FTIR studies of temperature influence on the DPPG model membrane

FTIR studies of temperature influence on the DPPG model membrane

Journal of Molecular Structure 887 (2008) 117–121 Contents lists available at ScienceDirect Journal of Molecular Structure journal homepage: www.els...

468KB Sizes 2 Downloads 74 Views

Journal of Molecular Structure 887 (2008) 117–121

Contents lists available at ScienceDirect

Journal of Molecular Structure journal homepage: www.elsevier.com/locate/molstruc

FTIR studies of temperature influence on the DPPG model membrane Feride Severcan a, Dana-Ortansa Dorohoi b,* a b

Faculty of Biology, Middle East Technical University of Ankara, 06531, Turkey Faculty of Physics, ‘‘Al. I. Cuza” University, 11 Carol I Bvd., Iasßi, RO 700056, Romania

a r t i c l e

i n f o

Article history: Received 14 September 2007 Received in revised form 23 January 2008 Accepted 7 February 2008 Available online 20 March 2008 Keywords: Dipalmitoyl phosphatidyl glycerol (DPPG) FTIR spectra Model membranes Degree of order

a b s t r a c t The thermal changes induced in dipalmitoyl phosphatidyl glycerol (DPPG) model membranes were studied by FTIR technique. When temperature increases, the DPPG model membrane passes from a high ordered gel phase to a few ordered liquid crystalline phase. The very quick passing between the two phases is characterized by the main phase transition temperature at which the bilayers contain equal percentages of ordered and disordered phospholipids. The main phase transition induces important changes in the symmetric and asymmetric stretching vibrations of the ACH2 groups of the DPPG acyl tails. Mathematical methods for determining the main phase transition and the degree of order of the DPPG model membrane at a given temperature are proposed in this paper. Ó 2008 Elsevier B.V. All rights reserved.

1. Introduction Biological membranes [1–3] are crucial cellular components having multiple roles. They maintain electrochemical gradients by controlling the diffusion of ions and biomolecules. They act as a supporting matrix for embedded enzymes or receptors and they are engaged in stabilizing interactions with skeleton proteins. The biological membrane fluidity is a natural condition of its functionality. In order to eliminate the multitude of factors influencing the biological membrane fluidity, some simplified model membranes, built from one of the main constituents of the biological membranes, are usually used. The main constituents of the biological membranes are the phospholipids as dipalmitoyl phosphatidyl coline (DPPC), dipalmitoyl phosphatidyl glycerol (DPPG) or cholesterol. The amphiliophilic nature of the phospholipids determines the bilayers formation when the phospholipids concentration in water is appropriate [4–7]. The bilayers have the hydrophilic heads of the phospholipids in contact with water and the hydrophobic tails, oriented to the middle of the bilayer. The DPPG model membranes can be received in a buffer solution [5–9] and studied by FTIR technique. The measurement of some spectral parameters like band frequency, width and intensity provided information regarding the possible structural interactions and conformational rearrangements taking place. The phospholipids model membrane stability is affected by temperature and by the impurities. A phase transition of the type

* Corresponding author. Tel.: +40 2 3220 1182; fax: +40 2 3220 1150. E-mail address: [email protected] (D.-O. Dorohoi). 0022-2860/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2008.02.039

(1) takes place when the model membrane temperature increases [7,10,11]. Gel phase () Liquid crystalline phase

ð1Þ

The model membranes are built as lamellar symmetric bilayers forming a gel phase with a high degree of order at low temperatures. However, at high temperatures, they become disordered and pass into a liquid crystalline phase. The transition (1) is very quick and it is characterized by the main phase transition temperature, Tm. The value of the main phase transition temperature is dependent on the chemical nature of the model membrane constituents and also on their concentration [11,12]. The bilayers contain equal percentages of ordered and disordered phospholipids at the main phase transition temperature. Important changes in the IR spectra such as spectral shifts or bandwidth modifications are induced by the phase transition from the gel to the liquid crystalline phase. FTIR spectra indicate the degree of order in the phospholipids bilayers by the shifts of the stretching symmetric and asymmetric vibration bands of the CH2 groups belonging to the acyl chains [7,8]. On the other hands, the APO 2 [5] and AC@O [4,9] groups of the phospholipids hydrophobic heads are potential recipients of hydrogen bonding interactions with the water layer separating the phospholipids bilayers. So, the stretching vibrations of these groups offer information about the temperature influence on the hydrogen bonding mechanism. Some observations regarding the temperature influence on the degree of order of the acyl chains in DPPG model membranes and a mathematical model permitting the calculation of the percentages of gel and liquid crystalline phases at a given temperature are developed here.

118

F. Severcan, D.-O. Dorohoi / Journal of Molecular Structure 887 (2008) 117–121

2. Experimental setup Dipalmitoyl phosphatidyl glycerol (DPPG) model membranes were used in this study. DPPG was purchased from Sigma Chemical Co., St. Louis, MO, and used without purification. Multilamellar vesicles were obtained from DPPG dried films and phosphate buffer solution, using the procedure proposed by Severcan et al. [10–13]. Fourier Transform Infrared (FTIR) spectra were registered in CaCl2 cells, using a Bomeme 157 FTIR spectrometer. The interferograms were averaged for 100 scans. An Unicam Digital Temperature Controller unit with a thermocouple located around the edge of the cell window was used for temperature monitoring. The samples were investigated in a large temperature range with increasing temperature from 26 to 61 °C. The temperature increasing rate was small enough to assure quasi-static processes [11]. Each spectrum was recorded after 15 min of temperature stabilization to make sure that the sample temperature is that displayed by the digital controller. The water vaporous influence was eliminated by subtracting the FTIR spectra of buffered solution from the model membrane spectrum, at each studied temperature. The FTIR spectra of the DPPG model membrane in the region 3000–3800 cm1 were also analysed without buffer spectrum subtraction in order to evidence the important modifications in the hydrogen bonds interactions when temperature increased [14]. The structural formula of DPPG and also the arrangement of the hydrocarbon tails in the gel and liquid crystalline phases of DPPG are given in Figs. 1 and 2, respectively. 3. Calculation methods Based on the idea that at the main phase transition temperature, the gel phase and the liquid crystalline phase coexist in equal proportions in sample, we determined the Pg and P lc percents of the molecules in the gel and in the liquid crystalline phases in a model membrane, at each temperature. Let us consider a system consisting from N amphiphilic molecules that can have only two thermodynamic phases gel and liquid crystalline ones. Let suppose that the transition between these phases is a reversible thermodynamic transformation (1). Relation (1) suggests that by the increasing temperature of the system, molecules can pass in the liquid crystalline phase and, by the system cooling, they can return in the gel phase. Let be N g the number of the molecules in gel phase and N lc the number of the molecules in liquid crystalline phase in the studied system. These numbers are dependent on temperature and satisfy the equation: N g þ N lc ¼ N

ð2Þ

At low temperatures, when the system is in its gel phase, we can consider that N g tends to N, while at the temperature higher than the melting point, N g is near zero, because the system passed in its liquid crystalline phase and N lc tends to N. Such a condition is satisfied by DPPG acyl chains from the DPPG model membrane. By temperature increasing, N lc increases and N g decreases as relation (1) predicts. Contrariwise, N g increases and

Fig. 2. Changes induced by temperature in the order degree of the long acyl chains of DPPG.

N lc decreases when the model membrane is cooled, in order to assure the returning of the system into its gel phase. The reversibility of the transformation (1) has been experimentally demonstrated [12] by usıng wavenumber modıfıcatıon of CH2 stretchıng modes measured in the DPPG model membranes. So, by temperature increasing the wavenumbers of the CH2 stretching modes increased, showing a sharp modification at 41.5 °C (the melting point of DPPG) and then, at the sample cooling, they decreased in the same way to the values of the wavenumbers corresponding to the gel phase. Supposing that all the phospholipids acyl chains are ordered in the gel phase of the sample one can write: Nðefg  eig Þ ffi h  cðmg  m0 Þ efg

ð3Þ

eig

where and are the interaction energies between (N  1) molecules in their ground vibration state and one molecule of the sample in its final (f) and initial (i) states of the spectral transition, when all molecules are in the gel phase (g): mg is the wavenumber measured in the gel phase of the system and m0 is the wavenumber measured in the gas phase of the same system. A similar relation can be written for the liquid crystalline phase of the system: Nðeflc  eilc Þ ffi h  cðmlc  m0 Þ

ð4Þ eflc

eilc

by using the interaction energies and between (N  1) molecules of the system in their ground vibration state and one molecule of this system in its vibration states (f and i) participating to the IR transition, when all molecules of the system are in liquid crystalline phase. The wavenumbers mlc and m0 correspond to the system in its liquid crystalline and gas phases, respectively. For the system at a given temperature T, different from Tm, one can define the average ratios of the molecules from the gel and liquid crystalline phases, by: Pg 

g N N

and

Plc 

 lc N N

ð5Þ

The ratios P g and P lc satisfy the relation: Pg þ Plc  1

ð6Þ

On the other hand, for the system at a given temperature T, having Ng molecules in the gel phase and Nlc molecules in liquid crystalline phase, one obtains:  g ðef  ei Þ þ N  lc ðef  ei Þ  h  cðm  m0 Þ N g g lc lc

ð7Þ

 g and N  lc , The averages values of Ng and Nlc , denominated by N respectively, were introduced because they determine the shifts

Fig. 1. Structural characteristics of DPPG.

F. Severcan, D.-O. Dorohoi / Journal of Molecular Structure 887 (2008) 117–121

119

in IR spectra of DPPG model membrane at a given temperature. In relation (7), m is the wavenumber measured when the studied sample is at temperature T. From Eqs. (7), (3), (4), and (5) one obtains:

Ozaki [15]. The calculus was made in the matrix form using Hilbert transform. The synchronous matrix is:

mlc  m Pg  mlc  mg m  mg P lc  mlc  mg

U¼ ð8Þ

1 MM T n1

and the asynchronous matrix is given by: ð9Þ W¼

It results that the ratios of the molecules in the gel phase and in the liquid crystalline phase corresponding to the system at temperature T can be estimated by the wavenumbers of the IR transition measured for the sample in the gel phase ðmg Þ, in the liquid crystalline phase ðmlc Þ and in an intermediate phase ðmÞ, consisting from a mixture of these phases. Wavenumber m increases with the temperature increasing. So, relations (8) and (9) are indicators of the Plc increasing and of the Pg decreasing by temperature increasing. It results that the ratios of the subsystems in the gel phase and in the liquid crystalline phase corresponding to the studied DPPG model membrane at temperature T can be estimated by the wavenumbers of the ACH2 IR symmetric and asymmetric stretching modes measured for the system in the gel phase ðmg Þ, in the liquid crystalline phase ðmlc Þ and in an intermediate phase ðmÞ, consisting of a mixture of these phases.

1 MðHMT Þ n1

4.1. Main phase transition reflected in stretching modes of the acyl chains of DPPG model membrane The spectral range {2817–2955} cm1 is widely studied in literature [5,8,9]. The temperature induced spectral modifications in the range 2800–3000 cm1 in FTIR spectra of the DPPG model membrane are illustrated in Fig. 3. One can observe the increase of the intensity and bandwidth of the symmetric and asymmetric stretching bands of DPPG model membrane when temperature increases. The bandwidth increase by temperature rising denotes the increase of the disorder in the hydrophobic tails of the DPPG (Fig. 3). Having a high degree of order, the gel phase of the model membrane offers fewer possibilities for the vibrational and rotational motions of the acyl chains than the liquid crystalline phase, where the low degree of order enlarges the dynamics of these motions. Consequently, the IR bands corresponding to the symmetric or asymmetric modes of the acyl chains of DPPG are shifted to the high frequencies when the model membranes pass from the gel to the liquid crystalline phase (Fig. 3). Membrane properties were investigated using the two-dimensional correlation spectroscopy methods developed by Noda and

Fig. 4. DPPG model membrane synchronous correlation spectra.

Absorbance (a.u)

1.0

2950

3000

45

m

2900

Te

2850

pe

ra

tu

re

(o C

0.0 25 27 31 33 39 43

)

0.5

-1

W avenumber (cm ) Fig. 3. FT-IR spectra of DPPG model membrane at some temperature.

ð11Þ

where M is the matrix of spectra; MT is transposed matrix M; n is number of spectra in series and H is Hilbert–Noda transform matrix (i.e. Hmn = 0 if m = n and Hmn = 1/p(n  m) otherwise). A program written in Matlab was developed [13] in order to obtain mean spectrum and dynamic spectra, as well as synchronous and asynchronous 2D correlation graphs (Figs. 4 and 5). The mean spectrum was obtained by averaging FTIR data for different temperatures. From the dynamic spectrum it results that for T < Tm, the intensities of 2850 and 2917 cm1 lines are bigger than in the mean spectrum. On the contrary, for T > Tm, the intensities of these bands are smaller than those in the main spectrum. This observation could be a confirmation of the changes in the relative positions of the CH2 groups in the acyl chains of DPPG.

4. Results and discussion

2800

ð10Þ

Fig. 5. DPPG model membrane asynchronous correlation spectra.

F. Severcan, D.-O. Dorohoi / Journal of Molecular Structure 887 (2008) 117–121

Using the two-dimensional correlation spectroscopy techniques we can obtain information about simultaneous or coincidental changes (using temperature as perturbation parameter) of two separate intensity variations measured at two different wavenumbers from synchronous spectra and out of phase changes from asynchronous spectra, respectively. From Figs. 4 and 5 it results a very good correlation between the symmetric and asymmetric vibrations in the CH2 groups of the acyl chains, showing simultaneous changes of the bands concerning these vibrations with the temperature modification. From these maps we can see that almost all the correlations between different FTIR bands are synchronous (maximum value for synchronous map is 0.2722 compare to the 0.02247 from asynchronous map). There are two strong autocorrelation peaks at about 2917 and 2848.5 cm1 (peaks from main diagonal in Fig. 4). The off diagonal peaks indicates that there are positive correlations for 2917 and 2864 cm1 pair bands and negative for 2864 and 2848.5 cm1. This is clear due to both temperature changes in the membranes and transition between gel phase and crystal liquid phase. 4.2. The degree of order in model membrane

12

11

10

9

8 25

30

35

40

45

50

55

60

65

t (°C) Fig. 7. Dependence of CH2 asymmetric stretching mode bandwidth on temperature.

Table 1 Regression coefficients from (4)

From the graphs of the wavenumbers or of the bandwidths corresponding to the symmetric and asymmetric stretching modes of the ACH2 groups of the hydrocarbon tails one can easily establish the main phase transition, because the phase transition of the type (1) is very quick and spontaneously installed. In Figs. 6 and 7 the temperature dependences of the wavenumbers and also of the bandwidths of the bands 2850 and 2917 cm1 are illustrated. A very quick transition of gel phase to the liquid crystalline phase takes place at the main phase transition temperature as it results from Figs. 6 and 7. This temperature is easily established from graphs and by using the formula [11]: dm ¼

13

Δνa (cm-1)

120

dmm  dmM   þ dmM m 1 þ exp dtdt dt mM

ð12Þ

DPPG DPPC

dmm ðcm1 Þ

dmM ðcm1 Þ

dtmM ð CÞ

dtm ð CÞ

sym.

asym.

sym.

asym.

sym.

asym.

sym.

asym.

0.22 0.15

0.51 0.24

3.62 3.01

5.81 6.10

0.41 0.29

0.53 0.47

14.82 13.85

14.89 13.95

In (13), m0 is the wavenumber registered at the main phase transition for the symmetric and asymmetric stretching modes of the ACH2 groups. As initial data were considered the following values: ti ¼ 26:5  C; mi ðsym:Þ ¼ 2850:0 cm1 and mi ðasym:Þ ¼ 2917:2 cm1

where dm ¼ m  mi is the spectral shift measured relatively to an initial value mi registered at initial temperature t i ; dt ¼ t  t i with t -current temperature; dtm ¼ t m  ti ; t m -the main phase transition temperature T m ¼ 273:15 þ t m ; dmm and dmM are the minim and maxim of the spectral shifts; dt m;M is the smallest temperature interval corresponding to the values dmm and dmM . When dt ¼ dt m , it results: dmm þ dmM dm ¼ ¼ dm0 ¼ m0  mi 2

ð13Þ

The results of statistical analysis of the data by using relation (4), are given in Table 1, where the data previously obtained for DPPC model membrane [11] are given in the last row, for comparison. From the data in Table 1 it results the coordinates of the main phase transition of DPPG model membrane. tm ¼ 41:36  Cm0 ðsymÞ ¼ 2852:2 cm1 and m0 ðasym:Þ ¼ 2920:16 cm1

2854.0 100

2853.5 2853.0

80

2852.0

60

NG

νs (cm-1)

2852.5

2851.5

40

2851.0 2850.5

20

2850.0 0

2849.5 25

30

35

40

45

50

55

60

65

t (°C)

25

30

35

40

45

50

55

60

65

t (°C) Fig. 6. Dependence of CH2 symmetric stretching mode wavenumber on temperature.

Fig. 8. The average percent of gel phase when temperature increases.

F. Severcan, D.-O. Dorohoi / Journal of Molecular Structure 887 (2008) 117–121

121

vibration modes offer information about the degree of order in the model membranes. The knowledge of the temperature influence on the model membrane fluidity is very important when the interactions drugs-model membranes are studied. So, this study is important to read-in the mechanism of pulmonary membrane disruption in lung diseases, influenced by temperature and drugs. Further work is the study of antibiotic influence on the phospholipids model membrane fluidity on the basis of the relations established in this paper.

100

80

N LC

60

40

20

Acknowledgements

0

The partial support by the NATO Tubitak Committee is gratefully acknowledged by the authors. 25

30

35

40

45

50

55

60

65

t (°C) Fig. 9. The average percent of liquid crystalline phase when temperature increases.

On the basis of formulas (8) and (9), the percentage of the gel and liquid crystalline phases were established at each current temperature. The results are given in Figs. 8 and 9. From these figures it results a decrease of the average number of the molecules in an ordered phase and an increase of the molecules in liquid crystalline phase, when temperature increases. 5. Conclusions The main phase transition temperature and the phase composition of the phospholipids membranes can be established by using the wavenumbers of the ACH2 symmetric and asymmetric modes from FTIR spectrum. The spectral modifications from the ACH2

References [1] M. Jackson, H. Mantsch, Spectrochim. Acta Rev. 15 (1993) 53. [2] M.K. Jain, Introduction to Biological Membranes, John Wiley and Sons, 1988. [3] C. Branden, J. Tooze, Introduction to Protein Structure, Gareand Publishing, Inc., New York, London, 1991. [4] A. Blume, W. Hubner, G. Messner, Biochemistry 27 (1988) 8239. [5] J.R. Arrondo, F.M. Gont, J.M. Macarulla, Biochim. Biophys. Acta 794 (1984) 165. [6] E. Egberts, S.J. Marrink, H.J.C. Berenson, Eur. Biophys. J. 22 (1994) 423. [7] B.G. Akinoglu, M. Gheith, F. Severcan, J. Mol. Struct. 565–566 (2001) 281. [8] S. Choi, W. Ware, S.R. Lauterbach, W.M. Philips, Biochemistry 30 (1991) 8563. [9] L.C. Stewart, M. Kates, Biochem. Cell Biol. 68 (1989) 266. [10] F. Severcan, H.O. Durmus, F. Eker, P.I. Haris, B.G. Akinoglu, Talanta 53 (2000) 205. [11] F. Severcan, S. Tokmak, C. Agheorghiesei, D.O. Dorohoi, Ann. Univ. Al. I. Cuza, Iasßi, s. Chimie 10 (2002) 259. [12] C. Stan, R. Cristescu, F. Severcan, D. Dorohoi, Mol. Cryst. Liq. Cryst. 457 (2006) 27. [13] F. Severcan, C. Agheorghiesei, D.O. Dorohoi, Rev. Chim. Bucuresßti 59 (3) (2008) 356. [14] F. Severcan, F. Kokmaz, M. Aflori, D. Dorohoi, Digest J. Nanomat. Biostruct. 3 (2) (2008) 55. [15] Isao Noda, Yukihiro Ozaki, Two-Dimensional Correlation Spectroscopy, John Wiley and Sons, 2004.