3 Membrane Interaction of Two Peptides Detected by Biosensor Measurement

a



b



c



Concentration Compound A

A



Compound A

Lipid A

Lipid B



1-Hexadecanethiol

Gold Layer

Quartz



Fig. 3 (a) Sensorgram of the phase signal derived from membrane binding of Compound A. Binding to the

membrane and corresponding deposition of mass on the sensor is accompanied by a shift in phase. Bound

molecules were rarely washed away by flow buffer, indicated by the continuously increasing signal. (Gray triangles: Starting injection of an individual concentration of Compound A. Black triangles: End of injection.)



Biosensors for Analyte-Membrane Interaction



153



The above assigned concentrations were reached by dilution of 1 mM DMSO stock with running buffer instead of

DMSO (see Note 7).

5. The injections (gray triangles in Figs. 3 and 4) start with the

lowest concentration and a binding event can be identified by

a phase shift establishing a higher baseline level, explained by a

permanently bound mass. It has to be mentioned that this is

not the case for most dynamic binding processes; where at the

end of injection (black triangles in Figs. 3 and 4) dissociation

of the attached analyte is the dominant process, leading to a

decrease in phase shift.

6. The steeper the phase shift increases at an individual concentration, the more total mass is immobilized on the sensor surface respectively the model membrane. The reversible peaks at

higher concentrations were mainly induced by viscosity changes

as a result of, e.g., solvent residues and should not be considered for evaluation.

7. Considering the amplitude signals, viscosity changes occurring at

the sensor surface can be monitored in real time. Binding of a

flexible ligand to a pure sensor surface leads to a decrease in

amplitude, while the rare event of increasing the amplitude would

represent a rigidification, e.g., a membrane condensing effect.

A pure POPC membrane provides an uncharged surface to an

analyte at pH 7.0. Since both Compounds A and B represent peptides with a net negative charge, we simulated the physiological

conditions by incorporating a negative charge into the membrane

by DOPG and supplementation of divalent cations (2.5 mM Ca2+).

This leads to an acceleration of the interaction with the model

membrane at pH 7.0, driven by ionic interaction.

Comparing the sensorgrams of the phase signals obtained for

Compound A (Fig. 3a) and Compound B (Fig. 4a), there is no

major difference detectable. Both compounds are able to bind the

membrane indicated by a step-wise increase in the phase shift, confirming a simultaneous mass deposition at the membrane.

Considering the single steps, only small amounts of bound compound appear to be washed away, indicating a tight binding.



Fig. 3 (continued) (b) Sensorgram of the amplitude signal influenced by Compound A. The binding of

Compound A to the membrane leads to an agglomerate that is more flexible than the pure membrane. These

viscoelastic changes are recorded with the amplitude signal and the more viscous state is represented by an

overall decrease. (Gray triangles: Starting injection of an individual concentration of Compound A. Black triangles: End of injection.) (c) Schematic illustration of the sensor surface during application of increasing

concentrations of Compound A. The amount of Compound A deposited to the membrane increases, which is

reflected in the increase in phase shift. Compound A only binds to the outer area of the model membrane and

therefore the additional layer on the membrane allows for more flexibility of the whole material on the sensor

surface. The more viscous properties are reflected in the decreasing amplitude



a



b



c



Concentration Compound B

B



Compound B

Lipid A

Lipid B



1-Hexadecanethiol

Gold Layer

Quartz



Fig. 4 (a) Sensorgram of the phase signal influenced by Compound B. In comparison to Compound A there is

virtually no difference in the binding behavior of Compound B, regarding only the deposited mass. (Gray triangles: Starting injection of an individual concentration of Compound B. Black triangles: End of injection.) (b)

Sensorgram of the amplitude signal influenced by Compound B. Regarding injections one to three, the

amplitude



Biosensors for Analyte-Membrane Interaction



155



Furthermore, with increasing concentrations, the steps become

smaller referring to an obvious saturation of the membrane interaction capacity (see also Note 8).

The amplitude signal allows deriving further information.

Compound A (Fig. 3b) influences the amplitude in the expected

manner as binding of the compound to the membrane surface

leads to a decrease in amplitude, which results from a higher overall flexibility of the whole construct on the sensor surface. This

molecular mode of action is depicted in Fig. 3c.

In case of Compound B the amplitude signal has a different

appearance (Fig. 4b). During the first three injections of Compound

B the amplitude increases. Increasing amplitudes can be observed

if a modification turns the whole bound mass on the sensor surface

more rigid. Starting with the fourth injection, a decrease in amplitude is evident which is similar to the shape resulting from

Compound A. Figure 4c provides a schematic model for this

behavior. The initially injected small amounts of Compound B are

able to incorporate into the membrane and thus change the viscoelastic membrane properties to a more rigid nature. With the buffer flow adding more and more Compound B, the process is

saturated and the molecules are thenceforward solely attached to

the membrane surface.

In conclusion, these two examples of peptide interaction with

model membranes illustrate that despite the obvious identical tendency for membrane binding, the SAW-biosensor approach is able

to discriminate different modes of membrane interaction that justifies this technology as an appropriate tool for an initial and rapid

characterization of compounds, like antibiotic drug candidates.



4



Notes

1. MOPS buffer was used instead of the common DPBS (Dulbecco’s

Phosphate-Buffered Saline) buffer based on optimization studies

showing a better reproducibility of the binding data.



Fig. 4 (continued) signal increases upon application of Compound B to the model membrane. Afterwards the

signal decreases and later shows a similar appearance as seen with Compound A. The initial increase of the

amplitude indicates a more rigid nature of the construct on the sensor surface in response to compound application and is explained in Fig. 4c, (Gray triangles: Starting injection of an individual concentration of Compound

B. Black triangles: End of injection.) (c) Schematic illustration of the sensor surface during application of increasing concentrations of Compound B. Compound B interacts besides an inherent binding ability and therefore

overall mass deposition (Compound A) in an additional manner with the membrane. Compound B is able to

integrate in the membrane and as a result changes their viscoelastic properties. At small concentrations

(Injections 1–3) the integration leads to a more rigid construct and the monitored increasing amplitude signal.

The ability of the membrane for integration of Compound B is limited and with increasing compound concentrations in the flow buffer, the mass attached to the surface predominates the viscoelastic properties, which leads

in sum to a decrease in amplitude



156



Sebastian G. Hoß and Gerd Bendas



2. The quality of the gold surface is directly related to the quality

of the established model membrane, as it is true for the physical

properties being reflected in the sensorgram. Therefore, special

care has to be taken to an intact and homogenous gold film.

Best results are obtained using new sensor quartzes, but the

regeneration procedure with piranha solution can be executed

1–2 times without influencing data in an inappropriate manner.

3. The lipid mixture can easily be prepared by aspirating the indicated volume from the stock solution with a glass microsyringe

(e.g., Hamilton®) and mixing with the syringe’s piston. Avoid

contact of the lipid containing chloroform solution with plastic

parts and execute aspiration and mixing procedures with glass

equipment.

4. If the quartz surface is obviously damaged (e.g., black pinholes

or scratches) one should discard it.

5. Liquid serving as flow buffer should always be properly

degassed because otherwise the chip surface (which is due to

the membrane that in general is rather lipophilic!) tends to

adsorb gas. The measurement is disturbed by these air-bubbles

because the true active surface is reduced and also leads to falsified phase signal. If only one or two of the five channels of the

sensor are disturbed by air, the remaining channels suffice to

interpret the phase shift in the sensorgram. No significant

influence on the buffer flow rate assumed as a precondition.

6. The concentration series being most suitable to be applied on

the sensor has to be determined empirically in preliminary tests

until binding occurs with appropriate signal intensities.

7. As the SAW technique is very sensitive to viscosity changes in

the measurement medium, one has to take special care when

performing dilution series. The ready diluted samples should

be comprised to most possible extent of flow buffer to avoid

artifacts in the sensorgram resulting from viscosity changes in

the running buffer.

8. Regarding the phase changes in the original sensorgrams, indicating the bound mass, they seem not to follow a linear principle according to the concentration series. As this is not an

uncommon finding, one should bear in mind that the concentration range in this kind of assay is very low and the detection

principle is rather sensitive even to small deviations in sample

preparation.



Acknowledgments

The authors would like to thank Hildegard Falkenstein-Paul for

excellent technical competences in performing the SAW measurements and providing the data presented here.



Biosensors for Analyte-Membrane Interaction



157



References

1. Nguyen HH, Park J, Kang S, Kim M (2015)

Surface plasmon resonance: a versatile technique for biosensor applications. Sensors 15:

10481–10510

2. Gronewold TMA (2007) Surface acoustic wave

sensors in the bioanalytical field: recent trends

and challenges. Anal Chim Acta 603:119–128

3. Gronewold TMA, Schlecht U, Quandt E (2007)

Analysis of proteolytic degradation of a crude

protein mixture using a surface acoustic wave

sensor. Biosens Bioelectron 22:2360–2365

4. Länge K, Rapp BE, Rapp M (2008) Surface

acoustic wave biosensors: a review. Anal

Bioanal Chem 391:1509–1519

5. Schlesinger M, Simonis D, Schmitz P, Fritzsche

J, Bendas G (2009) Binding between heparin

and the integrin VLA-4. Thromb Haemost

102:816–822

6. Baca JT, Severns V, Lovato D, Branch DW,

Larson RS (2015) Rapid detection of Ebola

virus with a reagent-free, point-of-care biosensor. Sensors 15:8605–8614

7. Slamnoiu S, Vlad C, Stumbaum M, Moise A,

Lindner K, Engel N et al (2014) Identification

and affinity-quantification of ß-amyloid and

α-synuclein polypeptides using on-line SAWbiosensor-mass spectrometry. J Am Soc Mass

Spectrom 8:1472–1481



8. Engelman DM (2005) Membranes are more

mosaic than fluid. Nature 438:578–580

9. Lee TH, Hall KN, Aguilar MI (2016)

Antimicrobial peptide structure and mechanism

of action: a focus on the role of membrane structure. Curr Top Med Chem 16:25–39

10. Schneider T, Sahl HG (2010) Lipid II and

other bactoprenol-bound cell wall precursors

as drug targets. Curr Opin Investig Drugs

11:157–164

11. Reder-Christ K, Schmitz P, Bota M, Gerber U,

Falkenstein-Paul H, Fuss C et al (2013) A dry

membrane protection technique to allow surface acoustic wave biosensor measurements of

biological model membrane approaches.

Sensors 13:12392–12405

12. Gerber U, Hoß SG, Shteingauz A, Jüngel E,

Jakubzig B, Ilan N et al (2015) Latent heparanase facilitates VLA-4-mediated melanoma cell

binding and emerges as a relevant target of heparin in the interference with metastatic progression. Semin Thromb Hemost 41:244–254

13. Girard-Egrot AP, Blum LJ (2006) LangmuirBlodgett technique for synthesis of biomimetic

lipid membranes. In: Martin D (ed)

Nanotechnology of biomimetic membranes,

1st edn. Springer Science+Business Media

LLC, New York, pp 23–74



Chapter 10

Measurement of Cell Membrane Fluidity by Laurdan GP:

Fluorescence Spectroscopy and Microscopy

Kathi Scheinpflug, Oxana Krylova, and Henrik Strahl

Abstract

Membrane fluidity is a critical parameter of cellular membranes which cells continuously strive to maintain

within a viable range. An interference with the correct membrane fluidity state can strongly inhibit cell

function. Triggered changes in membrane fluidity have been postulated to contribute to the mechanism of

action of membrane targeting antimicrobials, but the corresponding analyses have been hampered by the

absence of readily available analytical tools. Here, we provide detailed protocols that allow straightforward

measurement of antibiotic compound-triggered changes in membrane fluidity both in vivo and in vitro.

Key words Membrane fluidity, Membrane viscosity, Lipid domains, Lipid packing, Fatty acid disorder, Laurdan, Lipid adaptation, Membrane targeting antimicrobials



1



Introduction

Numerous antimicrobial compounds target bacterial cytoplasmic

membranes, and disrupt the normal function of this essential cellular structure. Membrane targeting compounds frequently unfold

their antimicrobial properties by interfering with the diffusion barrier function of the cytoplasmic membrane [1]. As a consequence,

a comprehensive set of tools has been developed to analyze cellular

parameters related to membrane permeability such as ion leakage

and membrane potential. However, not all membrane targeting

antimicrobials trigger permeabilization of cellular membranes. The

mechanisms of action of this category of antimicrobials are considerably less understood [1].

In addition to its permeability barrier function, a correct fluidity

state of the membrane is equally important in order to support the

multitude of membrane associated cellular processes [2]. Interference

with the fluidity state of the membrane by an antimicrobial compound, either by causing changes in the overall membrane fluidity,

or by triggering formation of abnormal lipid domains, has high

potential to inhibit cell growth [3]. The analysis of these important



Peter Sass (ed.), Antibiotics: Methods and Protocols, Methods in Molecular Biology, vol. 1520,

DOI 10.1007/978-1-4939-6634-9_10, © Springer Science+Business Media New York 2017



159



160



Kathi Scheinpflug et al.

10°C



fluorescence (a. u.)



600



25°C

37°C

400



200



0

400



500



600



λ (nm)



Fig. 1 Fluorescence emission spectrum of laurdan incorporated in small unilamellar vesicles (SUVs) formed of E. coli polar lipid extract. Note the spectral shift toward

higher wavelength in higher temperatures (indicating increased membrane fluidity). The wavelength ranges used for the ratiometric measurement of laurdan fluorescence (laurdan GP) are highlighted in light and dark gray, respectively



cellular parameters has been greatly hampered by the relative absence

of suitable, easy to adapt analytical tools. Here, we provide detailed

protocols for the analysis of membrane fluidity of bacterial cell membranes both on a global scale, and on a single cell level with spatial

resolution. The provided measurements can be carried out with

widely available standard laboratory equipment such as fluorescence

microplate reader and wide field epifluorescence microscope.

The protocols provided in this chapter make use of a fluorescent, fluidity-sensitive, and noninhibitory membrane dye laurdan

[4, 5]. The fluorescence emission spectrum of laurdan is sensitive

to the presence of H2O close to its chromophore. The ability of

H2O to penetrate the hydrophobic membrane interior is dominated by lipid head group packing density and fatty acid disorder

of lipid bilayers. As a consequence, the fluorescence emission spectrum of membrane embedded laurdan is sensitive to membrane

fluidity and disorder in its surrounding (see Fig. 1) [4–8].

The provided protocols are optimized for Gram-positive model

organism Bacillus subtilis but also offer a good starting point for

measurements in other Gram-positive microorganism such as

Staphylococcus aureus. We provide example measurements how

these methods can be applied to gain insight into mechanism of

action of membrane targeting antimicrobials.



2



Materials



2.1 Laurdan

Fluorescence

Spectroscopy In Vivo



1. 1 mM Laurdan (6-Dodecanoyl-2-Dimethylaminonaphthalene;

either from Molecular Probes or Sigma-Aldrich) stock solution

in 100 % DMF (Dimethylformamide), store in −20 °C, keep

always in dark.



Laurdan GP Measurement



161



2. 5 M benzyl alcohol stock by dilution with DMSO (Dimethyl

sulfoxide), store in −20 °C, cover stored aliquots with Argon

or N2 to prevent oxidation.

3. Fluorescence microplate reader. Both monochromator-based

plate readers, and a filter-based readers equipped with 350 nm

excitation filter and appropriate emission filters (ranges spanning 420–460 nm and 490–520 nm) are suitable.

4. Black, flat bottom 96-well plates; if reusable plates are used

ensure proper cleaning after use.

2.2 Laurdan

Fluorescence

Spectroscopy In Vitro



1. Phospholipids of choice. Either natural lipid extracts, or mixtures of synthetic or purified lipids can be used. We recommend

either Escherichia coli Polar Lipid Extract, or a mixture mimicking bacterial cytoplasmic membrane composed of a zwitterionic

1-palmitoyl-2-oleoyl- sn -glycero-3- phosphoethanolamine

(POPE) combined either with anionic cardiolipin or with

1-palmitoyl-2-oleoyl-sn-glycero-3[phosphor-rac-(1-glycerol)]

(POPG). All lipids mentioned above can be purchased from

Avanti Polar Lipids.

2. Laurdan (6-Dodecanoyl-2-Dimethylaminonaphthalene, Molecular

Probes or Sigma-Aldrich). Prepare a 0.2 mg/ml laurdan solution in

chloroform, store in −20 °C, keep dark.

3. 10 mM sodium phosphate (NaH2PO4/Na2HPO4) buffer containing 154 mM NaCl and 0.1 mM EDTA, pH 7.4. Or other

buffer of choice.

4. Chloroform and methanol of highest available purity.

5. Nitrogen or argon gas.

6. 1.5 ml and 2 ml reaction tubes and pipette tips siliconized if

necessary (see Note 1).

7. Round-bottomed glass vials (~5 ml) with tightly sealed caps.

Flat-bottomed glass vials (~2 ml) with caps.

8. Graduated glass pipettes (2 ml); Hamilton gastight syringe

(100–200 μl).

9. High-vacuum pump (10−2 to 10−4 mbar).

10. Mini-extruder and polycarbonate membranes with defined

pore size (see Note 2). Can be purchased from Avestin Inc. or

Avanti Polar Lipids.

11. Dry ice, ultrasonic bath with thermoregulation.

12. Fluorescence spectrometer (monochromator-based).

13. Disposable macro UV/VIS cuvettes (3 ml, 1 × 1 cm).

14. Magnetic stir bar (<10 mm in length).



162



Kathi Scheinpflug et al.



2.3 Laurdan

Fluorescence

Microscopy



1. 10 mM Laurdan (6-Dodecanoyl-2-Dimethylaminonaphthalene;

either from Molecular Probes or Sigma-Aldrich) stock solution

in 100 % DMF (Dimethylformamide), store in −20 °C, keep

always in dark.

2. PBS (8.0 g/L NaCl, 0.2 g/L KCl, 1.15 g/L Na2HPO4,

0.2 g/L KH2PO4, pH 7.3) supplemented with 0.1 %

d-glucose.

3. Agarose (electrophoresis grade).

4. Fluorescence microscope equipped with:

(a) A high quality 100× objective with good chromatic correction such as Nikon Plan Apo series, Zeiss Plan Apochromat

series, or equivalent.

(b) Appropriate filter sets (excitation at approx. 350 nm, emission at 430–460 and 500–530 nm) (see Note 3).

(c) Wide field illumination with strong light output at 350 nm.

We prefer Hg-vapor or metal halide light source for this

application.

(d) Temperature control.

(e) High sensitivity CCD, EM-CCD, or sCMOS camera with

maximally 8 × 8 μm pixel size.

5. High quality microscope slides, coverslips, and immersion oil.

6. 0.1 μm diameter TetraSpeck™ fluorescent microspheres

(Thermo Fisher Scientific).



3



Methods



3.1 Laurdan

Fluorescence

Spectroscopy In Vivo



1. Grow cells to an optical density at 600 nm (OD600) of approx.

0.5 in suitable growth medium supplemented with 0.1 % glucose at the desired temperature (see Notes 4–6).



3.1.1 Sample

Preparation and Data

Acquisition



2. Transfer the cell suspension to a 2 ml reaction tube and add

laurdan to a final concentration of 10 μM (from a 1 mM laurdan stock solution, see Note 7).

3. Incubate cells with laurdan for 5 min at the desired growth

temperature in a thermomixer. Cover tubes with aluminium

foil to avoid light exposure.

4. Wash cells 4× in 2 ml pre-warmed PBS/glucose (centrifuge for

1 min at 16,000 × g in a table top centrifuge, carefully remove

the supernatant by pipetting, resuspend in fresh PBS/glucose,

repeat 4 times). After the last wash, resuspend to obtain an

OD600 of approx. 0.5 (see Notes 8 and 9)

5. Remove 500 μl of the cell suspension, transfer to a new reaction tube, and centrifuge as described above. Carefully harvest

~450 μl of the supernatant, which serves as laurdan background



Laurdan GP Measurement



163



fluorescence in subsequent measurements (background of buffer + dye not associated with cells).

6. Immediately proceed with fluorimetric measurement by transferring the stained cell suspensions, and the background sample to a pre-warmed black, flat bottom black 96-well microtiter

plate (150 μl/well).

7. Depending on the antimicrobial compound of interest, and

the specific research question, three measurement options are

possible:

(a) Preincubation of the cell culture with the antibiotic of

choice, followed by staining and measurement. This measurement mode is suitable for slow acting but tightly

bound antimicrobials, or for an analysis of potential adaptation to subinhibitory concentrations. In this case, we

recommend a brief (2 min) shaking interval in the microplate reader before the fluorescence measurement.

(b) Incubation of stained cell suspension with the antibiotic of

choice for a given incubation time. In this case, incubate

the stained cell suspension with the compound directly in

the microtiter plate for a required time under shaking, followed by fluorescence measurement. In well-energized

untreated cells (PBS/glucose + shaking) laurdan GP values

were found to be stable for up to 45 min.

(c) For a kinetic measurement, laurdan fluorescence can be

measured before, and as a time series after addition of the

antibiotic of interest. In order to ensure sufficient energization of the cells, we recommend either continuous

shaking or a relatively low number of parallel samples.

Measurement intervals of 0.5–1 min are a good starting

point (see Note 10).

8. As a positive control, incubate cells with 50 mM membrane

fluidizer benzyl alcohol (see Note 11).

9. Measure laurdan fluorescent intensities upon excitation at

350 nm at two emission wavelengths. In a monochromatorbased fluorimeter, the optimal wavelengths (435 and 500 nm)

should be used. In a filter-based fluorimeter, filters with wavelengths centered at 430–460 nm, and 490–520 nm are

acceptable.

3.1.2 Data Analysis



1. Subtract values obtained from the background sample (fluorescence of unbound dye) from the cell suspension values for

each wavelength. The same background values are subtracted

from both treated and untreated samples (this assumes that the

compound of choice does not have fluorescent properties

itself).