Laser Raman study of internally perfused muscle fibers effect of Mg2+, ATP and Ca2+

Biochimica et Biophysica Acta, 758 (1983) 121-127 Elsevier

121

BBA 21492

LASER RAMAN STUDY OF INTERNALLY P E R F U S E D M U S C L E FIBERS E F F E C T O F Mg 2+, ATP AND Ca 2+ J E A N - P I E R R E CAILLI~ a, M A R I E P I G E O N - G O S S E L I N t, and M I C H E L PI~ZOLET b Dbpartement de Biophysique, Facultd de Mddecine, Universitd de Sherbrooke, Sherbrooke, P.Q. J I H 5N4 and /' Ddpartement de Chimie, Facult~ des Sciences et de Gdnie, Universitd Laval, Qudbec, P.Q. G I K 7P4 (Canada) (Received March 7th, 1983)

Key words." Laser Raman; Mg2 +,• ATP," Ca2+,• Muscle fiber

Raman spectra of an intact muscle fiber and of internally perfused fibers in capillary tubes have been obtained. The use of internal perfusion has insured a good control of the concentration of Ca 2+, Mg 2+ and ATP. The comparaison of the spectra obtained with the two types of fibers shows that the muscle structure is well preserved in capillary tubes. In addition, it appears that the sarcomere length has no significant effect on the Raman spectrum of muscle fibers. Our results on perfused fibers demonstrate that a fiber can be kept in the relaxed state for several hours, then displaying an intact fiber spectrum, when the concentration of ATP, Mg 2+ and Ca 2+ is maintained at 5, 2 and 0 mM, respectively. Therefore ATP and Mg 2+ do not affect the Raman spectrum of muscle fibers. When one of these components is removed, or when Ca 2+ is added, contraction occurs and causes major spectral changes. These results are interpreted as being due to strong electrostatic interactions between basic and acidic residues during contraction, and to a change of the a-belical content, or of the orientation, of some of the contractile proteins.

Introduction

Since vibrational spectra of molecules are sensitive of their geometry and to the strength of the chemical bonds, Raman spectroscopy has been used successfully to characterize the conformation of biological macromolecules [1-3]. For example, the conformation of isolated muscle proteins [4-5] and membrane proteins [6-7] has been investigated by Raman spectroscopy. Since the interference of water is much smaller in Raman than in infrared spectroscopy, Raman scattering studies have been performed on muscle cytoplasm [8-10], and recently the effect of Ca 2÷, Mg 2÷ and ATP on the contractile proteins of single muscle fibers [11] has been investigated. Following this report [11], we have pursued these experiments in an attempt to dissociate the joined effects of Ca 2÷, Mg 2÷ and ATP. 0304-4165/83/$03.00 © 1983 Elsevier Science Publishers B.V.

In order to control the concentration of these substances, the experimental model was slightly modified to perfuse the myoplasm with solutions of electrolytes. The experimental conditions were varied to observe, from internally perfused muscle fibers, a Raman spectrum similar to the one recorded from freshly isolated muscle cells. The resuits obtained show that the sarcomere length (Ls), if it is longer than 9/xm, has no significant effect on the Raman spectrum of fibers between 500 and 1800 cm -~, and that ATP and Mg 2+ are necessary in the solution of perfusion in order to maintain the fibers in the relaxed state. When one of these components is removed, or when Ca 2÷ is added, spontaneous contractions occur and result in major spectral changes, and particularly in an increase of the intensity of the amide I band at 1650 cm-~ and of the C-C stretching band at 936 cm-1. Both changes indicate that contraction af-

122

fects the a-helical structure of the contractile proteins.

teflon

cap---" l ~

~ 0 0 .

~ = - +--

+ t ' " ,+ ?::ous u e

Material and Methods

Muscle fibers. Freshly isolated muscle fibers from the depressor muscle of the barnacle (Balanus Nubilus) were used thoughout this study. Raman spectra were recorded from intact muscle fibers immersed in artificial sea water and from fiber-filled capillaries. Solutions. Solutions were prepared with highly purified chemicals and their compositions are summarized in Table I. Fiber-filled capillaries. This experimental model, which is well suited for Raman spectroscopy, has already been described [12]. Essentially, an isolated muscle fiber is gently pulled in the lumen of a glass capillary chosen to fit the diameter of the fiber. In order to perfuse the intracellular compartment, a small porous tube (150 # m diameter) was introduced in the capillary before the fiber was pulled in, so that the tube was squeezed between the capillary wall and the muscle cell (Fig. 1). Electrolyte solutions were perfused through the muscles cell by the porous tube and the glass capillary insured that the water content of the fiber was held constant during the perfusion. Raman spectra. The Raman spectra were excited with the 514.5 nm line of a Spectra Physics model 165 argon ion laser and recorded with a Spex 1400 computerized spectrometer that has

~-:-.~--~-~t" eo -- . . . . =--]'n

II

dialysis

•

.

• OAo ° •

~" •

Io

t +,+s+

muscle cell

Fig. 1. C a p i l l a r y cell used to perfused muscle fibers.

been described elsewhere [13]. The laser beam was alligned perpendicularly to the axis of the muscle fiber and the scattered light was collected at 90°C of the incident laser beam. All the spectra reproduced in this paper were obtained at approx. 10°C, with 125 mW laser power at the sample, with a spectral resolution of 5 cm-1 and with an integration period of 2 s/step of 1 cm-~. They were not smoothed but they were corrected for a slight luminescent background by substracting the appropriate polynomial function [13]. Sarcomere length. The sarcomere length is an important parameter for the physiological properties of muscle cells. This report contains some results on the effect of the sarcomere length on the Raman spectrum of muscle fibers. The sarcomere length was measured by light diffraction using an He-Ne laser beam ()~ = 632.8 nm). The diffracted patterns were recorded on a paper screen or on photographic films. To obtain samples of myoplasm with sarcomere length shorter than 8 /xm, shrivelled fibers were pulled in capillaries without any stretching.

TABLE I C O M P O S I T I O N O F T H E S O L U T I O N S I N m M A T p H 7.2 [NaC1] Artificial sea w a t e r Artificial sea water, low C a z + Solution I

[K CI]

[K Ac]

450

8

--

480

8

--

100

--

[CaCI2 ]

[MgC12]

[EGTA]

[THAM]

20

10

--

25

--

10

2

25

--

2

0.5

1

1

2

0.5

I

2

2

0.5

1

5

0.5

1

Solution II

--

I00

"Solution III

--

100

--

Solution IV

--

100

--

° "O~

! +o,u,,o°

[K2ATP ]

-

123

Results

Intact muscle fibers The Raman spectrum of an intact muscle fiber not inserted into a glass capillary tube [10] and immersed in artificial sea water (Table I) is shown in Fig. 2, where the major peaks are also identified. As previously observed, three features of this spectrum indicate that muscle proteins are mainly in the a-helical conformation: (1) the strong and relatively sharp amide I band at 1650 c m - l ; (2) the weak scattering in the amide III region between 1225 and 1280 cm-1; (3) the strong C-C skeletal stretching band at 939 c m - l . We also believe that the 1045 cm -1 band, that is tentatively assigned to a C-N stretching vibration, is sensitive to the conformation since it decreases in intensity when the fibers are denatured. The intensity of the amide I and III bands can lead to quantitative information on the conformation of proteins [14-16]. However, since the muscle cell is a complex system and many parameters of its structure have not been established, only a qualitative analysis of the intensity of the Raman band is presented. Intensities were measured as peak height from a line drawn between the two minima on each side of the bands, and were divided by the height of the 1450 cm -1 band assigned to the CH 2 and CH 3 bending modes. The use of this band as an internal intensity standard

~-

~o

~

~

co

o

is justified since it has been shown to be insensitive to the conformation of the proteins [5,17-18]. The Raman spectra of five intact muscle fibers, which present weak carotenoid bands at 1520 and 1156 cm-1, were analysed and the mean intensity ratios a r e : 11650/11450 = 1.23 + 0.02, 11415/11450 = 0 . 5 8 + 0 . 0 2 , 11250/11450 -----0.33 + 0.02, 1940/11450 0.64 + 0.03 and 1758/11450 = 0.49 + 0.04. The uncertainty on these ratios is the standard error of the mean. It should be pointed out that the amide I band was not corrected for the water bending vibration at 1635 c m - i since the contribution of this mode is weak because of the high concentration of proteins in muscle. =

Effect of the sarcomere length By changing the sarcomere length, the overlapping between the actin and myosin filaments is increased or reduced. Thus, it was important to verify that the relative intensity of the Raman bands is not dependent on the sarcomere length. At first, Raman spectra of muscle fibers in capillaries without perfusion tubes were recorded between 700 and 1500 cm-1, and the sarcomere length was measured. The intensity ratios of four important bands as a function of the sarcomere length are presented in Table II. It should be mentioned here that the sarcomere length is long in barnacle muscle cells compared to mamalian and frog muscles; the physiological range of sarcomere length for these fibers in between 9 and 15/~m [19]. The results given in table II show that no significant change of the intensity ratios was observed, except for very short sarcomere length.

T A B L E II

¢

® ~,o -

E F F E C T OF T H E S A R C O M E R E L E N G T H ON T H E RELATIVE INTENSITY OF T H E R A M A N BANDS A T 758, 940, 1250 A N D 1415 cm - l

---

The values are mean values + the standard error of the mean of five independent measurements. X)

I

i

I

i

/

IOCO

I

Frequency

I

Shift

I

I

I

1,500

I

I

(cm-')

Fig. 2. R a m a n spectrum of an intact muscle fiber in artificial sea water (Table I). Abbreviations: str, stretching; sym, symmetric; bend, bending; C-C, carbon-carbon bond; C-N, carbon-nitrogen bond; C-H, carbon-hydrogen bond;phe, phenylalanine; tyr, tyrosine; trp, tryptophan.

L~(Bm)

1758/11450

1940//1450 11250/11450

11415/11450

6.4+0.1 7.7+0.2 10.1_+0.3 11.8_+0.3 14.2+0.1

0.41 _+0.01 0.64-+0.04 0.26-+0.02 0.46_+0.01 0.61_+0.03 0.28_+0.01 0.53_+0.02 0.66+0.02 0.29_+0.01 0.48_+0.02 0.65_+0.02 0.30_+0.01 0.534-0.01 0.69_+0.03 0.30+0.01

0.50 + 0.02 0.52+0.02 0.56_+0.01 0.58_+0.01 0.58_+0.01

124

Similar results were obtained with intact muscle fibers in artificial sea water.

Internally perfused muscle fibers In order to observe from internally perfused muscle fibers a R a m a n spectrum similar to the one recorded from intact muscle cells, we have tried different procedures for the pretreatment of the muscle fibers. First, the extracellular space of the muscle fiber (4%) was cleaned out by soaking the fibers in an isotonic sucrose solution containing 2 m M EGTA. To increase the percentage of successful perfusions, it was preferable to let the fibers soak for 2 h in artificial sea water with low Ca e+ concentration, and then to rinse them for 30 min in an isotonic sucrose solution containing 0.5 mM EGTA. Even with this pretreatment, only 20% of the perfused fibers (solution I) displayed, after 3 h, a R a m a n spectrum similar to the one observed in intact muscle cells. Unsuccessful perfusions caused a marked decrease of the intensity of the 758, 1045 and 1415 cm-~ bands and thus induced contraction of the fibers [11]. The replacement of KC1 by potassium acetate in the dialysis solution (solution II) has increased the number of successful perfusions to approx. 75%. An example of the spectrum obtained after such a perfusion is presented in Fig. 3. As shown in this figure, the fiber was still in the relaxed state even after 4 h of perfusion with solution II. The R a m a n bands that are sensitive to the state of contraction of muscle fibers, that is the 758, 940, 1045, 1415 and 1650 c m - 1 bands, are very weak in the difference spectrum; the major peaks at 1520 and 1156 cm -~ come from carotenoid chromophores that are bleached after long exposure to the laser beam. The relative intensity of the 758, 1250, 1415 and 1650 cm 1 bands are given in Table III for six different measurements on fibers perfused with solution II after a minimum of 2 h perfusion. In order to study the effect of exogenous ATP and Mg z+ on the perfused fiber spectrum, we have recorded spectra of fibers perfused with solutions of various concentrations of these substances. U p o n increasing the concentration of ATP to 5 m M (solution III), it was easier to keep the fibers in the relaxed state and, as seen in table III, the intensity ratios obtained were very close to those of intact muscle fibers. On the other hand, without

A

)0

I

I

I

I

[

1000

I

I

l

Frequency Shift (cm-I)

I

15

~

I

i

Fig. 3. Raman spectra of an internally perfused muscle fiber. (A) Before perfusion; (B) after 4 h of perfusion with a solution containing 2 mM A T P / 2 m M Mg 2+ (solution II); (C) the difference between spectrum A and spectrum B (A-B) using the 1450 cm-~ band as an internal intensity standard. CAR. indicates carotenoid bands.

A T P and Mg 2+ in the perfusion solution (solution IV), contraction was induced as detected by the decrease of the intensity of the 758 and 1415 c m bands (Table III). Similar spectral changes were also observed when either ATP or Mg 2+ was added to solution IV. Thus, both substances are necessary in order to keep the fibers in the relaxed state. An interesting result that is shown in Table III is that the intensity of the amide I band increases upon contraction of the muscle cell. In order to confirm this finding, a fiber-filled capillary was first perfused with a solution containing Mg • ATP (solution I) to maintain the fiber in a relaxed state, and then the concentration of calcium in the solution of perfusion was increased to 10 -4 M to induce contraction of the sample. A rapid modification of the spectrum was then observed and as shown in Fig. 4 the use of the spectral substraction technique clearly reveals the changes induced by

125 TABLE III E F F E C T OF T H E CONCENTRATION OF ATP A N D Mg 2÷ ON THE RELATIVE INTENSITY OF THE 758, 1250, 1415 A N D 1650 C M - J BANDS The values are mean values _+the standard error of the mean; the number of independent measurements is given in parentheses.

Perfusion solution Intact muscle fibers: Artificial sea water (5) Fiber-filled capillaries: (20) Internally perfused fibers: Solution II (6) Solution III (9) Solution IV (8)

1758/11450

11250/11450

11415/11450

11650/11450

0.49+0.04

0.33+0.02

0.58_+0.02

1.23_+0.02

0.46 4- 0.01

0.30 -+ 0.01

0.59 -+ 0.01

1.22 -+0.02

0.37_+0.01 0.35+0.01 0.25-+0.01

0.33_+0.01 0.35+0.01 0.40-+0.01

0.51 _+0.02 0.52_+0.01 0.39-+0.01

1.35_+0.01 1.24+0.02 1.50_+0.01

contraction. As previously observed, contraction causes a marked decrease of the intensity of the carboxylate band at 1414 cm -1 and of the tryptophan band at 758 c m - L Simultaneously, the bands at 1045 and 899 cm-J, as well as the 1520

~.oo - - if)

~

-- --

~

--

0,1~

and 1156 cm-1 carotenoid bands, also decrease in intensity. On the other hand, the bands associated with the conformationally sensitive amide I mode at 1650 c m - ] and C-C stretching vibration at 940 c m - ~ increase in intensity when the concentration of Ca 2÷ is increased. Both changes show that contraction is accompanied by either an increase of the amount of the a-helical content or a change of the orientation of a-helical segments of the contractile proteins. This effect was not noticed in our previous study [11] although a close examination of the spectra reveals the same modifications. Discussion

:_ tO tO0o

~

~

,~

i 500

i

i

-

n --0-----

~

~-~

I

I I000 Frequency

0 N

~

/ O~ -- 0

|

~

-

=

-

i Shift

ii

i

L

_,.2

I 1500

i

*

(era -I)

Fig. 4. Effect of Ca 2 + on the Raman spectrum of an internally perfused muscle fiber. (A) Before perfusion; (B) after 1 h of perfusion with solution I containing 10 -4 M Can+; (C) the difference between spectrum A and spectrum B (A-B) using the 1450 c m - i band as an internal intensity standard; CAR. indicates carotenoid bands.

We have already demonstrated that the depressor muscle of the giant barnacles is well suited for Raman studies [9-11]. All the spectra that we have reported so far on this system were obtained with single fibers inserted into glass capillary tubes. Although these fiber-filled capillaries are very convenient, the physiological integrity of the preparation may be questioned, specially for short fiber segments, since the fiber is cut at both ends of the capillary. The comparaison of the spectrum of an intact muscle fiber (Fig. 2) still attached to the base plate and with an intact membrane, with the spectrum of a fiber-filled capillary (Figs. 3A or 4A) shows that these spectra are almost identical. This is also confirmed from the intensity ratios given in Table III. Therefore, the fibers structure is well preserved

126

in capillary tubes which in addition prevent any increase of the water content of the fibers. Our results on the effect of the sarcomere length on the Raman spectrum of muscle fibers show that the relative intensity of the 758, 1250 and 1415 c m - 1 bands is independent of the sarcomere length if it is longer than 9 /~m. However, for short sarcomere lengths, the relative intensity of the 758 and 1415 cm - t bands is slightly reduced. Although the origin of these changes is still unclear, they might be associated with the decrease of the length of the A band that corresponds to the length of the thick filaments. Baskin et al. [20] have already observed this effect on contracted barnacle myofibrils with sarcomere lengths shorter than 8/~m. The spectral changes observed on fibers perfused with solution I (100 mM KC1) were probably due to slow contractions of the myoplasm. Contractions were also observed on barnacle skinned muscle fibers [21 ]. As noted above, the use of potassium acetate (solution II) has led to much more successful perfusions. The same substitution in the embedding solution of a chemically-skinned frog skeletal muscle has also improved the X-ray low angle diffraction patterns [22]. The results obtained with perfused fibers demonstrate that Mg 2÷ (2 mM) and ATP (5 mM), which are essential to keep the myoplasm in a relaxed state [22,23], do not affect the Raman spectrum of the cytoplasm of these cells. On the other hand, the major changes of the Raman spectrum observed when the Ca 2+ concentration is increased to 10 -5 M, a condition that induces muscle contraction [22], show clearly that the side-chain environment as well as the secondary structure of the contractile proteins are affected by muscular contraction. For example, the 1412 c m - I (Fig. 4) band, associated with the symmetric C-O stretching vibration of the carboxylate groups of the acidic side-chains, decreases in intensity. We believe that this effect comes from the interaction between the positively- and the negatively-charged residues of the contractile proteins during contraction, since in tropomyosin there is a marked increase of the intensity of the carboxylate band at 1414 cm 1 when this highly helical protein is thermally denatured (Pezolet, M., Nadeau, J., Pigeon-Gosselin,

M. and Caill6, J.P., unpublished data). It is well known that because of the repetitive heptapeptide primary structure of tropomyosin, it forms a coilcoiled structure stabilized by the interactions between basic and acidic residues [24]. The intensity of the 759 cm-~ band also decreases during contraction. Although we believe that this band is mainly associated with a ringbreathing vibration of the tryptophan residues [11], acidic side-chains can also contribute to the scattering intensity in this region. A band assigned to the C - C O O - stretching mode has been observed at 758 c m - I in the Raman spectrum of some poly-carboxylic acids and its intensity was sensitive to the degree of neutralization of the acid groups [25]. Therefore, it is likely that the change of intensity of the 759 cm 1 is associated with that of the 1412 c m - i carboxylate band. A particularly important spectral change occurring during contraction is the increase of the scattering intensity of the characteristic bands of the a-helical conformation at 1650 and 940 cm The amide I band also shifts in frequency to 1647 cm 1, a value that is closer to the one observed for a-helical tropomyosin (Pezolet et al., unpublished data). These changes can have two different origins. First, they can result from an increase of the a-helical content of the conformation of the contractile proteins. This conformational change could be associated with the above discussed interaction between the positively- and negatively-charged side-chains of the proteins induced by contraction. Such a change of conformation has been observed for positively-charged poly-L-lysine when it is bound to acidic phospholipids [26]. On the other hand, since the intensity of the amide I band increases by 20% during contraction, it seems hard to believe that the a-helical content of the proteins increases that much. A second interpretation for the change observed in the amide I region is a partial reorientation of the a-helical segments of the contractile proteins during contraction. Snyder has demonstrated that induced polarizability changes along a bond may be quite different from that perpendicular to the bond [27,28]. Thus, if the a-helical segments, and therefore the peptide C = O groups of these segments, are parallel to the fiber axis, the amide I band of the a-helices should be stronger

127 w h e n the i n c i d e n t electric field of the laser b e a m is also parallel to the fiber axis. W e have in fact observed that there is a 25% decrease of the i n t e n sity of the amide I b a n d of relaxed fibers when the i n c i d e n t electric field is alligned p e r p e n d i c u l a r l y to the fibers (Pezolet et al., u n p u b l i s h e d data). Therefore, if more C = O groups b e c o m e parallel to the fiber axis d u r i n g contraction, one would expect an increase of the i n t e n s i t y of the amide I b a n d . This is indeed what we have observed (Fig. 4). I n conclusion, it should be emphasized that the use of i n t e r n a l l y perfused fibers in this study has i n s u r e d a good control of the c o n c e n t r a t i o n of C a 2÷, Mg 2+ a n d A T P in the myoplasm. O u r results show that muscle cells can be kept in the relaxed state when the level of ATP, Mg 2÷ a n d Ca 2÷ is kept at 5, 2 a n d 0 m M respectively. O n the other h a n d , c o n t r a c t i o n occurs rapidly when the c o n c e n t r a t i o n of either one of these substances is lowered or when Ca 2+ is added to the solution of perfusion. C o n t r a c t i o n is always characterized by the decrease of the i n t e n s i t y of some of the R a m a n b a n d s of the p r o t e i n side-chains a n d b y the increase of the a m i d e I a n d C-C stretching b a n d s .

Acknowledgements This research was s u p p o r t e d by the N a t u r a l Sciences a n d E n g i n e e r i n g Research C o u n c i l a n d the Medical Research C o u n c i l of C a n a d a . The authors t h a n k Mrs. Micheline D e l o r m e - F o u r n i e r for her technical assistance a n d acknowledge the help of Mrs. Gail Josephfowich in the p r e p a r a t i o n of the manuscript.

References 1 Thomas, G.J., Jr (1975) Raman spectroscopy of biopolymers. In Vibrational Spectra and Structure (Durig, J.R., ed.), pp. 239-315, Marcel Dekker, Inc. New York 2 Spiro, T.G. and Gaber, B.P. (1977) Annu. Rev. Biochem. 46, 553-572

3 Yu, N.-T. (1977) C.R. Biochem. 46, 229-280 4 Carew, E.B., Asher, I.M. and Stanley, H.E. (1975) Science 188, 933-936 5 Barrett, T.W., Peticolas, W.L. and Robson, R.C. (1978) Biophys. J. 23, 349-358 6 Nikkelsen, R.B., Verma, S.P. and Wallach, D.F.H. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 5478-5482 7 Lippert, J.L., Lindsay, R.M. and Schultz, R. (1981) J. Biol. Chem. 256, 12411-12416 8 Asher, I.M., Carew, E.B. and Stanley, H.E. (1976) In Physiology of Smooth Muscle (Bulbring, E. and Shuba, M.F., eds.), pp. 229-238, Raven Press, New York. 9 P6zolet, M., Pigeon-Gosselin, M. and Caill~, J.P. (1978) Biochim. Biophys. Acta 533, 263-269 10 P~zolet, M., Pigeon-Gosselin, M., Nadeau, J. and Caillb, J.P. (1980) Proc. of the VIIth International Conference on Raman Spectroscopy, (Murphy, W.F., ed), pp. 600-601, North-Holland, New York 11 P~zolet, M., Pigeon-Gosselin, M., Nadeau, J. and Caill& J.P. (1980) Biophys. J. 31, 1-8 12 Caill~, J.P. and Hinke, J.A.M. (1972) Can. J. Physiol. Pharrnacol. 50, 228-237 13 Savoie, R., Boul6, B., Genest, G. and P6zolet, M. (1979) Can. J. Spectrosc. 24, 112-117 14 Lippert, J.L., Tyminski, D. and Desmeules, P.J. (1976) J. Am. Chem. Soc. 98, 7075-7080 15 P~zolet, M., Pigeon-Gosselin,M. and Coulombe, L. (1976) Biochim. Biophys. Acta 453, 502-512 16 Williams, R.W. and Dunker, K. (1981) J. Mol. Biol. 152, 783-813 17 Frushour, B.G. and Koenig, J.L. (1974) Biopolymers 13, 1809-1819 18 Yu, T.S., Lippert, J.L. and Peticolas, W.L. (1973) Biopolymers 12, 2161-2176 19 Gayton, D.C. and Elliott, G.F. (1980) J. Muscle Res. Cell Mot. 1,391-407 20 Baskin, R.J. and Wiese, G.M. (1964) Science 143, 134-136 21 Ashley, C.C. (1978) Ann. N.Y. Acad. Sci. 307, 308-329 22 Magid, A. and Reedy, M.K. (1980) Biophys. J. 30, 27-40 23 Pemrick, S.M. and Edwards, C. (1974) J. Gen. Physiol. 64, 551-567 24 McLachlan, A.D. and Stewart, M. (1975) J. Mol. Biol. 98, 293-304 25 Koda, S., Nomura, H. and Nagasawa, M. (1982) Biophys. Chem. 15, 65-72 26 Hartmann, W. and Galla, H.-J. (1978) Biochim. Biophys. Acta 509, 474-490 27 Snyder, R.G. (1971) J. Mol. Spectrosc. 36, 222-231 28 Snyder, R.G. (1971) J. Mol. Spectrosc. 37, 353-365