journal of biophotonics - bph.ruhr-uni-bochum.de · j. biophotonics 5, no. 7, 582–591 (2012) /...

www. biophotonics-journal.orgJournal of

BIOPHOTONICS

REPRINT

FULL ARTICLE

Nanoscale distinction of membrane patches –a TERS study of Halobacterium salinarum

Tanja Deckert-Gaudig1, Rene Bohme2, Erik Freier3, Aleksandar Sebesta1,Tobias Merkendorf 3, Jurgen Popp1; 2, Klaus Gerwert3, and Volker Deckert*; 1; 2

1 Institute of Photonic Technology (IPHT), Albert-Einstein-Straße 9, 07745 Jena, Germany2 Institute of Physical Chemistry, Friedrich-Schiller-University Jena, Helmholtzweg 4, 07743 Jena, Germany3 Lehrstuhl fur Biophysik, Ruhr-Universitat Bochum, Universitatsstraße 150, 44780 Bochum, Germany

Received 2 December 2011, revised 20 January 2012, accepted 20 January 2012Published online 28 February 2012

Key words: Raman, TERS, Halobacterium salinarum, membrane patch

Æ Supporting information for this article is available free of charge under http://dx.doi.org/10.1002/jbio.201100131

1. Introduction

Cell membranes are part of every living organismand thus are the structural basis of life as we knowit. Mainly consisting of amphiphilic lipids and a vari-ety of proteins, the membrane acts as a selective bar-rier between the cytoplasm and the surroundingmedium. Many important mechanisms take place at

this site, ranging from transport of nutrients andwaste to cell signalling events [1–3]. For a long timethe so-called fluidic mosaic model, proposed by Sin-ger and Nicolson in 1972, as an organisational basisof the membrane remained unchallenged. In thismodel comparatively few proteins are immersed inlipids and every molecule diffuses randomly in twodimensions [4]. In reality, however, a manifold of

# 2012 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of

BIOPHOTONICS

The structural organization of cellular membranes hasan essential influence on their functionality. The mem-brane surfaces currently are considered to consist of var-ious distinct patches, which play an important role inmany processes, however, not all parameters such as sizeand distribution are fully determined. In this study, pur-ple membrane (PM) patches isolated from Halobacte-rium salinarum were investigated in a first step usingTERS (tip-enhanced Raman spectroscopy). The charac-teristic Raman modes of the resonantly enhanced com-ponent of the purple membrane lattice, the retinal moi-ety of bacteriorhodopsin, were found to be suitable asPM markers. In a subsequent experiment a single Halo-bacterium salinarum was investigated with TERS. Bymeans of the PM marker bands it was feasible to identi-fy and localize PM patches on the bacterial surface. Thesize of these areas was determined to be a few hundrednanometers.

1 2

1650 1350 1050 750 1650 1350 1050 750

retinal containing retinal free

Wavenumber / cm Wavenumber / cm-1 -1

Schematic TERS measurement on a Halobacterium sali-narum cell wall. The TERS probe is close to the majordomains of the inner membrane (purple and red mem-brane). TERS spectra are characterized accordingly.

* Corresponding author: e-mail: [email protected]

J. Biophotonics 5, No. 7, 582–591 (2012) / DOI 10.1002/jbio.201100131

proteins fill the membrane surface and the distribu-tion is nowhere near random. In fact, it is dominatedby structurally and functionally distinct patches con-sisting of domains with a specific composition. Ex-amples are so called lipid rafts, which are known tobe segregated regions in cell surfaces or even in purelipid vesicles [5]. Consequently, the new model pre-dicts membranes with variable patchiness, thicknessand higher protein occurrences than considered be-fore [6]. Since the exact composition, the typical sizeand variability of such segregated regions are mostlyunknown, the need of analytical tools to unravelthese parameters is obvious.

An attractive method to elucidate nanoscale sur-face domains is tip-enhanced Raman spectroscopy(TERS) [7–10]. By combining surface-enhanced Ra-man spectroscopy (SERS) [11–13] with atomic forcemicroscopy (AFM) or scanning tunnelling micro-scopy (STM), TERS provides a sensitive, moleculespecific and non-destructive approach with a spatialresolution in the range of at least 10 nm [14–16].The resolution is determined by a single metal nano-particle attached to the apex of a commercial AFMtip. When a laser beam is focused on such a particle,localized surface plasmon polaritons are excited,which can enhance the Raman signals by several or-ders of magnitude. Consequently, spectroscopic in-formation from a sample volume well below the op-tical diffraction limit is gained. Since spectral andtopographic information are collected simulta-neously, a structurally specific characterization of thesample surface is feasible. Compared to other techni-ques no additional labels are required, which favoursTERS over fluorescence based imaging techniqueswhere an interference due to the label is difficult torule out [17–19]. TERS has been already applied toseveral biological samples [20–24]. Of particular in-terest for the present studies are TERS investiga-tions on human cell surfaces [25], cell organelles [26]and membrane models [27, 28], the application tospecific membranes is straightforward.

In this contribution the target is the characteriza-tion and localization of specific membrane domainson the bacterial surface of the archean Halobacteriumsalinarum [29, 30] on the nanometer scale with TERS.Isolated purple membrane (PM) patches were investi-gated for comparison and as a reference prior to theactual experiments on the more complex bacterial sur-face. The inner membrane of the organism consists oftwo parts: the red membrane, the basis of the plasmamembrane, and the PM, which forms patches in theplasma membrane. However, a distinctive knowledgeregarding size and distribution does not exist yet.

To differentiate between specific membrane con-stituents and their environment the resonantly en-hanced retinal moiety of bacteriorhodopsin (bR)within the purple membrane is utilized. With a frac-tion of 75% the purple membrane consists of bR

surrounded by a phospholipid matrix [31, 32]. Thebacteriorhodopsin acts as a light-driven proton pumpand the resulting electrochemical gradient is appliedfor energy storage in the cell [33–37].

2. Experimental

2.1 Materials

Wild-type bacteriorhodopsin isolated from Halobac-terium salinarum strain S9 as purple membrane (pro-tein to lipid ratio �3 : 1 [32]) was purchased fromSigma Aldrich. Salts for the buffer solution (150 mMKCl, 10 mM Tris, 2 mM CaCl2 in water, pH 7,4)were obtained either from Sigma Aldrich or Merck.

2.2 Sample preparation

2.2.1 Bacteria and isolated purple membrane

Halobacterium salinarum S9 cells were grown ac-cording to Oesterheldt and Stoeckenius (1971) [38].After inoculation in 3 mL medium and growth for48 h at 42 �C and 110 rpm in a shaker the culturewas turbid. With this pre-culture 50 mL mediumwere inoculated and grown in the dark until a tur-bidity of OD600 0.9 was reached. This culture wastransferred to another shaker and exposed to lightfor 2 days to express sufficient amounts of bacterior-hodopsin. Small amounts (10–20 mL) of the halobac-terial cell culture were dropped on glass plates anddried. The evaporation of water must not happentoo fast, since it would stress the cells, yielding a saltcovered sample. Slow evaporation leads to areas ofvarying salt content. Low/no-salt areas alternate withhigh-salt areas. Large salt crystals were removed me-chanically. Suitable bacteria (5–10 mm long, 1–2 mmwide) for the measurements were found betweensalt dendrites in locations of low salt content.

For the direct Raman and TERS investigation ofisolated patches, PM fragments dissolved in aqueousbuffer solution (10 mL, 1 mM) were deposited on aglass cover slide and dried under argon.

2.3 Instrumental setups

2.3.1 TERS setup

General features of the setup used for TERS meas-urements on Halobacterium salinarum and on iso-lated purple membrane patches have been described

J. Biophotonics 5, No. 7 (2012) 583

FULLFULLARTICLEARTICLE

# 2012 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimwww.biophotonics-journal.org

in detail previously [9, 39]. Particular care was takento exclude fake signals by tip contamination duringthe TERS measurement by repeated reference meas-urements [40]. An inverted micro-Raman system inbackscattering mode (HR LabRam, Jobin YvonHoriba (France), 600 lines/mm grating, N2 – cooledCCD device (1024 � 256 pixels)) is combined withan AFM setup (NanoWizard AFM, JPK InstrumentAG, Germany). Spectra were recorded using an ex-citation wavelength of 568.2 nm (Krþ ion laser, Co-herent, USA) and a sampling time of 10 s (Halobac-teria) and 20 s (PM patch), respectively (P � 1 mW,measured at the sample). For all measurements thelaser was focused using an oil immersion microscopeobjective (60�, NA 1.45, Olympus) through sub-strate and sample onto the silver coated AFM tip.Therefore, the spot was slightly defocused to matchthe antenna condition of the TERS-probe. The sam-ple is scanned independently using the intermittentcontact mode (modulation amplitude �6 nm), tomaintain focusing on the TERS tip all the time. TheRaman scattered light is collected in transmissionmode through the same microscope objective. Thenon-contact AFM tips (NSG10, resonant frequency190–325 Hz, NT-MDT) were coated with 20 nm sil-ver (99,99% pure, Balzers Materials). These TERStips were used directly after evaporation within twodays. The diameter of the silver coated AFM tips isestimated to be approximately 20 nm, as typicalSEM pictures of similar TERS probes show [39].

2.3.2 FTIR and Raman measurements

600 mg purified purple membrane sheets were sus-pended in 1 M KCl and 100 mM Tris at pH 7. Theprotein pellet – after centrifugation at 200,000 g for2 h – was squeezed between CaF2 windows, trans-ferred into an airtight cuvette and stabilized at 20 �Cduring the whole measurement. The data was ac-quired using a modified Bruker IFS66V spectro-meter recording with 4 cm�1 spectral resolution anda time resolution of up to 30 ns (41). The photocyclewas triggered by an excimer laser (Lambda Physics305i) driven dye laser containing Coumarin153.

Resonance Raman measurements of isolated PMpatches were done using the setup described forTERS measurements, at 530.9 nm (P � 1.6 mW,sampling time ¼ 30s).

3. Results and discussion

In a first experiment, isolated purple membranepatches were investigated. After deposition of thePM patches on glass cover slides from an aqueous

buffer solution, TERS spectra were recorded onthese structures. Figure 1 shows four selected spectra[1–4] recorded on a PM patch with a tentative as-signment given in Table 1 [42–45]. The TERS spec-tra are reproducible with respect to their spectral po-sitions, but compared to Resonance-Raman or FTIR– difference-spectra the band widths and intensityratios can differ substantially. Such varying intensi-ties and intensity ratios can be attributed to specificnanoscale effects, which are related to the smallnumber of molecules in the tip-sample interactionregion resulting in the detection of specific molecu-lar orientations of the micro ensembles [20]. This be-haviour is reflected in the observed variation ofband intensity ratios. Each peak can be tentativelyassigned to be characteristic for bacteriorhodopsin(bR) or lipid (Lipid), respectively. Therefore the

1700 1600 1500 1400 1300 1200 1100 1000

1025

1136

124512

97

1363

1346

14581533

1599

1617

1022

1027

1136

125212

9913

43

1367

1387

14601540

1604

1638

1100

1299

1345

1387

1467

1540

1578

1638

1095

123012

9713

43

1382

145115

40

1635

Ram

an In

tens

ity

Wavenumber / cm -1

1685

1243

bR

Lipi

d

bR, L

ipidbR

bR, L

ipid

bR, L

ipidbR

bR, L

ipidbR bR bR

bR, L

ipid

1

2

3

4

Figure 1 (online color at: www.biophotonics-journal.org)Selected TERS spectra recorded on an isolated purplemembrane patch (lexc ¼ 568.2 nm, tacq ¼ 20 s, P � 1 mW).Significant bacteriorhodopsin marker bands are assignedas “bR”. Further spectral information from the lipid ma-trix is assigned as “Lipid”.

T. Deckert-Gaudig et al.: Nanoscale distinction of membrane patches584

Journal of

BIOPHOTONICS

# 2012 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org

TERS bands are listed in Table 1. Since the excita-tion wavelength (568.2 nm) is almost at the maxi-mum of the visible absorption profile of the protein,resonantly enhanced Raman modes of bR dominate.The main contribution is related to the retinal moi-ety, due to the resonant enhancement [42, 43]. Inparticular the C¼C and C¼NHþ stretching modesaround 1540 cm�1 and 1635 cm�1 serve as unique Ra-man markers for the protonated Schiff base Retinal[46]. The N––H and C––H deformation modes of theretinal around 1340–1380 cm�1 cannot directly serveas unique bR Raman marker bands, although theirhigh intensity indicates a resonance enhancementand points towards the presence of bR. The lipidmatrix shows typical bands at approximately1095 cm�1 and 1230 cm�1 in spectrum 1 and 2, whichcan be tentatively assigned to the phosphate residue.However, these modes are not always detected. Thiscorresponds to the estimated protein – lipid ratio(3 : 1) [32] and demonstrates the crucial influence ofan exact tip positioning with respect to the lateral

and vertical resolution capabilities. For example, re-cent TERS studies on isolated mitochondria re-vealed a distinct spectral contribution of non-reso-nant components compared to the resonantlyenhanced cytochrome c [26]. This finding could beexplained by the different sample-TERS probe dis-tances in z-direction. The same effects must be con-sidered for the present experiments as well. Here,the probe is able to approach the bacteriorhodopsinvery closely and the spectral characteristics of purplemembrane patches are dominated by the retinalmoiety of bR. Nevertheless still surprisingly manyfeatures can be found that are related to the lipidenvironment. These features are not observed instandard resonance Raman scattering and again sup-port the fact that molecules closer to the tip (herelipid and protein) experience a higher electromag-netic enhancement compared to molecules or frag-ments (Schiff base) further away. Hence, distance ef-fects can match electronic resonance effects and leadto the observed spectral features. To support ourfindings on the compositional assignment of the pur-ple membrane patches, conventional Raman andFourier Transform Infrared (FTIR) spectroscopymeasurements are performed. Since both techniqueshave been widely used to study bR and purple mem-brane [37, 47, 48] we present two selected spectra inFigure 2 for comparison with our TERS results ofisolated PM patches. Figure 2 shows a resonance Ra-man spectrum (black line) of purple membrane anda time resolved inverted M-bR infrared differencespectrum (red line) of purple membrane. The Ra-man and the FTIR difference spectra show niceagreement regarding the chromophore bands at1641 cm�1, 1530 cm�1, 1202 cm�1, 1166 cm�1 and1009 cm�1. In the FTIR difference spectra in addi-tion protein bands as for example protonation ofAsp85 at 1762 cm�1 is observed. However, the inten-sity patterns of the Resonance and FTIR spectra de-viate largely from the TERS spectra in Figure 1 ofthe PM. The resonance Raman spectrum is domi-nated by the bands of the chromophore, exhibitingthe same prominent Raman markers at approx.1530 cm�1 (C¼C) and 1641 cm�1 (C¼NH) as theTERS spectra in Figure 1, even though in a differentintensity ratio and slightly shifted in frequency,which can be understood by specific surface selec-tion rules [49]. Interestingly, TERS lacks the promi-nent band at 1200 cm�1. FTIR is equally sensitivefor all components, and the additional bands in theFTIR spectra can be assigned to molecular changesduring the bR to M transition (with ground state(bR) contributions pointing upwards), like the bandsat approx. 1166 cm�1, 1253 cm�1, 1348 cm�1

1556 cm�1, 1660 cm�1 and 1690 cm�1. Some of thesesignals can also be identified in the TERS spectra inFigure 1, hence, supporting the distance dependencediscussed above. For the subsequent TERS investi-

Table 1 Tentative assignment of TERS spectra (cm�1)shown in Figure 1 recorded on isolated purple membranepatches [42–45].

Spec 1 Spec 2 Spec 3 Spec 4 assignment

1022 w 1027 w 1025 m r CH3 (bR)

1095 m 1100 w ns PO2 (L)1136 w 1136 w n C––C (bR)

1230 m nas PO2 (L),amide III (bR)

1243 w 1252 m 1245 w n C––C/lysinerock (bR)

1297 m 1299 m 1299 m 1297 m d CH/amide III(bR),tip CH2 (L)

1343 vs 1345 s 1343 m 1346 m r NH, Trp (bR)1367 s 136 s d CH3 (bR, L)

1382 m 1387 m 1387 (S) 1387 (S) d CH3 (bR, L)

1451 m 1467 m 1460 m 1458 m d CH3 (bR),d CH2,3 (L)

1540 m 1540 m 1540 m 1533 m n C¼C (bR, L)

1578 m 1604 w 1599 w n C=C (bR, L),amide II (bR)

1635 vs 1638 s 1638 m 1617 w n C¼Nþ

n C¼N (bR)

1685 m amide I (bR)

w: weak, m: middle, s: strong, vs: very strong, (S):shoulder, L: lipid contribution, bR: bacteriorhodopsin con-tribution (retinal with protein envelope), Trp: Tryptophan,n: stretch, d: deformation, r: rocking, t: torsion, s: sym-metric, as: asymmetric, ip: in plane




gation of Halobacterium salinarum, the more com-plex composition of the bacterium compared to thebR patches has to be considered [32]. Here, purplemembrane sections are embedded in a so-called redmembrane, where they are presumably organized inpatch-like domains. The red membrane contains avariety of proteins and b-carotene. This ensembleforms the inner membrane which is covered by aprotein envelope, the so-called S-layer (50). This S-layer is not present on isolated PM-patches.

Prior to the actual TERS experiments single iso-lated Halobacterium salinarum cells were localizedusing the AFM topography. Figure 3(a) shows a typi-

cal AFM topography image of two halobacterialstructures recorded with a silver coated TERSprobe. In a next step TERS spectra were recordedusing 568.2 nm as excitation wavelength. This wave-length is also within the electronic absorption of b-carotene, the staining component of the red mem-brane. The TERS spectra were recorded on thesmall archaea structure on a 300 � 375 nm grid (4 �5 points), which is schematically displayed as an in-set in Figure 3(a). The distinct positions are sepa-rated by 75 nm. The chosen step size is a compro-mise between the expected size of the membranedomains, measurement time and the actual spatialresolution of TERS. Because the enhancement re-gion is smaller than 75 nm this results in a slight butdeliberate undersampling of the investigated area.The 20 spectra on the grid shown in Figure 3(a),

1800 1700 1600 1500 1400 1300 1200 1100 1000Wavenumber / cm -1

Ram

an In

tens

ity /

Diff

eren

ce A

bsor

banc

e

1641

1530

1202

1660

1556

1253

1166

1690

1009

1762

134814

4414

66

Figure 2 (online color at: www.biophotonics-journal.org)Comparison of the resonance Raman spectrum(lexc ¼ 530.9 nm, black line) and time-resolved FTIR dif-ference spectrum (red line, inverted M-intermediate spec-trum). While FTIR is basically equally sensitive to lipid,protein and chromophore absorption, resonance Ramanalmost exclusively detects the chromophore related signals.The comparison reveals that some signals e.g. around1253 cm�1, 1348 cm�1 and 1690 cm�1 are not detected byresonance Raman spectroscopy. As these bands are de-tected by TERS (see Figure 1) this supports a spatial com-petition between the resonance Raman and the pure elec-tromagnetic enhancement.

1 µm

1

23

4

(a)

(b)

1650 1350 1050 750

606

679

797

849 74

6

907

1051

1176

1247

1289

139715

4515

99

670

620

104111

1311

5311

991254

1294

1355

1366

1404

1452

1495

1523

1542

1583

1614

1638

Ram

an In

tens

ity

Wavenumber / cm-1

1688

1

2

PM patch

* **

*

* *

1622*

1650 1350 1050 750

799

620

833

1045

1075

1132

1314

1332

1350

1278

1394

1475

1513

161116

72

623

779

83793

999

010

4010

8111

32

1274

1330

1380

1484

1513

1566

1611

Wavenumber / cm-1

1664

3

4

PM free

801

945

1201

1409

1153

1153

1346

215 nm

0 nm

Figure 3 (online color at: www.biophotonics-journal.org)(a) AFM image of two isolated Halobacterium salinarumorganisms (256 � 256 pixels). Single points (4� 5) sepa-rated by 75 nm are measured on a 300 � 375 nm grid. Po-sitions characteristic for purple or red membrane are high-lighted in the enlarged colour map in purple or red,respectively. (b) Selected TERS spectra (lexc ¼ 568.2 nm,taqc ¼ 10 s, P � 1 mW) were recorded on the positions in-dicated in the colour map in (a). Intensive bands of thepurple membrane in spectra 1 and 2 are labelled with anasterisk. In spectra 3 and 4 characteristic red membranebands are labelled with a diamond.


Journal of

BIOPHOTONICS


were recorded by subsequent positioning of theTERS probe on the preselected grid while oscillat-ing. In addition to the characteristic PM signals,bands of proteins and lipids were detected, reflect-ing the increased complexity of the real bacterialsurface. A compositional assignment of each singlespectrum was accomplished depending on the pre-sence of the distinct PM marker bands and is sche-matically shown in the coloured grid of Figure 3(a).Each pixel represents the assignment of a singlespectrum and is either coloured purple (PM) orred (red membrane – no PM features), respec-tively. The colour map shows a contiguous spectraldistribution, which gives a hint to the organisationof the halobacterial surface. Figure 3(b) shows fourselected TERS spectra characterizing PM patches[1, 2] and domains without PM [3, 4]. The latter isused to classify the red membrane. A detailed as-signment of the spectra is provided in Table 2 [42–45].

The complete spectral data of the entire grid canbe found in the supplementary section. Here, the in-dividual positions of the TERS-probe are not shownin the AFM-image for a better visibility of the halo-bacterial surface. In spectra 1 and 2 the most promi-nent signals refer to PM marker bands. Consideringthe aforementioned position sensitive aspects ofTERS and focussing mainly on the band positions,the PM spectra on the cells in (Figure 3(b) left side)agree with the spectra of isolated PM (Figure 1).The PM marker bands are labelled with an asteriskand correspond to the very sensitive C=C and C=N/C=NH+ stretching modes of the retinal moiety at ap-proximately 1523 cm�1, 1542 cm�1, 1622 cm�1 and1638 cm�1. As already mentioned, the position ofthe C¼N stretching mode depends on its protona-tion state (1622 cm�1 (non protonated) or 1638 cm�1

(protonated)). Furthermore, very intensive peaks at1355 cm�1 and 1366 cm�1 can be regarded as addi-tional indicators of the retinal. An alternative assign-ment of the 1348 cm�1 band to the Trp adjacent tothe retinal group seems unlikely. While a similarband has been observed in the FT-IR differencespectra (Figure 2) and assigned to Trp (51), a similarband intensity for a non-resonant and a resonantlyenhanced molecule at a comparable TERS probe-molecule distance is not plausible. Further contribu-tions of lipids, non-resonant proteins and b-caroteneare weak, still, Raman modes in the amide I(1250 cm�1) and amide III (1688 cm�1) regions againreveal the sensitivity of TERS for non-resonant com-ponents, as highlighted previously with the combinedTERS/resonance Raman/FTIR investigation on theisolated PM patches. Varying signal intensities of thematrix (lipids/proteins) are caused by the previouslymentioned high resolution properties of TERS thatcan result in selective probing of even sub domainsof matrix proteins [25].

Table 2 Tentative assignment of TERS spectra (cm�1) shownin Figure 3 recorded on Halobacterium salinarum [42–45].

Spec 1 Spec 2 Spec 3 Spec 4 assignment

620 w 606 w 623 m 620 w Phe, Trp

670 w 679 w d CH

746 w 779 w Trp, ns O––P––O (L)

797 w 801 w 799 w ns O––P––O (L)

849 w 837 s 833 w nas O––P––O (L)

907 w 939 m 945 w n C––C (P),d C––H (L)

990 m n C––C (P)

1041 w 1051 w 1040 s 1045 w n C––O––P (L),d CH3 (bR)

1081 m 1075 w ns PO�2 (L)1113 w n C––C (L)

1132 s 1132 s n C––C (L), (P)

1153 m 1153 w 1153 w n C––C (bR)

1199 m 1176 w 1201 n C––C (bR),Tyr

1254 (S) 1247 s n (C––C)/lysinerock (bR),amide III (P)

1274 m 1278 m amide III (P)

1294 s 1289 s d (CH)/amide III(bR),tip ––CH2 (L)

1314 w wag CH2 (L)

1330 s 1332 w d CH2 (L)

1355 vs 1346 w 1350 w r NH, Trp (bR),d CH (P)

1366 vs 1380 m d CH3 (bR, L),Trp (bR)

1404 m 1397 w 1409 w 1394 m d CH3 (L)1452 w 1475 m d CH3 (bR),

d CH2,3 (L)1495 w 1484 m d CH3 (L),

amide II (P)1513 s 1513 w n C¼C (bR)

1523 m ns C¼C (bR)1542 s 1545 s nas C¼C (bR)

1566 m amide II1583 w 1599 m n C¼C (bR, L)

1614 w 1611 s 1611 w n C¼C (bR, L)Trp, Tyr

1638 s 1622 (S) n C¼NH(+) (bR)1688 w 1664 s 1672 m amide I (bR, P)

w: weak, m: middle, s: strong, vs: very strong, L: lipid con-tribution, bR: bacteriorhodopsin contribution (retinal withprotein envelope), P: protein contribution (without bR en-velope) n: stretch, d: deformation, r: rocking, t: torsion,wag: wagging, s: symmetric, as: asymmetric, ip: in plane,Phe: Phenylalanine, Trp: Tryptophan, Tyr: Tyrosine




Spectra 3 and 4 do not show any PM characteris-tics, based on the missing C¼C and the C¼N/C¼NHþ stretching modes of the retinal moiety. Dis-tinct lipid phosphate modes were detected at ap-proximately 800 cm�1 and 1080 cm�1. A b-carotenecontribution is mainly indicated by a shoulder at ap-proximately 1150 cm�1 and the C¼C stretching at1513 cm�1, both are labelled with a diamond. Typicalprotein Raman modes such as the amide I andamide III bands can be assigned to non-resonantproteins within the red membrane. All spectra as-signed to the retinal free domain in the grid do notshow the specific bR bands signals, as depicted by 2examples in Figure 2, right side.

In Figure 4 the results of the TERS measure-ments on halobacterial surface are summarized sche-matically. The protein envelope covering the innermembrane (S-layer), is not shown for simplification.Estimating an apex of approximately 20 nm the sil-ver covered TERS tip scans the inner membrane ofthe surface sequentially. The localized field enhance-ment is generated by the laser excitation at the nanoparticle and decays rapidly (after a few nanometers)with increasing distance of the particle surface. Thespectral information is based on the molecular com-position in the range of this evanescent field. Posi-tion 1 shows the tip localized on a purple membranedomain containing bR. Typically the proteins are or-ganized as bR trimers, ordered in a hexagonal lattice[32]. The red circles symbolize components presentin the red membrane, like b-carotene and proteins,which are randomly embedded in the lipid matrix.Hence, position 2 in Figure 4 refers to a purplemembrane free domain. Compositional differences

of the lipid molecules between purple and red mem-brane may cause a slightly differing matrix environ-ment, which is not the topic of the current investiga-tion.

The results confirm a patch like organization ofpurple and red membranes on the halobacterial sur-face. From the experiments the patch size of red andpurple membrane can be estimated to be on the or-der of a few hundred nanometer.

4. Conclusion

In this work purple membrane patches isolated fromHalobacterium salinarum cells and whole bacterialorganisms were investigated using tip-enhanced Ra-man spectroscopy. TERS experiments on bacterio-rhodopsin containing purple membrane (PM) patcheswere performed in order to identify specific PM Ra-man marker bands. These are the very specific C¼Cand C¼N stretching modes as well as intensive C––Hand N––H deformation modes of the resonantly en-hanced retinal moiety of bacteriorhodopsin. Thesedata, supported by resonance Raman and FTIRspectra, allowed the nanoscale differentiation of dis-tinct membrane domains on the halobacterial surfaceusing TERS. The spectra could be specifically as-signed to purple membrane (retinal containing) andretinal free domains, most likely red membrane, thetwo major components of the membrane surface ofHalobacterium salinarum. The established distribu-tion indicates a patch-wise organization of the com-ponents with typical length scales of several hundrednanometers. Considering the signal to noise ratio ofthe actual TERS-spectra, a shorter sampling time isfeasible. This allows smaller step sizes and would im-prove the resolution accordingly.

The present results signify an important step to-wards a nanoscale structural mapping of specific do-mains on membrane surfaces. The distribution aswell as the arrangement of cell surface compart-ments are important parameters and will broadenthe understanding of cell operations. Specifically na-noscale structural investigations are required to ad-dress the still unclear function of the so-called lipidrafts and TERS seems to be ideally suited for thistask.

Further progress in the instrumental develop-ment will allow also large scale structural studieswith nanometer lateral resolution which will enableto investigate the structural and functional composi-tion of an entire membrane system.

1

20 nm

2

Figure 4 (online color at: www.biophotonics-journal.org)Schematic TERS measurement on Halobacterium salina-rum. The TERS probe scans the inner membrane point-wise and the crescent shaped evanescent field penetratesthe surface on domains with (Pos 1) and without (Pos 2)purple membrane. Bacteriorhodopsin trimers are orderedin a hexagonal 2D lattice. Single components of the redmembrane (red circles) are randomly embedded in the li-pid bilayer.


Journal of

BIOPHOTONICS


Tanja Deckert-Gaudig stud-ied chemistry at the Uni-versity of Wurzburg. In 1992she received an Erasmus fel-lowship for a stay at theUniversity College of Swan-sea (Wales) and in 1994 aDAAD research grant atthe University of Tokyo (Ja-pan). She obtained her Ph.D.in Organic Chemistry in1997 with Prof. S. Hunig at

the University of Wurzburg. After a baby break shestarted working with Dr. V. Deckert at the Universityof Dresden in 2002 and moved to the Institute for Ana-lytical Sciences in Dortmund in the same year. Since2009 she works at the Institute of Photonic Technologyin Jena. Her main research interest is the characteriza-tion of protein related biopolymers with TERS.

Rene Bohme studied chem-istry at the Friedrich-Schil-ler-University in Jena from2003 until 2008. After he re-ceived his diploma hestarted a Ph.D. under thesupervision of Prof. JurgenPopp. His research is fo-cused on tip-enhanced Ra-man spectroscopy for the in-vestigation of biologicalmembrane systems.

Erik Freier studied physicsat the Ruhr-University Bo-chum, Germany. He finishedhis Pd.D. in 2011 at the de-partment of Biophysics,Ruhr-University Bochumunder the supervision ofKlaus Gerwert. His researchfocuses on time- and spa-tially resolved FTIR andRaman spectroscopy of bio-logical samples.

Aleksandar Sebesta studiedchemistry at ZHAW Win-terthur (BSc, 2003–2006)and then at ETH Zurich(MSc, 2007–2010), Switzer-land. Currently he is pursu-ing his DPhil in physicalchemistry in the Group ofDr. Philipp Kukura at theUniversity of Oxford. His

research focuses on fast microscopic methods to studythe kinetics of molecular motors.

Tobias Merkendorf studiedbiology at Ruhr-UniversityBochum, Germany andLund University, Sweden.He finished his studies 2008with his diploma thesis andstarted a Ph.D. in the groupof Klaus Gerwert in Bo-chum. His research is fo-cused on investigating thephotocycle of microbial rho-dopsins via time resolvedUV/Vis and FT-IR measure-ments in vitro and in vivo.

Jurgen Popp is Professor forPhysical Chemistry at theFriedrich-Schiller UniversityJena and Scientific Directorof the Institute of PhotonicTechnology Jena, Germany.He received his Ph.D. inchemistry from the Univer-sity of Wurzburg in 1995 andsubsequently spent a year inthe Department of AppliedPhysics at the Yale Univer-

sity, USA. After his habilitation in the group of Prof. Dr.W. Kiefer he became a full professor of the UniversityJena in 2002. In June 2006 he was also appointed scienti-fic director of the Institute of Photonic Technology. Hiswork was awarded by the faculty prize of chemistry(1995), the Bayerische Habilitationsforderpreis (1997),the Forderpreis der Wurzburger Korporationen (2001)and the Kirchhoff-Bunsen award (2002). His research in-terests are focussed on problems being resolved bymeans of innovative frequency-, time- and spatially re-solved laser spectroscopic methods. Thereby, bio- andmaterial photonics are setting up the two main prioritiesof the research activities.

Klaus Gerwert received hisPh.D. in biophysical chemis-try from the University ofFreiburg in 1985. From1986–1989 he conducted re-search at the Max-Planck-Institute in Dortmund, thenas Heisenberg-fellow from1990–1993 at the ScrippsResearch Institute, La Jolla,USA and also in Dortmund.




References

[1] M. Tanaka and E. Sackmann, Nature 437, 656–663(2005).

[2] A. Engel and H. E. Gaub, Annu. Rev. Biochem. 77,127–148 (2008).

[3] R. P. Richter, R. Berat, and A. R. Brisson, Langmuir22, 3497–3505 (2006).

[4] S. J. Singer and G. L. Nicolson, Science 175, 720–731(1972).

[5] F. R. Maxfield and I. Tabas, Nature 438, 612–621(2005).

[6] D. M. Engelman, Nature 438, 578–580 (2005).[7] J. Wessel, J. Opt. Soc. Am. B: Opt. Phys. 2, 1538–

1541 (1985).[8] R. M. Stockle, D. S. Yung, V. Deckert, and R. Zenobi,

Chem. Phys. Lett. 318, 131–136 (2000).[9] E. Bailo and V. Deckert, Chem. Soc. Rev. 37, 921–

930 (2008).[10] B. Pettinger, Mol. Phys. 108, 2039–2059 (2010).[11] K. Hering, D. Cialla, K. Ackermann, T. Dorfer,

R. Moeller, H. Schneidewind, R. Mattheis, W. Fritzsche,P. Rosch, and J. Popp, Anal. Bioanal. Chem. 390,113–124 (2008).

[12] K. A. Willets and R. P. Van Duyne, Annu. Rev. Phys.58, 267–297 (2007).

[13] J. Kneipp, H. Kneipp, and K. Kneipp, Chem. Soc.Rev. 37, 1052–1060 (2008).

[14] F. Festy, A. Demming, and D. Richards, Ultramicro-scopy 100, 437–441 (2004).

[15] A. Downes, D. Salter, and A. Elfick, J. Phys. Chem. B110, 6692–6698 (2006).

[16] M. Sackrow, C. Stanciu, M. A. Lieb, and A. J. Meix-ner, ChemPhysChem 9, 316–320 (2008).

[17] G. H. Patterson, Semin. Cell Dev. Biol. 20, 886–893(2009).

[18] G. Patterson, M. Davidson, S. Manley, and J. Lippin-cott-Schwartz, Annu. Rev. Phys. Chem. 61, 345–367(2010).

[19] T. Deckert-Gaudig, and V. Deckert, Phys. Chem.Chem. Phys. 12, 12040–12049 (2010).

[20] E. Bailo and V. Deckert, Angew. Chem., Int. Ed. 47,1658–1661 (2008).

[21] B.-S. Yeo, S. Maedler, T. Schmid, W. Zhang, andR. Zenobi, J. Phys. Chem. C 112, 4867–4873 (2008).

[22] D. Cialla, T. Deckert-Gaudig, C. Budich, M. Laue,R. Moller, D. Naumann, V. Deckert, and J. Popp, J.Raman Spectrosc. 40, 240–243 (2009).

[23] P. Hermann, A., Hermelink, V. Lausch, G. Holland,L. Moeller, N. Bannert, and D. Naumann, Analyst136, 1148–1152 (2011).

[24] U. Neugebauer, P. Rosch, M. Schmitt, J. Popp, C. Ju-lien, A. Rasmussen, C. Budich, and V. Deckert,ChemPhysChem 7, 1428–1430 (2006).

[25] R. Bohme, M. Richter, D. Cialla, P. Rosch, V. De-ckert, and J. Popp, J. Raman Spectrosc. 40, 1452–1457(2009).

[26] R. Bohme, M. Mkandawire, U. Krause-Buchholz,P. Rosch, G. Rodel, J. Popp, and V. Deckert, Chem.Commun. 47, 11453–11455 (2011).

[27] R. Bohme, D. Cialla, M. Richter, P. Rosch, J. Popp,and V. Deckert, J. Biophoton. 3, 455–461 (2010).

[28] L. Opilik, T. Bauer, T. Schmid, J. Stadler, and R. Ze-nobi, Phys. Chem. Chem. Phys. 13, 9978–9981 (2011).

[29] M. Kamekura and M. Kates, Biosci. Biotechnol. Bio-chem. 63, 969–972 (1999).

[30] C. Klein, C. Garcia-Rizo, B. Bisle, B. Scheffer,H. Zischka, F. Pfeiffer, F. Siedler, and D. Oesterhelt,Proteomics 5, 180–197 (2005).

[31] D. Oesterhelt and W. Stoeckenius, Proc. Nat. Acad.Sci. U.S.A. 70, 2853–2857 (1973).

[32] W. Stoeckenius, R. H. Lozier, and R. A. Bogomolni,Biochim. Biophys. Acta, Rev. Bioenerg. 505, 215–278(1979).

[33] S. Wolf, E. Freier, and K. Gerwert, ChemPhysChem9, 2772–2778 (2008).

[34] J. K. Lanyi, Biochim. Biophys. Acta, Bioenerg. 1757,1012–1018 (2006).

[35] F. Garczarek and K. Gerwert, J. Am. Chem. Soc. 128,28–29 (2006).

[36] F. Garczarek and K. Gerwert, Nature 439, 109–112(2006).

[37] E. Freier, S. Wolf, and K. Gerwert, Proc. Natl. Acad.Sci. U.S.A. 28, 11435–11439 (2011).

Since 1993 he has been appointed as full professor atthe Ruhr-University Bochum, where he holds a chair ofbiophysics; since 2009 he is director at the Max-Planck-CAS partner-institute in Shanghai (dual appointment)and Max-Planck-fellow. His research focuses on pro-teins using time-resolved FTIR-spectroscopy combinedwith x-ray structure analysis and computational biologyto elucidate molecular reaction mechanisms and inter-actions of membrane and membrane-bound proteins:microbial rhodopsins; Ras-superfamily; in addition vi-brational imaging of cells and tissues in translationalclinical research.

Volker Deckert holds a jointposition at the Institute ofPhysical Chemistry and theInstitute of Photonic Tech-nologies, both in Jena, Ger-many. He obtained his Di-ploma and Ph.D. from theUniversity of Wurzburg,Germany, working on Ra-man spectroscopy. As a post-doc he worked on non-linearand time-resolved laser spec-troscopy at the University of

Tokyo and KAST, in Kawasaki. During his habilitationat the ETH Zurich, he came into contact with near-field optical spectroscopy, a subject he still pursues andapplies to mainly biological samples in his current posi-tion.


Journal of

BIOPHOTONICS


[38] D. Oesterhelt and W. Stoeckenius, Nature 233, 149–152 (1971).

[39] A. Rasmussen and V. Deckert, J. Raman Spectrosc.37, 311–317 (2006).

[40] C. Budich, U. Neugebauer, J. Popp, and V. Deckert,J. Microsc. 229, 533–539 (2008).

[41] R. Rammelsberg, B. Hessling, H. Chorongiewski, andK. Gerwert, Appl. Spectrosc. 51, 558–562 (1997).

[42] S. O. Smith, J. A. Pardoen, J. Lugtenburg, and R. A.Mathies, J. Phys. Chem. 91, 804–819 (1987).

[43] S. O. Smith, M. S. Braiman, A. B. Myers, J. A. Pardoen,J. M. L. Courtin, C. Winkel, J. Lugtenburg, and R. A.Mathies, J. Am. Chem. Soc. 109, 3108–3125 (1987).

[44] I. Notingher, S. Verrier, S. Haque, J. M. Polak, andL. L. Hench, Biopolymers 72, 230–240 (2003).

[45] S. Stewart and P. M. Fredericks, Spectrochim. Acta,Part A 55A, 1641–1660 (1999).

[46] S. O. Smith, J. Lugtenburg, and R. A. Mathies, J.Membr. Biol. 85, 95–109 (1985).

[47] K. Gerwert, B. Hess, J. Soppa, and D. Oesterhelt,Proc. Natl. Acad. Sci. U.S.A. 86, 4943–4947 (1989).

[48] S. Wolf, E. Freier, M. Potschies, E. Hofmann, andK. Gerwert, Angew. Chem. Int. Ed. 49, 6889–6893(2010).

[49] T. Deckert-Gaudig, E. Rauls, and V. Deckert, J. Phys.Chem. C 114, 7412–7420 (2010).

[50] T. J. Beveridge, Curr. Opin. Struct. Biol. 4, 204–212(1994).

[51] R. G. Efremov, A. V. Feofanov, and I. R. Nabiev, J.Raman Spectrosc. 23, 69–73 (1992).

Register now for the freeWILEY-VCH Newsletter!www.wiley-vch.de/home/pas

Covering interface science from a novel surface scienceperspective, this unique handbook offers a comprehensiveoverview of this burgeoning field.Eight topical volumes cover basic concepts & methods,elemental and composite surfaces, solid-gas, solid-liquid andinorganic-biological interfaces, as well as applications ofsurface science in nanotechnology, materials science andmolecular electronics.With its broad scope and clear structure, it is ideal as areference for scientists working in the field, as well as anintroduction for newcomers.

WILEY-VCH • P.O. Box 10 11 61 • 69451 Weinheim, GermanyFax: +49 (0) 62 01 - 60 61 84e-mail: [email protected] • http://www.wiley-vch.de

KLAUS WANDELT, University of Bonn, Germany (ed.)

Surface and Interface ScienceVolume 1: Concepts and MethodsVolume 2: Properties of Elemental Surfaces

+++ Suggested Reading +++ Suggested Reading +++ Suggested Reading +++

2012. XLVIII, 956 pages, 504 figures25 in color, 23 tables. Hardcover.ISBN: 978-3-527-41156-6Prepublication price valid untilJune 30, 2012




more than 200 yearsof fi rst-class publishing

Five good reasonsto choose AdP� New content: best fundamental physics, top applied physics

and hot topics.

� Additional sections: Rapid Research Letters, brief historical essays, Expert Opinion.

� New team: Editor-in-Chief Guido W. Fuchs, supported by a new Advisory Board and an in-house Editorial Offi ce guarantee high quality standard.

� Modern layout: new front cover, new article layout, new everything.

� Rapid publication times: efficient reviewing and subsequent fast online publication.

Submit your work online now: http://mc.manuscriptcentral.com/andp

Advisory BoardB. Allen, Hannover, GER / D. D. Awschalom, Santa Barbara, USA /C. W. J. Beenakker, Leiden, NL / K. Blaum, Heidelberg, GER /I. Bloch, Munich, GER / Ch. Bruder, Basel, CH / A. Caldwell, Munich,GER / F. Capasso, Cambridge, USA / I. Cirac, Garching, GER /G. Dvali, Munich, GER / N. Gisin, Genève, CH / R. Fazio, Pisa, IT /R. Frésard, Caen, FR / T. Hänsch, Munich, GER / S. Hell, Göttingen,GER / A. Imamoglu, Zurich, CH / C. Kiefer, Cologne, GER / P. Kim, New York, USA / J. Mannhart, Stuttgart, GER / M. Orrit, Leiden, NL /G. Schön, Karlsruhe, GER / M. Schreiber, Chemnitz, GER / Y. Tokura, Tokyo, JP / V. Vedral, Oxford, UK / I. A. Walmsley, Oxford, UK /H. Zohm, Garching, GER / P. Zoller, Innsbruck, AT

ScopeThe journal covers the physics of

� Condensed Matter / Solid State / Materials

� Optics / Photonics / Quantum Information

� Cosmology / Gravitation / Relativity

� High Energy / Particles / Nuclear

� Atoms and Molecules / Plasma

� Biophysics / Biological and Medical Applications

� Geo / Climate / Environment

“Annalen der Physik is more than a journal, it is aninspiration! As author you can rely on an individualized,unbiased, fair and fast publication procedure. As editorI will count on your creativity and richness of ideas tomake the journal grow even further!”

Guido W. Fuchs, Editor-in-Chief AdP (2012)

Contact Annalen der PhysikWILEY-VCH Verlag GmbH & Co. KGaARotherstraße 21, 10245 Berlin, GermanyE-mail: [email protected]

NEW RENEWED IN 2012

journal of biophotonics - bph.ruhr-uni-bochum.de · j. biophotonics 5, no. 7, 582–591 (2012) /...

Documents