Neuroscience Letters, 141 (1992) 47 52 © 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
47
NSL 08723
Near-infrared FT-Raman spectra of the rat brain tissues A r i t a k e M i z u n o a, Takashi Hayashi", K o u i c h i T a s h i b u b, Shuichi M a r a i s h i c, K a z u a k i K a w a u c h i c a n d Yukihiro Ozaki d aDepartment of Biochemistry and bDepartment of Neurosurgery, The Jikei University School of Medicine, Tokyo (Japan), "Division of Analytical Instruments, JEOL Ltd. Tokyo (Japan) and dDepartment of Chemistry, School of Science, Kwansei Gakuin University, Nishinomiya, Hyogo (Japan) (Received 13 January 1992; Revised version received 11 March 1992; Accepted 31 March 1992)
Key words. Raman spectroscopy; Fourier transform; Rat brain; Gray matter; White matter; Caudate-putamen; Thalamus Near-infrared Fourier transform (FT) Raman spectroscopy was applied to brain tissues in situ. The spectra were obtained from the cerebral cortex, white matter of the cerebrum, caudate-putamen, thalamus, synaptosomal fraction, and myelin fraction. High-quality Raman spectra in the 400 to 2940 cm-~ range were measured without interference of autofluorescence. Common spectral bands were assigned. The ratios of the intensity at 1664 (amide I), 1442 (CH2 deformation), 2885 (CH2 asymmetric stretching), 2938 cm -~ (CH 3 symmetric stretching) could be used for differentiation between the gray and white matters.
The brain is composed of highly complex materials making it difficult to search its compositions in vivo or in situ at a molecular level. The Raman spectroscopic technique is one of the most powerful non-invasive probes for investigating biological materials [11, 14, 16]. Two major disadvantages for biological or clinical application of Raman spectroscopy have been pointed out: interference by internal fluorescence with relatively low Raman scattering intensity and laser beam-induced decomposition of biological materials [5, 11, 14]. Tashibu attempted to measure Raman spectra in situ from rat brain tissue using a conventional Raman spectrometer and Argon ion laser [13]. He succeeded in estimating the relative content of water from the spectral region of 2800 to 3800 cm -l in normal and edematous brain tissues. It was relatively easy to get good spectra in the 2800-3800 cm -1 region because the intensity of a band at 3390 cm -1 was strong due to the OH stretching mode of water. However, he could not obtain good spectra in situ from brain tissues in the region below 2800 cm -1 because of the interference of autofluorescence and relatively low sensitivity of the conventional Raman spectrometer with a 514.5 nm excited beam provided by an Argon ion laser [13]. The autofluorescent substances in living materials are generally deposited with aging or pathological effects Correspondence." A. Mizuno, Department of Biochemistry, The Jikei University School of Medicine, 3-25-8 Nishi-shinbashi, Minato-ku, Tokyo 105, Japan.
[11, 14]. Recently developed near-infrared Fourier transform (FT) Raman spectroscopy has nearly solved such disadvantages; using a 1064 nm laser beam from a NdYAG laser, very little fluorescence would come out from the fluorescence-rich biological materials, and light beam-induced decomposition would be minimum [9, 10, 12]. High-quality Raman spectra from rat brain tissues were measured to demonstrate that near-infrared FTRaman spectroscopy could be a powerful non-destructive structural probe for further medical or other practical applications. Male 8-week-old Wistar rats were used in the present study. The brain was taken out from the skull under pentobarbital anesthesia (60 mg/kg b.w., i.p.). It was placed on a glass dish filled with a Krebs-Ringer bicarbonate solution. The brain was sliced to examine the white matter, caudate-putamen and thalamus. The subcellular fractionation was performed according to the method of Whittacker to obtain synaptosomal and myelin fractions [15]. The fractions were used as pellets after centrifugation for the Raman measurements. The FT-Raman system used in this study was a JEOL JRS-FT 6500 N equipped with an InGaAs detector. Excitation light at 1064 nm wavelength was provided by a CW Nd:YAG laser, and laser power was 200-300 mW at the sample position. All data were collected at 8 cm-' spectral resolution and usually 1000-3000 scans were accumulated for an acceptable S/N ratio. For the Raman measurements the laser beam was fo-
48
40.00-
32.00
,?,
2
F~ 24.00
16.00
8.00
0.00
28b0
24b0
20bO
16b0
b
1200
sob
4oh
RAMAN S H I F T ( C M - 1 )
Fig. 1. Near-infrared FT R a m a n spectrum from rat cerebral cortex. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a summation of 1500 scans.
cused on the surface of a sample from upward. The spot size was about 0.1 mm in diameter. The scattered light was collected at 180 ° with an ellipsoidal mirror [12]. Raman spectra were measured from four brain structures, including the cerebral cortex, white matter of the cerebrum, caudate-putamen and thalamus, and synaptosomal and myelin fractions. Figs. I 4 show Raman spectra obtained from the cerebral cortex, white matter of the cerebrum, synaptosomal
fraction and myelin tYaction of the rat brain, respectively. The acceptable spectral region seen in these figures was between 400 and 2940 cm -~. The main peaks of the spectra were attributed mainly to the compositions of proteins and phospholipids. There were common bands among the spectra in these four figures. Bands at 1664, 1272, and 1006 cm ~were assigned to amide I, amide III, and the breathing mode of phenylalanine of proteins, respectively. Sharp peaks at 1442 cm -~ assigned to CH,
i 40.00
20.00.
i !,~i ?
0.00 2sbo
'
24bo
"
2obo RAMAN
'
~ebo
SHWT(CM-~)
12bo
sob
~oG
Fig. 2. Near-infrared FT R a m a n spectrum from white matter of rat cerebrum. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a summation of 1000 s c a n s
49 20.00-
15.00
10.00-
i -'TiT.,,!
5.00.
~'o
Ln
0.00
- 5 . O0
2,bo
24b0
zobo RAMAN
~6bo SHIFTtCM-1)
~2bo
80b
40b
Fig. 3. Near-infrared FT Raman spectrum from rat brain synaptosomalfraction. Excitationpower by Nd-YAGlaser was 200 mW at sample position. The spectrum is a summation of 2000 scans.
deformation modes originated from both lipids and proteins [5, 14]. The ratio of the intensity of R a m a n bands at 1442 and 1662 cm -1 (I1442/I166~) can be used as the relative content ratio of proteins and lipids, and therefore as a good marker for estimating the ratio of gray matter and white matter. The value of/1442/11662 in the gray matter region of the cerebral cortex was much less than that in
the white matter region of the cerebral cortex. Similar spectra were observed between synaptosomal and myelin fractions, indicating the content ratio of proteins and lipids. Therefore, the ratio of the intensities at 1664 and 1442 cm -1 may be a good marker to indicate the content ratio of proteins and lipids to differentiate gray matter from white matter.
12.00-
J, 8.00
4.00
0.00-
-4.00.
28b0
24b0
20bo RAMAN
lsbo
~2bo
80b
40b
SHIFT(CH-I)
Fig. 4. Near-inffared FT Raman spectrum ffomrat brain myelin fraction. Excitation power by Nd-YAGlaser was 200mWatthe sample position. The spectrum is a summation of3000scans.
50
t: O.¢O00-
z c~
z w >
.t, ,
~t.. ,
~o=
,
0.200~
\ 0.0000
28bo
246o
2obo
l~bo
12bo
8o1~
,rob
RAMAN SHIFT(OH~l)
Fig. 5. Near-infrared FT R a m a n spectrum from rat caudate-putamen. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a s u m m a t i o n of 2000 scans.
The appearances of the peak at 1271 c m -I of amide III and that at 1662 cm -I of amide I in Figs. 1-4 indicated that the main secondary structure of brain proteins was an a-helix [5, 14]. There were several bands from phospholipids at 1302 (CH2 twist and wagging), 1132 (C-C stretching), 1086 (C-C stretching and PO 2- symmetrical stretching), 1066 (C-O stretching and C-O-C sym-
metric stretching), and 746 cm -~ (P-O-P symmetrical stretching). Those bands appeared predominantly in the white matter (Fig. 2) and myelin fraction (Fig. 4) where the phospholipid components were enriched [6, 7, 14]. The spectral differences between the synaptosomal (Fig. 3) and the myelin fractions (Fig. 4) were found for the intensities of the bands at 525, 553, and 703 c m -1
r~¢O
I ®® i , = 1
i ,
O. 4000~
Z
>
n-
0.2ooo
0.0000
2abO
2J,bO
20bo ISbO RAHAN ~H l FT (¢H-1 i
12bo
8ob
4ob
Fig. 6. Near-infrared FT R a m a n spectrum from rat thalamus (ventromedial nucleus). Excitation power of Nd-YAG laser was 300 m W at the sample position. The spectrum is a s u m m a t i o n of 2000 scans.
51
which were assigned to the S-S stretching mode of the gauche-gauche-trans (G-G-T) form, that of the transgauche-trans (T-G-T) form [6], and the C-N stretching mode from the choline head in phosphatidylcholine and sphingomyelin, respectively [1]. In the synaptosomal fraction, unusual S-S bonds appeared at 525 and 553 cm -1. The frequency of 553 cm -I was rather high for the usUal T-G-T form (up to 545 cm -1) and, therefore, the conformation of some S-S bonds in the synaptosomal fraction was slightly different from the normal T-G-T form. The sharp peak at 703 cm x in the myelin fraction was thought to be due to choline-rich phospholipids. The peak at 704 cm -~ has two possibilities; one is the C-N stretching mode, mainly from phospholipid polar head choline [1], and another is the C-S stretching vibration, due to methionine, cysteine, and cystine residues of the proteins [14]. The major proteins of myelin were myelin basic protein and proteolipid protein. Methionine showed a sharp peak around 704 cm -1, whereas myelin basic protein contained 3 methionines in 195 amino acid residues [3, 4], and proteolipid protein contained only one methionine among 267 amino acid residues [2]. Since the concentration of protein was very small in the white matter of the myelin fraction, the C-S bond of methionine seemed not to contribute to a Raman band at 704 cm 1. It is most likely that the peak at 704 cm -~ in the white matter and myelin fraction was due to the C-N stretching mode of the polar choline in phosphatidylcholine and sphingomyelin [1]. Three peaks at 1060, 1090, and 1130 cm -1 varied in the fatty acid chain length and in the ratio of trans-gauche isomers of fatty acid chains [6, 7]. The intensity at 1090 cm -1 was the highest in the cerebral cortex (Fig. 1) and the second highest in the synaptosomal fraction (Fig. 3). The intensity at 1090 cm ~ was negligible in the white matter (Fig. 3) and low in the myelin fraction (Fig. 4). These results suggest that the trans isomers of fatty acid chains are enriched in the gray matter and synaptosomal fraction, whereas the gauche isomers are enriched in the white matter and myelin fraction. In the region of 2800-3000 cm -1, CH vibrational bands appeared at 2852 (CH2 symmetric stretching), 2885 (CH2 asymmetric stretching), 2938 (CH 3 symmetric stretching), and 2958 cm -~ (CH3 asymmetric stretching). The peak at 2938 cm -I was relatively high in all tissue specimen, and the peaks at 2852 and 2885 cm ~were high in the white matter and myelin fraction (Figs. 2 and 4) and low in the gray matter and synaptosomal fraction (Figs. 1 and 3). The ratio of the intensities of the bands between 2885 and 2938 cm -~ may also be a good marker for the discrimination between the gray matter and white matter. Figs. 5 and 6 show the spectra of the caudate-putamen
and thalamus. These spectra closely resemble each other. In the caudate-putamen and thalamus, the bands of amide I and amide III appear at 1659 and 1269 cm -1, respectively. The main structure of the proteins in these areas is an 0~-helix. The lipid and protein ratios (11659/ I1439) are similar to those of cerebral cortex (Fig. 1) and synaptosomal fractions (Fig. 3). In the region of 28003000 cm -l, CH 2 stretching symmetric and asymmetric modes were observed at 2873 and 2850 cm -l, while CH3 stretching symmetric and asymmetric modes at 2938 and 2958 cm -I were not observed in the spectra of the caudate-putamen and thalamus. The ratios o f I1084 to II126 o r I1065 in the spectra of the caudate-putamen and thalamus were similar to the ratio of the cerebral cortex or synaptosomal fraction which means that the fatty acids of these areas are enriched in trans type [5, 6]. In the spectrum of the caudate-putamen (Fig. 5), a small band of the T-G-T form of S-S bond was observed at 543 cm -1. The present study indicates that near-infrared FTRaman spectroscopy is promising in monitoring the changes of relative content of proteins and lipids or the changes of their structure or component content ratio in various portions of brain tissues.
1 Akutsu, H., Direct determination by Raman scattering of the conformation of the choline group in phospholipid bilayers, Biochemistry, 20 (1981) 7359-7366. 2 Culatolo, W., Verma, S.E, Sakura, J.D., Small, D.M., Shipley, G.G. and Wallach, D.F.H., Structural effects of myelin proteolipid apoprotein on phospholipids: a Raman spectroscopic study, Biochemistry, 17 (1978) 1802-1807. 3 Dunkley, ER. and Carnegie, ER., Amino acid sequence of the smaller basic protein from rat brain myelin, Biochem. J., 141 (1974) 243-255. 4 Ferra, F., Engh, H., Hudson, L., Kamholz, J., Puckett, C., Molineaux, S. and Lazzarini, R.A., Alternative splicing accounts for the four forms of myelin basic protein, Cell, 43 (1985) 721-727. 5 Krimm, S., Peptides and proteins. In T.G. Spiro (Ed.), Biological Applications of Raman Spectroscopy, Vol. l, Wiley, New York, 1987, pp. 1~6. 6 Mendelsohn, R., Laser Raman spectroscopic study of egg lecithin and egg lecithin-cholesterol mixtures, Biochem. Biophys. Acta, 290 (1972) 15-21. 7 Milanovich, F.E, Shore, B., Harney, R.C. and Tu, A.T., Raman spectroscopic analysis of Dutch belt rabbit erythrocyte hosts, Chem. Phys. Lipids, 17 (1976) 79-84. 8 Milner, B.J., Lai, C., Nave, K.-A., Lenoir, D., Ogata, J. and Sutcliffe, J.G., Nucleotide sequences of two mRNAs for rat brain myelin proteolipid protein, Cell, 42 (1985) 931-939. 9 Nie, S., Bergbauer, K.L., Kuck Jr., J.ER. and Yu, N.-T., Nearinfrared Fourier transform Raman spectroscopy in human lens research, Exp. Eye Res., 51 (1990) 619-623. 10 Nie, S., Marzilli, L.G. and Yu, N.-T., Near-infrared Fourier transform Raman spectroscopy of photolabile organ cobalt Bt2 and model compounds 1. Detection of cobalt-carbon stretching mode in solid state and in solution, J. Am. Chem. Soc., 111 (1989) 92569258.
52 11 Ozaki, Y. and Iriyama, K., Potential of Raman spectroscopy in medical science. In J. Spepanek, P. Anzenbacher and B. Sedlacek (Eds.), Laser Scattering Spectroscopy of Biological Objects, Elsevier, Amsterdam, 1987, pp. 559-582. 12 Ozaki, Y., Mizuno, A., Sato, H., Kawauchi, K. and Murauchi, S., Biomedical application of near-infrared Fourier-transform Raman spectroscopy I. The 1064 nm-excited Raman spectra of blood and met hemoglobin, Appl. Spectrosc., 46 (1992) 335-344. 13 Tashibu, K., Analysis of water content in rat brain using Raman spectroscopy, Brain Nerve, 42 (1990) 999-1004.
14 Tu, A.T., Raman Spectroscopy in Biology, Wiley, New York, 1982, 448 pp. 15 Wittacker, V.P. and Baker~ L.A., The subcellular fractionation of brain tissue with special reference to the preparation of synaptosomes and their component organelles. In R. Fried (Eds.), Method of Neurochemistry, Marcel Dekker, New York, 1972, pp. 1-52.
16 Yu, N.-T., DeNagel, D.C., Ho, D.J.-Y. and Kuck Jr., J.F.R., Ocular lenses. In T.G. Spiro (Ed.), Biological Applications of Raman Spectroscopy, Vol. 1, Wiley, New York, 1987, pp. 47 80.
47
NSL 08723
Near-infrared FT-Raman spectra of the rat brain tissues A r i t a k e M i z u n o a, Takashi Hayashi", K o u i c h i T a s h i b u b, Shuichi M a r a i s h i c, K a z u a k i K a w a u c h i c a n d Yukihiro Ozaki d aDepartment of Biochemistry and bDepartment of Neurosurgery, The Jikei University School of Medicine, Tokyo (Japan), "Division of Analytical Instruments, JEOL Ltd. Tokyo (Japan) and dDepartment of Chemistry, School of Science, Kwansei Gakuin University, Nishinomiya, Hyogo (Japan) (Received 13 January 1992; Revised version received 11 March 1992; Accepted 31 March 1992)
Key words. Raman spectroscopy; Fourier transform; Rat brain; Gray matter; White matter; Caudate-putamen; Thalamus Near-infrared Fourier transform (FT) Raman spectroscopy was applied to brain tissues in situ. The spectra were obtained from the cerebral cortex, white matter of the cerebrum, caudate-putamen, thalamus, synaptosomal fraction, and myelin fraction. High-quality Raman spectra in the 400 to 2940 cm-~ range were measured without interference of autofluorescence. Common spectral bands were assigned. The ratios of the intensity at 1664 (amide I), 1442 (CH2 deformation), 2885 (CH2 asymmetric stretching), 2938 cm -~ (CH 3 symmetric stretching) could be used for differentiation between the gray and white matters.
The brain is composed of highly complex materials making it difficult to search its compositions in vivo or in situ at a molecular level. The Raman spectroscopic technique is one of the most powerful non-invasive probes for investigating biological materials [11, 14, 16]. Two major disadvantages for biological or clinical application of Raman spectroscopy have been pointed out: interference by internal fluorescence with relatively low Raman scattering intensity and laser beam-induced decomposition of biological materials [5, 11, 14]. Tashibu attempted to measure Raman spectra in situ from rat brain tissue using a conventional Raman spectrometer and Argon ion laser [13]. He succeeded in estimating the relative content of water from the spectral region of 2800 to 3800 cm -l in normal and edematous brain tissues. It was relatively easy to get good spectra in the 2800-3800 cm -1 region because the intensity of a band at 3390 cm -1 was strong due to the OH stretching mode of water. However, he could not obtain good spectra in situ from brain tissues in the region below 2800 cm -1 because of the interference of autofluorescence and relatively low sensitivity of the conventional Raman spectrometer with a 514.5 nm excited beam provided by an Argon ion laser [13]. The autofluorescent substances in living materials are generally deposited with aging or pathological effects Correspondence." A. Mizuno, Department of Biochemistry, The Jikei University School of Medicine, 3-25-8 Nishi-shinbashi, Minato-ku, Tokyo 105, Japan.
[11, 14]. Recently developed near-infrared Fourier transform (FT) Raman spectroscopy has nearly solved such disadvantages; using a 1064 nm laser beam from a NdYAG laser, very little fluorescence would come out from the fluorescence-rich biological materials, and light beam-induced decomposition would be minimum [9, 10, 12]. High-quality Raman spectra from rat brain tissues were measured to demonstrate that near-infrared FTRaman spectroscopy could be a powerful non-destructive structural probe for further medical or other practical applications. Male 8-week-old Wistar rats were used in the present study. The brain was taken out from the skull under pentobarbital anesthesia (60 mg/kg b.w., i.p.). It was placed on a glass dish filled with a Krebs-Ringer bicarbonate solution. The brain was sliced to examine the white matter, caudate-putamen and thalamus. The subcellular fractionation was performed according to the method of Whittacker to obtain synaptosomal and myelin fractions [15]. The fractions were used as pellets after centrifugation for the Raman measurements. The FT-Raman system used in this study was a JEOL JRS-FT 6500 N equipped with an InGaAs detector. Excitation light at 1064 nm wavelength was provided by a CW Nd:YAG laser, and laser power was 200-300 mW at the sample position. All data were collected at 8 cm-' spectral resolution and usually 1000-3000 scans were accumulated for an acceptable S/N ratio. For the Raman measurements the laser beam was fo-
48
40.00-
32.00
,?,
2
F~ 24.00
16.00
8.00
0.00
28b0
24b0
20bO
16b0
b
1200
sob
4oh
RAMAN S H I F T ( C M - 1 )
Fig. 1. Near-infrared FT R a m a n spectrum from rat cerebral cortex. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a summation of 1500 scans.
cused on the surface of a sample from upward. The spot size was about 0.1 mm in diameter. The scattered light was collected at 180 ° with an ellipsoidal mirror [12]. Raman spectra were measured from four brain structures, including the cerebral cortex, white matter of the cerebrum, caudate-putamen and thalamus, and synaptosomal and myelin fractions. Figs. I 4 show Raman spectra obtained from the cerebral cortex, white matter of the cerebrum, synaptosomal
fraction and myelin tYaction of the rat brain, respectively. The acceptable spectral region seen in these figures was between 400 and 2940 cm -~. The main peaks of the spectra were attributed mainly to the compositions of proteins and phospholipids. There were common bands among the spectra in these four figures. Bands at 1664, 1272, and 1006 cm ~were assigned to amide I, amide III, and the breathing mode of phenylalanine of proteins, respectively. Sharp peaks at 1442 cm -~ assigned to CH,
i 40.00
20.00.
i !,~i ?
0.00 2sbo
'
24bo
"
2obo RAMAN
'
~ebo
SHWT(CM-~)
12bo
sob
~oG
Fig. 2. Near-infrared FT R a m a n spectrum from white matter of rat cerebrum. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a summation of 1000 s c a n s
49 20.00-
15.00
10.00-
i -'TiT.,,!
5.00.
~'o
Ln
0.00
- 5 . O0
2,bo
24b0
zobo RAMAN
~6bo SHIFTtCM-1)
~2bo
80b
40b
Fig. 3. Near-infrared FT Raman spectrum from rat brain synaptosomalfraction. Excitationpower by Nd-YAGlaser was 200 mW at sample position. The spectrum is a summation of 2000 scans.
deformation modes originated from both lipids and proteins [5, 14]. The ratio of the intensity of R a m a n bands at 1442 and 1662 cm -1 (I1442/I166~) can be used as the relative content ratio of proteins and lipids, and therefore as a good marker for estimating the ratio of gray matter and white matter. The value of/1442/11662 in the gray matter region of the cerebral cortex was much less than that in
the white matter region of the cerebral cortex. Similar spectra were observed between synaptosomal and myelin fractions, indicating the content ratio of proteins and lipids. Therefore, the ratio of the intensities at 1664 and 1442 cm -1 may be a good marker to indicate the content ratio of proteins and lipids to differentiate gray matter from white matter.
12.00-
J, 8.00
4.00
0.00-
-4.00.
28b0
24b0
20bo RAMAN
lsbo
~2bo
80b
40b
SHIFT(CH-I)
Fig. 4. Near-inffared FT Raman spectrum ffomrat brain myelin fraction. Excitation power by Nd-YAGlaser was 200mWatthe sample position. The spectrum is a summation of3000scans.
50
t: O.¢O00-
z c~
z w >
.t, ,
~t.. ,
~o=
,
0.200~
\ 0.0000
28bo
246o
2obo
l~bo
12bo
8o1~
,rob
RAMAN SHIFT(OH~l)
Fig. 5. Near-infrared FT R a m a n spectrum from rat caudate-putamen. Excitation power by Nd-YAG laser was 300 m W at the sample position. The spectrum is a s u m m a t i o n of 2000 scans.
The appearances of the peak at 1271 c m -I of amide III and that at 1662 cm -I of amide I in Figs. 1-4 indicated that the main secondary structure of brain proteins was an a-helix [5, 14]. There were several bands from phospholipids at 1302 (CH2 twist and wagging), 1132 (C-C stretching), 1086 (C-C stretching and PO 2- symmetrical stretching), 1066 (C-O stretching and C-O-C sym-
metric stretching), and 746 cm -~ (P-O-P symmetrical stretching). Those bands appeared predominantly in the white matter (Fig. 2) and myelin fraction (Fig. 4) where the phospholipid components were enriched [6, 7, 14]. The spectral differences between the synaptosomal (Fig. 3) and the myelin fractions (Fig. 4) were found for the intensities of the bands at 525, 553, and 703 c m -1
r~¢O
I ®® i , = 1
i ,
O. 4000~
Z
>
n-
0.2ooo
0.0000
2abO
2J,bO
20bo ISbO RAHAN ~H l FT (¢H-1 i
12bo
8ob
4ob
Fig. 6. Near-infrared FT R a m a n spectrum from rat thalamus (ventromedial nucleus). Excitation power of Nd-YAG laser was 300 m W at the sample position. The spectrum is a s u m m a t i o n of 2000 scans.
51
which were assigned to the S-S stretching mode of the gauche-gauche-trans (G-G-T) form, that of the transgauche-trans (T-G-T) form [6], and the C-N stretching mode from the choline head in phosphatidylcholine and sphingomyelin, respectively [1]. In the synaptosomal fraction, unusual S-S bonds appeared at 525 and 553 cm -1. The frequency of 553 cm -I was rather high for the usUal T-G-T form (up to 545 cm -1) and, therefore, the conformation of some S-S bonds in the synaptosomal fraction was slightly different from the normal T-G-T form. The sharp peak at 703 cm x in the myelin fraction was thought to be due to choline-rich phospholipids. The peak at 704 cm -~ has two possibilities; one is the C-N stretching mode, mainly from phospholipid polar head choline [1], and another is the C-S stretching vibration, due to methionine, cysteine, and cystine residues of the proteins [14]. The major proteins of myelin were myelin basic protein and proteolipid protein. Methionine showed a sharp peak around 704 cm -1, whereas myelin basic protein contained 3 methionines in 195 amino acid residues [3, 4], and proteolipid protein contained only one methionine among 267 amino acid residues [2]. Since the concentration of protein was very small in the white matter of the myelin fraction, the C-S bond of methionine seemed not to contribute to a Raman band at 704 cm 1. It is most likely that the peak at 704 cm -~ in the white matter and myelin fraction was due to the C-N stretching mode of the polar choline in phosphatidylcholine and sphingomyelin [1]. Three peaks at 1060, 1090, and 1130 cm -1 varied in the fatty acid chain length and in the ratio of trans-gauche isomers of fatty acid chains [6, 7]. The intensity at 1090 cm -1 was the highest in the cerebral cortex (Fig. 1) and the second highest in the synaptosomal fraction (Fig. 3). The intensity at 1090 cm ~ was negligible in the white matter (Fig. 3) and low in the myelin fraction (Fig. 4). These results suggest that the trans isomers of fatty acid chains are enriched in the gray matter and synaptosomal fraction, whereas the gauche isomers are enriched in the white matter and myelin fraction. In the region of 2800-3000 cm -1, CH vibrational bands appeared at 2852 (CH2 symmetric stretching), 2885 (CH2 asymmetric stretching), 2938 (CH 3 symmetric stretching), and 2958 cm -~ (CH3 asymmetric stretching). The peak at 2938 cm -I was relatively high in all tissue specimen, and the peaks at 2852 and 2885 cm ~were high in the white matter and myelin fraction (Figs. 2 and 4) and low in the gray matter and synaptosomal fraction (Figs. 1 and 3). The ratio of the intensities of the bands between 2885 and 2938 cm -~ may also be a good marker for the discrimination between the gray matter and white matter. Figs. 5 and 6 show the spectra of the caudate-putamen
and thalamus. These spectra closely resemble each other. In the caudate-putamen and thalamus, the bands of amide I and amide III appear at 1659 and 1269 cm -1, respectively. The main structure of the proteins in these areas is an 0~-helix. The lipid and protein ratios (11659/ I1439) are similar to those of cerebral cortex (Fig. 1) and synaptosomal fractions (Fig. 3). In the region of 28003000 cm -l, CH 2 stretching symmetric and asymmetric modes were observed at 2873 and 2850 cm -l, while CH3 stretching symmetric and asymmetric modes at 2938 and 2958 cm -I were not observed in the spectra of the caudate-putamen and thalamus. The ratios o f I1084 to II126 o r I1065 in the spectra of the caudate-putamen and thalamus were similar to the ratio of the cerebral cortex or synaptosomal fraction which means that the fatty acids of these areas are enriched in trans type [5, 6]. In the spectrum of the caudate-putamen (Fig. 5), a small band of the T-G-T form of S-S bond was observed at 543 cm -1. The present study indicates that near-infrared FTRaman spectroscopy is promising in monitoring the changes of relative content of proteins and lipids or the changes of their structure or component content ratio in various portions of brain tissues.
1 Akutsu, H., Direct determination by Raman scattering of the conformation of the choline group in phospholipid bilayers, Biochemistry, 20 (1981) 7359-7366. 2 Culatolo, W., Verma, S.E, Sakura, J.D., Small, D.M., Shipley, G.G. and Wallach, D.F.H., Structural effects of myelin proteolipid apoprotein on phospholipids: a Raman spectroscopic study, Biochemistry, 17 (1978) 1802-1807. 3 Dunkley, ER. and Carnegie, ER., Amino acid sequence of the smaller basic protein from rat brain myelin, Biochem. J., 141 (1974) 243-255. 4 Ferra, F., Engh, H., Hudson, L., Kamholz, J., Puckett, C., Molineaux, S. and Lazzarini, R.A., Alternative splicing accounts for the four forms of myelin basic protein, Cell, 43 (1985) 721-727. 5 Krimm, S., Peptides and proteins. In T.G. Spiro (Ed.), Biological Applications of Raman Spectroscopy, Vol. l, Wiley, New York, 1987, pp. 1~6. 6 Mendelsohn, R., Laser Raman spectroscopic study of egg lecithin and egg lecithin-cholesterol mixtures, Biochem. Biophys. Acta, 290 (1972) 15-21. 7 Milanovich, F.E, Shore, B., Harney, R.C. and Tu, A.T., Raman spectroscopic analysis of Dutch belt rabbit erythrocyte hosts, Chem. Phys. Lipids, 17 (1976) 79-84. 8 Milner, B.J., Lai, C., Nave, K.-A., Lenoir, D., Ogata, J. and Sutcliffe, J.G., Nucleotide sequences of two mRNAs for rat brain myelin proteolipid protein, Cell, 42 (1985) 931-939. 9 Nie, S., Bergbauer, K.L., Kuck Jr., J.ER. and Yu, N.-T., Nearinfrared Fourier transform Raman spectroscopy in human lens research, Exp. Eye Res., 51 (1990) 619-623. 10 Nie, S., Marzilli, L.G. and Yu, N.-T., Near-infrared Fourier transform Raman spectroscopy of photolabile organ cobalt Bt2 and model compounds 1. Detection of cobalt-carbon stretching mode in solid state and in solution, J. Am. Chem. Soc., 111 (1989) 92569258.
52 11 Ozaki, Y. and Iriyama, K., Potential of Raman spectroscopy in medical science. In J. Spepanek, P. Anzenbacher and B. Sedlacek (Eds.), Laser Scattering Spectroscopy of Biological Objects, Elsevier, Amsterdam, 1987, pp. 559-582. 12 Ozaki, Y., Mizuno, A., Sato, H., Kawauchi, K. and Murauchi, S., Biomedical application of near-infrared Fourier-transform Raman spectroscopy I. The 1064 nm-excited Raman spectra of blood and met hemoglobin, Appl. Spectrosc., 46 (1992) 335-344. 13 Tashibu, K., Analysis of water content in rat brain using Raman spectroscopy, Brain Nerve, 42 (1990) 999-1004.
14 Tu, A.T., Raman Spectroscopy in Biology, Wiley, New York, 1982, 448 pp. 15 Wittacker, V.P. and Baker~ L.A., The subcellular fractionation of brain tissue with special reference to the preparation of synaptosomes and their component organelles. In R. Fried (Eds.), Method of Neurochemistry, Marcel Dekker, New York, 1972, pp. 1-52.
16 Yu, N.-T., DeNagel, D.C., Ho, D.J.-Y. and Kuck Jr., J.F.R., Ocular lenses. In T.G. Spiro (Ed.), Biological Applications of Raman Spectroscopy, Vol. 1, Wiley, New York, 1987, pp. 47 80.
