2 3 5 7 9
the same specific activities as found in control cells. This suggests
that the recovery of permease activity does not simply result from the
synthesis of new protein. The time required for the recovery of
permease activity is directly related to alcohol concentration (figure 2A)
and inversely related to the growth rate (figure 3A). Changes in fatty
acid composition occur during this recovery period. The time required
for the completion of these changes was directly related to alcohol
concentration (figure 2B) and inversely related to growth rate (figure 3B).
In all cases, permease activity was restored prior to the completion of
these changes in bulk lipid composition. Thus the completion of these
changes in lipid composition is not essential for the recovery. Protein
synthesis in strain Kl-221 was inhibited by the addition of chloramphen-
ical (100mg/L) and by threonine starvation. Neither treatment prevented
the recovery of permease activity during incubation with ethanol,
although a longer time was required (figure 4). The decrease of the
permease activity with time in both the ethanol-treated and control
cells is presumably due to turnover of permease.
Effect of Ethanol on ATPase
The effect of ethanol on ATPase isolated from E. coli grown at 370 C
in the absence of ethanol is dramatically affected by the choice of
divalent metal cofactor (figure 5). With Mg ++ ethanol caused a dose-
dependent stimulation of activity. With Ca ++, ethanol caused the
opposite effect, a dose-dependent inhibition. This differential effect
suggests that ethanol is interacting in some way which alters the shape
C ) 00'rU "
u ( C
9 (a Q) _e ^ *^
C-o C o Cj
0 o 0 m 0 1
(6 C CL
Q E & r Oro
CCC>O 0 C)
tc" 3 38I
SC !O 0
iC C o C
C;o a a So
S- ^ o
*5i- o~ - C
ZCC) C)~~ '
C) C h. N C)O Q
NO ~ ~ I ON
iE 80 e
30 m c -
C C < 0 o nt
- D 0- 1
~u u t
3 o nO ^
0 0 0a c
0j Z 0 0 0
.00 wE 2(D
,4 -C0 o T
0 0 0 -- *o & 0
C o 2
>-i 0 0 *
O r C 0 0< 0 0
00-, 00 .2 o
U tt 1C 0 i, 1
r0 d. '-i1E 001
J NOI ISOdVOO3
GIOV AIIVJ '4DS/lDsun
o o o a
(Jnoq x slI a801 /salowOuDU)
C0 0 0
.C o 41 -
0 00 <
-- c I U
(Jqy ] WU/salIowouDu)
G 3 ZA"IOHG aH EdNO
Figure 5. Effect of ethanol on membrane-bound ATPase
from cells grown at 37 C with Mg++ or Ca+ as a cofactor
assayed at 37 C. Control cells with Mg'+,*; control
cells with Ca ,A; ethanol grown cells with Mg+ ,0;
and ethanol grown cells with Ca++, .
1 I I I I
2 6 10
of the catalytic site of the enzyme. The Km of ATPase was found to be
unaltered by the inclusion of ethanol in the assay mixture.
Growth in the presence of ethanol results in changes in the lipid
composition which could also influence enzyme function (table I).
ATPase from ethanol-grown cells was equally sensitive to inhibition by
ethanol using Ca++ as a cofactor (figure 5). However, ATPase from
ethanol-grown cells was much less stimulated by ethanol using Mg++ as
a cofactor (figure 5). Thus changes have occurred during growth in the
presence of ethanol which partially ameliorate the effect of ethanol on
ATPase activity assayed with Mg
Growth of E. coli strain CSH-2 at 300 C results in a fatty acid
composition similar to that produced during growth at 370 C in the
presence of ethanol (table I). When ATPase from cells grown at 30 C
was assayed at 370 C using Mg++ as a cofactor, ethanol caused a
modest stimulation of activity, intermediate between cells grown
without ethanol at 37 C and 37 C ethanol-grown cells. This suggests
that fatty acid changes may be involved in ethanol resistance and a
possible relationship between ethanol and low temperature.
Ethanol has been shown to have little effect on the activity of the
Mg++ stimulated ATPase of mitochondria from rat kidney when assayed
at high temperatures (Hosein et al., 1977). This enzyme was inhibited,
however, at lower temperatures (Hosein et al., 1977). The ATPase of
E. coli with Mgg as a cofactor is stimulated at 37 C (growth temperature)
by ethanol, however a slight inhibition in activity by ethanol was
C- C L . . Lo m 0 co
c. c o .j m .--
C* *o C-I CO CM CO * *
C- CO O C r C:O M cMt Col C )O rC
co c C C Cl C r) Cl O l O C
UO) C) m o c) m 3 CN C) C) CL
o lo 0 C c C) 0 cn c
(' U) C' C Co, C- Ic' C] C) C ]1
0C-]0.o -Clll C C <4'C' i I
o C C=
C C C C C CC (.
C' C C' C C' C,-401
0 o~g o~5oo
co n coCoomo o m
- C-- C- C0 C' C-
C'] C'] ',
VS ca C co m"
u ()u c) E
observed when assayed at a lower temperature (figure 6). This inhibition
at a lower temperature (150 C) was relieved following growth in the
presence of ethanol.
Arrhenius plots are frequently used to study membrane-bound enzymes
(for review see Lenaz et al., 1975). Arrhenius plots of membrane-
bound enzymes typically display one or two discontinuities. Massey
et al. (1966) have shown a direct correlation between the break tempera-
ture in Arrhenius plots and a conformational change in a soluble enzyme.
The position of the break in membrane-bound enzymes is presumed to
also be due to a conformatioral change in the enzyme. Break temperatures
are dependent upon the specific protein and its bound annular lipids
(Hesketh et al., 1976). Only one break was observed in the Arrhenius
plots in this report. Based upon previous reports (Linden et al., 1973;
Morrisett et al., 1975; and Thilo et al., 1977) the second break would
be expected to occur at or below our lowest measured temperatures.
In cells grown at 42 C and in cells at 37 C with exogenous palmitic
acid (16:0) to boost the saturated fatty acid content, we observed a
second break with the ATPase (data not shown).
Arrhenius plots of lac permease have break temperatures which are
dependent upon the lipid composition of the membrane (Overath et al.,
1971). Cells grown at 300C contain a lower proportion of saturated
fatty acids than cells grown at 370 C (table I). As expected, the break
Figure 6. Effect of ethanol on membrane-bound ATPase
from cells grown at 37 C with Mg++ as a cofactor and
assayed at 15C. Control cells,6; ethanol-grown cells, 0.
2 6 (
temperature of the 300 C-grown cells is lower than that for 370 C-grown
cells (figure 7A & B). Inclusion of ethanol in the assay mixture resulted
in an increase in the break temperature of these Arrhenius plots (figure
7A & B). This increase is similar to the effect caused by an increase
in the proportion of saturated fatty acids. The effects of ethanol on
Arrhenius plots of lac permease are summarized in table II. From these
results it appears that ethanol in some way stabilizes the form of the
lac permease with the higher energy of activation, increasing the tempera-
ture necessary for the conformation change to the form with the lower
energy of activation. Unlike the break temperature, however, the
energy of activation of the lac permease appears unaffected by the
addition of ethanol with one possible exception (table II), the lower Ea
of cells grown at 370 C in the absence of ethanol.
Growth in the presence of ethanol results in changes within the
membrane which prevent the ethanol-induced increase in break tempera-
ture (figure 7C). Arrhenius plots of ethanol-grown cells assayed in
the absence of ethanol display the same break temperature as control
cells assayed in the absence of ethanol, despite differences in fatty
acid composition (table I). The lipid composition of strain KI-221
grown at 30 oC without ethanol and ethanol-grown cells at 37 OC is very
similar (table I). However, the break temperature in these 30 OC-cells
is still increased by the inclusion of ethanol. These results suggest
that bulk lipid composition alone does not determine break temperature
or ethanol resistance.
o o 0
0 0 0 0
(inoq x Sllax O/SalOLUOUoU)
30 0 0o 0NO
0 I) 0
(JnoJ x !I~ OI/OIOWUDU
av~.foa r OdN
I 4 CI o i U)LOc
M C C 0
< 0 01mm) -m
i-;- CIM C) fn CM
\ ac r oi c Un m ^ O o n a
)+1 ^+ +1 +1+1 I r
ra I "S
< S -^ T
The break temperature in Arrhenius plots of ATPase is less affected
by changes in bulk membrane lipid composition than lac permease
(figure 8) although lipids are required for ATPase activity (Peter and
Ahlers, 1974). As with the lac permease, the inclusion of ethanol in
the assay mixture results in an increase in the break temperature of
membrane-bound ATPase (figure 8A & B). ATPase from ethanol-grown
cells exhibits no such increase in break temperature when assayed in
the presence of ethanol (figure 8C). The lipid composition of strain
CSH-2 grown at 300C without ethanol and ethanol-grown cells at 370C
are very similar (table I). However, the break temperature in these
300C-cells is still increased by the addition of ethanol, suggesting
that changes in bulk lipid composition per se do not provide protection
against ethanol. The effect of ethanol on Arrhenius plot characteristics
of ATPase are summarized in table III. As shown in this table, ethanol
did not alter the energies of activation below the break temperature in
ethanol-grown cells. The similarities between the effect of ethanol on
the break temperature of the lac permease and that of membrane-bound
ATPase may reflect a general effect of ethanol on lipid/protein inter-
ATPase can easily be solubilized by treatment with low ionic
strength buffers (Futai et al., 1974). Arrhenius plots of solubilized
ATPase exhibit the same break temperature as the membrane-bound form
(table III). Similar results were found by Sinerez et al. (1973). Inclu-
sion of ethanol in the assay mixture caused an increase in the break
Cd'~ u '-4
.- - 4-
o 0 0 n
o d 2 N i
o o o o 9 I
l I d ~ I
0 0 0 0
0 0 0
(Jnoq x ulaioJd 5wu/saIOLUoj3LUj)
O.0m 00 om N c NO N m c o
C- cNi 1 oC m r Coo-C
+ +1+ 1+1 + +1+41+1 + +1+1+1 I+1 +1+1+1+1 +1+1+1+1 I
( -c r Io n m 0 0 o3 3 0 m C
O U m L c'Jo"uooa enC N L0Lo
-)--+i ) ecnocN en cM cmeN cl c ci in nJ m mcc
I 1 c O ujjC' O CC C C Omu n Cooc
L- I C) M00 OOm(00 0aO)MW CMn ma)C 0mo00M
+ + I +I1+1+ +1+1+1+ +1+1+1+1 +II+ II+1 +1+l+1+1
S L .( CCM- e C0 C ] O C- ,-
Co 0 (0o 0 o co co o a) 0 o o 0 -i
)D D 000 0 000N- 0000 0000 0000 CD
- 1 : + : + ++ +14 +1+1 +1+ 1+1+ 1 + +I +1+1+ 4
0 0 0 0 0 0 0
cco do o(oo oo
---a 0 0 0 0 0 0 0 + OI
E-O -OM C
CC 0 <
temperature of soluble ATPase (table III) analogous to that observed
with membrane-bound ATPase. This indicates a direct effect of ethanol
on the soluble ATPase complex. Thus the effect of ethanol on Arrhenius
plots is not mediated through the interaction of ethanol with bulk
membrane lipids. Soluble ATPase isolated from ethanol-grown cells is
also resistant to the effect of ethanol (table III), indicating that bulk
membrane is not directly responsible for the protection against ethanol.
E. coli strain WN I fails to alter its lipid composition when grown
at different temperatures (Gelmann and Cronan, 1972) or in the presence
of ethanol (table I). The membrane-bound ATPase from this organism
displays a break temperature in its Arrhenius plot which is increased
by the inclusion of ethanol (figure 9A). The break temperature from
cells grown in the presence of ethanol is unaffected by ethanol (figure
9B) despite the failure of this strain to alter its bulk fatty acid composi-
tion. Again, changes in bulk lipid composition do not appear essential
for the amelioration of the ethanol effect on Arrhenius plots of ATPase.
Reconstitution of Membrane-bound ATPase
Membrane-bound ATPase from E. coli can easily dissociate into
two parts: the FO or transmembrane proton translocating portion and
the ATPase active F1 portion (Vogel and Steinhart, 1976) with bound
annular lipids (Peter and Ahlers, 1975). Reconstitution experiments
were performed using solubilized F1 ATPase and ATPase depleted
membranes to further examine the importance of bound lipids versus
bulk membrane in ethanol resistance. Depleted membranes from
(Jnno4 x u!aloid 6wlsalowoJcnw )
ethanol-grown cells were reconstituted with soluble F1 ATPase from
control cells. The break temperature of this preparation was increased
by the inclusion of ethanol (figure 10) indicating that the changes in
bulk membrane alone do not provide protection against ethanol.
Membranes from control cells were also reconstituted with F ATPase
from ethanol-grown cells. This reconstituted ATPase was resistant to
the effect of ethanol on break temperature (figure 10B) As a control,
ATPase and stripped membranes from ethanol-free cells were reconsti-
tuted. This preparation exhibited the same ethanol-induced increase in
break temperature as native membranes. Thus the sensitivity and
resistance of ATPase to the effects of ethanol on break temperature of
Arrhenius plots is a property of the soluble portion containing protein
and bound annular lipids, independent of the bulk membrane.
Cooperativity of ATPase Inhibition by Sodium
Membrane-bound ATPase of E. coli is inhibited by sodium ions with
Ca+ as a cofactor (Evans, 1969). This inhibition is cooperative with
a Hill coefficient (n) greater than one (Sinerez et al., 1973). Unlike
the break temperature observed in Arrhenius plots, the cooperativity of
this inhibition is strictly dependent upon the ATPase being membrane-
bound. Cooperativity of sodium inhibition is easily destroyed by
treatments such as freezing, assaying at a temperature above or below
the growth temperature or by solubilizing the enzyme (table IV).
Figure 11 shows the effect of ethanol on a Hill plot of ATPase.
Inclusion of ethanol in assays reduced the Hill coefficient. Dimethyl
Cd C Cf
Cd Cd C
0d~~ -, 0
Cd C~r C
(Jnoq x uialojd bUJ/salowuolUOuj)
I I I
o o 0
I I I
Table IV. Effect of a Variety of Physical Treatments on the
Cooperativity of Sodium Inhibition of ATPase from E. colia
Freeze and thawed membranes
Assayed at 42C
Assayed at 300C
1.26+ 0.08 (12)
1.03t 0.08 (3)
1.12 0.04 (2)
1.09 t 0.02 (2)
aE. coli TB-4 was grown at 370C and assayed at the
growth temperature unless otherwise specified.
Figure 11. Hill plots of the sodium inhibition of membrane-
bound ATPase. Assays were performed in the absence (*) and
the presence (0) of ethanol (16g/L). The sodium concentration
was varied over the range of 0-250nM.
In SODIUM CONCENTRATION
- i.0 o
sulfoxide, pentobarbital and chloropromazine also reduced the co-
operativity of sodium inhibition (table V). Cooperativity was not
reduced by ethanol in membrane-bound ATPase isolated from cells grown
in the presence of ethanol (table V) Additionally, ATPase from ethanol-
grown cells was resistant to dimethyl sulfoxide and pentobarbital but
remained sensitive to chloropromazine (table V). Pentobarbital (Ingram
et al., 1978) and dimethyl sulfoxide (Ingram, 1977) induce changes in
fatty acid composition similar to those caused by ethanol while
chloropromazine (Ingram et al., 1978) induces nearly opposite changes.
The protection provided against other agents by growth in the presence
of ethanol suggests that some adaptive change has occurred involving
lipids which compensates for the direct physical effect of these agents
To investigate whether changes in bulk membrane fatty acid
composition per se are sufficient to afford protection against the direct
effect of ethanol on the cooperativity of sodium inhibition, the fatty
acid composition of cells was artificially altered. E. coli was grown
in the presence of exogenous oleic acid (18:1) to increase the proportion
of unsaturated fatty acids within the membrane (table I). Under these
conditions, no protection was provided against the effect of ethanol
(table VI) in spite of the increased proportion of unsaturated fatty acids.
Growth of E. coli with exogenous palmitic acid (16:0) decreased the
proportion of unsaturated fatty acids within the membrane (table I) and
also had no effect on ethanol sensitivity (table VI) Thus changes
Table V. The Effect of Ethanol and other Membrane Agents on
the Hill Coefficients of Membrane-bound ATPase
Control Grown with
Assay Additive (n S. D.)b Ethanol (16g/L)
(n S. D.)a
None 1.26 + 0.08 (12) 1.28 0.06 (7)
Ethanol 16g/L 0.89 0.14 (9) 1.27 +0.09 (8)
Dimethyl Sulfoxide 33g/L 0.98 0.06 (4) 1.24 0.03 (2)
Pentabarbital 20g/L 0.93 + 0.03 (2) 1.26t 0.02 (2)
Chloropromazine 0.2g/L 1.03 0.05 (2) 1.01 0.04 (2)
aNumber within parentheses
to the right of S. D. denotes number
b. D., standard deviation.
I I I IC *0
M C) C '4 00 o 0 (.0 *F 0
_ C 0 0 "
00 0000. . -
S000 0 0000 o
a) j +0 + I + + I +I +I +1 4-
> Ua 0 0 0
0 D c> 3 0 1 c*, m L 0 0 0.
S1 o .0 a
S2 I i c i 4 C ) ,
00 I -0 0 0 0 ;
U'0 CD 0o
o Q o. . .0 O. A
. Ioo 1o a
1z u Lo -
F0 I 0 0 3
C) 1 0 a
. g 5i 0 g i
CO I D S 0 C^ 3 C! o
0 C-S 0C .0 0 ^ c>
2 u a j u ro 0 M .
gU~( 55 ( g
ccu - Q
in the bulk lipid composition are not sufficient to provide protection
against the effect of ethanol on the Hill coefficient.
The role of growth and new protein synthesis in the acquisition of
resistance to ethanol was investigated. Chloramphenicol was added
to a culture at a concentration sufficient to stop protein synthesis.
The culture was then split into two parts and incubation continued for
three hours. One part was supplemented with ethanol (16g/L) while the
other served as a control. Membranes were isolated from both of these
cultures for the determination of Hill coefficients in the presence and
absence of ethanol (table VI). ATPase from the control cells was
sensitive to the addition of ethanol while the cells which were exposed
to ethanol were resistant to the reduction in the Hill coefficient by
ethanol. This indicates that the acquisition of ethanol resistance does
not require new protein synthesis. In a similar experiment, a stationary
phase culture was split into two parts. Ethanol (16g/L) was added to
one part while the other served as a control. Following incubation for
an additional 16 hours at 370C, cells were harvested and the Hill
coefficients determined for each in the presence and absence of ethanol
(table VI). Stationary cells incubated in the absence of ethanol remained
sensitive while those incubated in the presence of ethanol acquired
resistance to ethanol. The bulk lipid composition of membranes in
stationary phase cells was not altered by the addition of ethanol (table I).
This provides further evidence that changes in bulk lipid composition
are not essential for the amelioration of ethanol effects.
In Vitro Acclimation of ATPase
Our previous experiments with Hill coefficients and Arrhenius plots
indicated that the adaption of membrane-bound ATPase to ethanol does
not require new protein synthesis or changes in the bulk lipid composi-
tion. This hypothesis was further examined in vitro. Isolated membranes
were incubated for 12 hours at 370 C in the presence and absence of
ethanol. The cooperativity of sodium inhibition was protected from the
ethanol effect following prior incubation in the presence of ethanol
(table VI), analogous to the results obtained in vivo with whole cells.
Likewise, the Arrhenius plots of the membrane-bound ATPase incubated
for 12 hours with ethanol were resistant to the inclusion of ethanol
(figure 12). Similar experiments, in vitro, were attempted with soluble
ATPase (7 & 12 hour incubations). Storage of the soluble ATPase in
ethanol, however, resulted in Arrhenius plots which did not exhibit a
distinct discontinuity in their slopes. The adaptation of membrane-
bound ATPase in vitro indicates that resistance can be acquired by a
rearrangement of ATPase with existing membrane components. The
failure of soluble ATPase to maintain its native thermodynamic
properties following prolonged incubation in the presence of ethanol
suggests that this rearrangement may involve other components in
addition to the bound lipids.
Figure 12. Arrhenius plots of membrane-bound ATPase
previously exposed to ethanol (16g/L) in vitro. Assays
were performed in the absence (e) and presence (0) of
16g/L of ethanol.
<. 5.0 -
1I I .
3.2 3.4 3.5 3.6
I/ K x 1000
Our results indicate that growth in the presence of ethanol results
in changes which confer resistance to the effects of ethanol on two
membrane-bound enzymes. The lac permease is inhibited in vivo by
the initial addition of ethanol to cells grown without ethanol. During
growth in the presence of ethanol, this inhibition is ameliorated (figure
2). Membrane-bound ATPase (Mg++) from cells grown without ethanol
is inhibited in vitro by the addition of ethanol at 150 C and stimulated
by the addition of ethanol at 370 C (figures 5 & 6). Again ATPase
preparations from cells grown in the presence of ethanol are substantially
resistant to both effects. Arrhenius plots of lac permease and ATPase
exhibit a discontinuity, the position of which is shifted to an elevated
temperature by the inclusion of ethanol in the assay mixture (figures 7
& 8). Enzyme preparations from cells grown in the presence of ethanol
are resistant to this effect of ethanol. Ethanol eliminates the
cooperativity of Na+ inhibition of ATPase from control cells. Again,
preparations from cells grown in the presence of ethanol are resistant
to ethanol. Thus adaptive changes occur during growth in the presence
of ethanol which ameliorate the direct effect of ethanol on these
The effect of alcohols on biological systems appears to result from
their interaction at hydrophobic sites (Ingram, 1976; and Seeman, 1972).
These hydrophobic sites need not be limited to the membrane lipid bilayer
but may also include the hydrophobic regions of proteins and their
annular lipids. Following prolonged exposure to ethanol two types of
membrane adaptation are frequently observed. The first involves changes
in physical properties of membrane-bound enzymes (Hosein et al. 1977;
and Israel and Kuriyama, 1971) and the second involves changes in the
bulk lipid composition of membranes (Ingram, 1976; Littleton and John,
1977; and Miceli and Ferrel, 1973).
Membrane-bound enzymes are affected by changes in their lipid
environment (Esfahani et al., 1971; Kimelberg and Papahadjopoulas,
1972; and Mavis and Vagelos, 1972). The interaction of drugs like
ethanol with the membrane alters the physical properties of the membrane
and affects the activities of membrane-bound enzymes (Lee, 1976).
The changes in bulk lipid composition following growth in the presence
of ethanol (table I) have been postulated as part of an adaptive response
to compensate for some of these effects (Ingram, 1976).
We have examined the role of the lipids in the adaptation of
membrane-bound enzymes. Our results with membrane-bound and
soluble ATPase (table III) indicates that the effect of ethanol on
Arrhenius plots of this enzyme is mediated through a direct interaction
of ethanol with the enzyme and perhaps annular lipid rather than through
changes in bulk membrane. Reconstitution experiments indicate that
sensitivity or resistance to ethanol are determined within the solubilized
portion containing bound lipids, independent of bulk membranes. The
adaptation of ATPase to ethanol in E. coli strain WN I (figure 9) as
well as the adaptation of stationary cells (table VI) and isolated
membranes (figure 12 and table VI) to ethanol indicates that the bulk
lipid changes are not essential for the adaptation of membrane-bound
enzymes to ethanol. However, changes in lipid composition may be of
significance for other cellular function in vivo.
Membrane-bound enzymes have been shown to exhibit some
selectivity for particular phospholipids from their environment which
allows maximal enzymatic activity (Warren et al., 1975). Indeed, the
isolation of different regions of the membrane has provided direct
physical evidence that membrane proteins may partition between specific
lipid domains in vivo (van Heerikhuisen et al., 1975; and Letellier et
al., 1977). These associations with particular annular lipids presumably
maintain the enzyme in the most thermodynamically favored configuration.
Thus membrane-bound enzymes appear to behave somewhat like a
cartesian diver in that they appear to select the proper environment
(i. e. annular lipid composition) to maintain the thermodynamically
favored state. The intercalation of ethanol results in changes in the
membrane environment. In the presence of ethanol, membrane-bound
enzymes may exchange their annular lipids for more unsaturated species
to maintain the most thermodynamically favored state.
This cartesian diver model explains how membrane-bound enzymes
could adapt in vitro to ethanol. The selection of a new environment by
membrane-bound enzymes would only require the proper unbound lipids
within the membrane with which the enzyme could reassociate. In
this model, the increase in unsaturated fatty acids following growth
in the presence of ethanol could maintain the proper microviscosity in
the bulk lipid bilayer which would have become more saturated in
nature due to the removal of unsaturated phospholipid species by this
protein/lipid reassociation. Alternatively, an increase in unsaturation
of molecular species could increase the proportion of the lipid environ-
ment with the physical properties which optimize enzyme function.
This model suggests that cells with naturally higher levels of unsatu-
rated fatty acids in their bulk lipids would be capable of coping with
ethanol better than cells with lower levels of unsaturated fatty acids.
Recently we have isolated fifteen mutants of E. coli which are resistant
to ethanol all of which contain higher proportions of unsaturated fatty
acids than the parental strain when grown in the absence of ethanol.
This model for the development of ethanol tolerance for membrane-
bound enzymes may be related to the observed consequence of prolonged
alcohol consumption. An initial acute tolerance occurs following
several hours of exposure to ethanol. This tolerance may be due to
rearrangement of membrane proteins with existing lipids as described
in our model. Chronic tolerance and its associated dependence could
be due to both this rearrangement and the ethanol-induced changes in
membrane composition (Ingram et al., 1978; and Littleton and John,
1977). These lipid changes need not be limited to phospholipids and
fatty acids. Other membrane components such as cholesterol and
carotenoids may also be involved.
Buttke, T. M., and Ingram, L. 0. (1978), Biochemistry 17:637-644.
Cantley, L. C. Jr., Josephson, L., Warner, R., Yanagisawa, M.,
Lechene, C., and Guidotti, G. (1977), L. Biol.Chem.252:
Chang, T., Lewis, J., and Glazko, A. J. (1967), Biochim. Biophys.
Cronan, J. E. Jr. (1975), i. Biol. Chem. 250:7074-7077.
Evans, D. J. Jr. (1969), J. Bacteriol. 114:239-248.
Esfahani, M., Lunbreck, A. R., Knutton, S. K., Oka, T., and Wakil,
S. J. (1971), Proc. Natl. Acad. Sci. USA 68:3180-3184.
Fox, J. E., Bourdages, R., and Beck, I. T. (1978), AM. J. Dig. Dis.
Fried, V. A., and Novick, A. (1973), J. Bacteriol. 114:239-248.
Fulco, A. J. (1974), Ann. Rev. Biochem. 43:215-241.
Futai, M., Sternweis, P. C., and Heppel, L. A. (1974), Proc. Natl.
Acad. Sci. USA 71:2725-2729.
Gelmann, E. P., and Cronan, J. E. Jr. (1972), J. Bacteriol. 112:
Grisham, C. M., and Barnett, R. E. (1972), Biochim. Biophvs. Acta
van Heerikhuisen, H., Kwah, E., van Brugzen, E. E. J. and Wilhott,
B. (1975), Biochim. Biophys. Acta 413:177-191.
Hesketh, T. R., Smith, G. A., Houslay, M. D., McGill, K. A.,
Birdsall, N. J. M., Metcalfe, J. C., and Warren, G. B. (1976),
Hill, M. W., and Bangham, A. D. (1975), In M. M. Gross (Ed.),
Alcohol Intoxication and Withdrawal. Experimental Studies. II,
pp 1-9, New York, Plenum Press.
Hosein, E. A., Hoffmann, I., and Linder, E. (1977), Arch. Biochem.
Hoyumpa, A. M., Nicols, S. G., Wilson, F. A., and Schenker, S.
(1977), Lab.Clin. Med. 90:1086-1095.
Ingram, L. 0. (1977), A-o. Env. Microbiol. 33:1233-1236.
Ingram, L. 0. (1976), J. Bacteriol. 125:670-678.
Ingram, L. O., Ley, K. D., and Hoffmann, E. M. (1978), Life Sciences
Israel, M., and Kuriyama, K. (1971), Life Sciences 10:591-599.
Israel, Y., Salazar, I., and Rosenmann, E. (1968), J. Nutr. 96:499-504.
Kanfer, I., and Kennedy, E. P. (1963),J. Biol. Chem. 238:2919-2922.
Kepes, A. (1971), J. Memb. Biol. 4:87-112.
Kimelberg, H. K., and Papahadjopoulas, D. (1972), Biochim. Biophvs.
Lee, A. G. (1977), Mol. Pharmacol. 13:474-485.
Lee, A. G. (1976), Nature 262:545-548.
Lenaz, G., Parenti-Castelli, G., and Sechi, A. M. (1975), Arch.
Biochem. Biophys. 167:72-79.
Letellier, L., Moudden, H., and Shechter, E. (1977), Proc. Natl.
Acad. Sci. USA 74:452-456.
Linden, C. D., Wright, K. L., McConnell, H. M., and Fox, C. F.
(1973), Proc. Natl. Acad. Sci. USA 70:2271-2275.
Littletcn, J. M., and John, G. (1977),J. Pharm. Pharmac. 29:579-580.
Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.
(1951), J. Biol. Chem. 193:265-275.
Luria, S. E., and Delbruck, M. (1943), Genetics 28:491-511.
Massey, V., Curti, B., and Ganther, H. (1966), J. Biol. Chem.
Mavis, R. D., and Vagelos, P.R. (1972), J. Biol. Chem. 247:652-659.
Miceli, J. M., and Ferrel, W. J. (1973), Lipids 8:722-727.
Miller, J. H. (1972) Experiments in Molecular Genetics, pp. 301,
Cold Spring Harbor Laboratory, Cold Spring Harbor, N. Y.
Mitjavila, S., Lacombe, C., and Carrera, G. (1976), Biochem.
Morrisett, J. D., Pownall, H. I., Plumlee, R. T., Smith, L. C.,
Zehner, Z., Esfahani, M., and Wakil, S. J. (1975), i. Biol.
Overath, P., Hill, F. F., and Lannek-Hirsch, I. (1971), Nature
New Biology 234:264-267.
Peter, H. W., and Ahlers, J. (1974), Arch. Biochem. Biophys.
Post, R. L., Hegyvary, C., and Kume, S. (1972),J. Biol. Chem.
Putman, S., and Koch, A. L. (1975), Anal. Biochem. 63:350-360.
Rathburn, W. B., and Betlach, M. V. (1969), Anal. Biochem. 28:436-445.
Rosen, B. P., and Adler, L. A. (1975), Biochim. Biophys. Acta
Seeman, P. (1972), Pharmacol. Rev. 24:583-655.
Sinensky, M. (1974), Proc. Natl. Acad. Sci. USA 71:522-525.
Sinerez, F., Farias, R. N., and Trucco, R. E. (1973), FEBS Letters
Sullivan, K. H., Jain, M. K., and Koch, A. L. (1974), Biochim.
Biophys. Acta 352:287-297.
Sun, A. Y. (1976), Ann. N. Y. Acad. Sci. 273:295-302.
Thilo, L., Trauble, H., and Overath, P. (1977), Biochemistry
Vogel, G., and Steinhart, R. (1976), Biochemistry 15:208-216.
Warren, G. B., Houslay, M. D., Birdsall, N. J. M., and Metcalfe,
J. C. (1975), Nature (London) 255:684-687.
Worthington, B. S., Meserole, L., and Syrotuck, J. A. (1978),
Am. I. Dig. Dis. 23:23-32.
Benjamin Fisher Dickens Jr. was born on August 1, 1951, in
Jacksonville, Florida. He lived in Fernandina Beach, Florida, with
his parents, Dr. and Mrs. B. F. Dickens, from 1951 until 1969 when
he graduated from Fernandina Beach Senior High School. In September
1969 he began attending college at the University of Florida in
Gainesville, Florida. He obtained his Bachelor of Science in June
1973 and in March 1976 he obtained his Master of Science from the
University of Florida. He married Bonnie Estelle Quattlebaum on July
1, 1972, and their son, Brian Fisher, was born on April 21, 1975.
Benjamin is a candidate for the degree of Doctor of Philosophy in
I certify that I have read this study and that in my opinion It
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Lonnie O. Ingram, Chairm n
Associate Professor of Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Edward M. Hoffmanr /i
Associate Professor of Microbiology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Charles M. Allen Jr.
Associate Professor of Biochemistry
This dissertation was submitted to the Graduate Faculty of the Depart-
ment of Microbiology and Cell Science in the College of Liberal Arts and
Sciences and to the Graduate council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
Dean, Graduate School