Chinese Chemical Letters  2018, Vol. 29 Issue (3): 357-360   PDF    
Ordering effects of cholesterol on sphingomyelin monolayers investigated by high-resolution broadband sum-frequency generation vibrational spectroscopy
Yiyi Lia,b, Rongjuan Fenga, Lu Lina, Minghua Liuc, Yuan Guoa,b,1, Zhen Zhanga,1    
a Beijing National Laboratory for Molecular Sciences, CAS Research/Education Center for Excellence in Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China;
b University of Chinese Academy Sciences, Beijing 100049, China;
c National Center for Nanoscience and Technology, Beijing 100190, China
Abstract: This report investigated the ordering of the alky chain of sphingomyelin (SMs) monolayers induced by cholesterol at the air/water interface using high-resolution broadband sum frequency generation vibrational spectroscopy (HR-BB-SFG-VS). The SFG spectra of the three nature sphingomyelin/cholesterol mixture monolayers with two concentrations of the cholesterol at the air/water interface are performed under different polarization combination. A new resolved CH2 symmetric stretching (d+, ~2834 cm-1) and the CH3 symmetric stretching (r+, ~2874 cm-1) mode are applied to characterize the conformational order in the sphingomyelin/cholesterol mixture monolayers. It was found that the cholesterol make the sphingosine backbones more conformational order. During this process, the conformational order of the N-linked acyl chain remains unaltered. Moreover, the sphingosine backbones of SMs have much larger contributions to gauche defects of SMs than one in the N-linked acyl chain. These results presented here not only shed lights on understanding of the interactions of sphingomyelin molecules with cholesterol molecules at interface but also demonstrates the ability of HR-BB-SFG to probe such complicated molecular systems.
Key words: Sphingomyelin     Cholesterol     Ordering effect     Air/liquid interface     High-resolution broadband sum frequency     generation     Vibrational spectroscopy    

Sphingomyelin (SM) along with cholesterol (Chol), as a composition of the biological membranes, play a critical role in biological processes: signaling pathways, differentiation, multiple cell functions, cell apoptosis, cellular calcium homeostasis and lipid rafts, and accordingly has received much attention [1-4]. In order to understand those biological functions, it is necessary to study the effect of the cholesterol on SMs molecules, which mainly involves the two ordering effects, orientational order and conformational order. Previous studies have revealed that the cholesterol in SMs membrane layers can lead to an increased ordering of the hydrocarbon chains of SMs [5-16]. In the studying processes, many spectroscopic methods, including NMR, fluorescence, FTIR and electron paramagnetic resonance, have been used to investigate the ordering effects of acyl chains by counting the number of gauche rotamers or comparing the average molecular area in SM-Chol monolayer [5, 10, 11]. However, up to date, there has been far less studies to quantify the effect of Chol on SM monolayer at the air/water interface by directly probing the orientation information on the alky chain at the molecular level by these techniques and the detail of this effect is still vague. For example, SMs consist of two hydrophobic acyl chains: sphingosine backbone chain and N-linked fully saturated acyl chain (Fig. 1). Which chains of SMs can be affected by Chol at this monolayer and how the concentration of the Chol affects the structure of the SMs remain unknown. In fact, vibrational sum frequency generation (SFG) technique is more suitable than that mentioned above for studying the order effect at the air/liquid interface. It is sensitive to the structure, structural variations and the complex local environment at the interface, and thus it has been used to determine the functional group orientation and conformation structure of monolayers at interfaces [7, 17-19].

Download:
Fig. 1. Molecular structures of ESM, BSM, MSM and cholesterol-d7. Three nature SMs consist of two hydrophobic acyl chains, one is sphingosine backbone (red color) and another is N-linked fully saturated acyl chain (blue color). In cholesterol-d7, the CH3 group was deuterated in order to get rid of the CH3 contribution to the SFG spectra in the mixture of SMs and Chol.

There are numerical SFG studies to quantify the order and orientation of lipids in planar membranes such as monolayers and supported lipid bilayers. Bonn et al. investigated the effect of Chol on phospholipid structures using SFG and demonstrated that Chol can affect the lipid conformational and orientational order even at very low concentration [19]. Recently, one of important developments of SFG techniques is high-resolution broadband sum frequency generation vibrational spectroscopy (HR-BB-SFG-VS), which can obtain unambiguous interfering fine spectral features and nearly intrinsic spectral lineshapes of the molecular vibrational modes because of the high resolution and large signal-tonoise ratio (SNR) [20-22].

In this report, we applied HR-BB-SFG-VS to detect the mixed monolayers from three nature SMs and two concentration of Chol for the purpose of studying the influence of the Chol on SM molecules. All three nature SMs contain a sphingosine backbone of equal length and an acyl chain of unequal length (Fig. 1). We found that the cholesterol makes sphingosine backbones more orintational order, in which case, the conformation of the N-linked acyl chain remains unaltered. And the cholesterol has a same effect on the three nature SMs mixtures. The detailed experimental description can be found in the Supporting information.

Fig. 2 shows the HR-BB-SFG-VS spectra of ESM and cholesterold7 (Chol-d7) monolayers at the pure water under ssp polarization combinations (ssp indicates that the SFG signal, the visible beam, and the IR beam are in the s, s, and p polarizations, respectively) at surface pressures 15 mN/m, in which case the monolayer lies in the LE-LC phase [7, 23]. From Fig. 2, one can observe nine vibrational modes from the spectra of ESM monolayer and eleven vibrational modes from the spectra of cholesterol-d7 monolayer, which are consistent with the observation in previous reports [20, 24-26]. The detailed assignment of those peaks can be found in Tables S1-S4 in supporting information, all of which are fitted using multiLorentzian lineshape. The HR-BB-SFG-VS spectra can resolve a number of additional peaks comparing with common BB-SFG-VS. For example, more CH2 symmetric stretching vibration modes can be resolved by HR-BB-SFG-VS spectra, which are obscured in the lower resolution SFG-VS spectra [21, 23, 24, 26]. The peaks observed here have been assigned in the literatures. All the peaks centered at ~2834, 2850, 2861 cm-1 were assigned to the CH2 symmetric stretching (d+) in the hydrophobic alky chains [21, 23, 24, 26]. The peaks of 2874 cm-1, 2915 cm-1, 2926 cm-1, 2942 cm-1 and 2964 cm-1 can be assigned to the CH3 symmetric stretching (r+), CH2 asymmetric stretching (d-), CH2 Fermi resonance, CH3 Fermi resonance and CH3 asymmetric stretching (r-), respectively [7, 17, 19, 27-34]. The peak at 2953 cm-1 is the CH2 symmetric stretching in the headgroup [35].

Download:
Fig. 2. HR-BB-SFG-VS spectra recorded using an ssp polarization combination for ESM and cholesterol-d7 monolayers in 15 mN/m surface pressure at the air/water interface. The red arrow shows the CH2-ss peak at d+ (~2834 cm-1) and the blue arrow shows the CH3-ss peak at r+ (~2874 cm-1) for ESM monolayer which are not overlap with Chol-d7 SFG spectra. Those two peaks have been used to quantify the ordering effect of the SMs chain affected by adding Chol. Experimental data are shown as points and fitted as solid lines.

In Fig. 2, the SFG spectra shows that there exist d+ (~2834 cm-1) and r+ (2874 cm-1) in the spectra of ESM monolayer but there are no those in the spectra of cholesterol-d7. As recently reported, the new CH2 spectral feature centered at ~2834 cm-1 observed here often overlaps with d+ (~2850 cm-1) in common BB-SFG-VS and the picosecond scanning SFG system, instead, they can be well resolved by the new HR-BB-SFG-VS technique [23]. Therefore, in the following, the d+ (~2834 cm-1) and r+ (2874 cm-1) modes were chosen to avoid the overlap of the peaks centered at ~2850 and 2861 cm-1 between ESM and chol-d7 in the sphingomyelin/chol monolayers. It is known that the intensity of the r+ (~2874 cm-1) is sensitive to the orientation of the acyl chains and the intensity of d+ (~2834 cm-1) is proportional to the chain gauche defects which can be used to quantify the conformational disorder in the chains [19]. The d+ and r+ modes can be thus used to illustrate the conformational change in the SMs and cholesterol mixture monolayers. Therefore, in this work, we will obtain accurately conformational change in the sphingomyelin hydrocarbon chains by studying the change of the d+ at ~2834 cm-1 instead of d+ at ~2850 cm-1, in spite of which the latter has stronger signal than that of the former for both cholesterol-d7 monolayers and sphingomyelin monolayers.

Previous studies have shown that in the phospholipid monolayers the intensity of the CH2 symmetric stretch at ~2850 cm-1 decreases once cholesterol is added to the phospholipid monolayers, from which we can draw the conclusion that the cholesterol could make phospholipid more ordered at the air/solid interface [7, 19]. However, this strategy may be not very rigorous because it neglects possible interference effects and the phase differences between the CH2 signals of cholesterol and phospholipid. The similar case also appears in Stolow's works on the effect of Chol on SMs monolayers at the air/solid interface by means of the common BB-SFG-VS spectra, in which they used the corrected CH2 symmetric stretch intensity to evaluate the effect. In order to obtain the corrected CH2 symmetric stretch intensity, they had to neglect the phase difference of Chol-d7 and SM molecules, and had to consider the total CH2-ss intensity of the mixture as the sum of the individual CH2-ss intensities of SM and Chol-d7, which is also not rigorous. However, this problem no longer exists in our case because the HR-BB-SFG-VS is a direct and reliable measurement for the detailed spectral information on complex molecular surfaces and interfaces [24].

As we discussed above, we will use the peak at d+ (~2834 cm-1) and r+ (~2874 cm-1) to illustrate the ordering of cholesterol on SMs at the air/water interface. Fig. 3 shows the HR-BB-SFG-VS spectra for three nature SMs monolayers and their mixtures with cholesterol-d7 under ssp and ppp (p-SFG, p-visible, p-IR) polarization combinations, in which there are almost the same line shape in these monolayers at first glance. However, the fitting results show that the peak width of the r+ gradually becomes broad. Moreover, the fitting results also show an interesting fact that in case of the MSM spectra, there are two CH3 symmetric stretching vibration modes at 2872 and 2875 cm-1 clearly resolved by the HR-BB-SFG-VS. The two peaks are likely to be attributed to the vibration of CH3 of the MSM hydrocarbon chains located at different chain length. The detailed interpretation for these two peaks will be discussed in our follow-up works.

Download:
Fig. 3. HR-BB-SFG-VS spectra under ssp and ppp polarization combination for SMs, SMs + 10 mol% Chol-d7 and SM + 25 mol% Chol-d7 monolayers at the air/water interface: (A) ESM, ESM + 10 mol% Chol-d7 and ESM + 25 mol% Chol-d7, (B) BSM, BSM + 10 mol% Chol-d7 and BSM + 25 mol% Chol-d7, (C) MSM, MSM + 10 mol% Chold7 and MSM + 25 mol% Chol-d7. Experimental data are shown as points and fitted as solid lines. The present of 10% to 25% cholesterol-d7 clearly show that the intensity of the symmetric CH3 stretch increase gradually for three nature SMs, and the intensity of the symmetric CH2 stretch decrease gradually. The relative intensity of the CH2-ss and CH3-ss signals provide information on the conformational order and gauche defects for the acyl chains.

In what follows, we focus on the intensity changes of the d+ (~2834 cm-1) and r+ (~2874 cm-1) modes for the three nature SMs and its mixture with cholesterol-d7 to clarify the conformational order of SMs acyl chains at the air/water interface. It is found in Fig. 3 that when varying Chol concentrations from 10% to 25% at the Chol/SM mixture monolayers interface, the r+ (~2874 cm-1) peaks become stronger; inversely, the d+ (~2834 cm-1) peaks become slightly weaker, in which case the ratio of the CH3 and CH2 symmetric stretch oscillator strengths became larger. It is known that this ratio provides an estimate for the lipid molecular order. Therefore, the larger ratio of the CH3 and CH2 symmetric stretch oscillator strengths is an indication that the acyl chains of SM at the Chol/SM mixture monolayers interface are more ordered than that in the pure SM monolayers. And the order affected by 25% Chol is more than that affected by Chol 10%. Such observation is in agreement with previous research of lipids reported in DPPC/Chol and SMs/Chol monolayers [7, 19, 30, 33, 36, 37].

The extracted intensity of CH2-ss (~2834 cm-1) and CH3-ss (~2874 cm-1) are shown in Fig. 4 for SMs monolayers as a function of Chol concentration. Fig. 4A shows that the CH2-ss (~2834 cm-1) peaks of the three nature SMs for the SMs/Chol mixtures are lower than that of pure the three nature SMs monolayer. Such decrease in the intensity indicates that there must be a lower density of gauche defects and the more ordered in acyl chains because the CH2 signal intensity (d+, ~2834 cm-1) is related to the number of gauche defects in the chains [38]. In addition, Fig. 4A also shows that the peak intensity at 2834 cm-1 for ESM is lower than those for BSM and MSM at 15 mN/m surface pressure, which indicates a better ordering for the ESM acyl chains and abundant gauche conformers for both BSM and MSM [39]. This is reasonable because for ESM monolayer there are two acyl chains in equivalent lengths between sphingosine and N-linked alky chain, which is more tightly packed; while for MSM and BSM monolayers, there are two different acyl chain lengths respectively. The larger differences of this two chain lengths result in a smaller steric hindrance effects, contributing to more gauche defects in the chains.

Download:
Fig. 4. Comparison of CH2-ss (~2834 cm-1) and CH3-ss (~2874 cm-1) intensity for SMs monolayers as a function of Chol concentration. Panel (A) shows the CH2-ss (~2874 cm-1) intensity for ESM, ESM + 10 mol% Chol-d7 and ESM + 25 mol% Chol-d7 gradually decrease with the addition of cholesterol-d7, which means the gauche defects in the chain decrease. Panel (B) shows the CH3-ss (~2874 cm-1) intensity for BSM, BSM + 10 mol% Chol-d7 and BSM + 25 mol% Chol-d7 increased with the addition of cholestol-d7, which means the better orientational order of the chain.

In order to further clarify the different chain length effects, we compared the two peaks at the r+ (~2874 cm-1) and d+ (~2834 cm-1) on the ESM/Chol, BSM/Chol and MSM/Chol mixed monolayer (Fig. 4B). The 2874 cm-1 peak intensities of the three nature SMs/Chol monolayers are larger than one of the pure SMs monolayers under the same surface pressure of 15 mN/m and the same concentration of cholesterol-d7 at the air/water interface. Fig. 4B also show the similar increasing trends of the r+ (~2874 cm-1) intensity with Chol concentration for the three nature SMs, which is important to understand how the two hydrophobic acyl chains interact with the cholesterol-d7.

As mentioned above, SMs have two hydrophobic acyl chains, sphingosine backbone and N-linked fully saturated acyl chain (Fig. 1). The only difference of the three nature SMs lies in the length of N-linked fully saturated acyl chain. If the Chol affected the ordering of N-linked fully saturated acyl chain, this acyl chain's conformational ordering in the three nature SMs should be different, because they have different N-linked fully saturated acyl chains lengths. In fact, the intensity changes in CH3 symmetric stretch for ESM, BSM and MSM, shown in Fig. 4B, are almost identical at 15 mN/m surface pressure as a function of concentration of Chol, which means that the increase results from the identical sphingosine backbones, rather than different N-linked fully saturated acyl chains. We can conclude thus that cholesterol must make the sphingosine backbones more ordered and the conformation of the N-linked acyl chain remains unaltered during these processes. This conclusion also agrees well with our previous results, in which the sphingosine backbone of ESM became ordered and the conformation of the N-linked acyl chain was still unchanged with addition of Ca2+ ions in the subphase at the air/ water interface [23].

In conclusion, we use high-resolution broadband sum frequency generation vibrational spectroscopy (HR-BB-SFG-VS) to investigate the interaction of the sphingomyelin-cholesterol mixed monolayers at the air/water interface. We found that the intensity changes in CH2 (~2834 cm-1) and CH3 (~2874 cm-1) symmetric stretching can be used to characterize the ordering of three nature sphingomyelin mixture monolayers as a function of cholesterol concentration. The results reveal that cholesterol causes sphingosine backbones more ordered and the conformation of the N-linked acyl chain remains unaltered at the SMs/Chol mixed monolayer. And the cholesterol can cause better ordering effects for the ESM acyl chains and much gauche conformations for both BSM and MSM. This study provides fundamental information for SMs and SMs/Chol membranes at the molecular level. Furthermore, it also shows the advantages of applicability of HR-BB-SFG-VS to study complicated lipid systems. Further studies are underway to investigate the effect of cholesterol on strong hydrogen bonding capacity of ESM monolayer at the air/water interface.

Acknowledgements

Y. Guo and M.H. Liu are grateful for funding from the National Natural Science Foundation of China (No. 21227802). Z.Z. and Y. G. are grateful for funding from the National Natural Science Foundation of China (Nos. 21503235, 21673251) and the ICCAS for Start-up Funding to support this work. The authors declare no competing financial interests.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.cclet.2017.11.006.

References
[1]
K. Simons, E. Ikonen, Nature 387 (1997) 569-572. DOI:10.1038/42408
[2]
Y.A. Hannun, L.M. Obeid, Trends Biochem. Sci. 20 (1995) 73-77. DOI:10.1016/S0968-0004(00)88961-6
[3]
S. Jayadev, B. Liu, A.E. Bielawska, et al., J. Biol. Chem. 270 (1995) 2047-2052. DOI:10.1074/jbc.270.5.2047
[4]
E.A. Sweeney, J. Inokuchi, Y. Igarashi, Febs Lett. 425 (1998) 61-65. DOI:10.1016/S0014-5793(98)00198-7
[5]
T. Rog, M. Pasenkiewicz-Gierula, I. Vattulainen, M. Karttunen, Biochim. Biophys. Acta 1788 (2009) 97-121. DOI:10.1016/j.bbamem.2008.08.022
[6]
H. Ohvo-Rekila, B. Ramstedt, P. Leppimaki, J.P. Slotte, Prog. Lipid Res. 41 (2002) 66-97. DOI:10.1016/S0163-7827(01)00020-0
[7]
C. Weeraman, M.H. Chen, D.J. Moffatt, et al., Langmuir 28 (2012) 12999-13007. DOI:10.1021/la301332e
[8]
E. London, Biochim. Biophys. Acta 1746 (2005) 203-220. DOI:10.1016/j.bbamcr.2005.09.002
[9]
S.L. Veatch, S.L. Keller, Biochim. Biophys. Acta 1746 (2005) 172-185. DOI:10.1016/j.bbamcr.2005.06.010
[10]
J.M. Smaby, H.L. Brockman, R.E. Brown, Biochemistry 36 (1997) 2338. DOI:10.1021/bi975000z
[11]
X.M. Li, M.M. Momsen, J.M. Smaby, H.L. Brockman, R.E. Brown, Biochemistry 40 (2001) 5954-5963. DOI:10.1021/bi002791n
[12]
A. Bjorkbom, T. Rog, K. Kaszuba, et al., Biophys. J. 99 (2010) 3300-3308. DOI:10.1016/j.bpj.2010.09.049
[13]
K. Cheng, M.H. Ropers, C. Lopez, Food Chem. 224 (2017) 114-123. DOI:10.1016/j.foodchem.2016.12.035
[14]
O. Engberg, H. Nurmi, T.K.M. Nyholm, J.P. Slotte, Langmuir 31 (2015) 4255-4263. DOI:10.1021/acs.langmuir.5b00403
[15]
K. Hac-Wydro, P. Wydro, P. Dynarowicz-Latka, M. Paluch, J. Colloid Interface Sci. 329 (2009) 265-272. DOI:10.1016/j.jcis.2008.09.057
[16]
P. Wydro, Colloids Surf. B Biointerfaces 93 (2012) 174-179. DOI:10.1016/j.colsurfb.2011.12.035
[17]
M. Sovago, G.W.H. Wurpel, M. Smits, M. Muller, M. Bonn, J. Am. Chem. Soc. 129 (2007) 11079-11084. DOI:10.1021/ja071189i
[18]
N.N. Casillas-Ituarte, X.K. Chen, H. Castada, H.C. Allen, J Phys. Chem. B 114 (2010) 9485-9495. DOI:10.1021/jp1022357
[19]
M. Bonn, S. Roke, O. Berg, et al., J. Phys. Chem. B 108 (2004) 19083-19085. DOI:10.1021/jp0452249
[20]
L. Velarde, H.F. Wang, Phys. Chem. Chem. Phys. 15 (2013) 19970-19984. DOI:10.1039/c3cp52577e
[21]
L. Velarde, H.F. Wang, J. Chem. Phys. 139 (2013) 084204. DOI:10.1063/1.4818996
[22]
L. Velarde, X.Y. Zhang, Z. Lu, et al., J. Chem. Phys. 135 (2011) 241102. DOI:10.1063/1.3675629
[23]
R.J. Feng, L. Lin, Y.Y. Li, et al., Biophys. J. 112 (2017) 2173-2183. DOI:10.1016/j.bpj.2017.04.026
[24]
L. Fu, S.L. Chen, H.F. Wang, J. Phys. Chem. B 120 (2016) 1579-1589. DOI:10.1021/acs.jpcb.5b07780
[25]
H.F. Wang, Prog. Surf. Sci. 91 (2016) 155-182. DOI:10.1016/j.progsurf.2016.10.001
[26]
A.L. Mifflin, L. Velarde, J.M. Ho, J. Phys. Chem. A 119 (2015) 1292-1302. DOI:10.1021/jp510700z
[27]
S. Roke, J. Schins, M. Muller, M. Bonn, Phys. Rev. Lett. 90 (2003) 128101. DOI:10.1103/PhysRevLett.90.128101
[28]
C. Ohe, Y. Ida, S. Matsumoto, et al., J. Phys. Chem. B 108 (2004) 18081-18087. DOI:10.1021/jp0404347
[29]
X.K. Chen, H.C. Allen, J. Phys. Chem. A 113 (2009) 12655-12662.
[30]
P.J.N. Kett, M.T.L. Casford, P.B. Davies, J. Phys. Chem. B 115 (2011) 6465-6473. DOI:10.1021/jp1112685
[31]
J. Liu, J.C. Conboy, Langmuir 21 (2005) 9091-9097. DOI:10.1021/la051500e
[32]
M.R. Watry, T.L. Tarbuck, G.I. Richmond, J. Phys. Chem. B 107 (2003) 512-518.
[33]
E.H.G. Backus, D. Bonn, S. Cantin, S. Roke, M. Bonn, J. Phys. Chem. B 116 (2012) 2703-2712. DOI:10.1021/jp2074545
[34]
K.L. Harper, H.C. Allen, Langmuir 23 (2007) 8925-8931. DOI:10.1021/la7006974
[35]
R.J. Feng, X. Li, Z. Zhang, Z. Lu, Y. Guo, J. Chem. Phys. 145 (2016) 244707. DOI:10.1063/1.4972564
[36]
C. Ohe, T. Sasaki, M. Noi, Y. Goto, K. Itoh, Anal. Bioanal. Chem. 388 (2007) 73-79. DOI:10.1007/s00216-006-1030-0
[37]
M. Gurau, S. Lim, E. Castellana, et al., J. Am. Chem. Soc. 126 (2004) 10522. DOI:10.1021/ja047715c
[38]
D. Zhang, R.S. Ward, Y.R. Shen, G.A. Somorjai, J. Phys. Chem. B 101 (1997) 9060-9064. DOI:10.1021/jp9718358
[39]
D. Zhang, Y.R. Shen, G.A. Somorjai, Chem. Phys. Lett. 281 (1997) 394-400. DOI:10.1016/S0009-2614(97)01311-0