Investigations on the rapid transbilayer movement of phospholipids in
biogenic membranes

Dissertation

zur Erlangung des akademischen Grades
doctor rerum naturalium
(Dr. rer. nat.)
im Fach Biophysik

eingereicht an der
Mathematisch-Naturwissenschaftlichen Fakultät I
der Humboldt-Universität zu Berlin

von
Diplom Biophysiker Janek Kubelt
geborem am 07. März 1972 in Neustrelitz

Präsident der Humboldt-Universität zu Berlin:
Prof. Dr. Jürgen Mlynek

Dekan der Mathematisch-Naturwissenschaftlichen Fakultät I:
Prof. Dr. Michael Linscheid

Gutachter:
1. Prof. Dr. A. Herrmann
2. Prof. Dr. T. Pomorski
3. Prof. Dr. A.K. Menon

Tag der mündlichen Prüfung:16.04.2004

Table of contents

• 1  Introduction
  • 1.1  Phospholipid transmembrane movement in eukaryotic cells
  • 1.2  The architecture and dynamic of the bacterial envelope
    • 1.2.1 The membrane organization of E.coli
    • 1.2.2 Glycerophospholipids in E.coli
    • 1.2.3 Phospholipid movement across bacterial membranes
  • 1.3  Methods for the characterization of transmembrane distribution and movement of phospholipids in biological membranes
    • 1.3.1 Assays for the determination of transmembrane movement and distribution of endogenous phospholipids
    • 1.3.2  Assays for the determination of transmembrane movement and distribution of phospholipid analogues
• 2  Scope
• 3  Material and Methods
  • 3.1 Chemicals
  • 3.2  Preparation of inverted inner membrane vesicles from E.coli
  • 3.3  Reconstitution of IIMV derived from E.coli
  • 3.4  Incorporation of NBD-labeled phospholipids into IIMV
  • 3.5  The BSA back-exchange assay
    • 3.5.1 Extraction of fluorescent labeled phospholipid analogues by BSA – cuvette experiments
    • 3.5.2  The stopped-flow assay
  • 3.6  The dithionite assay
  • 3.7  Ion exchange chromatography
  • 3.8  SDS-PAGE analysis
  • 3.9 Methods for the determination of protein concentration
    • 3.9.1 The Lowry method modified by Peterson
    • 3.9.2 The bicinchoninic acid (BCA) Method
  • 3.10  The lipid extraction procedure
  • 3.11 Quantitation of phospholipids
  • 3.12 Detergent determination
  • 3.13  The measurement of the purity of isolated IIMV
  • 3.14  Kinetic analysis
• 4  Results
  • 4.1  Incorporation of fluorescent phospholipid analogues into IIMV
  • 4.2  Transbilayer movement of fluorescent phospholipid analogues across IIMV membranes
  • 4.3  Transbilayer movement of fluorescent phospholipid analogues across the membrane of reconstituted proteoliposomes derived from IIMV
  • 4.4  Effect of proteins on the transbilayer movement of phospholipids
    • 4.4.1  Extraction of M-C6-NBD-PE from IIMV membranes
    • 4.4.2  Reduction of M-C6-NBD-PE in IIMV-derived membranes by dithionite
  • 4.5  Effect of the chain length of fluorescent phospholipid analogues on the transbilayer movement across IIMV-derived membranes
  • 4.6  Protein modifying treatment of reconstituted proteoliposomes
  • 4.7  Ion exchange chromatography (IEC) with Triton extracts derived from IIMV of E.coli
    • 4.7.1  Efficiency of the separation of proteins from E.coli with IEC
    • 4.7.2  Enrichment of flippase activity of inner membrane proteins of E.coli by anion exchange chromatography (AEC)
    • 4.7.3  Successive fractionation of solubilized proteins from IIMV with anion exchange chromatography
• 5  Discussion
  • 5.1  Transbilayer movement of short-chain, fluorescent phospholipid analogues in IIMV and reconstituted proteoliposomes derived from IIMV
  • 5.2  Effect of proteins on the transport of fluorescent phospholipid analogues
  • 5.3  Ion exchange chromatography - Attempts to enrich flippase activity
  • 5.4  Are specific proteins required for phospholipid flip-flop?
• 6  Future perspectives
• 7  Summary
• 8  Zusammenfassung
• Literature
• Abbreviations
• Acknowledgement
• Publications
• Erklärung

Tables

• Table 1 : Half-times of transbilayer movement of NBD-labeled phospholipids across the membrane of IIMV and of extraction of analogues from the outer leaflet.
• Table 2 : Half-times of transbilayer movement of NBD-labeled phospholipids across the membrane of reconstituted IIMV-derived proteoliposomes and extraction of analogues from the outer leaflet
• Table 3: Half-times of transbilayer movement of M-C6-NBD-PE across E.coli -derived proteoliposomes determined by the dithionite assay. Three independent experiments (n=3) were analyzed by the modified three-compartment model.
• Table 4: Partial purification of phospholipid flippase activity from E.coli inner membrane TE. The data are representative for five independent experiments.
• Table 5: Partial purification of phospholipid flippase activity from E.coli inner membrane TE. The data are representative for two independent experiments and were calculated from data shown in Figure 21.
• Table 6 : Flippase activity in fractions of anion exchange chromatography. The data are representative for three independent experiments. pTE refers to proteoliposomes prepared from TE of IIMV. pQF, pQE4 and pQE designated for proteoliposomes reconstituted from AEC separated proteins of the flow-through and eluates, respectively.

Images

• Figure 1: Schematic representation of the E.coli envelope adapted from Raetz and Withfield with minor modifications (Raetz and Whitfield, 2002). Abbreviations: LPS, lipopolysaccharide; MDO, membrane-derived oligosaccharides; Kdo, 3-deoxy-d-manno-oct-2-ulosonic acid; PPEtn, phosphoethanolamine
• Figure 2: Structures of fluorescent short-chain phospholipid analogues used in the course of this thesis.
• Figure 3: Chemical reaction of dithionite with the fluorescent NBD-group. The fluorescence of NBD is quenched by generation of a non-fluorescent derivative due to chemical interaction of the dithionite radical with the nitroxide group of the NBD-molecule.
• Figure 4: Representative experiment of detection of the PLA 2 activity in IIMV and outer membrane vesicles (OMV) derived from E.coli . The white arrow indicates the lyso derivatives of the PLA 2 activity. The blue arrow indicates the non-cleaved N-Rh-PE molecules by PLA 2 .
• Figure 5 : Model of the transbilayer movement of phospholipid (analogues) across membranes and extraction of fluorescent NBD-lipid analogues by BSA. The rate constants k -1 and k +1 indicate the inward and outward movement of phospholipid analogues across the membrane, respectively. The extraction of fluorescent analogues is described by k +2 and the movement from BSA to the membrane is indicated by k -2 .
• Figure 6 :  Kinetics of incorporation of NBD-labeled phospholipid analogues into IIMV membranes. Two molpercent of M-C6-NBD-PE, M-C6-NBD-PC or M-C6-NBD-PG in 10mM HPS, with respect to the total lipid content of the IIMV, was added to IIMV (12.5µM final phospholipid concentration), and the kinetics of membrane incorporation of the analogues were monitored. At time point zero, the IIMV were added to the label suspension. The value of 100% corresponds to incorporation of all fluorescent labeled analogues. All experiments were performed at room temperature. Due to the low time resolution, the initial fluorescence increase upon mixing of IIMV with lipid analogues could not be adequately resolved.
• Figure 7 : Kinetics of extraction of M-C6-NBD-PE from IIMV by BSA. An aliquot (25 µl) of IIMV (2.89 mg protein/µmol phospholipid) were incubated with two milliliters of a buffered suspension of M-C6-NBD-PE (two molpercent of the lipid content) for 30 min at room temperature. Subsequently, the labeled IIMV were rapidly mixed with an equal volume of 4% (w/v) BSA in a stopped-flow accessory. The fluorescence decay was recorded with a time resolution of 0.2 s. The kinetics were normalized as follows: The initial fluorescence intensity (before BSA extraction) was set to one, the intensity after 300 s to zero. The curve represents the average of five measurements (A). The solid yellow line represents the fit obtained by fitting the data to the three-compartment model. The dashed blue line was obtained by fitting the data to a monoexponential function. The residuals of the model fit (B) according to the three-compartment model ( 3.14 ) and the monoexponential fit (C) are depicted.
• Figure 8 : Time dependence of extraction of M-C6-NBD-PE at room temperature. An aliquot of IIMV was mixed with buffer containing two molpercent of M-C6-NBD-PE of the total lipid content and the mix was immediately transferred to the stopped-flow device. Subsequently, the IIMV/label suspension was rapidly mixed with an equal volume of 4% (w/v) BSA in a stopped-flow accessory and the fluorescence decay was recorded with a time resolution of 0.5 s at the indicated time points. The kinetics were normalized as described before, and the data were fitted to the three-compartment model. The transbilayer distributions were elucidated from the fits and are displayed as function of incubation time of the analogues with IIMV. The open symbols refer to the concentration of analogues in the inner leaflet, the closed symbols to the concentration of analogues in the outer leaflet of the IIMV. The solid lines display simple one-phase regressions.
• Figure 9 : Complete extraction of M-C6-NBD-PE from IIMV by BSA. (A) IIMV (60 µM final phospholipid concentration) were labeled with two mol% M-C6-NBD-PE for 30 min, (B) subsequently, the analogues were extracted by 2% (w/v) BSA in 10 mM HPS for 300 s. Longer incubation (>300 s) did not change the fluorescence intensity. (C) BSA (2% (w/v))-bound M-C6-NBD-PE (two mol% of total phospholipid content of IIMV) in the absence of IIMV. (D) An aliquot of IIMV was added to BSA-bound M-C6-NBD-PE (two mol% of total phospholipid content of IIMV). (E) IIMV (equal to the amount used in (A) and (D)) in two ml 10 mM HPS. (F) 2% (w/v) BSA in two milliliters 10 mM HPS. The error bars indicate the standard deviation of two experiments. The experiments were performed at room temperature.
• Figure 10 : Stopped-flow kinetics of extraction of M-C6-NBD-PE from reconstituted IIMV by BSA. IIMV-derived proteoliposomes were labeled with two molpercent of the fluorescent phospholipid analogues during the reconstitution as described in 3.3 . In brief, equal volumes of proteoliposomes and 4% (w/v) BSA were mixed in the stopped-flow apparatus and the fluorescence decrease was recorded at room temperature (A). The curve represents the average of five separate kinetic traces. Kinetics were corrected for scattering. The solid yellow line represents the fit of the experimental data to the three-compartment model shown Figure 5 . The dashed blue line was obtained by a monoexponential fit of the data. The residuals for the fit of the data to the three-compartment model (B) and of a monoexponential fit (C) of the extraction kinetics are displayed.
• Figure 11: Extraction kinetics of pure ePC-vesicles. Control vesicles were prepared in parallel to proteoliposomes but without bacterial protein as described in 3.3 . (A) Pre-labeled ePC-vesicles (black line) were prepared by adding two molpercent M-C6-NBD-PE during the reconstitution procedure, whereas post-labeled ePC-vesicles (red line) were incubated with two molpercent of the same analogue prior the fluorescence measurement for 30 min. The stopped-flow back-exchange assay was performed as described for proteoliposomes (see Text and Figure 10 ). In panels B and C the residuals resulting from monoexponential fits from pre-labeled and post-labeled ePC-vesicles, respectively, are shown.
• Figure 12 : Stopped-flow-kinetics of extraction of M-C6-NBD-PG from reconstituted IIMV by BSA. IIMV-derived proteoliposomes were labeled with two molpercent of the fluorescent phospholipid analogues during the reconstitution as described in 3.3 . Equal volumes of proteoliposomes and 4% (w/v) BSA were mixed in the stopped-flow apparatus, and the fluorescence decrease was recorded at room temperature. A typical fluorescence kinetics of extraction of the fluorescent PG analogue is shown. The fluorescence intensity did not reach a stable plateau after 300 s of BSA incubation. In the inset, representative extraction kinetics of M-C6-NBD-PG with an elongated incubation time to 30 min is shown. Essentially all PG analogues were extracted by BSA after 30 min. The experiments were carried out at room temperature.
• Figure 13 : Effect of protein concentration on flip-flop and extraction of M-C6-NBD-PE. (A) The proteoliposomes were prepared with 4.5 mM ePC, increasing amounts of the TE and labeled with M-C6-NBD-PE during the reconstitution procedure. Subsequently, the labeled proteoliposomes made from IIMV were analyzed by stopped-flow assay as described before. The kinetics represent the mean of at least six records corresponding to pure ePC-liposomes (trace a), and to proteoliposomes with 40 µg protein/ml (trace b), 100 µg protein/ml (trace c) and 180 µg protein/ml (trace d). (B) The final amplitudes of fluorescence decrease (see (A)) of two independent experiments with increasing amounts of proteins are shown. The error bars depict the standard deviation. The letters a-d next to each data point correspond to the traces (for one of the two experiments) in panel A.
• Figure 14: Protein dependence of transmembrane movement and extraction of M-C6-NBD-PE. Proteoliposomes containing fluorescent phospholipid analogues and increasing amounts of TE were assayed by stopped-flow as described in the text. The percentage of fluorescence decrease of six independent experiments is depicted.
• Figure 15: Stopped-flow kinetics of fluorescence quenching of M-C6-NBD-PE from reconstituted IIMV by dithionite. IIMV-derived proteoliposomes were labeled with one molpercent of the fluorescent phospholipid analogues during the reconstitution as described in 3.3 . Equal volumes of proteoliposomes and 20 mM freshly prepared dithionite were mixed in the stopped-flow apparatus, and the fluorescence decrease was recorded at room temperature. A typical fluorescence kinetics of fluorescence reduction of the fluorescent PE analogue is shown. The protein content of the reconstituted proteoliposomes was 150 µg/ml.
• Figure 16: Protein dependence of transmembrane transport and extraction of M-C6-NBD-PE by the dithionite assay. Proteoliposomes containing one molpercent M-C6-NBD-PE and increasing amounts of TE were assayed by stopped-flow dithionite assay as described in 3.6 . Each data point represents the mean of five dithionite kinetics. The percentage of fluorescence decrease of three independent experiments is summarized. The dotted line indicates the linear range and the plateau of protein content-dependent fluorescence intensity amplitudes.
• Figure 17 : Influence of the chain length of analogues on the rapid transbilayer movement of fluorescent phospholipid analogues across the membrane of proteoliposomes derived from IIMV. The proteoliposomes were labeled with 0.5 mol% of the respective analogue during the reconstitution procedure (see 3.3 ). Equal volumes of the vesicle suspension and 20 mM dithionite were mixed in the stopped-flow accessory, and the fluorescence decrease was monitored at room temperature. The fluorescence traces represent the average of three separate kinetics. Each kinetics was normalized to the maximum fluorescence intensity.
• Figure 18 : Kinetics of the extraction of M-C6-NBD-PE from reconstituted IIMV by BSA prior to and after proteinase K treatment. Reconstituted IIMV-derived proteoliposomes were labeled with two molpercent M-C6-NBD-PE during the reconstitution (see chapter 3.3 ), and a stock was treated with proteinase K (1 mg/ml) for 0 min (black trace), 30 min (red trace) and 60 min (blue trace). Each aliquot was incubated with three mM (final) PMSF for five minutes to terminate the reaction prior to analysis by stopped-flow BSA back-exchange. Traces were corrected for scattering. For comparison, the extraction kinetics measured with protein-free liposomes (gray line) is shown.
• Figure 19 : Qualitative analysis of proteins derived from TE of IIMV. A representative polyacrylamide gel is shown. 2.9 mg protein from solubilized IIMV (TE) were applied to an anion exchange column. The resulting flow-through (QF) and eluate (QE) were analyzed by a 15% SDS-PAGE.
• Figure 20 : Comparison of flippase activity of reconstituted protein fractions after IEC. An aliquot TE (1.74mg protein) was applied to an anion exchange (Q-Sepharose) resin and the resulting flow-through (QF) and eluate (QE) were dialyzed. Fractions were taken for reconstitution to generate proteoliposomes with equal protein contents. Subsequently, the accessibility of M-C6-NBD-PE of proteoliposomes and liposomes was analyzed by the dithionite assay. 100 µl of each sample were diluted into 1.9 ml HPS in a cuvette, the fluorescence was recorded for 30 s. Then 10 mM (final) dithionite in 40 mM TRIS pH 8 was added and the resulting decay of fluorescence intensity was recorded for 600 s for each vesicle preparation. The protein concentration of the reconstituted proteoliposomes were 23.2µg/ml, 23.9µg/ml and 14.9µg/ml for pTE, pQF and pQE, respectively.
• Figure 21 : Comparison of flippase activities of flow-through and eluate from Q-Sepharose column and TE. An aliquot of TE was applied to an anion exchange (Q-Sepharose) resin, and the resulting flow-through (QF) and eluate (QE) were dialyzed. Proteoliposomes with equal protein contents and 0.5 mol% M-C6-NBD-PE were reconstituted. Additionally, respective amounts of the originally TE applied to the column, were taken to obtain proteoliposomes (pTE) with the same protein concentration compared to the pQF and pQE and proteoliposomes with two times higher concentration of protein (2xpTE) The protein concentrations of the proteoliposomes were 48.2 µg/ml, 50.9 µg/ml, 47.1 µg/ml and 93.2 µg/ml for pTE, pQF, pQE and 2xpTE, respectively. Subsequently, aliquots of proteoliposomes and control liposomes were analyzed by the dithionite assay (see legend of Figure 20 ).
• Figure 22:Flippase activities of Q-Sepharose column fractions, liposomes and TE after successive elution. An aliquot of TE was applied to an anion exchange (Q-Sepharose) resin and the resulting flow-through (QF) and eluates (QE4, QE) were dialyzed. Proteoliposomes with equal protein contents and 0.5 mol% M-C6-NBD-PE were generated. For comparison a respective amount of the originally TE applied to the column was reconstituted (pTE) to create proteoliposomes with the same protein concentration compared to the fractions. The protein concentration of the proteoliposomes were 22.8 µg/ml, 15.3 µg/ml, 14.4 µg/ml and 23.5 µg/ml for pTE, pQF, pQE4 and pQE, respectively. Subsequently, aliquots of proteoliposomes and control liposomes were analyzed by the dithionite assay. To this end, 100 µl of sample were diluted in 1.9 ml 10 mM HPS and the fluorescence decrease by reduction with 10 mM dithionite was monitored for 300 s using the stopped-flow accessory. All experiments were carried out at room temperature.