Supplementary Information:Liquid-liquid phase coexistence in lipid membranes observed by natural abundance 1H–13C

solid-state NMR

L. Loser,1 K. Saalwachter,1 and T. M. Ferreira1, ∗

1Institut fur Physik - NMR, Martin-Luther-Universitat Halle-Wittenberg, Germany(Dated: March 23, 2018)

Sample preparation

The phospholipid 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and 1,2-dioleoyl-sn-glycero-3-phosphocholine(DOPC) were purchased from Avanti Polar Lipids, while cholesterol and ethanol were purchased from Sigma-Aldrich. EachDPPC/DOPC/cholesterol sample was prepared by first mixing ethanol equimolar solutions of DPPC, DOPC and cholesterol inthe right amounts and then rapidly evaporating the solvent in a 1.5 ml EPPENDORF tube under a N2 gas stream. During theevaporation (which took no more than 1-2 min) the tube was immersed in a water bath at 55 oC and under sonication. Afterevaporation observed with the naked eye, the dried lipid mixtures were left overnight under vacuum. The resulting lipid film(about 20 mg in total) was then hydrated such that samples had 40 %wt of deuterium oxide in total weight. For hydrating thesystems the EPPENDORF tubes were placed in an EPPENDORF THERMOMIXER set to 55 oC and water was added. Themixture was then centrifuged, heated and mixed manually with a thin metal tip repeatedly, as many times necessary to attain anhomogeneous sample. Each sample was then centrifuged from the EPPENDORF tube to a KEL-F insert for solid-state NMR 4mm rotors with a sample volume of 25 µl.

Solid-state Nuclear Magnetic Resonance

All experiments were performed using a Bruker Avance III 400 spectrometer operating at a 1H Larmor frequency of 400.03MHz equipped with a standard 4 mm CP-MAS HXY probe. The temperature was calibrated with a methanol sample spinning atthe MAS frequency used [1]. The composition of each sample was determined by performing 13C direct polarisation experiments(13C 90o followed by acquisition under SPINAL64 1H decoupling [2]) and fitting gaussians to the spectra in order to determinearea fractions between components. The fits and calculations are shown in figures S5-S8 for all samples.

For describing the set up of the R-PDLF experiments we define the pulse sequence parameters according to Figures 1c and2c of the original reference for the R-PDLF experiment [3]. The MAS frequency used was 5.15 kHz and t1 increments usedwere six 180o pulses for each of the two R1871 blocks. The refocused-INEPT [4, 5] delays τ1 and τ2 were multiples of the MASrotation period, namely 1.94 ms and 0.97 ms, respectively. The RF pulses used had the following nutation frequencies: 46.35kHz (R187

1 pulses), 63.45 kHz (13C 90o and 180o pulses) and 50 kHz (SPINAL64 1H decoupling [2]). A recycle delay of 5 swas used. The total acquisition time for each scan in the direct dimension was 0.1 s using a spectral width of 200 ppm (a totalnumber of 256 scans was used). For a description of the processing of data see e.g. [6].

For a given C–H bond its order parameter magnitude, |SCH|, relates to its R-PDLF dipolar splitting by

∆νR−PDLF = 0.315dmaxCH |SCH| (1)

where 0.315 is a scaling factor specific of the dipolar recoupling sequence used and dmaxCH the rigid dipolar coupling constant

equal to 21.5 kHz [3].

∗ tiago.ferreira@physik.uni-halle.de[1] A. L. V. Geet, Anal. Chem. 42, 679 (1970).[2] B. Fung, A. Khitrin, and K. Ermolaev, J. Mag. Res. 142, 97 (2000).[3] S. V. Dvinskikh, H. Zimmermann, A. Maliniak, and D. Sandstrom, J. Mar. Res. 168, 194 (2004).[4] G. Morris and R. Freeman, J. Am. Chem. Soc. 101, 760 (1979).[5] D. Burum and R. Ernst, J. Mag. Res. 39, 163 (1980).[6] T. M. Ferreira, B. Medronho, R. W. Martin, and D. Topgaard, Phys. Chem. Chem. Phys. 10, 6033 (2008).

Electronic Supplementary Material (ESI) for Physical Chemistry Chemical Physics.This journal is © the Owner Societies 2018

2

Supplementary Figures

45505560657075

125130135140145150155160165170175

1015202530354045

γ

αβg1g3g2

C=CC=O

C15DPPC

C17DOPC

C16DPPC

C18DOPC

C3DPPC

C3DOPC

C11

DOPC

C2DPPC

C2DOPC

C14DPPC

C16DOPC

C8

δ(13C) / ppm

O

OOP

O

N

O

O

OO

-

O

OOP

O

N

O

O

OO

-

O

OH

O

OOP

O

N

O

O

OO

FFFFFFHO

OP

OO

OO

-

O

base

R1

R2

Labels used for bases:

N+

choline inositol

HO

HO

OHOH

OHHO

HOglycerol

Glycerophospholipid

Most abundant lipids in Lung Surfactant:

(sn-1)

(sn-2)

R1

DPPC R2

base: choline

O

O

R1

POPC R2

base: choline

O

O

H

+

-

FF

FFFFFFF FF

O

O

O

+

OP

O

N

O

OO

H

+

-

+

O

OOP

O

N

O

O

OO

FFFFFFHO

H

+

-

FF

FFFFFFF FF

HO

DPPC

γ

γγ

βα

g2g1g3 2

16

216

8:2 FTOH

Cholesterol

OOO

R1

H

R2R3

O

OOP

O

N

O

O

OO

-

O

OOP

O

N

O

O

OO

-

O

OH

O

OOP

O

N

O

O

OO

FFFFFFHO

OP

OO

OO

-

O

base

R1

R2

Labels used for bases:

N+

choline inositol

HO

HO

OHOH

OHHO

HOglycerol

Glycerophospholipid

Most abundant lipids in Lung Surfactant:

(sn-1)

(sn-2)

R1

DPPC R2

base: choline

O

O

R1

POPC R2

base: choline

O

O

H

+

-

FF

FFFFFFF FF

O

O

O

+

OP

O

N

O

OO

H

+

-

+

O

OOP

O

N

O

O

OO

FFFFFFHO

H

+

-

FF

FFFFFFF FF

HO

DPPC

γ

γγ

βα

g2g1g3 2

16

216

8:2 FTOH

Cholesterol

OOO

R1

H

R2R3

C2

C18

C2

C16

αβ

αβ

g1g3 g2g1g3 g2

C2 C2

C18C16

DOPC DPPC

γγγ

γγγ

FIG. 1: Molecular structures of DPPC and DOPC with carbon labels and full 13C spectrum for the samples with DPPC/DOPC/cholesterolcompositions 32:57:11 (black) and 44:37:19 (blue).

3

2929.53030.53131.53232.53333.53434.535 2727.5 24.52525.5 2222.523 13.514

49:29:22

δ(13C) / ppm

ΔνR

-PD

LF /

kHz

lo1.5

1

0.5

0

-0.5

-1

-1.5

FIG. 2: Regions of the 1H–13C R-PDLF spectrum for the sample with DPPC/DOPC/cholesterol composition 49:29:22. Spectra acquired usinga MAS frequency of 5.15 kHz and SPINAL64 decoupling at an RF frequency of 50 KHz.

0 2 4 6 8x 10−3

−0.5

0

0.5

113.73

0 2 4 6 8x 10−3

−0.5

0

0.5

122.54

0 2 4 6 8x 10−3

−0.5

0

0.5

124.95

0 2 4 6 8x 10−3

−0.5

0

0.5

127.12

0 2 4 6 8x 10−3

−0.5

0

0.5

127.31

0 2 4 6 8x 10−3

−0.5

0

0.5

129.19

0 2 4 6 8x 10−3

−0.5

0

0.5

129.38

0 2 4 6 8x 10−3

−0.5

0

0.5

129.72

0 2 4 6 8x 10−3

−0.5

0

0.5

130.01

0 2 4 6 8x 10−3

−0.5

0

0.5

130.49

0 2 4 6 8x 10−3

−0.5

0

0.5

131.89

0 2 4 6 8x 10−3

−0.5

0

0.5

133.91

t1 / s

FIG. 3: Time dependence over the dipolar coupling dimension of the R-PDLF experiments for samples with DPPC/DOPC/cholesterol compo-sition 32:57:11 (black) and 44:37:19 (blue) at the 13C chemical shifts indicated on top of each plot.

4

−1 0 1

13.73

−1 0 1

22.54

−1 0 1

24.95

−1 0 1

27.12

−1 0 1

27.31

−1 0 1

29.19

−1 0 1

29.38

−1 0 1

29.72

−1 0 1

30.01

−1 0 1

30.49

−1 0 1

31.89

−1 0 1

33.91

ΔνR-PDLF / kHz

FIG. 4: Dipolar slices of the R-PDLF spectra for samples with DPPC/DOPC/cholesterol composition 32:57:11 (black) and 44:37:19 (blue) atthe 13C chemical shifts indicated on top of each plot (i.e. the Fourier transforms of the time domain curves shown above).

5

Cα

CII

Cγ

C50

C56

– 38 / 42 / 20

C C fchol

C50 0.20 0.20 0.20

C56 0.20 0.20

C C fDOPC

CII 0.42 0.42 0.42

FIG. 5: Regions of the 13C direct polarization spectrum for sample 38:42:20. The composition was determined by using the areas of the fitsshown for each carbon indicated. The upper table shows the mol fraction of cholesterol as determined from using the γ and α carbons of thephospholipids and using two distinct carbons for cholesterol with 13C chemical shifts at approx. 50 and 56.5 ppm. The bottom table showsthe fraction of DOPC by using the fraction of cholesterol calculated above and the areas from the γ, α, and the carbons connected by doublebonds in the DOPC acyl chains within 129-130 ppm.

6

– 32 / 57 / 11

C C fchol

C50 0.13 0.13 0.11

C56 0.09 0.10

C C fDOPC

CII 0.57 0.57 0.57

Cα

CII

Cγ

C50

C56

FIG. 6: Regions of the 13C direct polarization spectrum for sample 32:57:11. The composition was determined by using the areas of the fitsshown for each carbon indicated. The upper table shows the mol fraction of cholesterol as determined from using the γ and α carbons of thephospholipids and using two distinct carbons for cholesterol with 13C chemical shifts at approx. 50 and 56.5 ppm. The bottom table showsthe fraction of DOPC by using the fraction of cholesterol calculated above and the areas from the γ, α, and the carbons connected by doublebonds in the DOPC acyl chains within 129-130 ppm.

7

– 44 / 37 / 19

C C fchol

C50 0.18 0.19 0.19

C56 0.20 0.21

C C fDOPC

CII 0.36 0.38 0.37

Cα

CII

Cγ

C50

C56

FIG. 7: Regions of the 13C direct polarization spectrum for sample 44:37:19. The composition was determined by using the areas of the fitsshown for each carbon indicated. The upper table shows the mol fraction of cholesterol as determined from using the γ and α carbons of thephospholipids and using two distinct carbons for cholesterol with 13C chemical shifts at approx. 50 and 56.5 ppm. The bottom table showsthe fraction of DOPC by using the fraction of cholesterol calculated above and the areas from the γ, α, and the carbons connected by doublebonds in the DOPC acyl chains within 129-130 ppm.

8

– 49 / 29 / 22

C C fchol

C50 0.20 0.20 0.22

C56 0.23 0.23

C C fDOPC

CII 0.29 0.28 0.29

Cγ

Cα

CII C50

C56

FIG. 8: Regions of the 13C direct polarization spectrum for sample 49:29:22. The composition was determined by using the areas of the fitsshown for each carbon indicated. The upper table shows the mol fraction of cholesterol as determined from using the γ and α carbons of thephospholipids and using two distinct carbons for cholesterol with 13C chemical shifts at approx. 50 and 56.5 ppm. The bottom table showsthe fraction of DOPC by using the fraction of cholesterol calculated above and the areas from the γ, α, and the carbons connected by doublebonds in the DOPC acyl chains within 129-130 ppm.