Supporting Information - PNAS · Starbar Japan), fixed in a metal chamber (Attofluor; Life...
11
Supporting Information Ando et al. 10.1073/pnas.1418088112 SI Materials and Methods Experimental Procedures for Raman/Fluorescence Imaging and Spectroscopy. Experimental procedures used for Fig. 2. Supported monolayer mem- branes of single lipid composition, including SM, diyne-SM, DOPC, and chol, were prepared on a quartz substrate (0.17-mm thickness; Starbar Japan), fixed in a metal chamber (Attofluor; Life Tech- nologies), and placed on the sample stage of the microscope (Raman-11; Nanophoton). For Raman spectroscopic measure- ment, the 532-nm laser was focused at a point on the membrane with a water immersion objective lens (CFI Plan Apo IR 60XWI; Nikon). The laser intensity at the sample plane was 340 mW. Exposure time was 6 s. Slit width for the spectrophotometer was set as 50 μm. Raman measurement was performed 15 times at dif- ferent positions on the same membrane. Averaged Raman spectra of each membrane are shown in Fig. 2. The Raman peak of diyne- SM appeared at 2,263 cm −1 . Experimental procedures used for Fig. 3. Supported monolayer mem- brane of diyne-SM/DOPC/chol ternary monolayer with the com- position ratio 1:1:1 was prepared on a quartz substrate (0.17-mm thickness; Starbar Japan). The membrane was fixed in a metal chamber (Attofluor; Life Technologies) and placed on the sample stage of the microscope (Raman-11; Nanophoton). For Raman spectroscopic measurement, the 532-nm laser was focused at a point on the membrane with a water immersion objective lens (CFI Plan Apo IR 60XWI; Nikon). Focus drift was suppressed during imaging by using a real-time feedback system (PFS; Nikon). The laser intensity at the sample plane was 360 mW. Exposure time was 6 s/pixel. Slit width for the spectrophotometer was set as 50 μm. Each Raman spectrum was smoothed using a moving average method. Raman images were reconstructed using peak intensity of diyne at 2,264 cm −1 and peak bottom intensity at 2,222 cm −1 . The Raman image of diyne was obtained after subtraction of the signal intensity between 2,264 and 2,222 cm −1 , whereas the Raman image of the peak bottom (2,222 cm −1 ) was obtained after subtraction of the signal intensity between 2,222 and 2,180 cm −1 . Experimental procedures used for Fig. 4. Supported monolayer mem- branes of diyne-SM/DOPC/chol ternary monolayer with com- position ratios of 1:1:1, 3:7:3, and 1:0:0 were prepared on a quartz substrate (0.17-mm thickness; Starbar Japan). Each membrane on the substrate was fixed in a metal chamber (Attofluor; Life Technologies) and placed on the sample stage of the microscope (Raman-11; Nanophoton). For Raman spectroscopic measure- ment, a 532-nm laser was focused as a point on the membrane with a water immersion objective lens (UPLSAPO 60XW; Olympus). The laser intensity at the sample plane was 370 mW. Exposure time was 6 s/pixel. Slit width for the spectrophotometer was set as 50 μm. Smoothing of Raman data was done using a moving average. Raman images were reconstructed using peak intensity of diyne at 2,263 cm −1 after subtraction of peak bot- tom intensity. Experimental procedures used for Fig. 5. The experimental setup was the same as used previously (1). Briefly, a frequency-doubled Nd:YVO 4 laser (Verdi; Coherent Inc.) was used as the excitation laser source. A line-shaped laser beam formed by a cylindrical lens was focused on the sample by a water immersion objective lens (CFI Plan Apo IR 60XWI; Nikon). Scattering light from the sample was collected by the same objective lens, chromatically separated by an edge filter (LP03-532RU-25; Semrock) to trans- mit Raman scattering light, and focused on the spectrophotometer slit (MK-300; Bunko Keiki) for detection with a cooled CCD camera (Pixis 400B; Princeton Instruments). Laser scanning was performed with a single-axis galvanometer mirror (710-745825, 000-3014016; GSI Lumonics). For Raman imaging in Fig. 5A, a supported 1:1:1 ternary monolayer membrane of diyne-SM/DOPC/chol was prepared on a quartz substrate (0.17-mm thickness; Starbar Japan). The mem- brane on the substrate was fixed in a metal chamber (Attofluor; Life Technologies) and placed on the sample stage of the home-built slit-scanning Raman microscope. For Raman spectroscopic mea- surement, a 532-nm laser was focused as a line on the membrane with a water immersion objective lens (CFI Plan Apo IR 60XWI; Nikon). Focus drift was suppressed during imaging by using a real- time feedback system (PFS; Nikon). The laser intensity at the sample plane was 10.5 mW/μm 2 . Exposure time was 100 s per line. Slit width for the spectrophotometer was set as 50 μm. The scanning number was 240 lines for Raman imaging, with an image size of 240 × 400 pixels. For noise reduction, the obtained Raman data were subjected to singular value decomposition (SVD) (1). We used a spectral region (1,799–2,301 cm −1 ) in the calculation procedure of SVD. We chose four loading vectors that signifi- cantly contributed to the image contrast. After SVD processing, Raman images were reconstructed using the peak intensity of diyne at 2,262 cm −1 based on the averaged peak top intensity between 2,259 and 2,265 cm −1 after subtraction of the averaged peak bottom intensity between 2,223 and 2,229 cm −1 . To com- pensate for aspect ratio, the image in the scanning (horizontal) direction was extended 1.72 times by linear interpolation. The final image consists of 412 × 400 pixels. For Raman and fluorescence imaging in Fig. 5B, a supported 1:1:1 ternary monolayer membrane of diyne-SM/DOPC/chol containing 0.2 mol% Bodipy-PC was prepared on a quartz sub- strate (0.17-mm thickness; Starbar Japan). The membrane on the substrate was fixed in a metal chamber (Attofluor; Life Technologies) and placed on the sample stage of the home-built slit-scanning Raman microscope. For both Raman and fluores- cence imaging, a 532-nm laser was focused as a line on the membrane with a water immersion objective lens (CFI Plan Apo IR 60XWI; Nikon). First, fluorescence imaging was performed, and second, Raman imaging was performed at the same area of the same sample. Fluorescence background was suppressed during Raman imaging by photobleaching of Bodipy-PC under 532-nm laser exposure. Focus drift was suppressed during im- aging by using a real-time feedback system (PFS; Nikon). The laser intensity at the sample plane was 14.1 mW/μm 2 , and the exposure time was 60 s per line for Raman imaging. The laser intensity for fluorescence imaging was 0.3 mW/μm 2 , and the exposure time was 0.5 s per line. Slit width of the spectropho- tometer was set as 50 μm. The scanning number was 225 lines for both Raman and fluorescence imaging, with the image size of 225 × 250 pixels. For noise reduction, the obtained Raman data were subjected to SVD. We used a spectral region (2,000– 2,314 cm −1 ) in the calculation procedure of SVD. We chose four loading vectors that significantly contributed to the image contrast. After SVD processing, the Raman image was reconstructed using the peak intensity of diyne at 2,264 cm −1 based on the averaged peak top intensity between 2,261 and 2,267 cm −1 after subtraction of the averaged peak bottom intensity between 2,225 and 2,231 cm −1 . Fluorescence images were reconstructed using average intensity at 542–603 nm. To compensate for aspect ratio, the image in the scanning (horizontal) direction was extended 1.72 times by linear interpolation. The final image consists of 387 × 250 pixels. Line profiles in Fig. 5C were calculated from the average in- tensity of 4 pixels along the vertical direction, which is indicated Ando et al. www.pnas.org/cgi/content/short/1418088112 1 of 11
Transcript of Supporting Information - PNAS · Starbar Japan), fixed in a metal chamber (Attofluor; Life...
Supporting InformationAndo et al 101073pnas1418088112SI Materials and MethodsExperimental Procedures for RamanFluorescence Imaging andSpectroscopyExperimental procedures used for Fig 2 Supported monolayer mem-branes of single lipid composition including SM diyne-SMDOPCand chol were prepared on a quartz substrate (017-mm thicknessStarbar Japan) fixed in a metal chamber (Attofluor Life Tech-nologies) and placed on the sample stage of the microscope(Raman-11 Nanophoton) For Raman spectroscopic measure-ment the 532-nm laser was focused at a point on the membranewith a water immersion objective lens (CFI Plan Apo IR 60XWINikon) The laser intensity at the sample plane was 340 mWExposure time was 6 s Slit width for the spectrophotometer was setas 50 μm Raman measurement was performed 15 times at dif-ferent positions on the same membrane Averaged Raman spectraof each membrane are shown in Fig 2 The Raman peak of diyne-SM appeared at 2263 cmminus1Experimental procedures used for Fig 3 Supported monolayer mem-brane of diyne-SMDOPCchol ternary monolayer with the com-position ratio 111 was prepared on a quartz substrate (017-mmthickness Starbar Japan) The membrane was fixed in a metalchamber (Attofluor Life Technologies) and placed on the samplestage of the microscope (Raman-11 Nanophoton) For Ramanspectroscopic measurement the 532-nm laser was focused at apoint on the membrane with a water immersion objective lens(CFI Plan Apo IR 60XWI Nikon) Focus drift was suppressedduring imaging by using a real-time feedback system (PFS Nikon)The laser intensity at the sample plane was 360 mW Exposure timewas 6 spixel Slit width for the spectrophotometer was set as 50 μmEach Raman spectrum was smoothed using a moving averagemethod Raman images were reconstructed using peak intensityof diyne at 2264 cmminus1 and peak bottom intensity at 2222 cmminus1 TheRaman image of diyne was obtained after subtraction of the signalintensity between 2264 and 2222 cmminus1 whereas the Raman imageof the peak bottom (2222 cmminus1) was obtained after subtraction ofthe signal intensity between 2222 and 2180 cmminus1Experimental procedures used for Fig 4 Supported monolayer mem-branes of diyne-SMDOPCchol ternary monolayer with com-position ratios of 111 373 and 100 were prepared on a quartzsubstrate (017-mm thickness Starbar Japan) Eachmembrane onthe substrate was fixed in a metal chamber (Attofluor LifeTechnologies) and placed on the sample stage of the microscope(Raman-11 Nanophoton) For Raman spectroscopic measure-ment a 532-nm laser was focused as a point on the membranewith a water immersion objective lens (UPLSAPO 60XWOlympus) The laser intensity at the sample plane was 370 mWExposure time was 6 spixel Slit width for the spectrophotometerwas set as 50 μm Smoothing of Raman data was done using amoving average Raman images were reconstructed using peakintensity of diyne at 2263 cmminus1 after subtraction of peak bot-tom intensityExperimental procedures used for Fig 5 The experimental setup wasthe same as used previously (1) Briefly a frequency-doubledNdYVO4 laser (Verdi Coherent Inc) was used as the excitationlaser source A line-shaped laser beam formed by a cylindricallens was focused on the sample by a water immersion objectivelens (CFI Plan Apo IR 60XWI Nikon) Scattering light from thesample was collected by the same objective lens chromaticallyseparated by an edge filter (LP03-532RU-25 Semrock) to trans-mit Raman scattering light and focused on the spectrophotometerslit (MK-300 Bunko Keiki) for detection with a cooled CCDcamera (Pixis 400B Princeton Instruments) Laser scanning was
performed with a single-axis galvanometer mirror (710-745825000-3014016 GSI Lumonics)For Raman imaging in Fig 5A a supported 111 ternary
monolayer membrane of diyne-SMDOPCchol was prepared ona quartz substrate (017-mm thickness Starbar Japan) The mem-brane on the substrate was fixed in a metal chamber (Attofluor LifeTechnologies) and placed on the sample stage of the home-builtslit-scanning Raman microscope For Raman spectroscopic mea-surement a 532-nm laser was focused as a line on the membranewith a water immersion objective lens (CFI Plan Apo IR 60XWINikon) Focus drift was suppressed during imaging by using a real-time feedback system (PFS Nikon) The laser intensity at thesample plane was 105 mWμm2 Exposure time was 100 s per lineSlit width for the spectrophotometer was set as 50 μm The scanningnumber was 240 lines for Raman imaging with an image size of240 times 400 pixels For noise reduction the obtained Raman datawere subjected to singular value decomposition (SVD) (1) Weused a spectral region (1799ndash2301 cmminus1) in the calculationprocedure of SVD We chose four loading vectors that signifi-cantly contributed to the image contrast After SVD processingRaman images were reconstructed using the peak intensity ofdiyne at 2262 cmminus1 based on the averaged peak top intensitybetween 2259 and 2265 cmminus1 after subtraction of the averagedpeak bottom intensity between 2223 and 2229 cmminus1 To com-pensate for aspect ratio the image in the scanning (horizontal)direction was extended 172 times by linear interpolation Thefinal image consists of 412 times 400 pixelsFor Raman and fluorescence imaging in Fig 5B a supported
111 ternary monolayer membrane of diyne-SMDOPCcholcontaining 02 mol Bodipy-PC was prepared on a quartz sub-strate (017-mm thickness Starbar Japan) The membrane onthe substrate was fixed in a metal chamber (Attofluor LifeTechnologies) and placed on the sample stage of the home-builtslit-scanning Raman microscope For both Raman and fluores-cence imaging a 532-nm laser was focused as a line on themembrane with a water immersion objective lens (CFI Plan ApoIR 60XWI Nikon) First fluorescence imaging was performedand second Raman imaging was performed at the same area ofthe same sample Fluorescence background was suppressedduring Raman imaging by photobleaching of Bodipy-PC under532-nm laser exposure Focus drift was suppressed during im-aging by using a real-time feedback system (PFS Nikon) Thelaser intensity at the sample plane was 141 mWμm2 and theexposure time was 60 s per line for Raman imaging The laserintensity for fluorescence imaging was 03 mWμm2 and theexposure time was 05 s per line Slit width of the spectropho-tometer was set as 50 μm The scanning number was 225 lines forboth Raman and fluorescence imaging with the image size of225 times 250 pixels For noise reduction the obtained Raman data weresubjected to SVD We used a spectral region (2000ndash2314 cmminus1)in the calculation procedure of SVD We chose four loading vectorsthat significantly contributed to the image contrast After SVDprocessing the Raman image was reconstructed using the peakintensity of diyne at 2264 cmminus1 based on the averaged peak topintensity between 2261 and 2267 cmminus1 after subtraction of theaveraged peak bottom intensity between 2225 and 2231 cmminus1Fluorescence images were reconstructed using average intensity at542ndash603 nm To compensate for aspect ratio the image in thescanning (horizontal) direction was extended 172 times by linearinterpolation The final image consists of 387 times 250 pixelsLine profiles in Fig 5C were calculated from the average in-
tensity of 4 pixels along the vertical direction which is indicated
Ando et al wwwpnasorgcgicontentshort1418088112 1 of 11
by a red dotted line in the Raman image in Fig 5B Upper and agray dotted line in the fluorescence image in Fig 5B Lower Theprofile from the Raman image was smoothed by use of themoving averageExperimental procedures used for Fig S5 In Fig S5A the experi-mental setup was the same as used in Fig 5 For Raman imaginga supported membrane of diyne-SM monolayer was preparedon a quartz substrate (017-mm thickness Starbar Japan) Themembrane on the substrate was fixed in a metal chamber (At-tofluor Life Technologies) and placed on the sample stage ofthe home-built slit-scanning Raman microscope For Ramanspectroscopic measurement a 532-nm laser was focused as a lineon the membrane with a water immersion objective lens (CFIPlan Apo IR 60XWI Nikon) Focus drift was suppressed duringimaging by using a real-time feedback system (PFS Nikon) Thelaser intensity at the sample plane was 139 mWμm2 Exposuretime was 60 s per line Slit width for the spectrophotometer wasset as 50 μm The scanning number was 200 lines for Ramanimaging with an image size of 200 times 190 pixels For noise re-duction the obtained Raman data were subjected to SVD (1)We used a spectral region (1800ndash2310 cmminus1) in the calculationprocedure of SVD We chose four loading vectors that signifi-cantly contributed to the image contrast After SVD processingRaman images were reconstructed using the peak intensity ofdiyne at 2263 cmminus1 based on the averaged peak top intensitybetween 2260 and 2266 cmminus1 after subtraction of the averagedpeak bottom intensity between 2212 and 2218 cmminus1 To com-pensate for aspect ratio the image in the scanning (horizontal)direction was extended 172 times by linear interpolation Thefinal image consists of 344 times 190 pixels In Fig S5B the imagewas extracted from lower left portion in Fig 5A which consist of344 times 190 pixels In Fig S5C line profiles were calculated fromthe average intensity of 7 pixels along the vertical directionwhich is indicated by a dotted linesExperimental procedures used for Fig S6 A supported monolayermembrane of 111 diyne-SMDOPCchol ternary monolayercontaining 02 mol Bodipy-PC was prepared on a quartz sub-strate (017-mm thickness Starbar Japan) The membrane on thesubstrate was fixed in a metal chamber (Attofluor Life Tech-nologies) and placed on the sample stage of the microscope(Raman-11 Nanophoton) For both fluorescence and Ramanspectroscopic measurement a 532-nm laser was focused as apoint on the membrane with a water immersion objective lens(CFI Plan Apo IR 60XWI Nikon) Focus drift was suppressedduring imaging by using a real-time feedback system (PFSNikon) For fluorescence imaging the laser intensity at thesample plane was 05 mW and the exposure time was 1 spixelFor Raman imaging the laser intensity at the sample was 360 mWand the exposure time was 6 spixel First fluorescence imaging wasperformed and second Raman imaging was performed at thesame area of the same sample Slit width for the spectrophotometerwas set as 50 μm For reconstruction of fluorescence images av-erage intensity between 555 and 605 nm was used Smoothing ofRaman data was done using a moving average Raman imageswere reconstructed using peak intensity of diyne at 2264 cmminus1
after subtraction of peak bottom intensity
DSC Thermograms of Diyne-SM Bilayers The thermal-phase be-havior of SM bilayers was examined with a nanodifferentialscanning calorimeter (Calorimetry Science Corp) Bilayer sam-ples were prepared by a conventional method Briefly SM dis-solved in chloroformmethanol (41) was dried under a flowof nitrogen and then reduced pressure for at least 24 h Theresulting lipid film was dispersed into distilled and deionizedwater (Simplicity UV Merck Millipore) and incubated forsim30 min at 60 degC with intermittent vortexing The final con-centration was 247 mM Then 330 μL sample was placed inthe DSC immediately before measurement A scanning rate of
05 degCmin was used The main transition temperatures of diyne-SM and SM bilayers were found to be 395 degC and 445 degC re-spectively as shown by the arrows in Fig S7
π-A Isotherm Measurements and Supported Monolayer PreparationMonolayers of lipid mixtures were prepared on a computer-controlled Langmuir film balance (USI System) calibrated usingstearic acid (Sigma Aldrich) The subphase which consisted ofdistilled freshly deionized water was obtained using a Milli-QSystem The sample solution was prepared by mixing the ap-propriate amount of each lipid solution in a microvial A total of30 μL lipid dissolved in chloroformmethanol (41 volvol 1 mgmL)was spread onto the aqueous subphase (100 times 290 mm2) using aglass micropipette (Drummond Scientific Company) After aninitial delay period of 10 min for evaporation of the organicsolvent the monolayers were compressed at a rate of 20 mm2sThe subphase and ambient temperatures were controlled at250 degC plusmn 01 degC and 25 degC plusmn 2 degC respectively The π-A isothermmeasurements were repeated three to five times under the sameconditions These measurements provided the molecular area ata corresponding pressure within an error of sim plusmn1 Aring2 The in-fluence of oxidation on the unsaturated chains at the airndashwaterinterface was checked by intentionally exposing pure SM andpure DOPC monolayers to air for 10ndash30 min before compres-sion The change in the isotherm after prolonged exposure of SMor DOPC monolayer to air was within the error described above
Analysis In Fig 6 we evaluated the intermolecular interaction inlipid binary mixtures at the surface pressure of 5 mNm on thebasis of the deviations of experimentally obtained mean molec-ular areas (Amean) from those of ideal mixtures (A12)
A12 =A1eth1minus xTHORN+A2x
where A1 and A2 are the molecular areas of pure components 1and 2 (eg SM and chol) respectively and x is the molar frac-tion of component 2 Thus the value of A12 corresponds to themean molecular area in the mixture constituted of noninterac-tive or completely immiscible molecules According to previousliterature (2) PMAs of components 1 (A1
PMA) and 2 (A2PMA) can
be defined as
A1PMA =
partNA1
mean
partN1
and
A2PMA =
partNA2
mean
partN2
where N N1 and N2 are the total amounts of all constituents andcomponents 1 and 2 respectively On the basis of the additivityrule the Amean also can be expressed as
AmeanethxTHORN= eth1minus xTHORNA1PMAethxTHORN+ xA2
PMAethxTHORN
Here denoting derivatives with respect to x by prime yields thefollowing equations are obtained
A1PMAethxTHORN=AmeanethxTHORNminus xAprimemeanethxTHORN
and
A2PMAethxTHORN=AmeanethxTHORN+ eth1minus xTHORNAprimemeanethxTHORN
Ando et al wwwpnasorgcgicontentshort1418088112 2 of 11
Areal compressibility (Cs) at the surface pressure of 5 mNmwas calculated from the π-A isotherm using
Cs =minus1
Amean
partAmean
partπ
π
The compressibility in ideal mixtures (C12) is calculated accord-ing to Ali et al (3)
C12 =
1A12
fethCs1A1THORNeth1minus xTHORN+ ethCs2A2THORNxg
where Cs1 and Cs2 are the areal compressibilities of the purecomponents 1 and 2 respectively They suggested that C12 isadditive with respect to the product of Csi and Ai rather thanCsi for either ideal or completely nonideal mixing Areal com-pressibility (Cs) was expressed in term of areal compressionalmodulus (Cs
minus1) for easy comparison with previous data
Fluorescence Observation of Ordered Domains in Diyne-SMDOPCChol and SMDOPCChol Ternary Monolayers Fig S8 shows fluo-rescence images of diyne-SMDOPCchol (111 molmolmol)(Fig S8A) and SMDOPCchol (111 molmolmol) (Fig S8B)quartz-supported monolayers in the presence of 02 mol Bodipy-PC at 12 mNm and 25 degC Fluorescence observations were con-ducted using a confocal laser-scanning microscope (FV1000-DIX81 Olympus) with an air objective lens with a long workingdistance (LUCPLFLN 60X Olympus) A wavelength of 488 nmwas used for excitation of Bodipy A laser-scanning rate of 40 or80 μspixel was used for acquisition of confocal images (1024 times1024 pixels)
2H NMR Measurements A mixture of lipids comprising 100 μmold2-SM or d2-diyne-SM (Fig 7) 100 μmol chol and 100 μmolDOPC was dissolved in chloroformmethanol (11 volvol) Afterremoving the solvent in vacuo for 20 h the dried membrane filmwas hydrated with 1 mL distilled water and vigorously vortexedat 65 degC to make multilamellar vesicles The sample was frozen andthawed three times lyophilized and rehydrated with deuterium-depleted water to make 50 (wtwt) water Then the mixturewas again frozen and thawed 10 times The sample was trans-ferred into a glass tube (5 times 26 mm) which was sealed with epoxyglue 2H NMR measurements were recorded on a 300-MHzCMX300 Spectrometer (Chemagnetics Varian) with a 5-mm 2Hstatic probe (Otsuka Electronics) using a quadrupolar echo se-quence (4) The 90deg pulse width was 2 μs interpulse delay was30 μs and repetition rate was 05 s The sweep width was 200 kHzand the number of scans was around 100000
General Information for the Synthesis of Diyne-SM Chemicals andsolvents were purchased from Nacalai Tesque Aldrich TCI orKanto Chemicals Inc and used without additional purificationunless otherwise noted TLC was done onMerck Precoated SilicaGel 60 F-254 Plates Spots on TLC plates were stained withphosphomolybdic acid NMR spectra were collected on a JEOLECA 500 (500 MHz) using deuterated solvent as the lockChemical shift is given in parts per million (δ) and couplingconstant (J) is in hertz The following terms are used to designatemultiplicity singlet (s) doublet (d) triplet (t) quartet (q) quin-tuplet (quint) multiplet (m) and broad (b) High-resolution massspectra (HRMS) were recorded on an LTQ-Orbitrap XL
Synthetic Procedures for Diyne-SM and d2-Diyne-SM(2S3RE)-(2-stearoylamino-3-hydroxyoctadec-4-en-1-yl)-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (diyne-SM)To a solutionof propargyl alcohol S1 (300 mg 535 mmol) in acetone (15 mL)N-bromosuccinimide (102 g 578 mmol) and silver nitrate (91 mg054 mmol) were added at room temperature The reaction mix-ture was stirred at room temperature for 2 h and then concen-trated The residue was extracted two times with diethyl ether Thecombined organic layers were dried with Na2SO4 filtered andconcentrated to afford S2 (645 mg 90 yield) as a pale yellow oilwhich was used directly in the next step 1H NMR (500 MHzCDCl3) δ 163 (brs 1H) 430 (s 2H) CuCl (03 mg 31 μmol)i-PrNH2 (51 μL 006 mmol) and NH2OHHCl (10 mg 145 μmol)were added to MeOH (1 mL) at room temperature under argonThe mixture was cooled to 0 degC and then S3 (5) (compound S3was the intermediate for the synthesis of SM head group analogs40 mg 005 mmol) was added forming a yellow acetylide suspen-sion A solution of S2 (21 mg 016 mmol) in MeOH (02 mL) wasadded immediately The reaction mixture was stirred at the sametemperature for 30 min concentrated and extracted with CHCl3The combined organic layers were dried with Na2SO4 filtered andconcentrated The mixture was passed through a short bed of silicagel eluting with CHCl3MeOH (51) The crude product was puri-fied on a Cosmosil 5C18-AR-II column (10 times 150 mm NacalaiTesque) with MeOH as the eluent to give diyne-SM (51 mg 12)as a white solid TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 124ndash142(m 50H) 153ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H) 213ndash220 (m 2H) 324 (s 6H) 370 (t J = 55 Hz 2H) 387ndash398(m 2H) 403 (dd J = 80 80 Hz 1H) 406ndash413 (m 1H) 427(s 2H) 425ndash430 (m 2H) 456 (s 2H) 543 (ddt J = 155 8015 Hz 1H) 569 (dtd J = 155 70 10 Hz 1H) 790 (d J = 90 Hz1H) 13C NMR (125 MHz CD3OD) δ 1311 2241 2585 29152918 2937 2944 2974 3175 3215 3605 4955 5086 53935579 5890 6422 6461 6564 6654 7122 7542 8127 1298913382 17458 HRMS (electrospray ionization) calculated forC46H86N2O7P [M + H]+ 8096167 found 8096184 1H and 13CNMR spectra of diyne-SM are shown in Figs S9 and S10 Thesynthetic scheme is summarized in Scheme S1(2S3RE)-[2-(1010-dideuteriumstearoylamino)-3-hydroxyoctadec-4-en-1-yl]-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (d2-diyne-SM)This compound was prepared from S4 (a full account of thesynthesis of compound S4 will be reported elsewhere) in 14yield as a white solid by following the same procedure as de-scribed for S3 TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 122ndash143 (m 48H) 150ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H)213ndash220 (m 2H) 325 (s 6H) 371 (t J = 50 Hz 2H) 388ndash399(m 2H) 402 (dd J = 80 80 Hz 1H) 406ndash412 (m 1H) 427(s 2H) 421ndash434 (m 2H) 456 (s 2H) 543 (ddt J = 150 8012 Hz 1H) 569 (dtd J = 150 70 06 Hz 1H) 790 (d J = 90Hz 1H) 13C NMR (125 MHz CD3OD) δ 1316 2243 25872918 2921 2933 2947 2952 2955 2958 3177 3179 32183606 4956 5087 5387 5393 5580 5890 5994 6422 64286461 6465 6567 6656 7121 7543 8129 12993 1337917454 HRMS (electrospray ionization) calculated forC46H83D2N2O7PNa [M + Na]+ 8336112 found 8336124 1Hand 13C NMR spectra of d2-diyne-SM are shown in Figs S11and S12 Synthetic scheme is summarized in Scheme S2
1 Palonpon AF et al (2013) Raman and SERS microscopy for molecular imaging of livecells Nat Protoc 8(4)677ndash692
2 Edholm O Nagle JF (2005) Areas of molecules in membranes consisting of mixturesBiophys J 89(3)1827ndash1832
3 Ali S Smaby JM Brockman HL Brown RE (1994) Cholesterolrsquos interfacial interactionswith galactosylceramides Biochemistry 33(10)2900ndash2906
4 Davis JH Jeffrey KR Bloom M Valic MI Higgs TP (1976) Quadrupolar echo deuteronmagnetic resonance spectroscopy in ordered hydrocarbon chains Chem Phys Lett42(2)390ndash394
5 Goretta SA Kinoshita M Mori S Tsuchikawa H Matsumori N Murata M (2012) Effectsof chemical modification of sphingomyelin ammonium group on formation of liquid-ordered phase Bioorg Med Chem 20(13)4012ndash4019
Ando et al wwwpnasorgcgicontentshort1418088112 3 of 11
Fig S1 Raman spectrum of the diyne-SM monolayer supported on a quartz substrate shown in Fig 2A Raman peaks of N2 O2 and diyne are indicated by redarrows Lower is a five times enlarged view of Upper in terms of intensity (vertical axis) Strong Raman scattering by N2 and O2 in air was always detectedduring Raman measurement of lipid membranes For ease of picking out the important Raman peaks such as diyne the two peaks of N2 and O2 have beenremoved in Figs 2 and 4 and Figs S3 and S4
Fig S2 (A) Raman image of the diyne-SMDOPCchol ternary monolayer with a 111 ratio reconstructed using the intensity of the diyne peak at 2263 cmminus1The image is the same as Fig 4A Mean value of the image is 564 The two phases were separated by taking pixel regions (B) below the mean value asdisordered domains and (C) above the mean value as ordered domains Raman images of diyne-SMDOPCchol ternary monolayers with (D) a 373 ratio and(E) a 100 ratio The images are the same as Fig 4 B and C respectively (Scale bar 10 μm)
Ando et al wwwpnasorgcgicontentshort1418088112 4 of 11
Fig S3 Raman spectra obtained along the red lines in the Raman images of the diyne-SMDOPCchol ternary monolayer with 111 373 and 100 ratios Theimages are the same as in Fig 4 AndashC with 36 times 24 pixels Raman peak of N2 at sim2330 cmminus1 was removed so that it would be easier to see Raman peaks fromlipid molecules Pix pixel (Scale bar 10 μm)
Fig S4 Comparison between Raman spectra of supported lipid monolayers with single-component composition (diyne-SM DOPC or chol) and Raman spectraof diyne-SMDOPCchol ternary monolayer at ordered and disordered domains The spectra of the ternary monolayer were obtained from five representativeareas (3 times 3 pixels) calculated using the Raman image in Fig 4A Characteristic Raman peaks are displayed in the image including diyne stretching vibration(2263 cmminus1) CH2 symmetricasymmetric stretching vibration (28502883 cmminus1) and CH3 symmetric stretching vibration (2936 cmminus1) Raman peak of N2 atsim2330 cmminus1 was removed so that it would be easier to see Raman peaks from lipid molecules
Ando et al wwwpnasorgcgicontentshort1418088112 5 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
by a red dotted line in the Raman image in Fig 5B Upper and agray dotted line in the fluorescence image in Fig 5B Lower Theprofile from the Raman image was smoothed by use of themoving averageExperimental procedures used for Fig S5 In Fig S5A the experi-mental setup was the same as used in Fig 5 For Raman imaginga supported membrane of diyne-SM monolayer was preparedon a quartz substrate (017-mm thickness Starbar Japan) Themembrane on the substrate was fixed in a metal chamber (At-tofluor Life Technologies) and placed on the sample stage ofthe home-built slit-scanning Raman microscope For Ramanspectroscopic measurement a 532-nm laser was focused as a lineon the membrane with a water immersion objective lens (CFIPlan Apo IR 60XWI Nikon) Focus drift was suppressed duringimaging by using a real-time feedback system (PFS Nikon) Thelaser intensity at the sample plane was 139 mWμm2 Exposuretime was 60 s per line Slit width for the spectrophotometer wasset as 50 μm The scanning number was 200 lines for Ramanimaging with an image size of 200 times 190 pixels For noise re-duction the obtained Raman data were subjected to SVD (1)We used a spectral region (1800ndash2310 cmminus1) in the calculationprocedure of SVD We chose four loading vectors that signifi-cantly contributed to the image contrast After SVD processingRaman images were reconstructed using the peak intensity ofdiyne at 2263 cmminus1 based on the averaged peak top intensitybetween 2260 and 2266 cmminus1 after subtraction of the averagedpeak bottom intensity between 2212 and 2218 cmminus1 To com-pensate for aspect ratio the image in the scanning (horizontal)direction was extended 172 times by linear interpolation Thefinal image consists of 344 times 190 pixels In Fig S5B the imagewas extracted from lower left portion in Fig 5A which consist of344 times 190 pixels In Fig S5C line profiles were calculated fromthe average intensity of 7 pixels along the vertical directionwhich is indicated by a dotted linesExperimental procedures used for Fig S6 A supported monolayermembrane of 111 diyne-SMDOPCchol ternary monolayercontaining 02 mol Bodipy-PC was prepared on a quartz sub-strate (017-mm thickness Starbar Japan) The membrane on thesubstrate was fixed in a metal chamber (Attofluor Life Tech-nologies) and placed on the sample stage of the microscope(Raman-11 Nanophoton) For both fluorescence and Ramanspectroscopic measurement a 532-nm laser was focused as apoint on the membrane with a water immersion objective lens(CFI Plan Apo IR 60XWI Nikon) Focus drift was suppressedduring imaging by using a real-time feedback system (PFSNikon) For fluorescence imaging the laser intensity at thesample plane was 05 mW and the exposure time was 1 spixelFor Raman imaging the laser intensity at the sample was 360 mWand the exposure time was 6 spixel First fluorescence imaging wasperformed and second Raman imaging was performed at thesame area of the same sample Slit width for the spectrophotometerwas set as 50 μm For reconstruction of fluorescence images av-erage intensity between 555 and 605 nm was used Smoothing ofRaman data was done using a moving average Raman imageswere reconstructed using peak intensity of diyne at 2264 cmminus1
after subtraction of peak bottom intensity
DSC Thermograms of Diyne-SM Bilayers The thermal-phase be-havior of SM bilayers was examined with a nanodifferentialscanning calorimeter (Calorimetry Science Corp) Bilayer sam-ples were prepared by a conventional method Briefly SM dis-solved in chloroformmethanol (41) was dried under a flowof nitrogen and then reduced pressure for at least 24 h Theresulting lipid film was dispersed into distilled and deionizedwater (Simplicity UV Merck Millipore) and incubated forsim30 min at 60 degC with intermittent vortexing The final con-centration was 247 mM Then 330 μL sample was placed inthe DSC immediately before measurement A scanning rate of
05 degCmin was used The main transition temperatures of diyne-SM and SM bilayers were found to be 395 degC and 445 degC re-spectively as shown by the arrows in Fig S7
π-A Isotherm Measurements and Supported Monolayer PreparationMonolayers of lipid mixtures were prepared on a computer-controlled Langmuir film balance (USI System) calibrated usingstearic acid (Sigma Aldrich) The subphase which consisted ofdistilled freshly deionized water was obtained using a Milli-QSystem The sample solution was prepared by mixing the ap-propriate amount of each lipid solution in a microvial A total of30 μL lipid dissolved in chloroformmethanol (41 volvol 1 mgmL)was spread onto the aqueous subphase (100 times 290 mm2) using aglass micropipette (Drummond Scientific Company) After aninitial delay period of 10 min for evaporation of the organicsolvent the monolayers were compressed at a rate of 20 mm2sThe subphase and ambient temperatures were controlled at250 degC plusmn 01 degC and 25 degC plusmn 2 degC respectively The π-A isothermmeasurements were repeated three to five times under the sameconditions These measurements provided the molecular area ata corresponding pressure within an error of sim plusmn1 Aring2 The in-fluence of oxidation on the unsaturated chains at the airndashwaterinterface was checked by intentionally exposing pure SM andpure DOPC monolayers to air for 10ndash30 min before compres-sion The change in the isotherm after prolonged exposure of SMor DOPC monolayer to air was within the error described above
Analysis In Fig 6 we evaluated the intermolecular interaction inlipid binary mixtures at the surface pressure of 5 mNm on thebasis of the deviations of experimentally obtained mean molec-ular areas (Amean) from those of ideal mixtures (A12)
A12 =A1eth1minus xTHORN+A2x
where A1 and A2 are the molecular areas of pure components 1and 2 (eg SM and chol) respectively and x is the molar frac-tion of component 2 Thus the value of A12 corresponds to themean molecular area in the mixture constituted of noninterac-tive or completely immiscible molecules According to previousliterature (2) PMAs of components 1 (A1
PMA) and 2 (A2PMA) can
be defined as
A1PMA =
partNA1
mean
partN1
and
A2PMA =
partNA2
mean
partN2
where N N1 and N2 are the total amounts of all constituents andcomponents 1 and 2 respectively On the basis of the additivityrule the Amean also can be expressed as
AmeanethxTHORN= eth1minus xTHORNA1PMAethxTHORN+ xA2
PMAethxTHORN
Here denoting derivatives with respect to x by prime yields thefollowing equations are obtained
A1PMAethxTHORN=AmeanethxTHORNminus xAprimemeanethxTHORN
and
A2PMAethxTHORN=AmeanethxTHORN+ eth1minus xTHORNAprimemeanethxTHORN
Ando et al wwwpnasorgcgicontentshort1418088112 2 of 11
Areal compressibility (Cs) at the surface pressure of 5 mNmwas calculated from the π-A isotherm using
Cs =minus1
Amean
partAmean
partπ
π
The compressibility in ideal mixtures (C12) is calculated accord-ing to Ali et al (3)
C12 =
1A12
fethCs1A1THORNeth1minus xTHORN+ ethCs2A2THORNxg
where Cs1 and Cs2 are the areal compressibilities of the purecomponents 1 and 2 respectively They suggested that C12 isadditive with respect to the product of Csi and Ai rather thanCsi for either ideal or completely nonideal mixing Areal com-pressibility (Cs) was expressed in term of areal compressionalmodulus (Cs
minus1) for easy comparison with previous data
Fluorescence Observation of Ordered Domains in Diyne-SMDOPCChol and SMDOPCChol Ternary Monolayers Fig S8 shows fluo-rescence images of diyne-SMDOPCchol (111 molmolmol)(Fig S8A) and SMDOPCchol (111 molmolmol) (Fig S8B)quartz-supported monolayers in the presence of 02 mol Bodipy-PC at 12 mNm and 25 degC Fluorescence observations were con-ducted using a confocal laser-scanning microscope (FV1000-DIX81 Olympus) with an air objective lens with a long workingdistance (LUCPLFLN 60X Olympus) A wavelength of 488 nmwas used for excitation of Bodipy A laser-scanning rate of 40 or80 μspixel was used for acquisition of confocal images (1024 times1024 pixels)
2H NMR Measurements A mixture of lipids comprising 100 μmold2-SM or d2-diyne-SM (Fig 7) 100 μmol chol and 100 μmolDOPC was dissolved in chloroformmethanol (11 volvol) Afterremoving the solvent in vacuo for 20 h the dried membrane filmwas hydrated with 1 mL distilled water and vigorously vortexedat 65 degC to make multilamellar vesicles The sample was frozen andthawed three times lyophilized and rehydrated with deuterium-depleted water to make 50 (wtwt) water Then the mixturewas again frozen and thawed 10 times The sample was trans-ferred into a glass tube (5 times 26 mm) which was sealed with epoxyglue 2H NMR measurements were recorded on a 300-MHzCMX300 Spectrometer (Chemagnetics Varian) with a 5-mm 2Hstatic probe (Otsuka Electronics) using a quadrupolar echo se-quence (4) The 90deg pulse width was 2 μs interpulse delay was30 μs and repetition rate was 05 s The sweep width was 200 kHzand the number of scans was around 100000
General Information for the Synthesis of Diyne-SM Chemicals andsolvents were purchased from Nacalai Tesque Aldrich TCI orKanto Chemicals Inc and used without additional purificationunless otherwise noted TLC was done onMerck Precoated SilicaGel 60 F-254 Plates Spots on TLC plates were stained withphosphomolybdic acid NMR spectra were collected on a JEOLECA 500 (500 MHz) using deuterated solvent as the lockChemical shift is given in parts per million (δ) and couplingconstant (J) is in hertz The following terms are used to designatemultiplicity singlet (s) doublet (d) triplet (t) quartet (q) quin-tuplet (quint) multiplet (m) and broad (b) High-resolution massspectra (HRMS) were recorded on an LTQ-Orbitrap XL
Synthetic Procedures for Diyne-SM and d2-Diyne-SM(2S3RE)-(2-stearoylamino-3-hydroxyoctadec-4-en-1-yl)-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (diyne-SM)To a solutionof propargyl alcohol S1 (300 mg 535 mmol) in acetone (15 mL)N-bromosuccinimide (102 g 578 mmol) and silver nitrate (91 mg054 mmol) were added at room temperature The reaction mix-ture was stirred at room temperature for 2 h and then concen-trated The residue was extracted two times with diethyl ether Thecombined organic layers were dried with Na2SO4 filtered andconcentrated to afford S2 (645 mg 90 yield) as a pale yellow oilwhich was used directly in the next step 1H NMR (500 MHzCDCl3) δ 163 (brs 1H) 430 (s 2H) CuCl (03 mg 31 μmol)i-PrNH2 (51 μL 006 mmol) and NH2OHHCl (10 mg 145 μmol)were added to MeOH (1 mL) at room temperature under argonThe mixture was cooled to 0 degC and then S3 (5) (compound S3was the intermediate for the synthesis of SM head group analogs40 mg 005 mmol) was added forming a yellow acetylide suspen-sion A solution of S2 (21 mg 016 mmol) in MeOH (02 mL) wasadded immediately The reaction mixture was stirred at the sametemperature for 30 min concentrated and extracted with CHCl3The combined organic layers were dried with Na2SO4 filtered andconcentrated The mixture was passed through a short bed of silicagel eluting with CHCl3MeOH (51) The crude product was puri-fied on a Cosmosil 5C18-AR-II column (10 times 150 mm NacalaiTesque) with MeOH as the eluent to give diyne-SM (51 mg 12)as a white solid TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 124ndash142(m 50H) 153ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H) 213ndash220 (m 2H) 324 (s 6H) 370 (t J = 55 Hz 2H) 387ndash398(m 2H) 403 (dd J = 80 80 Hz 1H) 406ndash413 (m 1H) 427(s 2H) 425ndash430 (m 2H) 456 (s 2H) 543 (ddt J = 155 8015 Hz 1H) 569 (dtd J = 155 70 10 Hz 1H) 790 (d J = 90 Hz1H) 13C NMR (125 MHz CD3OD) δ 1311 2241 2585 29152918 2937 2944 2974 3175 3215 3605 4955 5086 53935579 5890 6422 6461 6564 6654 7122 7542 8127 1298913382 17458 HRMS (electrospray ionization) calculated forC46H86N2O7P [M + H]+ 8096167 found 8096184 1H and 13CNMR spectra of diyne-SM are shown in Figs S9 and S10 Thesynthetic scheme is summarized in Scheme S1(2S3RE)-[2-(1010-dideuteriumstearoylamino)-3-hydroxyoctadec-4-en-1-yl]-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (d2-diyne-SM)This compound was prepared from S4 (a full account of thesynthesis of compound S4 will be reported elsewhere) in 14yield as a white solid by following the same procedure as de-scribed for S3 TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 122ndash143 (m 48H) 150ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H)213ndash220 (m 2H) 325 (s 6H) 371 (t J = 50 Hz 2H) 388ndash399(m 2H) 402 (dd J = 80 80 Hz 1H) 406ndash412 (m 1H) 427(s 2H) 421ndash434 (m 2H) 456 (s 2H) 543 (ddt J = 150 8012 Hz 1H) 569 (dtd J = 150 70 06 Hz 1H) 790 (d J = 90Hz 1H) 13C NMR (125 MHz CD3OD) δ 1316 2243 25872918 2921 2933 2947 2952 2955 2958 3177 3179 32183606 4956 5087 5387 5393 5580 5890 5994 6422 64286461 6465 6567 6656 7121 7543 8129 12993 1337917454 HRMS (electrospray ionization) calculated forC46H83D2N2O7PNa [M + Na]+ 8336112 found 8336124 1Hand 13C NMR spectra of d2-diyne-SM are shown in Figs S11and S12 Synthetic scheme is summarized in Scheme S2
1 Palonpon AF et al (2013) Raman and SERS microscopy for molecular imaging of livecells Nat Protoc 8(4)677ndash692
2 Edholm O Nagle JF (2005) Areas of molecules in membranes consisting of mixturesBiophys J 89(3)1827ndash1832
3 Ali S Smaby JM Brockman HL Brown RE (1994) Cholesterolrsquos interfacial interactionswith galactosylceramides Biochemistry 33(10)2900ndash2906
4 Davis JH Jeffrey KR Bloom M Valic MI Higgs TP (1976) Quadrupolar echo deuteronmagnetic resonance spectroscopy in ordered hydrocarbon chains Chem Phys Lett42(2)390ndash394
5 Goretta SA Kinoshita M Mori S Tsuchikawa H Matsumori N Murata M (2012) Effectsof chemical modification of sphingomyelin ammonium group on formation of liquid-ordered phase Bioorg Med Chem 20(13)4012ndash4019
Ando et al wwwpnasorgcgicontentshort1418088112 3 of 11
Fig S1 Raman spectrum of the diyne-SM monolayer supported on a quartz substrate shown in Fig 2A Raman peaks of N2 O2 and diyne are indicated by redarrows Lower is a five times enlarged view of Upper in terms of intensity (vertical axis) Strong Raman scattering by N2 and O2 in air was always detectedduring Raman measurement of lipid membranes For ease of picking out the important Raman peaks such as diyne the two peaks of N2 and O2 have beenremoved in Figs 2 and 4 and Figs S3 and S4
Fig S2 (A) Raman image of the diyne-SMDOPCchol ternary monolayer with a 111 ratio reconstructed using the intensity of the diyne peak at 2263 cmminus1The image is the same as Fig 4A Mean value of the image is 564 The two phases were separated by taking pixel regions (B) below the mean value asdisordered domains and (C) above the mean value as ordered domains Raman images of diyne-SMDOPCchol ternary monolayers with (D) a 373 ratio and(E) a 100 ratio The images are the same as Fig 4 B and C respectively (Scale bar 10 μm)
Ando et al wwwpnasorgcgicontentshort1418088112 4 of 11
Fig S3 Raman spectra obtained along the red lines in the Raman images of the diyne-SMDOPCchol ternary monolayer with 111 373 and 100 ratios Theimages are the same as in Fig 4 AndashC with 36 times 24 pixels Raman peak of N2 at sim2330 cmminus1 was removed so that it would be easier to see Raman peaks fromlipid molecules Pix pixel (Scale bar 10 μm)
Fig S4 Comparison between Raman spectra of supported lipid monolayers with single-component composition (diyne-SM DOPC or chol) and Raman spectraof diyne-SMDOPCchol ternary monolayer at ordered and disordered domains The spectra of the ternary monolayer were obtained from five representativeareas (3 times 3 pixels) calculated using the Raman image in Fig 4A Characteristic Raman peaks are displayed in the image including diyne stretching vibration(2263 cmminus1) CH2 symmetricasymmetric stretching vibration (28502883 cmminus1) and CH3 symmetric stretching vibration (2936 cmminus1) Raman peak of N2 atsim2330 cmminus1 was removed so that it would be easier to see Raman peaks from lipid molecules
Ando et al wwwpnasorgcgicontentshort1418088112 5 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Areal compressibility (Cs) at the surface pressure of 5 mNmwas calculated from the π-A isotherm using
Cs =minus1
Amean
partAmean
partπ
π
The compressibility in ideal mixtures (C12) is calculated accord-ing to Ali et al (3)
C12 =
1A12
fethCs1A1THORNeth1minus xTHORN+ ethCs2A2THORNxg
where Cs1 and Cs2 are the areal compressibilities of the purecomponents 1 and 2 respectively They suggested that C12 isadditive with respect to the product of Csi and Ai rather thanCsi for either ideal or completely nonideal mixing Areal com-pressibility (Cs) was expressed in term of areal compressionalmodulus (Cs
minus1) for easy comparison with previous data
Fluorescence Observation of Ordered Domains in Diyne-SMDOPCChol and SMDOPCChol Ternary Monolayers Fig S8 shows fluo-rescence images of diyne-SMDOPCchol (111 molmolmol)(Fig S8A) and SMDOPCchol (111 molmolmol) (Fig S8B)quartz-supported monolayers in the presence of 02 mol Bodipy-PC at 12 mNm and 25 degC Fluorescence observations were con-ducted using a confocal laser-scanning microscope (FV1000-DIX81 Olympus) with an air objective lens with a long workingdistance (LUCPLFLN 60X Olympus) A wavelength of 488 nmwas used for excitation of Bodipy A laser-scanning rate of 40 or80 μspixel was used for acquisition of confocal images (1024 times1024 pixels)
2H NMR Measurements A mixture of lipids comprising 100 μmold2-SM or d2-diyne-SM (Fig 7) 100 μmol chol and 100 μmolDOPC was dissolved in chloroformmethanol (11 volvol) Afterremoving the solvent in vacuo for 20 h the dried membrane filmwas hydrated with 1 mL distilled water and vigorously vortexedat 65 degC to make multilamellar vesicles The sample was frozen andthawed three times lyophilized and rehydrated with deuterium-depleted water to make 50 (wtwt) water Then the mixturewas again frozen and thawed 10 times The sample was trans-ferred into a glass tube (5 times 26 mm) which was sealed with epoxyglue 2H NMR measurements were recorded on a 300-MHzCMX300 Spectrometer (Chemagnetics Varian) with a 5-mm 2Hstatic probe (Otsuka Electronics) using a quadrupolar echo se-quence (4) The 90deg pulse width was 2 μs interpulse delay was30 μs and repetition rate was 05 s The sweep width was 200 kHzand the number of scans was around 100000
General Information for the Synthesis of Diyne-SM Chemicals andsolvents were purchased from Nacalai Tesque Aldrich TCI orKanto Chemicals Inc and used without additional purificationunless otherwise noted TLC was done onMerck Precoated SilicaGel 60 F-254 Plates Spots on TLC plates were stained withphosphomolybdic acid NMR spectra were collected on a JEOLECA 500 (500 MHz) using deuterated solvent as the lockChemical shift is given in parts per million (δ) and couplingconstant (J) is in hertz The following terms are used to designatemultiplicity singlet (s) doublet (d) triplet (t) quartet (q) quin-tuplet (quint) multiplet (m) and broad (b) High-resolution massspectra (HRMS) were recorded on an LTQ-Orbitrap XL
Synthetic Procedures for Diyne-SM and d2-Diyne-SM(2S3RE)-(2-stearoylamino-3-hydroxyoctadec-4-en-1-yl)-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (diyne-SM)To a solutionof propargyl alcohol S1 (300 mg 535 mmol) in acetone (15 mL)N-bromosuccinimide (102 g 578 mmol) and silver nitrate (91 mg054 mmol) were added at room temperature The reaction mix-ture was stirred at room temperature for 2 h and then concen-trated The residue was extracted two times with diethyl ether Thecombined organic layers were dried with Na2SO4 filtered andconcentrated to afford S2 (645 mg 90 yield) as a pale yellow oilwhich was used directly in the next step 1H NMR (500 MHzCDCl3) δ 163 (brs 1H) 430 (s 2H) CuCl (03 mg 31 μmol)i-PrNH2 (51 μL 006 mmol) and NH2OHHCl (10 mg 145 μmol)were added to MeOH (1 mL) at room temperature under argonThe mixture was cooled to 0 degC and then S3 (5) (compound S3was the intermediate for the synthesis of SM head group analogs40 mg 005 mmol) was added forming a yellow acetylide suspen-sion A solution of S2 (21 mg 016 mmol) in MeOH (02 mL) wasadded immediately The reaction mixture was stirred at the sametemperature for 30 min concentrated and extracted with CHCl3The combined organic layers were dried with Na2SO4 filtered andconcentrated The mixture was passed through a short bed of silicagel eluting with CHCl3MeOH (51) The crude product was puri-fied on a Cosmosil 5C18-AR-II column (10 times 150 mm NacalaiTesque) with MeOH as the eluent to give diyne-SM (51 mg 12)as a white solid TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 124ndash142(m 50H) 153ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H) 213ndash220 (m 2H) 324 (s 6H) 370 (t J = 55 Hz 2H) 387ndash398(m 2H) 403 (dd J = 80 80 Hz 1H) 406ndash413 (m 1H) 427(s 2H) 425ndash430 (m 2H) 456 (s 2H) 543 (ddt J = 155 8015 Hz 1H) 569 (dtd J = 155 70 10 Hz 1H) 790 (d J = 90 Hz1H) 13C NMR (125 MHz CD3OD) δ 1311 2241 2585 29152918 2937 2944 2974 3175 3215 3605 4955 5086 53935579 5890 6422 6461 6564 6654 7122 7542 8127 1298913382 17458 HRMS (electrospray ionization) calculated forC46H86N2O7P [M + H]+ 8096167 found 8096184 1H and 13CNMR spectra of diyne-SM are shown in Figs S9 and S10 Thesynthetic scheme is summarized in Scheme S1(2S3RE)-[2-(1010-dideuteriumstearoylamino)-3-hydroxyoctadec-4-en-1-yl]-2-[(6-hydroxyhexa-24-diyn-1-yl)dimethylammonio]ethylphosphate (d2-diyne-SM)This compound was prepared from S4 (a full account of thesynthesis of compound S4 will be reported elsewhere) in 14yield as a white solid by following the same procedure as de-scribed for S3 TLC Rf = 018 (CH2Cl2MeOHNH4OH 70303)1H NMR (500 MHz CD3OD) δ 088 (t J = 75 Hz 6H) 122ndash143 (m 48H) 150ndash163 (m 2H) 201 (dt J = 70 70 Hz 2H)213ndash220 (m 2H) 325 (s 6H) 371 (t J = 50 Hz 2H) 388ndash399(m 2H) 402 (dd J = 80 80 Hz 1H) 406ndash412 (m 1H) 427(s 2H) 421ndash434 (m 2H) 456 (s 2H) 543 (ddt J = 150 8012 Hz 1H) 569 (dtd J = 150 70 06 Hz 1H) 790 (d J = 90Hz 1H) 13C NMR (125 MHz CD3OD) δ 1316 2243 25872918 2921 2933 2947 2952 2955 2958 3177 3179 32183606 4956 5087 5387 5393 5580 5890 5994 6422 64286461 6465 6567 6656 7121 7543 8129 12993 1337917454 HRMS (electrospray ionization) calculated forC46H83D2N2O7PNa [M + Na]+ 8336112 found 8336124 1Hand 13C NMR spectra of d2-diyne-SM are shown in Figs S11and S12 Synthetic scheme is summarized in Scheme S2
1 Palonpon AF et al (2013) Raman and SERS microscopy for molecular imaging of livecells Nat Protoc 8(4)677ndash692
2 Edholm O Nagle JF (2005) Areas of molecules in membranes consisting of mixturesBiophys J 89(3)1827ndash1832
3 Ali S Smaby JM Brockman HL Brown RE (1994) Cholesterolrsquos interfacial interactionswith galactosylceramides Biochemistry 33(10)2900ndash2906
4 Davis JH Jeffrey KR Bloom M Valic MI Higgs TP (1976) Quadrupolar echo deuteronmagnetic resonance spectroscopy in ordered hydrocarbon chains Chem Phys Lett42(2)390ndash394
5 Goretta SA Kinoshita M Mori S Tsuchikawa H Matsumori N Murata M (2012) Effectsof chemical modification of sphingomyelin ammonium group on formation of liquid-ordered phase Bioorg Med Chem 20(13)4012ndash4019
Ando et al wwwpnasorgcgicontentshort1418088112 3 of 11
Fig S1 Raman spectrum of the diyne-SM monolayer supported on a quartz substrate shown in Fig 2A Raman peaks of N2 O2 and diyne are indicated by redarrows Lower is a five times enlarged view of Upper in terms of intensity (vertical axis) Strong Raman scattering by N2 and O2 in air was always detectedduring Raman measurement of lipid membranes For ease of picking out the important Raman peaks such as diyne the two peaks of N2 and O2 have beenremoved in Figs 2 and 4 and Figs S3 and S4
Fig S2 (A) Raman image of the diyne-SMDOPCchol ternary monolayer with a 111 ratio reconstructed using the intensity of the diyne peak at 2263 cmminus1The image is the same as Fig 4A Mean value of the image is 564 The two phases were separated by taking pixel regions (B) below the mean value asdisordered domains and (C) above the mean value as ordered domains Raman images of diyne-SMDOPCchol ternary monolayers with (D) a 373 ratio and(E) a 100 ratio The images are the same as Fig 4 B and C respectively (Scale bar 10 μm)
Ando et al wwwpnasorgcgicontentshort1418088112 4 of 11
Fig S3 Raman spectra obtained along the red lines in the Raman images of the diyne-SMDOPCchol ternary monolayer with 111 373 and 100 ratios Theimages are the same as in Fig 4 AndashC with 36 times 24 pixels Raman peak of N2 at sim2330 cmminus1 was removed so that it would be easier to see Raman peaks fromlipid molecules Pix pixel (Scale bar 10 μm)
Fig S4 Comparison between Raman spectra of supported lipid monolayers with single-component composition (diyne-SM DOPC or chol) and Raman spectraof diyne-SMDOPCchol ternary monolayer at ordered and disordered domains The spectra of the ternary monolayer were obtained from five representativeareas (3 times 3 pixels) calculated using the Raman image in Fig 4A Characteristic Raman peaks are displayed in the image including diyne stretching vibration(2263 cmminus1) CH2 symmetricasymmetric stretching vibration (28502883 cmminus1) and CH3 symmetric stretching vibration (2936 cmminus1) Raman peak of N2 atsim2330 cmminus1 was removed so that it would be easier to see Raman peaks from lipid molecules
Ando et al wwwpnasorgcgicontentshort1418088112 5 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S1 Raman spectrum of the diyne-SM monolayer supported on a quartz substrate shown in Fig 2A Raman peaks of N2 O2 and diyne are indicated by redarrows Lower is a five times enlarged view of Upper in terms of intensity (vertical axis) Strong Raman scattering by N2 and O2 in air was always detectedduring Raman measurement of lipid membranes For ease of picking out the important Raman peaks such as diyne the two peaks of N2 and O2 have beenremoved in Figs 2 and 4 and Figs S3 and S4
Fig S2 (A) Raman image of the diyne-SMDOPCchol ternary monolayer with a 111 ratio reconstructed using the intensity of the diyne peak at 2263 cmminus1The image is the same as Fig 4A Mean value of the image is 564 The two phases were separated by taking pixel regions (B) below the mean value asdisordered domains and (C) above the mean value as ordered domains Raman images of diyne-SMDOPCchol ternary monolayers with (D) a 373 ratio and(E) a 100 ratio The images are the same as Fig 4 B and C respectively (Scale bar 10 μm)
Ando et al wwwpnasorgcgicontentshort1418088112 4 of 11
Fig S3 Raman spectra obtained along the red lines in the Raman images of the diyne-SMDOPCchol ternary monolayer with 111 373 and 100 ratios Theimages are the same as in Fig 4 AndashC with 36 times 24 pixels Raman peak of N2 at sim2330 cmminus1 was removed so that it would be easier to see Raman peaks fromlipid molecules Pix pixel (Scale bar 10 μm)
Fig S4 Comparison between Raman spectra of supported lipid monolayers with single-component composition (diyne-SM DOPC or chol) and Raman spectraof diyne-SMDOPCchol ternary monolayer at ordered and disordered domains The spectra of the ternary monolayer were obtained from five representativeareas (3 times 3 pixels) calculated using the Raman image in Fig 4A Characteristic Raman peaks are displayed in the image including diyne stretching vibration(2263 cmminus1) CH2 symmetricasymmetric stretching vibration (28502883 cmminus1) and CH3 symmetric stretching vibration (2936 cmminus1) Raman peak of N2 atsim2330 cmminus1 was removed so that it would be easier to see Raman peaks from lipid molecules
Ando et al wwwpnasorgcgicontentshort1418088112 5 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S3 Raman spectra obtained along the red lines in the Raman images of the diyne-SMDOPCchol ternary monolayer with 111 373 and 100 ratios Theimages are the same as in Fig 4 AndashC with 36 times 24 pixels Raman peak of N2 at sim2330 cmminus1 was removed so that it would be easier to see Raman peaks fromlipid molecules Pix pixel (Scale bar 10 μm)
Fig S4 Comparison between Raman spectra of supported lipid monolayers with single-component composition (diyne-SM DOPC or chol) and Raman spectraof diyne-SMDOPCchol ternary monolayer at ordered and disordered domains The spectra of the ternary monolayer were obtained from five representativeareas (3 times 3 pixels) calculated using the Raman image in Fig 4A Characteristic Raman peaks are displayed in the image including diyne stretching vibration(2263 cmminus1) CH2 symmetricasymmetric stretching vibration (28502883 cmminus1) and CH3 symmetric stretching vibration (2936 cmminus1) Raman peak of N2 atsim2330 cmminus1 was removed so that it would be easier to see Raman peaks from lipid molecules
Ando et al wwwpnasorgcgicontentshort1418088112 5 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S5 (A) Raman image of a pure diyne-SM monolayer taken with slit-scanning Raman microscopy The image was reconstructed using the diyne peakintensity at 2263 cmminus1 Exposure time and laser power were 60 s per line and 139 mWμm2 respectively The image consists of 344 times 190 pixels (Scale bar10 μm) (B) Raman image of a 111 diyne-SMDOPCchol ternary monolayer extracted from the lower left portion in Fig 5A The image consists of 344 times 190pixels (Scale bar 10 μm) (C) Line profiles of lipid membranes calculated along the dotted lines of the images in A and B
Fig S6 (A) Fluorescence and (B) Raman images of a 111 diyne-SMDOPCchol ternary monolayer containing 02 mol Bodipy-PC The fluorescence imagewas reconstructed using the average fluorescence intensity at 555ndash605 nm (blue) The Raman image was reconstructed using the diyne peak intensity at2264 cmminus1 (red) Both images were obtained in the same imaging area of the same sample Fluorescence background during Raman imaging was suppressedby photobleaching of Bodipy-PC under 532-nm laser exposure (C) Superimposed image of fluorescence and Raman images obtained in Fig S5 A and B Theimages consist of 54 times 28 pixels (Scale bar 20 μm)
Fig S7 DSC heating thermograms of (Upper) diyne-SM and (Lower) SM bilayers immediately after preparation The main transitions are indicated by arrows
Ando et al wwwpnasorgcgicontentshort1418088112 6 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S8 Fluorescence images of (A) diyne-SMDOPCchol and (B) SMDOPCchol quartz-supported monolayers in the presence of 02 mol Bodipy-PC at12 mNm and 25 degC (Scale bar 50 μm)
Fig S9 1H NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 7 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S10 13C NMR spectrum of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 8 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S11 1H NMR spectrum of d2-diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 9 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Fig S12 13C NMR spectrum of d2-diyne-SM
Scheme S1 Synthesis of diyne-SM
Ando et al wwwpnasorgcgicontentshort1418088112 10 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
Scheme S2 Synthesis of d2-diyne-SM
Table S1 Mean value of Raman intensity of the diyne peak indiyne-SMDOPCchol ternary monolayer calculated from theimages shown in Fig S2
Diyne-SMDOPCchol Pixel no Mean (plusmn SD) Relative
111 (ordered) 406 801 (plusmn 160) 522111 (disordered) 458 354 (plusmn 123) 231373 864 529 (plusmn 160) 345100 864 1535 (plusmn 148) 100
Relative value was calculated using diyne peak intensity of the membranewith a 100 ratio as 100
Ando et al wwwpnasorgcgicontentshort1418088112 11 of 11
