305

Biochimica et Biophysica Acta, 511 (1978) 305--319 © Elsevier/North-Holland Biomedical Press

BBA 78092 FORMATION OF L A R G E , ION-PERMEABLE MEMBRANE CHANNELS BY THE MATRIX PROTEIN (PORIN) OF E S C H E R I C H I A COLI

R. BENZ, K. JANKO, W. BOOS and P. LAUGER Department of Biology, University of Konstanz, D-775 Konstanz (G.F.R.)

(Received December 28th, 1977)

Summary One of the major proteins of the outer membrane of Escherichia coli, the matrix protein (porin), has been isolated by detergent solubilisation. When the protein is added in concentrations of the order of 10 ng/cm 3 to the outer phases of a planar lipid bilayer membrane, the membrane conductance increases by m a n y orders of magnitude. At lower protein concentrations the conductance increases in a stepwise fashion, the single conductance increment being about 2 nS (1 nS = 10 -9 siemens = 10 -9 ~-~) in 1 M KC1. The conductance pathway has an ohmic current vs. voltage character and a poor selectivity for chloride and the alkali ions. These findings are consistent with the assumption that the protein forms large aqueous channels in the membrane. From the average value of the single-channel conductance a channel diameter of about 0.9 nm is estimated. This channel size is consistent with the sugar permeability which has been reported for lipid vesicles reconstituted in the presence of the protein.

Introduction The cell envelope of gram-negative bacteria such as Salmonella t y p h i m u r i u m or Escherichia coli consists of the cytoplasmic membrane, a rigid peptidoglycan layer and the outer membrane [1,2]. The outer membrane is virtually impermeable to hydrophilic solutes of molecular weight greater than 1000, but allows the passage of sugars and other water-soluble compounds up to a molecular weight of 500--600 [3--8]. The transmembrane diffusion of small sugar molecules presumably occurs through aqueous channels, since both the activation energy and the specificity of the transport are low [2,3]. The structure and composition of the outer membrane have been extensively studied [9--31]. Apart from phospholipids, the outer membrane contains lipopolysaccharide [11] and a class of major proteins. The best characterized of these proteins

306 is a lipoprotein of molecular weight of a b o u t 7000 [17,18]. Recently, evidence has accumulated that the permeability properties of the outer membrane depend on the presence of another major protein, which has been variously described as protein I or 1 [11,13], matrix protein [14] and porin [32]. In E. coli B the matrix protein consists of a single polypeptide chain of molecular weight 36 500 [14], whereas the protein from K12 strain can be separated electrophoretically into two bands which correspond to two very closely related polypeptides [23,24,33--35]. The matrix protein is present in a b o u t 105 copies per cell and covers a substantial fraction of the cell surface [14,29]. Electron microscopic analysis shows that the protein is arranged on a hexagonal lattice whose repeat distance is 7.7 nm [29]. The unit cell of the lattice probably contains three molecules of the matrix protein; in negatively stained specimens the center of the unit cell appears as a pit which is filled with stain and which may represent the opening of a pore. These structural findings are consistent with cross-linking experiments [25,26,31] carried o u t with whole cells or isolated cell walls (outer membrane plus peptidoglycan layer), in which the matrix protein was specifically cross-linked to itself; from the analysis of the resulting protein complexes it was proposed that the basic unit is a trimer [31]. The matrix protein appears to span the outer membrane. It is tightly associated with the underlying peptidoglycan layer [14,24,29,30] and is at the same time accessible from the outside, acting as a phage receptor [23,36,37]. Evidence that the major proteins of the outer membrane may be involved in passive transmembrane diffusion comes from experiments of Nikaido et al. [38] showing that mutants of Salmonella typhimurium which were deficient in the major proteins had reduced permeabilities towards cephaloridin, a hydrophilic solute of molecular weight 415. Further, incorporation of isolated protein c o m p o n e n t s of the outer membrane into phospholipid vesicles rendered the vesicle membrane permeable to small sugar molecules [32,39,40]. In particular, Nakae [32] was able to demonstrate that phospholipid/lipopolysaccharide vesicles became permeable to hydrophilic solutes up to molecular weight of a b o u t 550 when the matrix protein from E. coli was incorporated into the vesicle membrane. If the matrix protein forms aqueous pores of a size large enough to allow the passage of di- and trisaccharides, then these pores should also be permeable to a variety of ions. In the following we show that incorporation of the matrix protein into planar lipid bilyer membranes creates electrically conducting pores of the expected size. Experimental

(a) Isolation of the matrix protein [32,40] E. coli K 12, strain pop 1730, was grown for 12 h in L-broth (bacto-tryptone 10 g/l, bacto yeast extract 5 g/l, NaC1 10 g/l) or for 5 h in Di YT medium (bactet r y p t o n 16 g/l, bacta yeast extra 10 g/l, NaC1 5 g/l). Strain pop 1730 has a deletion in lamB and lacks the h-receptor [43]. The cells were harvested, resuspended in 20 vols. of potassium phosphate buffer, pH 7.6, containing 0.1% fl-mercaptoethanol, and were passed three times through a French pres-

307 sure cell at 600 atm at 0°C. The extract was centrifuged for 110 min at 27 000 X g. The pellet was washed by resuspension in 10 mM Tris • HC1, pH 7.5, and centrifuged for 60 min at 100 000 X g. The material was stored at --70°C until used. A portion of the pellet containing 400 mg protein was extracted with 50 ml 2% sodium dodecyl sulfate (SDS) (Serva, Heidelberg, reinst.) at 32°C for 30 min and the suspension centrifuged at 100 000 X g for 30 min at 15°C. The pellet was washed 5 times by resuspension in distilled water and centrifugation. The suspension was then dialysed against 3 mM NaN3 at 4 ° C for 4 days. After centrifugation (100 000 X g, 30 min) the pellet was resuspended in 50 cm 3 5 mM Tris. HC1 buffer, pH 8, containing 3 mM NAN3, and was dispersed by sonication (Branson sonifier, 90 s). The suspension was incubated for 2 h at 37°C after addition -of 4 mg trypsin (Boehringer, zur Analyse). In this treatment the murein lipoprotein is hydrolysed [17]; the matrix protein, as long as it is associated with the peptidoglycan layer is resistant to trypsin [14,32]. The sonication and addition of trypsin was repeated twice. The insoluble material was spun down and the pellet was resuspended in 40 ml Tris. HC1, pH 8, containing 2% SDS. After incubation at 25°C for 30 min a slightly opaque solution was obtained. This mixture was applied to a Sepharose-4B column (2.5 X 50 cm) equilibrated with 5 mM Tris • HC1, pH 8, containing 0.1% SDS and 3 mM NAN3, and the column was eluted by the same buffer at a rate of about 20 cm3/h. The protein c o n t e n t was continuously monitored by measuring the optical density at 254 nm in an Uvicord spectrop h o t o m e t e r (LKB, Stockholm). In a second preparation, cholate (Merck, zur Analyse) was used t h r o u g h o u t instead of SDS, while the rest of the procedure remained the same. The results obtained with both detergents were almost identical (see below), but the yield of matrix protein in the cholate preparation was higher. A sugar test using the anthrone reaction [52] showed that the protein contained less than 1% (w/w) sugar. From this result one can conclude that the protein contained if anything only traces of the peptidoglycan layer. The binding of SDS to porin was found not to be very tight [14]. In 1% SDS solution the protein contained about 10% (w/w) SDS. But this a m o u n t has been shown to be easily reduced to 0.2% (w/w) by several washing procedures [14]. The Sepharose-4B chromatogram which is shown in Fig. 1 for a cholate preparation is similar to t h a t obtained by Nakae [32], although the relative peak heights are somewhat different. The main peak (I) corresponds to the peak I in Nakae's preparation which was found to be highly active in the permeability test with lipid vesicles [32]. According to Nakae [32] the protein of peak I is an aggregate of a molecular weight of about 800 000. SDS polyacrylamide gel electrophoresis was carried out after heating the protein in 2% SDS at 100°C for 2 min [41]. Peak I gave a single band corresponding to a molecular weight of about 37 000, in close agreement with the value given by Rosenbusch (36 500); for calibration lactate dehydrogenase (mol. wt. 35 000) and aldolase (mol. wt. 40 000) were used (The two subfractions of the matrix protein present in E. coli strain K12 are n o t separated in gel electrophoresis under the conditions given).

308

0 100

200 V/cm3

Fig. 1. Gel c h r o m a t o g r a p h y ( S e p h a r o s e - 4 B ) of o u t e r m e m b r a n e proteins (cholate preparation). The e l u t i o n m e d i u m w a s 5 m M Tris • HCI, p H 8, c o n t a i n i n g 3 m M N a N 3 a n d 0.1% c h o l a t e . T h e SDS p r e p a r a t i o n gave similar e h r o m a t o g r a m s , b u t w i t h a r e d u c e d h e i g h t o f p e a k I r e l a t i v e t o p e a k IV.

A sample of the matrix protein prepared by the m e t h o d of Rosenbusch [58] (which was kindly given to us by Dr. Rosenbusch) was indistinguishable from our preparation in gel electrophoresis. Protein determinations were carried o u t according to the m e t h o d of Lowry et al. [42]. For the membrane experiments, protein of peak I was used. Material from peak IV was also found to be active to some extent; however, the increase in membrane conductance was of an unspecific nature in this case and discrete conductance steps (as seen with peak I) were never observed (see below). The protein was stored either in lyophilized form at --25°C or in 5 mM Tris • HC1 buffer, pH 8, containing 0.1% SDS and 3 mM NAN3. In this solution the protein remained active in membrane experiments for several months. At high ionic strength (1 M NaC1), however, the protein became inactive within 20 h; for this reason fresh solutions for membrane experiments were prepared every day using small aliquots of the stock solution.

(b ) Membrane experiments Optically black lipid membranes were formed in the usual way [44] in a circular hole in a Teflon septum separating two aqueous solutions. The membrane area was 2 • 10 -2 cm 2 for the measurements of macroscopic conductance and about 10 .4 cm 2 for the conductance-fluctuation experiments. The Teflon cell was contained in a thermostated metal block; the temperature was 25°C throughout. The aqueous solutions bathing the membrane were unbuffered and had a pH between 5.5 and 6. In separate experiments it was established that the results were virtually pH-independent in the range pH 5.5--7. In some experiments the protein was already present prior to the formation of the membrane; in other experiments the protein was added with stirring after the membrane had become completely black. The following lipids were used for membrane formation: monoolein (NuCheck Prep, Elysian, Minn., U.S.A.); 1,2-dioleoyl-sn-glycerol-3-phosphorylcholine (dioleoyl phosphatidylcholine) synthesized as in ref. 45; egg lecithin, egg phosphatidylethanolamine, brain phosphatidylserine, brain phosphatidylinositol (purified by standard methods) [46,47]. Oxidized cholesterol was prepared by boiling a 4% suspension of cholesterol (Eastman, reagent grade) in n-octane for 4 h under reflux and bubbling oxygen through the suspension [48,49]. All lipids, except oxidized cholesterol, gave a single spot in the thin-

309 layer chromatogram. Where n o t otherwise indicated, the lipids were used as a 2% (w/v) solution in n-decane (Merck, standard for gas c h r o m a t o g r a p h y ) . F o r the electrical measurements the m e m b r a n e cell was c o n n e c t e d to the external circuit through Ag/AgCl electrodes or (for the d e t e r m i n a t i o n of the zero-current m e m b r a n e potential) t hr ough agar bridges via Ag/AgC1 electrodes. F o r the c o n d u c t a n c e - f l u c t u a t i o n experiments a Keithley 427 preamplifier was used in c o n n n e c t i o n with a T e c t r o n i x 5 1 1 1 / 5 A 2 2 storage oscilloscope. The amplified signal at the o u t p u t of the oscilloscope was recorded with a strip-chart recorder. In some cases the signal was stored on tape. The bandwidth o f the fluctuation measurements was 100 Hz--3 kHz. For the measurem e n t o f macroscopic c o n d u c t a n c e a Keithley 150 B Microvolt A m m e t e r or a Keithley 610 C e l e c t r o m e t e r were used. Membrane potentials were measured with the Keithley 610 C Electrometer. For the study o f current transients a b a t t e r y - o p e r a t e d pulse generator was used and the current was measured as a voltage d r o p across an external resistor with a T e c t r o n i x 5 1 1 5 / 5 A 2 2 storage oscilloscope. Results

Time behaviour o f membrane conductance When matrix protein from a stock solution in SDS or cholate was added in small quantities to the aqueous solutions bathing the m em brane, the m e m b r a n e c o n d u c t a n c e started to increase in a stepwise fashion (Fig. 2). T he occurrence o f these c o n d u c t a n c e steps is specific to the porin and is n o t seen when an

.j

•

i

:

,,

lOs

finSi10pA

F i g . 2. S t e p w i s e i n c r e a s e o f m e m b r a n e c u r r e n t a f t e r a d d i t i o n o f m a t r i x p r o t e i n . A s m a l l a m o u n t o f a stock solution c o n t a i n i n g 0.2 m g / c m 3 porin and 1 m g / c m 3 SDS was a d d e d to aqueous solutions on both s i d e s o f t h e m e m b r a n e to g i v e a n o m i n a l p r o t e i n c o n c e n t r a t i o n o f a b o u t 0 . 5 n g / c m 3. T h e a q u e o u s s o l u t i o n c o n t a i n e d 0 . 3 M K C I a n d 1 /~g/cm 3 S D S ; T = 2 5 ° C . T h e m e m b r a n e w a s f o r m e d f r o m a 2% ( w / v ) s o l u t i o n o f o x i d i z e d c h o l e s t e r o l i n n - d e c a n e . T h e a p p l i e d v o l t a g e w a s 20 m V , t h e c u r r e n t p r i o r t o t h e a d d i t i o n of porin was about 0.6 pA. The b a n d w i d t h o f the measurement was 1 kHz. The record starts at the left of the lower trace and continues in the upper trace.

310 equivalent a m o u n t of detergent alone is added. F u r t h e r m o r e , after heating the stock solution (0.2 mg/cm 3 porin, 1 mg/cm 3 SDS, 5 mM Tris • HC1, 3 mM NAN3, pH 8) to 100°C for 5 min, the activity o f the protein is completely lost. The c o n d u c t a n c e steps are observed irrespective w h e t h e r the porin is added to the aqueous solution on one side of the m e m b r a n e only or to both solutions. As seen from Fig. 2, m os t steps in the current record are directed upward, whereas terminating events are rarely observed. From records extending over prolonged periods, the average lifetime o f the conductive unit may be estimated to be at least 5 min. It was i n d e p e n d e n t o f salt concentration. The risetime o f the single c o n d u c t a n c e step was less than 0.5 ms; within this time resolution the c o n d u c t a n c e rise was always s m o o t h w i t h o u t any indication o f smaller intermediate steps. The c o n d u c t a n c e increments are n o t uni f orm in size but distributed over a certain range. Histograms of the c o n d u c t a n c e steps observed with matrix protein solubilized in SDS and in cholate are shown in Fig. 3. As can be seen the distribution o f the c o n d u c t a n c e increments is virtually t he same for both protein preparations; for protein solubilized in SDS the average c o n d u c t a n c e increm e n t in (1 M KC1) is A -~ 1.9 nS, and A -~ 2.0 nS for protein solubilized in cholate. A close agreement be t w een A values f r o m the two protein preparations was also observed with o t h e r salt solutions (0.1 M KC1 or 1 M NaC1). The

0.3 A 0.2

0.1

P (.A_) 0

°I

'T

0.2 ~

0

0

1

2

3

z,

5

IY/nS

Fig. 3. Probability P(A) of t h e o c c u r r e n c e

of a c o n d u c t a n c e s t e p o f m a g n i t u d e A. P(A) is t h e n u m b e r of o b s e r v e d s t e p s w i t h i n a n i n t e r v a l o f w i d t h AA = -+0.17 n S c e n t e r e d at A, divided by t h e t o t a l n u m b e r (n) of steps. The membranes w e r e m a d e from 2% (w/v) oxidized cholesterol in n-decane. The aqueous phase contained 1 M KC1; T = 25°C. T h e a p p l i e d v o l t a g e w a s 20 mY. A: porin solubilized in SDS (n = 321); B: porin solubilized in cholate (n = 231). T h e c o n d i t i o n s o f p r o t e i n a d d i t i o n w e r e s i m i l a r t o t h o s e i n d i c a t e d i n t h e legend of Fig. 2.

311 average value A of the conductance increment was usually by about 20--30% larger than the most frequently assumed A value in the distribution. When larger concentrations of the porin are present, the conductance quickly rises after formation of the bilayer and reaches values m a n y orders of magnitude above the level of the untreated membrane. The time course of the specific conductance X after protein addition is shown in Fig. 4. With' membranes made of monoolein in n-decane the rate of change of X increased with time in the initial phase of the experiment, whereas the reverse behaviour was found for oxidized-cholesterol/n-decane membranes. Under both conditions a steady conductance level was n o t reached, the membranes usually broke at a conductance of approx. 5--20 mS • cm -2 (1 M NaC1). When the membrane was formed in a solution of salt and detergent of the same concentration, but w i t h o u t porin, only a insignificant conductance increase was observed. (X* in Table III). Using valinomycin in 1 M KC1 as a probe for the surface potential, it was also tested t h a t 1 pg/cm 3 SDS or 1 pg/ cm 3 cholate did not change the surface potential appreciably. Reducing the concentration of SDS resulted in a steeper conductance rise with time, as Fig. 4 shows. A possible explanation for this p h e n o m e n o n would be t h a t at a lower detergent concentration the hydrophobic surface of the porin is more exposed to water. This would increase the tendency of the protein to become associated with the lipid membrane. In order to test whether or not the rate of conductance rise depends on voltage, X(t) was recorded under two different conditions. First the membrane was maintained at zero voltage during most of the time and only brief voltage pulses of 10 mV were applied in order to measure X. Second a constant voltage of 70 mV was applied. In both cases the values of k(t) were virtually the same.

mSc~ 2 6

/o

4

A o ox.cholestorol 01 ~g/cm 3 SDS

/

B ox.cho[esterol I ~g/cm 3 SOS

./'/'/"

./

o

o /

/

2

. / "

"

/

=/," I= I

c

monoolein 1IJ.g/cm 3 SDS I

0

10

20

30

40 t/rain

Fig. 4. S p e c i f i c m e m b r a n e c o n d u c t a n c e k as a f u n c t i o n o f t i m e t. T h e m e m b r a n e w a s f o r m e d at t i m e z e r o in a 1 M NaC1 s o l u t i o n c o n t a i n i n g 2 0 n g / c m 3 p o r i n a n d 0.1 o r 1 /~g/cm 3 S D S . C u r v e A : o x i d i z e d c h o l e s t e r o l m e m b r a n e , 0.1 ~ g / c m 3 S D S ; C u r v e B: o x i d i z e d c h o l e s t e r o l m e m b r a n e , 1 ~tg/cm 3 S D S ; C u r v e C: monoolein membrane, I ~g/cm 3 SDS. At time t ~ 1 min the membrane was completely black. The a p p l i e d v o l t a g e w a s 10 m Y . T h e m e m b r a n e c o n d u c t a n c e in a s o l u t i o n c o n t a i n i n g 1 M N a C l a n d 1 ~ g / c m 3 S D S ( b u t n o p r o t e i n ) w a s o f t h e o r d e r o f 3 • 1 0 -8 S • c m - 2 . I n a series o f e x p e r i m e n t s w i t h d i f f e r e n t m e m b r a n e s t h e r e p r o d u c i b i l i t y o f k ( t ) w a s a b o u t -+30%.

312 As m e n t i o n e d above, porin could be inactivated by heating the stock solution to 100°C for 5 min. A partial inactivation was observed after maintaining the same stock solution at 70°C for 3 min. In this case a b o u t 10 times m ore p ro t e i n had to be added in order to obtain a given macroscopic conductance. The appearance o f the c o n d u c t a n c e fluctuations, however, was very similar to those observed with the fully active preparation and the value of the average c o n d u c t a n c e i n c r e m e n t was virtually the same in bot h cases. The only difference was th at the fluctuating c u r r e n t signal became increasingly noisy with time. This noise possibly results f r om the adsorption o f inactivated protein t o t h e memb r an e.

Conductance as a function o f protein and salt concentration Since a steady c o n d u c t a n c e level could n o t be reached (Fig. 4), the dependence o f c o n d u c t a n c e on porin c o n c e n t r a t i o n could n o t be accurately determined. But a meaningful comparison was possible on the basis o f h(t) curves which were reproducible to -+30%. In experiments with oxidized cholesterol membranes in 1 M NaC1 and 1 pg/cm 3 SDS h(t) was f o u n d to be virtually i n d e p e n d e n t o f porin c o n c e n t r a t i o n in the range of 2--200 ng/cm3; below 2 ng/cm 3 the c o n d u c t a n c e rise became small and was less reproducible. Such a behaviour is reminiscent o f systems exhibiting a critical micellar concentration. It could result f r om the existence of an equilibrium between porin monomers and aggregates in water if only the m o n o m e r s are able to enter the membrane. A n o t h e r explanation would be t ha t the m e m b r a n e surface becomes saturated with adsorbed pr ot ei n at aqueous protein concent rat i ons above 2 ng/cm 3 and t h a t p o r e f o r m a t i o n occurs f r om the adsorbed state. T h e ef f ect o f salt c o n c e n t r a t i o n on X was studied by measuring the conductance at a given time after the f o r m a t i o n o f the bilayer; this time was chosen to be 30 min in all cases. As can be seen from Fig. 5 that the c o n d u c t a n c e X

x

/

S c m "z

IJ i

10"2

./'.

|0 "3

./i/"

10-~

IO-s

i |0 -3

/i

i

IO-2

10-I

I 1

cM/M

F i g . 5. S p e c i f i c m e m b r a n e conductance as a f u n c t i o n o f N a C 1 c o n c e n t r a t i o n c M in w a t e r . T h e a q u e o u s phase contained 200 ng/cm 3 porin and 1 #g/cm 3 SDS, T = 25°C. The membrane w a s f o r m e d f r o m 2% (w/v) of oxidized cholesterol in n-decane. The conductance w a s m e a s u r e d 30 r a i n a f t e r m e m b r a n e form a t i o n ( a p p l i e d v o l t a g e : 10 m V ) . E a c h p o i n t c o r r e s p o n d s t o a d i f f e r e n t m e m b r a n e .

313

Y~ nS

/

f / 0.1

/ I

I

0.01

0.I

Fig. 6. Average conductance T h e applied voltage was 50 in SDS. With the exception of A ranged b e t w e e n 1 8 0 and

I

I

cJM

increment ~. as a function of K C I concentration c M in water; T = 25°C. inV. Oxidized cholesterol/n-decane m e m b r a n e s . T h e porin was solubilized of c M = 0.03 M (n = 19) the n u m b e r n of events used for the evaluation 320.

is a linear function of salt concentration within the limits of experimental uncertainty. The interpretation of this experiment is complicated by the possibility that the rate of incorporation of the protein into the membrane may depend on aqueous salt concentration so that the observed concentration dependence of k may not only reflect the intrinsic concentration dependence of the single conductive unit b u t also an effect of salt concentration on the number of conductive units in the membrane. In order to study this questions further, the dependence of the average conductance increment A on salt concentration was directly determined. From Fig. 6 it is seen that A is a linear function of KC1 concentration between 0.03 and 3 M. This result suggests that in the experiment represented in Fig. 5 the rate of porin incorporation was independent of salt concentration. Conductance as a function o f voltage The dependence_of the average conductance increment ~, on voltage is represented in Table I. A was found to be virtually constant over the experimental voltage range between 0 and 150 mV. Current and voltage measurements over a larger voltage range were performed by applying short voltage pulses

TABLE I D E P E N D E N C E OF T H E A V E R A G E C O N D U C T A N C E I N C R E M E N T ~ ON V O L T A G E w a s o b t a i n e d f r o m c u r r e n t r e c o r d s s u c h as s h o w n in Fig. 2 b y d i v i d i n g t h e c u r r e n t i n c r e m e n t b y t h e a p p l i e d v o l t a g e V a n d a v e r a g i n g o v e r a large n u m b e r n o f e v e n t s . O x i d i z e d c h o l e s t e r o l / n - d e c a n e m e m b r a n e s in 0 . 3 M KCI, T = 25°C. T h e p o r i n w a s s o l u b i l i z e d in 1 ~ g / e m 3 SDS. V n

(mV) (nS)

20 0.72 192

40 0.79 120

50 0.69 315

60 0.70 183

100 0.79 208

150 0.85 53

314 I pAcrn -2

o

o/j VJ /2v. /:/

150

,oo

o//y

0

o

I

I

1

ioo

2oo

300

vlmv

F i g . 7. C u r r e n t vs. v o l t a g e c h a r a c t e r i s t i c o f a n o x i d i z e d c h o l e s t e r o l m e m b r a n e d o p e d w i t h m a t r i x p r o t e i n . T h e a q u e o u s p h a s e c o n t a i n e d 2 0 n g / c m 3 p o r i n , 1 p g / c m 3 S D S a n d 1 M KCII T = 2 5 ° C . V o l t a g e pul§es o f 1 m s d u r a t i o n were applied and the current measured after the decay of the capacitive transient. The results f r o m t w o d i f f e r e n t m e m b r a n e s a r e s h o w n .

to a m e m b r a n e in the presence of higher porin concentrationsi; for voltage pulses o f 1 ms duration {or less) the m e m b r a n e was stable up to several h u n d r e d millivolts. T he result of such experiments is represented in Fig. 7. As can be seen the m e m b r a n e cur r ent is a linear funct i on of voltage up to at least 350 mV. It was also f o u n d t hat the current remained const ant during the pulse length o f 1 ms after the decay of the capacitive transient. In separate experiments at lower voltages (100 mV) it was established t hat the conductance after a voltage j um p was nearly constant in the time range between 0.1 ms and several seconds. These experiments show t hat the properties of the single conductive units are virtually i n d e p e n d e n t of voltage.

Ion specificity In one series o f experiments the ion specificity o f the induced c o n d u c t a n c e was studied by measuring t he zero-current m e m b r a n e potential V0 in the presence o f a salt c o n c e n t r a t i o n gradient. In the case of an oxidized cholesterol m e m b r a n e and for a c o n c e n t r a t i o n ratio o f 10 : I the m e m b r a n e potential V0 was a b o u t 25 mV for KC1 and a b o u t 15 mV for NaC1, with the positive side in the dilute solution. For these experiments the m e m b r a n e c o n d u c t a n c e had to be kept low in order to avoid effects o f unstirred layers in the aqueous solution adjacent to the m e m b r a n e surface. The observed values o f V0 correspond to permeability ratios PK/Pcl ~--3.5 and PNa]Pcl "--2.1. This slight preference for the cation m a y result from negative charges on the porin protein or on the oxidized cholesterol membrane. In a second series o f experiments the average c o n d u c t a n c e i n c r e m e n t ~ was d e ter min ed in di f f er e nt electrocytes. As can be seen f r o m Table II t hat in different 1 M alkali chloride solutions (and NH4C1) A varies only within a fa c t o r o f a b o u t 3 (between 0.72 and 2.3 mS). Also given in Table II is the ratio A/a where a is the specific c o n d u c t a n c e o f a 1 M solution of the corresponding salt in water. This ratio varies by a factor of 2. The low ion specificity

315 T A B L E II AVERAGE

CONDUCTANCE

INCREMENT A IN DIFFERENT

1 M ELECTROLYTE

SOLUTIONS

O x i d i z e d cholesterol/n-decane m e m b r a n e s ; T = 25°C. The a q u e o u s solutions c o n t a i n e d 1 p g / c m 3 SDS. T h e a p p l i e d v o l t a g e w a s 50 m Y . n is t h e n u m b e r o f e v e n t s u s e d f o r t h e e v a l u a t i o n o f A; o is t h e s p e c i f i c conductance of a 1 M a q u e o u s solution of the electrolyte (at 18°C). Salt

LiC1 NaCI KC1 RbC1 CsC1 NH4C1

~

A/o

(nS)

(10-10 m)

n

0.72 1.2 1.9 2.1 2.3 2.0

1.0 1.4 1.7 1.8 2.0 1.8

418 206 321 251 194 213

of A is consistent with the assumption that the protein forms large water-filled pores in the membrane.

Effect of membrane composition The magnitude of the conductance increase in the presence of porin depends strongly on lipid composition of the membrane (Table III). With membranes made of oxidized cholesterol or of monoolein the conductance k in the presence of porin is by several orders of magnitude larger than the conductance of the u n d o p e d membrane. A much smaller effect (in the given time of the experiment) was observed with phosphatidylserine, phosphatidylinositol, phosphatidylcholine {lecithin) and phosphatidylethanolamine. Discrete conductance steps such as observed with oxidized cholesterol or monoolein membranes were also found with phosphatidylserine and phosphatidylinositol membranes (see below); they could n o t be detected, however, with lecithin and with phosphatidylethanolamine membranes. For this reason it cannot be excluded that in the case o f lecithin membranes the observed conductance

TABLE III SPECIFIC MEMBRANE BLACK FILM

CONDUCTANCE

kMEASURED

t

MIN

AFTER

FORMATION

OF

THE

T h e m e m b r a n e w a s f o r m e d f r o m a 2% ( w / v ) s o l u t i o n o f t h e lipid in n - d e e a n e . T h e a q u e o u s p h a s e c o n t a i n e d 1 M NaC1, 1 ~ g / c m 3 S D S a n d 2 0 n g / c m 3 p r o t e i n ; T = 2 5 ° C . k * is t h e m e m b r a n e c o n d u c t a n c e m e a s u r e d at t i m e t in t h e s a m e s o l u t i o n b u t w i t h o u t p o r i n . T h e a p p l i e d v o l t a g e w a s 10 i n V . M o n o o l e i n d i s s o l v e d in n - d o d e c a n e or n - h e x a d e c a n e g a v e s i m i l a r v a l u e s o f k as m o n o o l e i n in n - d e c a n e . Lipid

Oxidized cholesterol Monoolein Brain phosphatidylserine Brain phosphatidylinositol Dioleoyliecithin Egg lecithin Egg phosphatidylethanolaraine

t

k

k *

(rain)

(S • cm -2)

(S • c m - 2 )

40 12 40 30 50 30 20

5 8 9 6 4 7 5

3 2 5 3 4 5 1

• 10 -3 • 10 -4 • 10 -6 • 10 -6 • 10 -7 " 10- 7 • 10 -7

10 - 8 10-7 10-7 10-7 10-8 10- 8 10- 7

316

T A B L E IV A V E R A G E C O N D U C T A N C E I N C R E M E N T A F O R M E M B R A N E S O F D I F F E R E N T C O M P O S I T I O N IN A Q U E O U S KC1 S O L U T I O N ( C O N C E N T R A T I O N , CM) T = 2 5 ° C , V = 50 inV. T h e p o r i n w a s solubilized in 1 p g / c m 3 SDS. d is t h e t h i c k n e s s of t h e h y d r o p h o b i c p o r t i o n o f t h e m e m b r a n e , as c a l c u l a t e d f r o m c a p a c i t a n c e m e a s u r e m e n t s using a v a l u e of 2.1 for t h e dielectric c o n s t a n t [ 4 5 , 5 0 , 5 1 ]. n is t h e n u m b e r o f e v e n t s f r o m w h i c h ~ has b e e n c a l c u l a t e d . Lipid/solvent

Oxidized cholesterol]n-decane Monoolein/n-decane M o n o o l e i n / n - h e x a d e c ane Brain p h o s p h a t i d y l s e r i n e ] n - d e c a n e Brain p h o s p h a t i d y l s e r i n e / n - d e c a n e Brain p h o s p h a t i d y l i n o s i t o l

d

cM

~

(nm)

(M)

(nS)

3.4 * 4.8 3.2 5.3 5.3 5.2

1 1 1 1 0.1 0.1

1.9 1.9 1.8 2.5 0.30 0.29

n

321 142 252 42 35 40

* Benz, R., u n p u b l i s h e d results.

increase is of an unspecific nature. An explanation of the striking influence of lipid structure on the protein-induced conductance seems difficult at the moment. As a true equilibrium distribution of porin between water and membrane was not reached under the conditions of the conductance experiments, the observed effects of lipid composition may be mainly of kinetic nature. That is, the activation energy for the incorporation of the porin into the membrane may be much larger for the phospholipids than for oxidized cholesterol or monoolein. In contrast to the pronounced lipid effects on macroscopic conductance, the magnitude of the discrete conductance steps did not depend much on the composition of membrane. As can be seen from Table IV that the average conductance increment A varies only by a factor of 1.5 in a series of different lipids (oxidized cholesterol, monoolein, phosphatidylserine, phosphatidylinositol). The observation that A is n o t much different in neutral membranes (monoolein) and in negatively charged membranes (phosphatidylserine, phosphatidylinositol) is consistent with the finding that the conductive unit has a poor cation-to-anion selectivity (see above). Furthermore, A was virtually the same in membranes made from monoolein in n-decane and monoolein in n-hexadecane, despite a considerable difference in membrane thickness. Discussion The experiments described above show that the matrix protein from the outer membrane of E. coli is able to increase the ion permeability of lipid bilayer membranes by several orders of magnitude. At low porin concentration the change of membrane conductance occurs in descrete steps, suggesting that the conductive pathways consist of localized structures. These findings may be compared with recent studies of Nakae [32] who showed that the matrix protein renders lipid vesicles permeable to sugars up to a molecular weight of about 550. However, in contrast to Nakae's results [32] no lipopolysaccharide was needed in lipid bilayers to receive channel activity. If the protein forms wide aqueous pores in the bilayer (as Nakae proposed) then the size of the pore may

317 be estimated from the magnitude of the discrete conductance increment which is about 1.9 nS in 1 M KCI. Assuming t h a t the pore is filled with a solution of the same specific conductivity as the external solution and assuming a pore length of 4 nm, the cross-section of the pore is calculated to be 0.68 nm 2. This corresponds to a diameter of 0.93 nm if the cross-section is circular. The minim u m diameter that a circular pore must have in order to allow the passage of sucrose (which was found to be permeable in Nakae's experiments) is about 0.9 nm, as may be estimated from molecular models. Similar minimum pore diameters apply to rod-shaped trisaccharides such as raffinose which also have been used in the permeability studies with vesicles. The average pore size obtained from the conductance experiments is thus of the right magnitude to account for the nonelectrolyte permeability of outer membranes [4] and of reconstit u t e d vesicles [32]. However, at the present time we cannot exclude the possibility that the larger conductance steps in the histogram arise from the formation of aggregates containing more than one pore. More information bearing on this question should be obtained from specificity studies with large organic ions. The additional results are also consistent with the notion that the porin forms wide water-filled channels in the membrane. Such channels should have a poor cation-to-anion discrimination as well as a low selectivity for small ions in general. Furthermore, the current vs. voltage character of a wide, unselective channel should be ohmic. The observation t h a t the conductance of the single channel does not depend much on the composition and thickness of the membrane (Table IV) suggests t h a t the pore has a fixed length which is determined by the porin molecule itself. In order to accommodate a pore of fixed length, the lipid structure has to be distorted locally to a lesser or greater extent, depending on the membrane thickness [ 53]. So far, the available experimental results provide only limited information on the structure of the channel. From the histograms of the conductance fluctuations (Fig. 3) it is clear that under the conditions of our experiments the channel can assume a broad range of conductance values. This broad distribution of single-channel conductances may have a number of different origins. The channel may be an oligomeric structure composed of several of the 36 500dalton subunits of the matrix protein [14], and the number o f subunits may vary from one channel to the other. A well studied example of such an oligomeric channel with different conductance states is the alamethicin channel [54--57]. In t h a t case, however, several peaks in the histogram should be found, corresponding to different discrete states. Another possibility would be that the porin is chemically (or conformationally) heterogeneous. The matrix protein of E. coli strain K12 in known to consist of at least two subfractions representing two closely related peptides [ 23,24,33--35 ]. Furthermore, the possibility t h a t the isolation procedure which includes a trypsin t r e a t m e n t and solubilisation with detergent m a y lead to a spectrum of peptides with slightly modified structure has been shown to be rather unlikely [32]. In addition, on polyacrylamide gel electrophoresis no heterogeneity of the protein was observed. Control experiments carried o u t in the absence of porin have shown t h a t the detergent in the aqueous solution does n o t interfere with the experiments.

318 On the other hand, detergent molecules are probably associated with the solubilized protein and thus a certain a m o u n t of detergent may be incorporated into the membrane together with the porin. The question therefore arises whether the observed conductance effects are influenced by detergent introduced into the membrane. From the finding that the histogram of conductance fluctuation is virtually the same for cholate and for SDS preparations this possibility seems rather unlikely, although it cannot be completely excluded. From gel filtration experiments it is known that the solubilized porin is highly aggregated, the molecular weight being of the order of 800 000 [32]. This corresponds to an aggregate of a b o u t 20 monomers of molecular weight 36 500. It is not known whether in aqueous solution an equilibrium between monomers and aggregates exists and in what form the porin is incorporated into the membrane. Even if only monomers are able to enter the lipid, the porin m a y form aggregates of various sizes in the lipid and this may result, as discussed above, in a broadening of the conductance histogram. It is therefore not clear at the m o m e n t whether channels are formed singly and the distribution of conductance values results from chemical (or conformational) heterogeneity of the individual channels, or whether each reductance step represents the formation of an aggregate of channels of variable size.

Note added in p r o o f (Received May 29th, 1978} Schindler and Rosenbusch [59], using a different method, have recently incorporated porin into planar bilayer membranes and observed discrete conductance steps of a similar size as those reported here. Acknowledgement The authors wish to thank Drs. Alpes, Bamberg and Weidekamm for interesting discussions and Dr. Rosenbusch (Basel) for kindly providing a sample of the matrix protein. The excellent technical assistance of Mrs. R. Dieterle, Mr. G. Ehmann and Mr. P. Jauch is gratefully acknowledged. This work was financially supported by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 138}. References 1 2 3 4 5 6 7 8 9 10 11

G l a u e r t , A.M. a n d T h o r n l e y , M.J. ( 1 9 6 9 ) A n n u . R e v . M i c r o b i o l . 2 3 , 1 5 9 - - 1 9 9 C o s t e r t o n , J.W., I n g r a i n , J . M . a n d C h e n g , K . - J . ( 1 9 7 4 ) B a c t e r i o l . Rev. 3 8 , 8 7 - - 1 1 0 N a k a e , T. a n d N i k a i d o , H. ( 1 9 7 5 ) J. Biol. C h e m . 2 5 0 , 7 3 5 9 - - 7 3 6 5 D e c a d , G . M . a n d N i k a i d o , H. ( 1 9 7 6 ) J. B a c t e r i o l . 1 2 8 , 3 2 5 - - 3 3 6 Z i m m e r m a n n , W. a n d R o s s e l e t , A. ( 1 9 7 7 ) J. B a c t e r i o l . , s u b m i t t e d f o r p u b l i c a t i o n N i k a i d o , H., S o n g , S . A . , S h a l t i e l , L. a n d N u r m i n e n , M. ( 1 9 7 7 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 76,324--330 L u t k e n h a u s , J . F . ( 1 9 7 7 ) J. B a c t e r i o l . 1 3 1 , 6 3 1 - - 6 3 7 N i k a i d o , H. ( 1 9 7 6 ) B i o c h i m . B i o p h y s . A c t a 4 3 3 , 1 1 8 - - 1 3 2 M i n r a , T. a n d M i z u s h i m a , S. ( 1 9 6 9 ) B i o c h i m . B i o p h y s . A c t a 1 9 3 , 2 6 8 - - 2 7 6 S c h n a i t m a n , C . A . ( 1 9 7 1 ) J. B a c t e r i o l . 1 0 8 , 5 5 3 - - 5 6 3 H e n n i n g , U.° H 5 h n , B. a n d S o n n t a g , J. ( 1 9 7 3 ) E u r . J. B i o c h e m . 3 9 , 2 7 - - 3 6

319

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

S c h n a i t m a n , C. ( 1 9 7 3 ) A r c h . B i o c h e m . B i o p h y s . 1 5 7 , 5 5 3 - - 5 6 0 S c h n a i t m a n , C . A . ( 1 9 7 4 ) J. B a c t e r i o l . 1 8 8 , 4 4 2 - - 4 5 3 R o s e n b u s c h , J . P . ( 1 9 7 4 ) J. Biol. C h e m . 2 4 9 , 8 0 1 9 - - 8 0 2 9 H a n e r , J. a n d H e n n i n g , U. ( 1 9 7 4 ) P r o c . N a t l . A c a d . Sci. U.S. 7 1 , 2 0 1 8 - - 2 0 2 1 M f i h l r a d t , P . F . ( 1 9 7 4 ) E u r . J. B i o c h e m . 4 3 , 5 5 3 - - 5 6 9 B r a u n , V. ( 1 9 7 5 ) B i o c h i m . B i o p h y s . A c t a 4 1 5 , 3 3 5 - - 3 7 7 I n o u y e , M. ( 1 9 7 5 ) in M e m b r a n e B i o g e n e s i s , M i t o c h o n d r i a , C h l o r o p l a s t s , B a c t e r i a ( T z a g o l o f f , A., e d . ) , p p . 3 5 1 - - 3 9 1 , P l e n u m Press, N e w Y o r k H e n n i n g , U. a n d H a l l e r , J. ( 1 9 7 5 ) F E B S L e t t . 55, 1 6 1 - - 1 6 4 G a r t e n , W., H i n d e n n a c h , J. a n d H e n n i n g , U. ( 1 9 7 5 ) E u r . J. B i o c h e m . 6 0 , 3 0 3 - - 3 0 7 O v e r a t h , P., B r e n n e r , M., G u l i k - K r z y w i c k i , T., S h e c h t e r , E. a n d L e t e l l i c r , L. ( 1 9 7 5 ) B i o c h i m . B i o p h y s . Acta 389, 358--369 S m i t , J., K a m i o , Y. a n d N i k a i d o , H. ( 1 9 7 5 ) J. B a c t e r i o l . 1 2 4 , 9 4 2 - - 9 5 8 S c h m i t g e s , C.J. a n d H e n n i n g , U. ( 1 9 7 6 ) E u r . J. B i o c h e m . 6 3 , 4 7 - - 5 2 H a s e g a w a , Y., Y a m a d a , H. a n d M i z u s h i m a , S. ( 1 9 7 6 ) J. B i o c h e m . 8 0 , 1 4 0 1 - - 1 4 0 9 Palva, E.T. a n d R a n d a l l , L . L . ( 1 9 7 6 ) J. B a c t c r i o l . 1 2 7 , 1 5 5 8 - - 1 5 6 0 R e i t h m e i e r , R . A . F . a n d B r a g g , P.D. ( 1 9 7 7 ) B i o c h i m . B i o p h y s . A c t a 4 6 6 , 2 4 5 - - 2 5 6 V a n A l p h e n , L., L u g t e n b e r g , B., V a n B o x t e l , R. a n d V e r h o e f , K. ( 1 9 7 7 ) B i o c h i m . B i o p h y s . A c t a 4 6 6 , 257--268 V e r k l e i j , A., V a n A l p h e n , L., Bijvelt, J. a n d L u g t e n b e r g , B. ( 1 9 7 7 ) B i o c h i m . B i o p h y s . A c t a 4 6 6 , 269--282 S t e v e n , A . C . , t e n H e g g e l e r , B., Miiller, R . , K i s t l e r , J. a n d R o s e n b u s c h , J.P. ( 1 9 7 7 ) J. Cell Biol. 72, 292--301 Y u , F. a n d M i z u s h i m a , S. ( 1 9 7 7 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 7 4 , 1 3 9 7 - - 1 4 0 2 Palva, E.T. a n d R a n d a l l , L . L . ( 1 9 7 8 ) J. B a c t e r i o l . , in t h e p r e s s N a k a e , T. ( 1 9 7 6 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 7 1 , 8 7 7 - - 8 8 4 B a s s f o r d , J r . , P . J . , D i e d r i c h , D . L . , S c h n a l t m a n , C . A . a n d R e e v e s , P. ( 1 9 7 7 ) J. B a c t e r i o l . 1 3 1 , 6 0 8 - - 6 2 2 V a n A l p h e n , W. a n d L u g t e n b e r g , B. ( 1 9 7 7 ) J. B a c t e r i o l . 1 3 1 , 6 2 3 - - - 6 3 0 L u g t e n b e r g , B., B r o n s t e i n , H., V a n S e l m , N. a n d P e t e r s , R. ( 1 9 7 7 ) B i o c h i m . B i o p h y s . A c t a 4 6 5 , 571--578 S e h n a i t m a n , C., S m i t h , D. a n d De Salsas, M . F . ( 1 9 7 5 ) J. Virol. 1 5 , 1 1 2 1 - - 1 1 3 0 D i e d r i c h , D . L . , S u m m e r s , A . O . a n d S c h n a i t m a n , C . A . ( 1 9 7 7 ) J. B a c t e r i o l . 1 3 1 , 5 9 8 - - 6 0 7 N i k a i d o , H., S o n g , S . A . , S h a l t i e l , L. a n d N u r m i n e n , M. ( 1 9 7 7 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 7 6 , 324--330 N a k a e , T. ( 1 9 7 5 ) B i o c h e m . B i o p h y s . Res. C o m m u n . 6 4 , 1 2 2 4 - - 1 2 3 0 N a k a e , T. ( 1 9 7 6 ) J. Biol. C h e m . 2 5 1 , 2 1 7 6 - - 2 1 7 8 Laemmli, U.K. (1970) Nature 227,680--685 L o w r y , O . H . , R o s e b r o u g h , N . J . , F a r r , A . L . a n d R a n d a l l , R . J . ( 1 9 5 1 ) J. Biol. C h e m , 1 9 3 , 2 6 5 - - 2 7 5 R a n d a l l , L. a n d S c h w a r t z , M. ( 1 9 7 3 ) J. B a c t e r i o l . 1 1 6 , 1 4 3 6 - - 1 4 4 6 B e n z , R . , S t a r k , G., J a n k o , K. a n d L~/uger, P. ( 1 9 7 3 ) J. M e m b r a n e Biol. 1 4 , 3 3 9 - - 3 6 4 B e n z , R . a n d J a n k o , K. ( 1 9 7 6 ) B i o c h i m . B i o p h y s . A c t a 4 5 5 , 7 2 1 - - 7 3 8 S a n d e r s , H. ( 1 9 6 7 ) B i o c h i m . B i o p h y s . A c t a 1 4 4 , 4 8 5 - - 4 8 7 S i n g l e t o n , W.S., G r a y , M.S., B r o w n , M.L. a n d W h i t e , J . L . ( 1 9 6 5 ) J. A m . Oil C h e m . S o c . 4 2 , 5 3 - - 5 6 S m i t h , L . L . , M a t t h e w s , W.S., Price, J . C . , B a c k m a n n , R . C . a n d R e y n o l d s , B. ( 1 9 6 7 ) J. C h r o m a t o g r . 2 7 , 187--205 T i e n , H . T . , C a r b o n e , S. a n d D a w i d o w i c z ( 1 9 6 6 ) N a t u r e 2 1 2 , 7 1 8 - - 7 1 9 B e n z , R . , F r b h l i e h , O . , L ~ u g e r , P. a n d M o n t a l , M. ( 1 9 7 5 ) B i o c h i m . B i o p h y s . A c t a 3 9 4 , 3 2 3 - - 3 3 4 B e n z , R. a n d Gisin, B . F . ( 1 9 7 6 ) J. M e m b r a n e Biol., in t h e p r e s s Viles, Jr., F . J . , S i l v e r m a n , L. ( 1 9 4 9 ) A n a l . C h e m . 2 1 , 9 5 0 - - 9 5 3 H l a d k y , S.B. a n d H a y d o n , D . A . ( 1 9 7 2 ) B i o e h i m . B i o p h y s . A e t a 2 7 4 , 2 9 4 - - 3 1 2 MueUer, P. ( 1 9 7 5 ) A n n . N . Y . A c a d . Sci. 2 6 4 , 2 4 7 - - 2 6 4 E i s e n b e r g , M., H a l l , J . E . a n d M e a d , C.A. ( 1 9 7 3 ) J. M e m b r a n e Biol. 1 4 , 1 4 3 - - 1 7 6 B o h e i m , G. ( 1 9 7 4 ) J. M e m b r a n e Biol. 1 9 , 2 7 7 - - 3 0 3 G o r d o n , L . G . M . a n d H a y d o n , D . A . ( 1 9 7 5 ) T r a n s . R. Soc. B 2 7 0 , 4 3 3 - - 4 4 7 R o s e n b u s c h , J . P . a n d MiiUer, R . ( 1 9 7 7 ) in L i p i d - P r o t e i n C o m p l e x e s ( P e e p e r s , H. a n d Massu6, J.P., eds.), E u r o p e a n Press, G e n t S c h i n d l e r , H. a n d R o s e n b u s c h , J . P . ( 1 9 7 8 ) P r o c . N a t l . A e a d . Sci. U.S., in t h e p r e s s