Direct observation of the nanoscale dynamics of membrane lipids in a living cell.
ABSTRACT Cholesterol-mediated lipid interactions are thought to have a functional role in many membrane-associated processes such as signalling events. Although several experiments indicate their existence, lipid nanodomains ('rafts') remain controversial owing to the lack of suitable detection techniques in living cells. The controversy is reflected in their putative size of 5-200 nm, spanning the range between the extent of a protein complex and the resolution limit of optical microscopy. Here we demonstrate the ability of stimulated emission depletion (STED) far-field fluorescence nanoscopy to detect single diffusing (lipid) molecules in nanosized areas in the plasma membrane of living cells. Tuning of the probed area to spot sizes approximately 70-fold below the diffraction barrier reveals that unlike phosphoglycerolipids, sphingolipids and glycosylphosphatidylinositol-anchored proteins are transiently ( approximately 10-20 ms) trapped in cholesterol-mediated molecular complexes dwelling within <20-nm diameter areas. The non-invasive optical recording of molecular time traces and fluctuation data in tunable nanoscale domains is a powerful new approach to study the dynamics of biomolecules in living cells.
- SourceAvailable from: Francesco Cardarelli[Show abstract] [Hide abstract]
ABSTRACT: The translational motion of molecules in cells deviates from what is observed in dilute solutions. Theoretical models provide explanations for this effect but with predictions that drastically depend on the nanoscale organization assumed for macromolecular crowding agents. A conclusive test of the nature of the translational motion in cells is missing owing to the lack of techniques capable of probing crowding with the required temporal and spatial resolution. Here we show that fluorescence-fluctuation analysis of raster scans at variable timescales can provide this information. By using green fluorescent proteins in cells, we measure protein motion at the unprecedented timescale of 1 μs, unveiling unobstructed Brownian motion from 25 to 100 nm, and partially suppressed diffusion above 100 nm. Furthermore, experiments on model systems attribute this effect to the presence of relatively immobile structures rather than to diffusing crowding agents. We discuss the implications of these results for intracellular processes.Nature Communications 12/2014; 5:5891. · 10.74 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We have previously shown that GFP-tagged human ABCG1 on the plasma membrane (PM) and in late endosomes (LE) mobilizes sterol on both sides of the membrane lipid bilayer, thereby increasing cellular cholesterol efflux to lipid surfaces. In the present study, we examined ABCG1-induced changes in membrane cholesterol distribution, organization, and mobility. ABCG1-GFP expression increased the amount of mobile, non-sphingomyelin(SM)-associated cholesterol at the PM and LE, but not the amount of SM-associated-cholesterol or SM. ABCG1-mobilized non-SM-associated-cholesterol rapidly cycled between the PM and LE and effluxed from the PM to extracellular acceptors, or, relocated to intracellular sites of esterification. ABCG1 increased detergent-soluble pools of PM and LE cholesterol, generated detergent-resistant, non-SM-associated PM cholesterol, and increased resistance to both amphotericin B-induced (cholesterol-mediated) and lysenin-induced (SM-mediated) cytolysis, consistent with altered organization of both PM cholesterol and SM. ABCG1 itself resided in detergent-soluble membrane domains. We propose that PM and LE ABCG1 residing at the phase boundary between ordered (Lo) and disordered (Ld) membrane lipid domains alters SM and cholesterol organization thereby increasing cholesterol flux between Lo and Ld, and hence, the amount of cholesterol available for removal by acceptors on either side of the membrane bilayer for either efflux or esterification.Biology 12/2014; 3(4):866-891.
- [Show abstract] [Hide abstract]
ABSTRACT: The molecular in-plane and out-of-plane structure of dimyristoylphosphatidylcholine (DMPC) membranes containing up to 60 mol% of cholesterol was studied using X-ray diffraction. Up to 37.5 mol% cholesterol could be dissolved in the membranes, resulting in a disordered lateral membrane structure. Highly ordered cholesterol structures were observed at cholesterol concentrations of more than 40 mol% cholesterol. These structures were characterized as immiscible cholesterol plaques, i.e., bilayers of cholesterol molecules coexisting with the lipid bilayer. The cholesterol molecules were found to form a monoclinic structure at 40 mol% cholesterol, which transformed into a triclinic arrangement at the highest concentration of 60 mol%. Monoclinic and triclinic structures were found to coexist at cholesterol concentrations between 50 and 55 mol%.Soft Matter 01/2013; 9(39):9342. · 4.15 Impact Factor
Direct observation of the nanoscale dynamics of
membrane lipids in a living cell
Christian Eggeling1*, Christian Ringemann1*, Rebecca Medda1, Gu ¨nter Schwarzmann2, Konrad Sandhoff2,
Svetlana Polyakova1, Vladimir N. Belov1, Birka Hein1, Claas von Middendorff1, Andreas Scho ¨nle1& Stefan W. Hell1
Cholesterol-mediated lipid interactions are thought to have a
functional role in many membrane-associated processes such as
signalling events1–5. Although several experiments indicate their
existence, lipid nanodomains (‘rafts’) remain controversial owing
to the lack of suitable detection techniques in living cells4,6–9. The
controversy is reflected in their putative size of 5–200nm, span-
ning the range between the extent of a protein complex and the
resolution limit of optical microscopy. Here we demonstrate the
ability of stimulated emission depletion (STED) far-field fluor-
escence nanoscopy10to detect single diffusing (lipid) molecules
in nanosized areas in the plasma membrane of living cells.
Tuningoftheprobed areatospotsizes 70-foldbelowthediffrac-
tion barrier reveals that unlike phosphoglycerolipids, sphingoli-
pids and glycosylphosphatidylinositol-anchored proteins are
transiently ( 10–20ms) trapped in cholesterol-mediated molecu-
lar complexes dwelling within ,20-nm diameter areas. The non-
data in tunable nanoscale domains is a powerful new approach to
study the dynamics of biomolecules in living cells.
inositol (GPI)-anchored proteins are assumed to form molecular
complexes or integrate, assisted by cholesterol, into,200-nm-sized
lipid nanodomains impeding their diffusion in the plasma mem-
brane11. Featuring a diameter d.200nm, the commonly used detec-
tion spot of a confocal microscope averages over such details of
molecular diffusion (Fig. 1a). Electron microscopy provides the
could in principle provide an answer, but the direct visualization of
lipid nanodomains rather than that of clustered proteins is challenged
by the rapid diffusion of the lipids4. This rapid diffusion also prevents
the fast tracking of single lipid molecules in their native cargo-free
state16–18. Although fluorescence recovery after photobleaching19,20
and confocal fluorescence correlation spectroscopy21–23are able to
map the diffusion, their resolution is diffraction-limited which
In contrast, emergent stimulated emission depletion (STED) fluor-
escence microscopy10,25,26(see Methods) opens up a new avenue to
determine lipid diffusion, because its detection area can uniquely be
downscaled in size by suppressing the fluorophore excitation at the
outer parts of the focal spot (Fig. 1b).
Therefore we labelled sphingolipids, represented by sphingomye-
integrating’ phosphoethanolamine with the fluorophore Atto647N
and studied their dynamics in the plasma membrane of living cells
(see Methods). Control experiments indicated that the dye label had
GPI-anchor (Supplementary Fig. 8). An average of,10 labelled
molecules per mm2membrane area ensured the presence of single
fluorescent molecules in the detection area. The measurements were
Specifically, we targeted arbitrary points on the plasma membrane
and recorded the fluorescence bursts of molecules crossing the focal
spot. Figure 1c and e show single-molecule fluorescence time traces
of phosphoethanolamine, measured using the regular confocal and
the subdiffraction detection area downsized by STED to d < 40nm,
respectively. Besides demonstrating the ability of the STED micro-
scope to detect single molecules in living cells, the traces show a
uniform reduction of the burst length, accounting for the reduced
transit time of a phosphoethanolamine molecule through the sub-
diffraction sized spot.
Next we recorded a similar pair of confocal and STED traces of
sphingomyelin molecules (Fig. 1dandf).Theconfocal measurement
does not show obvious differences between the sphingomyelin and
phosphoethanolamine diffusion. Containing ,1,000 photons per
molecule at an average transit time of 10ms, the fluorescence bursts
are rather homogeneous in height and length for both lipids. In
contrast, the STED microscope reveals a whole range of burst dura-
tions for sphingomyelin and hence a substantial difference to phos-
phoethanolamine. Whereas the phosphoethanolamine traces feature
sharp peaks only, in the case of sphingomyelin the sharp peaks are
accompanied by longer bursts. Thus the reduction of the detection
area far below that provided by confocal microscopy reveals that the
diffusion of sphingomyelin is strongly heterogeneous.
An important observation is that STED microscopy reduces the
fluorescence burst length but hardly its peak height. The detected
photon count-rate of a typical molecular trace peaks at 100kHz,
for phosphoethanolamine and sphingomyelin (Fig. 1g, h and
Supplementary Fig. 2). For phosphoethanolamine we found just a
single type of diffusion, characterized by burst durations ,1ms.
Phosphoethanolamine shows no observable sign of heterogeneous
diffusion in the plasma membrane. In contrast, the sphingomyelin
histogram reveals a broad distribution of burst durations from short
events that are similar to that of phosphoethanolamine, to longer
events with burst durations ranging up to.50ms.
To quantify the lipid dynamics further, we applied fluorescence
*These authors contributed equally to this work.
1Department of Nanobiophotonics, Max Planck Institute for Biophysical Chemistry, Am Fassberg 11, 37077 Go ¨ttingen, Germany.2LIMES Membrane Biology and Lipid Biochemistry
Unit, University of Bonn, Gerhard-Domagk-Strasse 1, 53121 Bonn, Germany.
©2008 Macmillan Publishers Limited. All rights reserved
through the focal area (Supplementary Information). The confocal
compared to those recorded for phosphoethanolamine (Fig. 2a).
Because both curves can be described by normal diffusion of a single
species, we cannot infer whether the diffusion of sphingomyelin is
inhomogeneous on small spatial scales, or just slightly slower than
phosphoethanolamine. This is different in the correlation curves
apparent (Fig. 2b). Owing to the smaller detection area, the phos-
represents a single normal diffusion. In contrast, the sphingomyelin
model assuming two dissimilar modalities of focal transits or, alter-
natively, anomalous diffusion of the lipids (Supplementary
Information). To investigate whether the longer focal transit time of
sphingomyelin is associated with cholesterol, we added cholesterol
oxidase to deplete the cholesterol in the membrane. Figure 2c shows
that the longer transit times disappear.
STED microscopy enabled us to scale the diameter of the focal
detection area from d5250nm down to 30nm after increasing the
decrease in the transit time reveals the multifaceted dynamics of the
membrane molecules23,24. The transit time for phosphoethanolamine
decreases in proportion to the ,70-fold reduction in focal detection
area (Fig. 3), indicating that phosphoethanolamine is indeed freely
diffusing with a diffusion coefficient D5(562) 1029cm2s21. This
diameter d.160nm, the diffusion of sphingomyelin appears normal
and just 1.5–2.5 times slower than that of phosphoethanolamine.
Likewise, for 80nm,d,160nm, the spot size is not small enough
d,80nm, we can separate free from hindered diffusion and analyse
modalities of focal transits: free and hindered diffusion. Considering
that the free diffusion exhibits the same transit time tD1as phos-
phoethanolamine (compare Fig. 1g and h), we calculated that a frac-
tion of A2< 60% of sphingomyelin molecules crosses the d530–
80nm large areas within a transit time tD2which is.10 times longer
than tD1, reaching a plateau at ,10ms.
This observation of a restricted minimum transit time through
nanoscale areas rules out the possibility that sphingomyelin diffuses
freely (but just more slowly) in the plasma membrane, because the
tD1. Therefore the hindered diffusion of sphingomyelin and the
longer transit time tD2must be due to a brief trapping of these
molecules. The average trapping period ttrap,10ms is calculated
by subtracting tD1from tD2. The fact that ttrapis constant for
d,60nm indicates a single trapping event within the probed areas.
(Fig. 3) and the ganglioside GM1 (Supplementary Fig. 6) with D < 3
and 531029cm2s21, ttrap< 18 and 11ms, and A2< 35 and 45%,
Notably, the addition of the cholesterol-depleting agent choles-
terol oxidase causes sphingomyelin, the GPI-anchor, and GM1 to
exhibit a linear dependence of the transit time on d2, which is indi-
cative of the fact that the trapping has been reduced. Although cho-
lesterol oxidase may also induce other changes in the membrane, it
primarily depletes the cholesterol. Alternative cholesterol depletion
by b-cyclodextrin also reduces the trapping (Supplementary Fig. 5).
Figure 1 | STED microscopy time traces of single-molecule diffusion in live
cell plasma membrane. a, b, Molecules may move freely and/or be
transiently trapped on small spatial scales. The large detection area of a
confocal microscope (a) cannot discern such details. However, the
subdiffraction spot created by STED (b) is able to discriminate between
lipids that diffuse freely (I) and those that are hindered (II) during their
passage. c–h, Fluorescence bursts from single-diffusing Atto647N-labelled
phosphoethanolamine (PE) and sphingomyelin (SM) lipids detected with a
confocal(c, d)andaSTEDspot(e, f),andafrequencyplotofthevaluepairs
of the STED recording for phosphoethanolamine (g, 497 bursts) and
sphingomyelin (h, 539 bursts). By ensuring that a freely diffusing molecule
hindered (II) from free diffusion (I).
©2008 Macmillan Publishers Limited. All rights reserved
The hindered sphingolipid and GPI-anchor diffusion may be
caused by various molecular interactions. However, we can exclude
the floating of temporally stable lipid complexes or domains through
the membrane, because this would manifest itself as a linear depend-
ence of tD2on the reduced focal area d2created by STED. Moreover,
our data demonstrate that during the trapping, the sphingolipids or
GPI-anchored molecules remain within a,20-nm diameter area.
This conclusion is on the basis of the fact that for 30–60nm large
spots, the transit time tD2and the fraction A2of diffusion-hindered
events is constant (Supplementary Information). These findings are
supported by alternative evaluations of the FCS data applying anom-
alous diffusion models (Supplementary Fig. 3), by the STED images
recorded for sphingomyelin (Supplementary Fig. 10), and by Monte
Altogether, the evolving picture of the observed sphingolipid and
GPI-anchor dynamics is that of transient formations of cholesterol-
assisted molecular complexes, such as lipid-protein binding or lipid
shells (compare Supplementary Information). Although an entire
complex of for example several lipid and protein molecules may
(temporarily) be of larger spatial scales, the trapped molecule dwells
only within,20-nm diameter areas. The diffusion of the complexes
a time period that is longer than the trapping time, that is, with a
diffusion constant D=0.23d2/ttrap<10210cm2s21.
The linear dependence of the transit time of the freely diffusing
lipids on the subdiffraction detection area (and their average particle
compromised by photobleaching, local heating or radical formation.
Even if some molecules were bleached on passing the doughnut of
is less affected. In fact, the confocal measurements are more prone to
bleaching artefacts due to the inherently longer transit time through
the detection area. Thus the traces in Figs 1 and 2 show the ability of
STED microscopy to provide a unique and hitherto unrecognized
cells, thereby greatly expanding the potential of the popular FCS
technique. STED-FCS is complementary to single-molecule tracking
because several molecules per subdiffraction area may contribute to
the signal, allowing a fast gathering of statistically accurate data.
Besides, the spot size of a STED microscope is physically predefined
andthespotcanbedirectedtoarbitrary coordinates within acell.All
these features have provided a unique combination of temporal and
spatial resolution required to quantify the nanoscale dynamics of the
small lipid molecules in the cellular membrane. Thus the trapped
molecular diffusion revealed in this work is only a prelude of a new
class of nanoscale biomolecular studies that are to follow.
In our STED microscope, the diffraction-limited focal spot of excitation light is
overlapped with a doughnut-shaped spot of STED light that switches off the
ability of the label to fluoresce, thus confining the origin of the fluorescence to
STED focal area d2 (nm2)
STED focal diameter d (nm)
Transit time τD (ms)
Figure 3 | Molecular transit through nanoscale areas in live cell plasma
ranging from d5250nm (conf., confocal; grey area) down to 30nm in
open squares) decrease linearly with the area d2, confirming free diffusion
(solid line). Whereas large areas.180nm in diameter yield just a single
diameter reveal two distinct modalities of molecular transits (grey arrows),
with tD15tD(phosphoethanolamine) (solid line) and tD2demonstrating
hindered diffusion. The tD2plateau found for small d indicates transient
trapping of sphingomyelin. Slowed-down free diffusion of sphingomyelin
tDon d2, as exemplified for D52310210cm2s21(dashed line).
b, c, Fraction A2(a) and trapping time ttrap(c) of transits of sphingomyelin
(black dots) and GPI-anchor (grey triangles) exhibiting transient trapping.
thirty FCS measurements on different cells.
Correlation time (ms)
Figure 2 | FCS of Atto647N-labelled phosphoethanolamine and
sphingomyelin plasma membrane diffusion. a–c, Normalized correlation
data of phosphoethanolamine (PE, red dots) and sphingomyelin (SM, grey
cholesterol depletion by the addition of cholesterol oxidase (COase, black
dots) for STED recording. The STED but not the confocal data reveal
cholesterol-assisted heterogeneous diffusion of sphingomyelin. Red lines in
a and b denote single-species fit; tD519 (phosphoethanolamine, confocal),
28 (sphingomyelin, confocal) and 0.45ms (phosphoethanolamine, STED).
Blue lines denote fit assuming two dissimilar modalities of focal transits:
tD150.45ms, tD2510ms and A2564% (sphingomyelin, in b) and
tD254ms and A2515% (sphingomyelin 1 cholesterol oxidase, in c).
©2008 Macmillan Publishers Limited. All rights reserved
the doughnut centre (Supplementary Fig. 1). With the intensity Isbeing a
characteristic of the fluorophore and the STED light, and I?Isdenoting
the doughnut peak intensity, the diameter of the detection area can be tuned
=250nm, with NA denoting the numerical aper-
of 250–300ps pulses for STED at 770–780nm, and an average focal power of up
to 380mW. Imprinting a helical phase delay on the STED laser wavefront ren-
laser diode for fluorescence excitation with average powers of 5–8mW. Laser
focusing and collection of the fluorescence was performed by an NA1.42 oil
the STED microscope were calibrated by imaging 20-nm diameter fluorescent
beads (Supplementary Fig. 1). Labelling of the lipids with the organic dye
Atto647N was accomplished by chemical synthesis and of the GPI-anchor by
acylcarrierprotein(ACP)-tagging, respectively. Incubationwith thefluorescent
lipid–BSA complexes or GPI–ACP plasmid transfection resulted in molecular
insertion into the plasma membrane of living PtK2 cells. Measurements were
performed in cell culture medium at 27–37uC, using a temperature-controlled
microscope randomly on the lower plasma membrane.
Full Methods and any associated references are available in the online version of
the paper at www.nature.com/nature.
Received 6 September; accepted 29 October 2008.
Published online 21 December 2008.
1.Simons, K. & Ikonen, E. Functional rafts in cell membranes. Nature 387, 569–572
rich membrane rafts. J. Biol. Chem. 275, 17221–17224 (2000).
Fielding, C. J. Lipid Rafts and Caveolae (Wiley-VCH, 2006).
Jacobson, K., Mouritsen, O. G. & Anderson, G. W. Lipid rafts: at a crossroad
between cell biology and physics. Nature Cell Biol. 9, 7–14 (2007).
Hanzal-Bayer, M. F. & Hancock, J. F. Lipid rafts and membrane traffic. FEBS Lett.
581, 2098–2104 (2007).
Munro, S. Lipid rafts: elusive or illusive? Cell 115, 377–388 (2003).
Lommerse, P. H. M., Spaink, H. P. & Schmidt, T. In vivo plasma membrane
organization: results of biophysical approaches. Biochim. Biophys. Acta 1664,
Mol. Cell Biol. 7, 456–462 (2006).
Shaw, A. S. Lipid rafts: now you see them, now you don’t. Nature Immunol. 7,
10. Hell, S. W. & Wichmann, J. Breaking the diffraction resolution limit by stimulated
function. J. Lipid Res. 47, 1597–1598 (2006).
12. Fujita, A. et al. Gangliosides GM1 and GM3 in the living cell membrane form
clusters susceptible to cholesterol depletion and chilling. Mol. Biol. Cell 18,
13. Binnig, G., Quate, C. F. & Gerber, C. Atomic force microscope. Phys. Rev. Lett. 56,
14. Pohl, D. W., Denk, W. & Lanz, M. Optical stethoscopy: Image recording with
resolution l/20. Appl. Phys. Lett. 44, 651–653 (1984).
15. Zacharias, D. A., Violin, J. D., Newton, A. C. & Tsien, R. Y. Partitioning of lipid-
modified monomeric GFPs into membrane microdomains of live cells. Science
296, 913–916 (2002).
16. Saxton, M. J. & Jacobson, K. Single particle tracking: applications to membrane
dynamics. Annu. Rev. Biophys. Biomol. Struct. 26, 373–399 (1997).
17. Schu ¨tz, G. J., Kada, G., Pastushenko, V. P. & Schindler, H. Properties of lipid
microdomains in a muscle cell membrane visualized by single molecule
microscopy. EMBO J. 19, 892–901 (2000).
18. Fujiwara, T., Ritchie, K., Murakoshi, H., Jacobson, K. & Kusumi, A. Phospholipids
undergo hop diffusion in compartmentalized cell membrane. J. Cell Biol. 157,
19. Yechiel, E. & Edidin, M. Micrometer-scale domains in fibroblast plasma
membranes. J. Cell Biol. 105, 755–760 (1987).
20. Feder, T. J., Brust-Mascher, I., Slattery, J. P., Baird, B. A. & Webb, W. W.
Biophys. J. 70, 2767–2773 (1996).
21. Fahey, P. F. et al. Lateral diffusion in planar lipid bilayers. Science 195, 305–306
22. Schwille, P., Korlach, J. & Webb, W. W. Fluorescence correlation spectroscopy
with single-molecule sensitivity on cell and model membranes. Cytometry 36,
23. Wawrezinieck, L., Rigneault, H., Marguet, D. & Lenne, P.-F. Fluorescence
correlation spectroscopy: diffusion laws to probe the submicron cell membrane
organization. Biophys. J. 89, 4029–4042 (2005).
24. Wenger, J. et al. Diffusion analysis within single nanometric apertures reveals the
ultrafine cell membrane organization. Biophys. J. 92, 913–919 (2007).
25. Willig, K. I., Rizzoli, S. O., Westphal, V., Jahn, R. & Hell, S. W. STED-microscopy
reveals that synaptotagmin remains clustered after synaptic vesicle exocytosis.
Nature 440, 935–939 (2006).
26. Hell, S. W. Far-field optical nanoscopy. Science 316, 1153–1158 (2007).
27. Magde, D., Elson, E. L. & Webb, W. W. Thermodynamic fluctuations in a reacting
system - measurement by fluorescence correlation spectroscopy. Phys. Rev. Lett.
29, 705–708 (1972).
Supplementary Information is linked to the online version of the paper at
Acknowledgements We thank J. Jethwa, B. Rankin and M. Hilbert for critical
reading, K. Willig for help with the setup, T. Lang, R. Wagner and H. Rigneault for
valuable discussions, R. Machinek and H. Frauendorf for recording the NMR and
mass spectra, and S. Yan for help with the synthesis.
data, R.M. prepared samples and performed washing experiments, G.S., K.S., S.P.
and V.N.B. synthesized fluorescently labelled lipids and performed
chromatography, C.v.M., A.S., C.R. and C.E. realized and analysed simulated data,
C.E. and S.W.H. designed experiments and wrote the paper. All authors discussed
the results and commented on the manuscript.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. Correspondence and requests for materials should be
addressed to S.W.H. (firstname.lastname@example.org) or C.E. (email@example.com).
©2008 Macmillan Publishers Limited. All rights reserved
Cell culture. The epithelial cell line PtK2 was grown as previously described28.
Thecells wereseeded on number1 thicknessstandardglass coverslips(diameter
25mm) to a confluence of about 80% and grown at 37uC in a water-saturated
atmosphere of 5% CO2in air. All media and supplements were purchased from
Atto647N-labelled lipids. We used the organic dye Atto647N (fluorescence
excitation and emission maxima at 645 and 670nm, respectively; Atto-Tec) as a
fluorescence marker. N-(Atto647N)-1,2-dipalmitoyl-sn-glycero-3-phosphoetha-
nolamine (Atto647N-phosphoethanolamine, referred to as phosphoethanola-
mine) and N-(Atto647N)-sphingosylphosphocholine (N-Atto647-sphingomyelin,
Atto647N-sphingomyelin, referred to as sphingomyelin) were purchased from
Atto-Tec. The Atto647N-labelled ganglioside GM1 was synthesized as outlined in
the Supplementary Information starting from GM1 (ALEXIS Biochemicals). We
either with lipophilic or hydrophilic fluorescent labels. The control experiments
characteristics of the labelled lipids.
Plasma membrane insertion of the fluorescent lipids. Complexes of the
labelled lipids and bovine serum albumin (BSA) were prepared and incubated
with the cells according to a slightly modified protocol than that previously
described29,30; as detailed in the Supplementary Information.
Glycosylphosphatidylinositol anchor. We used an ACP tag (Covalys
Bioscience) for labelling of the GPI-anchor with the organic dye Atto647N.
Twenty-four hours after transfection with the plasmid pAEMXT-ACPwt-GPI
(Covalys), we incubated PtK2 cells expressing GPI-anchored ACP for 30min at
37uC in complete medium supplemented with 5mM CoA-Atto647N (AttoTec;
CoA, coenzyme A), 5mM MgCl2(Fluka) and 1mM ACP synthase (Covalys).
Afterwards, the cells were washed three times and prepared for measurements
using DMEM containing no phenol red. The cells were measured 24h after
transfection. Control experiments with different fluorescent markers precluded
an observable influence of the label on GPI trapping (Supplementary
Treatment with cholesterol oxidase and b-cyclodextrin. To modify the cho-
lesterol contents of the plasma membrane, the cells were either treated with
Streptomyces sp. cholesterol oxidase (Sigma, stock solution
34Uml21in 50mM KH2PO4, pH7.5) in HDMEM (HEPES 1 DMEM) and
then washed in HDMEM (see Supplementary Information), or treated with
10mM b-cyclodextrin (Sigma) in HDMEM (without phenol red) for 30min
at 37uC. The cholesterol oxidase or b-cyclodextrin treatment was performed
either before or after the insertion of the fluorescent lipid analogues into the
STED microscope. We implemented a confocalized setup in which the excita-
tion of the dye was performed with a 633nm pulsed laser diode (< 80ps pulse
width, LDH-P-635, Picoquant). The STED beam was provided by a titanium:
sapphire lasersystem(MaiTai, Spectra-Physics) operatingat770–780nm with a
repetition rate of 76MHz; this beam also provided the trigger for the excitation
laser. The STED laser power was controlled and stabilized by a power controller
unit (LPC, Brockton Electronics). After passing a 30-cm optical glass rod, the
linear-polarized light was coupled into a 120m long polarization maintaining
single-mode fibre (AMS Technologies) for stretching the pulses to 250–300ps.
The pulse timing of both lasers was adjusted using a home-built electronic delay
unit. After spatial overlay of both laser beams with appropriate dichroic filters
(AHF Analysentechnik) they were directed to a beam-scanning device (mirror
tilting system PSH 10/2, Piezosystem Jena) and then directed into the micro-
scope (DMIRBE, Leica Microsystems). We used an oil immersion objective
(PLAPON 360, NA51.42, Olympus) to focus the laser light to a spot on the
sample and to collect the fluorescence. The beam-scanning device allowed an
exact control of the lateral position of the focal spots on the sample and enabled
scanning. The axial position of the focal spots was adjusted by an objective lens
spot of the STED beam featuring a central zero intensity was produced by
introducing a phase-modifying plate (RPC Photonics) into the beam path,
imprinting on the wave front a helical phase ramp exp(iQ) with 0#Q#2p. A
l/4-plate ensured circular polarization of the STED and of the excitation
beam31,32. The fluorescence was imaged back over the beam-scanning device
and coupled into a multi-mode fibre splitter (Fibre Optic Network
Technology) with an aperture size corresponding to 31.4the magnified excita-
tion spot. The 50:50 split fluorescence signal was then detected by two single-
photon counting units (avalanche photo diode SPCM-AQR-13-FC, Perkin
light or unwanted autofluorescence by appropriate emission filters (AHF
correlator card (Flex02-01D, Correlator.com) for FCS measurements or by a
single-photon counting PC card (SPC 830, Becker & Hickl GmbH) for direct
single-molecule analysis. The focal intensity distribution of the excitation and
STED light were measured by scanning a scattering gold bead of sub-diffraction
diameter (80nm gold colloid, En.GC80, BBinternational) using a non-confocal
detector (MP 963 Photon Counting Module, Perkin Elmer). The (pulse) inten-
sity oftheexcitationlightwas 3–5MWcm22(,25kWcm22on average),stem-
ming from P55–8mW on the confocal spot of diameter (full-width-at-half
Measurements. We analysed the lipid dynamics by placing the foci on random
positions in the lower plasma membrane facing the coverslip and completed all
measurements before disruptive internalization, or any morphological changes
in the cell could take place. The measurement times were kept short (,15s) to
avoid biasingdistortion of thecorrelationdata dueto veryinfrequent transitsof
bright particles such as cell debris. The coverslips were mounted in a special
an objective heater (Bioptechs Inc.) allowed a precise control of the sample’s
temperature. We performed most measurements at 27–37uC in HDMEM as
outlined in detail in the Supplementary Information. We excluded diffusion
of non-integrated lipids (or dye tags) by control measurements in between the
28. Osborn, M., Franke, W. W. & Weber, K. Visualization of a system of filaments
7–10 nm thick in cultured cells of an epithelioid line (PtK2) by
immunofluorescence microscopy. Proc. Natl Acad. Sci. USA 74, 2490–2494
29. Martin, O. C.&Pagano, R. C.Internalization andsorting ofafluorescentanalogue
of glucosylceramide to the Golgi apparatus of human skin fibroblasts: utilization
of endocytic and nonendocytic transport mechanisms. J. Cell Biol. 125, 769–781
30. Schwarzmann, G., Hofmann, P., Pu ¨tz, U. & Albrecht, B. Demonstration of direct
glycosylation of nondegradable glucosylceramide analogsincultured cells. J. Biol.
Chem. 270, 21271–21276 (1995).
31. Keller, J., Scho ¨nle, A. & Hell, S. W. Efficient fluorescence inhibition patterns for
RESOLFT microscopy. Opt. Express 15, 3361–3371 (2007).
32. Willig, K. I. et al. Nanoscale resolution in GFP-based microscopy. Nature Methods
3, 721–723 (2006).
©2008 Macmillan Publishers Limited. All rights reserved