University of Groningen

Membrane reconstitution and functional analysis of a sugar transport systemKnol, Jan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:1999

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Knol, J. (1999). Membrane reconstitution and functional analysis of a sugar transport system Groningen:s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 03-04-2018

Chapter 5

MEMBRANE RECONSTITUTION MEDIATED BY DDMOR TRITON X-100: ASYMMETRIC KINETICS OF THE

LACTOSE TRANSPORT SYSTEM OF STREPTOCOCCUS

THERMOPHILUS

Jan Knol, Klaas Sjollema, and Bert Poolman

Biochemistry 37, 16410-16415 (1998)

SUMMARY

The lactose transport protein (LacS) of Streptococcus thermophilus was reconstituted into preformedliposomes that were treated with commonly employed detergents. The transport activity was markedlyhigher with Triton X-100 than with DDM and to rationalize the difference, the lipid/detergent structuresthat are formed were studied by cryo-transmission electron microscopy. Surprisingly, the two non-ionicdetergents Triton X-100 and DDM affected the liposome structures in a completely different manner. Atdetergent to lipid ratios of ~1, the liposomes titrated with Triton X-100 remained intact, whereas largebilayer sheets were formed when DDM was used. More strikingly the bilayer structure was completely lostat higher DDM to lipid ratios (1 - 1.5) and long thread-like micelles were formed. Triton X-100 on theother hand left the liposome and bilayer structure intact over a wide detergent concentration range. Thedifferent lipid-detergent structures formed by Triton X-100 and DDM readily explain why membraneproteins such as LacS reconstitute randomly when the reconstitution is mediated by DDM andunidirectionally when mediated by Triton X-100. The orientation of the reconstituted LacS protein explainsthe differences in transport activity obtained with DDM and Triton X-100 as the kinetics are very differentfor the opposite directions of transport.

- 48 - Chapter 5

INTRODUCTION

Although many integral membrane proteins have been reconstituted into liposomes, the molecularmechanisms of membrane reconstitution are largely unknown and, generally, the efficacy of reconstitutionis only qualitatively assessed. Structural data on the complexes formed by membrane proteins anddetergents as well as the process of protein insertion into lipid bilayers is limited. Such information is notonly crucial for the optimization of functional assays but also for the formation of ordered proteinassemblies (2-D crystals) for structure analysis of the proteins by electron crystallography. We have usedthe purified lactose transport protein (LacS) of Streptococcus thermophilus as an example to follow theprocess of membrane reconstitution with different detergents, but the methodology appears applicable toother proteins as well (Rigaud et al., 1995; Gaillard et al., 1996; Hagting et al., 1997). The LacS proteinis a polytopic membrane protein that belongs to a large family of secondary transport proteins (Poolmanet al., 1996). It differs from most other secondary transport proteins by the presence of a hydrophiliccarboxyl-terminal regulatory domain (Poolman et al., 1995b).

To optimize the reconstitution process for the purified LacS protein, the strategy of Rigaud et al. wasfollowed which involves the stepwise solubilization of preformed liposomes, and protein incorporationat the different stages of liposome solubilization (Rigaud et al., 1995; Knol et al., 1996). The physical stateof the liposomes during the titration with detergent was followed by measuring the optical density at 540nm. The solubilization curve can be divided into three stages (Lichtenberg, 1985; De La Maza and Parra,1997). During Stage I detergent molecules partition between the aqueous buffer and the bilayer. Stage IIstarts when liposomes are saturated with detergent (onset of solubilization = Rsat) and continues upon afurther increase of the detergent concentration, inducing liposome solubilization and the formation ofmicelles. At stage III, when the OD has reached its minimal value (Rsol), the mixture consists of micellesat varying detergent/lipid ratios. Although the optical density offers a convenient tool to follow thesolubilization, the actual effects of the detergents on the liposomes are not revealed. Here, we describe themacromolecular structures that are formed by the lipid/detergent mixtures in the process of membranereconstitution mediated by the non-ionic detergents Triton X-100 and DDM. Based on these findings amodel is proposed for the unidirectional incorporation of LacS into liposomes mediated by Triton X-100,and for the random incorporation mediated by DDM.

EXPERIMENTAL PROCEDURESMaterials - [D-glucose-1-14C]lactose (2.11 TBq/mol) and 3-(Trifluoromethyl)-3-(m-[125I]iodophenyl)diazirine

(370GBq/mmol) were obtained from the Radiochemical Centre Amersham, England. Biobeads SM-2 were from Bio-Rad, Triton

X-100, from Boehringer Mannheim and n-dodecyl-β-D-maltoside from Anatrace. Total E. coli lipids and L-α-phosphatidyl-

choline from egg yolk were from Avanti Polar Lipids.

Bacterial strains and growth conditions - S. thermophilus ST11(∆lacS) carrying plasmid pGKHis was grown semi-

anaerobically at 42 oC in Elliker broth supplemented with 0.5% beef extract, 20 mM lactose plus 5 µg/ml of erythromycin (Knol

et al., 1996).

Titrations of liposomes with detergent - Liposomes of acetone/ether washed E. coli lipids and L-α-phosphatidyl choline

from egg yolk in a ratio of 3:1 (weight/weight) were made by dissolving lipids (20 mg/ml) in 50 mM potassium phosphate, pH

7.0, plus 2 mM MgSO4, followed by three freeze/thaw cycles and extrusion through polycarbonate filters with 400 nm pore size

(Mayer et al., 1986). Liposomes were diluted to 4 mg of PL/ml and titrated by stepwise addition of ~0.5 mM of Triton X-100

or DDM. The effect of the detergents on the liposomes was followed by measuring the optical density at 540 nm using a SLM-

Aminco spectrophotometer (Lichtenberg, 1985).

Cryo-Transmission Electron Microscopy - A small droplet (ca. 5µl) of (detergent treated)-liposome suspension was

applied on one side of a Formvar film-coated, 300 or 500 mesh copper grid. The grid was mounted in a home-made plunging

device, carefully blotted with filter paper, and immediately plunged into liquid propane cooled by liquid nitrogen. Specimen

were subsequently stored under liquid nitrogen. The prepared frozen-hydrated specimens were analyzed in a Phillips CM10

Asymmetric Kinetics of the Lactose Transport System of S. thermophilus - 49 -

Transmission Electron Microscope equipped with a model 626 Gatan cryo stage at a temperature of –160 oC to –170 oC. Images

were recorded on a Kodak FGP film at 100 kV.

Purification and reconstitution of LacS - The protein was purified essentially as described (Knol et al., 1996). For the

reconstitution, preformed liposomes (4 mg of PL/ml) were equilibrated for 60 min with the appropriate amounts of DDM or

Triton X-100, mixed with the purified protein in 200 mM imidazole, pH 7.0, 100 mM NaCl, 10% (v/v) glycerol, plus 0.05%

of detergent and incubated for 30 min at 20 oC under gentle agitation. The final protein to lipid ratio was always 1:100-150

(w/w) and the final glycerol concentration never exceeded 1% (v/v). Polystyrene beads were added at a wet weight of 80 mg/ml

and the samples were incubated for another 2 hours at room temperature to remove the detergent. Fresh biobeads were added

twice and the samples were incubated at 4oC for 3 hours and overnight, respectively. The proteoliposomes were washed with

50 mM potassium phosphate, pH 7.0, harvested by centrifugation and stored in liquid nitrogen.

Lactose counterflow in proteoliposomes - Proteoliposomes were resuspended in 50 mM potassium phosphate, pH 7.0,

2 mM MgSO4 plus varying concentrations of lactose and frozen in liquid nitrogen. After thawing the samples at room

temperature, the liposomes were extruded through 400 nm polycarbonate filters. After centrifugation, aliquots of concentrated

proteoliposome suspension were diluted into 200 µl of potassium phosphate, pH 7.0, 2 mM MgSO4 (KPM) plus [14C]lactose

at concentrations indicated. The transport reactions were performed at 30 oC and processed as described previously (Knol et

al., 1996).

Efficiency of protein reconstitution - To determine the efficiency of protein reconstitution, LacS was labeled with the

photo-activatable reagent [125I]TID. Membrane vesicles (4 mg of protein/ml) were incubated with 65 µM TID and the probe

was crosslinked by incubating the samples under UV light (254 nm) for 2 min at 0 oC. Subsequently, the membrane vesicles

were solubilized and LacS was purified and reconstituted as described. The proteoliposomes were separated from non-

incorporated protein by sucrose gradient centrifugation (In’t Veld et al., 1992). To visualize the proteoliposomes in the gradient,

octadecylrhodamine-β-chloride (R18) was incorporated at 50 nmol per ml of liposomes (20 mg PL/ml) prior to the

freeze/thaw/extrusion cycles. Non-incorporated R18 was removed by washing with 50 mM potassium phosphate, pH 7.0.

Fractions from the sucrose gradient containing lipid were pooled and the amount of protein associated with the lipid was

compared to the total amount of protein used for the reconstitution.

Orientation of LacS in proteoliposomes - The orientation was determined with antibodies raised against the C-terminal

part of the protein, i.e., the cytoplasmically located IIA domain. (Proteo)liposomes were washed twice with 0.5 % BSA in PBS

and resuspended at 50 µg of LacS/ml. Subsequently, they were incubated for 2 h with antibodies raised against the purified IIA

domain of LacS (Gunnewijk and Poolman, unpublished results) or pre-immune serum at a 1000-fold dilution. After the

proteoliposomes were washed 4 times with 0.5% BSA in PBS, the proteins were separated by SDS-PAGE. The proteins were

blotted onto polyvinylidene difluoride membranes and the amount of bound antibody was determined by densitometry, following

detection with a secondary antibody-alkaline phosphatase conjugate using CSPDTM as a substrate.

RESULTS

Lactose transport activity - To optimize the detergent/lipid ratio for reconstitution, liposomes weretitrated stepwise with DDM. The solubilization of the liposomes was followed by measuring the OD540,and, after equilibration with the appropriate amount of detergent, LacS was added at a 1:100 protein tolipid ratio and reconstituted as described. The dependence of the transport activity on the DDMconcentration, i.e., the physical state of the liposomes, is shown in Fig 1. In case of DDM, transport activitywas only obtained when liposomes were destabilized with a certain minimal concentration of detergent(>3.5 mM). This coincides with the maximum optical density in the titration curve, which is indicative forthe onset of solubilization (Rsat; detergent to lipid ratio ~1). The transport activity was highest at Rsat andwhen the detergent/lipid ratios were increased further the transport activity declined. At the highest DDMconcentrations used (Rsol; detergent/lipid ~ 3), the activity had dropped more than 5-fold.

Also in case of Triton X-100, the liposomes used for the reconstitution required a minimal detergentconcentration for activity (Fig. 1, inset). Importantly, the Triton X-100 mediated reconstitutions yieldedtransport activities that were relatively constant over a broad detergent concentration range, and theseactivities were always higher than those of DDM mediated reconstitution (see also Knol et al., 1996). Anexplanation for the observed differences might be found in the way these detergents interact with theliposomes and affect the macromolecular lipid/detergent structures.

- 50 - Chapter 5

[n-Dodecyl-β-D-Maltoside], mM0 2 4 6 8 10 12

Co

unte

rflo

w a

ctiv

ity (

nm

ol/m

g o

f p

rote

in x

min

)

0

50

100

150

200

Ab

sorb

anc

e a

t 5

40 n

m

0.10

0.20

0.30

0.40

0.50

[Triton X-100], mM0 2 4 6 8

0

200

400

600

800

1000

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Rsat

Rsol

Rsat Rsol

Fig. 1. Lactose counterflow activity in proteoliposomes made viaDDM- or Triton X-100-mediated reconstitution. Purified LacSprotein was added to preformed liposomes (4 mg PL/ml)equilibrated with different concentrations of DDM or Triton X-100. The physical state of the liposomes was followed bymeasuring the OD540 (•). After removal of detergent, theproteoliposomes were washed and resuspended into potassiumphosphate, pH 7.0, 2 mM MgSO4 (KPM) plus 10 mM lactose. Thecounterflow reaction was started by diluting concentratedproteoliposome suspensions 200-fold into KPM containing 54 µM[ 14C]lactose, final concentration.

Cryo-transmission electron microscopy of liposomes titrated with Triton X-100 or DDM - Cryo-TEM pictures were taken of intact liposomes after extrusion through 400 nm filters. In general theliposomes were unilamellar, regularly shaped and >90% had a diameter of 100-300 nm. Equilibration ofthe liposomes with Triton X-100 at Rsat did not result in observable differences in the macromolecularstructures as assessed by cryo-TEM, whereas the OD540 did increase (Fig 2D). At these detergent/lipidratios, the initial increase in OD probably reflects a swelling of the liposomes due to partitioning ofdetergent molecules into the lipid bilayers and/or fusion of the liposomes. Surprisingly, a significantfraction of the liposomes remained intact at stage II of the solubilization, even at Triton X-100concentrations that are far beyond Rsat (Fig. 2E). The decrease in OD540 corresponds with a decrease in thesize of the liposomes. It seems that during the solubilization, lipids are extracted by the detergent withoutdisrupting the macromolecular structure of the liposomes. Only at concentrations above Rsol (7 mM)liposomes were not observed any longer and all the lipid seemed to be present in mixed micelles of lipidand detergent. However, even at relatively high detergent/lipid ratios, the lipids still tended to formliposomal structures despite the relative high concentration of Triton X-100 in the bilayers.

These observations are in sharp contrast to the results obtained with liposomes titrated with DDM.Starting with the same liposome suspension, the Rsat was reached at higher detergent concentrations (4mM) and the equilibration of DDM with the lipids was slow (2-3 hours). Cryo-transmission electronmicroscopy shows that only a few liposomes are still intact at this point. Aberrantly shaped structures oflipid sheets and open liposomes predominated under these conditions (Fig. 4B). Higher concentrations ofDDM led to a further rearrangement of the lipids, and long threads (>200 nm) composed of stretchedmicelles of lipid and detergent were formed during stage II. Intact liposomes or even sheets of lipids werebarely detectable under these conditions (Fig. 4C). The rearrangement of liposomes into these uniquestructures is probably slow and explains the slow equilibration of DDM and lipid during Stage I. In StageII the OD540 decreased further but equilibration was reached more rapid (several minutes), and the electronmicroscopic pictures showed that the threads break and become smaller. Ultimately, at concentrationsabove Rsol, all the lipid and DDM is present in micelles.

Asymmetric Kinetics of the Lactose Transport System of S. thermophilus - 51 -

Fig. 2. Cryo-transmission electron micro-graphs of liposomes titrated with DDM orTriton X-100. Liposomes were diluted to 4mg/ml and titrated with DDM or Triton X-100.Preformed liposomes without detergent (A);liposomes treated with 3.8 mM DDM (Rsat) (B),5.7 mM DDM (C), 1.8 mM Triton X-100 (Rsat)(D), or 2.7 mM Triton X-100 (E). Barrepresents 100 nm.

Efficiency of protein reconstitution - To assess the efficiency of reconstitution into the differentstructures formed during the solubilization of the liposomes with Triton X-100 and DDM, the amount ofradiolabeled LacS associated with liposomes was determined after fractionation of the protein and(proteo)liposomes on sucrose gradients. The reconstitution with Triton X-100 was very efficient and atleast 95% of the protein was incorporated using different concentrations of Triton X-100 to mediate thereconstitution. For DDM the efficiency of reconstitution at Rsat and Rsol were 85 and 80%, respectively,but this does not explain the lower transport activities observed. At concentrations of DDM above Rsol, thereconstitution was less reproducible and the fraction of protein inserted into the proteoliposomesdiminished in some cases to ~50%.

Orientation of LacS in proteoliposomes - The orientation of the incorporated LacS protein wasdetermined for reconstitutions at Triton X-100 and DDM concentrations that gave the typicalmacromolecular lipid-detergent structures as shown in Fig. 2 (BC and DE, respectively). The binding ofantibodies specifically recognizing the hydrophilic IIA domain of LacS was compared at Rsat and Rsol ofDDM and Triton X-100 (Fig. 3). In case of Triton X-100 mediated reconstitution, the antibody bindingwas maximal and equivalent to total immune precipitated protein used for the reconstitution at bothdetergent concentrations. Since all the IIA domain was available for the antibodies, the LacS protein mustbe reconstituted in an inside-out orientation. About 35% of the maximal amount of antibody bound, whenthe reconstitution was performed at DDM concentrations equivalent to Rsat and Rsol (Fig. 3; lane 3 and 4).The low amount of antibody binding can not only be explained by a lower amount of LacS in theproteoliposomes as the reconstitution efficiency was only 10-15% lower for DDM. It can be concluded that

- 52 - Chapter 5

[Lactose]out (mM)

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Rat

e of

Lac

tose

Upt

ake

(nm

ol /

min

* m

g of

pro

tein

)

0

100

200

300

400

500

600

Lacin:10 mM

Lacin:5 mM

Lacin:1 mM

Lacin:0.05/0.5 mM

LacS incorporates unidirectionally (inside-out) into liposomes when Triton X-100 is used to mediate thereconstitution, whereas the incorporation is random when DDM is used.

Fig. 3. Orientation of LacS in proteoliposomes.Proteoliposomes were resuspended at 50 µg ofLacS/ml in PBS plus 0.5 % BSA and,subsequently, incubated for 2 h with antibodiesraised against the IIA domain of LacS (lanes 1-6) or pre-immune serum (lanes 7 and 8).Immune-precipitated LacS, set at 100% (1);liposomes (2); proteoliposomes made with 4 mMDDM (3); 8 mM DDM (4); 1.8 mM Triton X-100(5); 4 mM Triton X-100 (6); 4 mM DDM (7);and 1.8 mM Triton X-100 (8).

Kinetic parameters of lactose transport - Although secondary transport proteins generally operatereversibly in the two opposing directions of transport, i.e., the direction of transport is determined by thedirectionality of the driving force and not by the orientation of the protein (Poolman and Konings, 1993),the kinetic parameters can be different for outside to inside and inside to outside translocation. Theapparent affinity constant (Km

app) for lactose exchange at the outer surface of the membrane in vivo is about10 mM (Poolman et al., 1995). Since Triton X-100 mediated reconstitution results in an inside-outorientation of the LacS protein, the counterflow activity should not saturate up to an internal lactoseconcentration of 10 mM. To study the kinetic parameters of LacS-mediated counterflow transport, theinternal and external lactose concentrations were varied over a wide range. The data clearly show that uponincreasing the outside lactose concentration the uptakesaturated with Km

app values in the submillimolar range,whereas the uptake rate increased almost linearly withinternal lactose concentrations up to 10 mM (Fig. 4).The latter value is consistent with the low affinity siteat the outer surface that was reported in the in vivostudies. The Km

app in the submillimolar range, i.e., 0.1- 0.3 mM, determined for the outer surface of themembrane in the proteoliposomes must therefore reflectthe internal surface in vivo.

Fig. 4. Kinetics of lactose counterflow. Proteoliposomes made byTriton X-100 (2 mM) mediated reconstitution, were equilibratedwith different concentrations of lactose and subsequently diluted200-fold to start the reaction at the [14C]-lactose concentrationsindicated. Initial uptake rates were determined in triplicate, andthe averaged data were fitted to the Michaelis-Menten equation.

Asymmetric Kinetics of the Lactose Transport System of S. thermophilus - 53 -

DISCUSSION

For a better understanding of membrane protein reconstitution and 2D-crystallization, it is necessaryto study the interactions between lipids, detergent and proteins. The success of membrane proteinreconstitution in active form and the formation of protein arrays for 2D-crystallization largely depends onthese interactions. Varying results have been obtained depending on the protein, lipid and detergent usedas well as the method of reconstitution. The activity of proteoliposomes has in most cases been optimizedby empirical methods, but the steps that have led to incremental increases in activity have not been studied.Important contributions have been made, however, by the groups of Rigaud and Møller who studied theprocess of membrane protein reconstitution in detail (Rigaud et al., 1995 ; Lund et al., 1989). In this paper,the interactions of DDM and Triton X-100 with liposomes made of E. coli lipids and egg PC (3:1) werestudied by cryo-transmission electron microscopy. From turbidity measurements, it was already clear thatsolubilization of liposomes is reproducible and that three stages can be distinguished during the processof solubilization (Lichtenberg, 1985; Rigaud et al., 1995). First, the detergent incorporates into the lipidbilayers (Stage I), which results in a peak of the turbidity at the onset of solubilization (Rsat). Subsequently,the liposomes disintegrate (Stage II) and a transition takes place from bilayer structures to mixed micelles.Finally, at Rsol, the liposomes are completely solubilized (Stage III) and further changes only involve theratio of detergent and lipids in the micelles. Although the turbidity measurements offer a good method tofollow the solubilization process, it does not give any information on the exact mechanism and thestructures formed during the solubilization. By employing cryo-TEM on the lipid-detergent mixtures, itis now demonstrated that the second stage in the solubilization is completely different for Triton X-100and DDM. The implications for protein reconstitution are schematically drawn in Fig. 5 and will beexplained in more detail.

In case of Triton X-100, intact liposomes were observed even near the end of Stage II. An equilibriumexists between Triton X-100-saturated liposomes and mixed micelles that diffuse in and out of the bilayer(Fig. 5). When more Triton X-100 is added the equilibrium shifts from the liposomal towards the micellarstate, and the opposite occurs when biobeads are used to lower the detergent concentration. The decreasein OD540 with increasing Triton X-100 concentrations parallels a decrease in the diameter of the liposomes,but the bilayer structure remains intact. We speculate that Triton X-100 has a molecular shape thatstabilizes membranes, in particular those that are largely composed of nonbilayer lipids such as PE andcardiolipin (Madden and Cullis, 1982). Notice that the E. coli lipid/egg PC mixture used consists of about50% PE and 10% cardiolipin. Patches of Triton X-100 molecules enable protein molecules, surroundedby detergent, to diffuse from the solution into the membrane. By removal of Triton X-100 with thehydrophobic beads, the equilibrium is shifted towards protein insertion. At the highest concentrations ofdetergent, when only mixed micelles are present, the addition of the polystyrene beads will (quickly) drivethe formation of bilayers into which the protein will insert. Because Triton X-100 is supporting a bilayerstructure membrane proteins such as LacS, of which the inner and outer surface differ significantly inhydrophobicity, will insert unidirectionally as it is unfavorable for the hydrophilic (IIA-)domain topenetrate the lipid bilayer once a closed vesicular structure is formed. Moreover, the liposomes initiallyformed from the mixed micelles are small in diameter, which induces a strong curvature of the membraneand a lipid asymmetry that might favor the unidirectional insertion even more (Huang and Mason, 1978).

With DDM, liposomal structures are only sustained until the onset of solubilization (Rsat). At thispoint a large fraction of the liposomes have lost their integrity and are present as membrane sheets. Thehighest transport activities were obtained when LacS was mixed with these structures. It seems likely thatmembrane proteins insert into these structures from both surfaces, without the need of the hydrophilic IIAdomain to cross the bilayer, leading to a random orientation. DDM is characterized by an hydrophobic tailof intermediate length and a bulky hydrophilic headgroup (Cullis et al., 1983). The alignment of these

- 54 - Chapter 5

molecules at saturating concentrations will cause curvature of the membrane and disruption of the bilayerstructure. Unlike Triton X-100, the physical properties of DDM, support the formation of long micellarthreads, at intermediate detergent/lipid ratios (Stage II). While this work was in progress, thread-likestructures were also reported for liposomes composed of egg phosphatidyl choline and egg phosphatidicacid that were treated with DDM (Lambert et al., 1998). Similar threads, with about the thickness of aphospholipid bilayer, have also been described for mixtures of octylglucoside and phospholipids (Vinsonet al., 1989). It is hard to imagine that a membrane protein will incorporate into such threads since theydo not seem to represent bilayer structures but rather form micellar structures (Fig. 5). We propose that,with DDM to mediate the reconstitution, membrane proteins are only incorporated into the membranesheets that are present at Rsat, i.e., when the stretched micelles are converted into liposomes. It is obviousthat the characteristics of DDM and Triton X-100 are also relevant for 2D-crystallization of membraneproteins. For stable arrays of protein and lipids, Triton X-100 may be a much more suitable detergent thanDDM as it permits protein insertion over a much broader range of detergent to lipid ratios and,consequently, exposes the protein to shorter periods that may lead to inactivation and/or aggregation.

Fig 5. Model for membraneprotein reconstitution mediatedby DDM or Triton X-100. A, B,C, D and E correspond to thestructures observed by cryo-TEM(Fig. 2). :, LacS protein.

The more or lessrandom orientation of LacSin proteoliposomes obtainedwith DDM also explains inpart the lower transportactivity measured in thecounterflow assay. Forproteoliposomes obtained viaTriton X-100 mediatedreconstitution, the affinityconstant for lactose at the outer surface of the membrane was ~0.2 mM, whereas it is at least 10 mM at theinner surface. In case of the DDM mediated reconstitution about 50% of the molecules have the lowaffinity side exposed to the outside of the proteoliposomes. At a substrate concentration of 54 µM, abouthalf of the molecules will not contribute significantly to the overall transport activity. The difference intransport activity at ‘low’ and ‘high’ concentrations of DDM reflects, in our opinion, loss of lipid, whichis more pronounced for higher concentrations of DDM. The activity of LacS appears to depend stronglyon the final lipid/protein ratio (unpublished results) and will thus be lower when more lipid is lost.