Human spectrin. V. A comparative electro-optic study of heterotetramers and heterodimers

74

Biochimica et Biophysica Acta, 668 (1981) 74--80

© Elsevier/North-Holland Biomedical Press BBA 38637 HUMAN SPECTRIN V. A COMPARATIVE ELECTRO-OPTIC STUDY OF H E T E R O T E T R A M E R S AND H E T E R O D I M E R S

ARNE MIKKELSEN and ARNLJOT ELGSAETER Institute o f Biophysics, University o f Trondheim, N-7034 Trondheim-NTH (Norway)

(Received June 24th, 1980) (Revised manuscript received October 23rd, 1980) Key words: Spec trin; Mem brane pro tein; Elec tro-op tic s tudy ; Birefringence

S u m mar y The electrically induced birefringence of hum an spectrin h e t e r o t e t r a m e r and h e t e r o d i m e r solutions at 5°C has been studied. 1. The steady~state birefringence, 5, was found to be approximately proportional to the electric field strength, E, when E ~> 0.2 k V / m m . For spectrin solutions the specific linear coefficient, 6 / ( E . c), t herefore is a more relevant parameter for describing birefringence saturation behavior when E/> 0.2 kV/ m m than the c o m m o n l y used Kerr constant. At 5°C we measured 5 / ( E . c) = (27 + 5) • 10 -8 m 4 • V -1 • kg -1 for heterodimers and heterotetramers. 2. At 5°C b o th het er ot e t r am e r s and heterodimers exhibited m ore than one birefringence relaxation time and the shortest of these was for bot h molecules fo u n d to be 4.2 +- 1.0 ~s. This indicates that the spectrin molecules are highly flexible. The birefringence build-up time for het erot et ram ers and heterodimers was f o u n d to be 20 _+ 7 ~s and 15 + 5 ~s, respectively. Introduction We have previously [1] studied the electro-optic properties of hum an spectrin h e t e r o d i m e r solutions at r o o m temperature. T hat study yielded a heterodimer birefringence relaxation time Td = 2.0-+ 0.3 /~S at 22°C, indicating a h e t e r o d i m e r c o n t o u r length of at least 40--50 nm. Spectrin h e t e r o t e t r a m e r solutions were n o t included in the previous study [1] since heterotetramers axe n o t t h e r m o d y n a m i c a l l y stable at r o o m t e m perat ure [2,3]. S h o t t o n et al. [4,5] presented electron microscopic evidence indicating that spectrin heterotetramers are f or m e d by an end-to-end association of heterodimers. This conclusion is supported also by sedimentation, lateral diffusion and viscosity studies [5--7]. The birefringence relaxation time of stiff mole-

75 cules is very sensitive to the length of the molecules; a doubling of the length of a stiff molecule results in an approximate 6-fold increase in the birefringence relaxation time [8,9]. Such an increase in the relaxation time is experimentally easily detected and a comparative electro-optic study of spectrin heterodimers and heterotetramers thus becomes particularly revealing. We here report on a comparative electro-optic study of human spectrin heterotetramers and heterodimers at 5°C. Methods

Preparation of human spectrin heterodimers. Human spectrin heterodimers were prepared according to procedure A described by Elgsaeter [10], except for the following modifications; (a) 120--150 ml of freshly drawn blood was added to an anticoagulant solution containing 22 g/1 trisodium citrate, 8.0 g/1 citric acid and 24.5 g/1 dextrose until a blood/anticoagulant ratio of 7 : 1 was obtained. (b) The final packing of the ghosts before spectrin extraction was done by centrifugation at 35 000--40 000 X g (18 000--20 000 rev./min in a Beckman SW25.1 rotor) for 120 min at 0--4°C. (c) The diameter of the gel filtration columns was 1.6 cm. (d) The final dialysis was carried out at 2--4°C for 40--70 h against three times 1 1 of the desired salt solutions. (e) The final spectrin preparations were not optically clarified. However, in order to check the effect of such a clarification some of the preparations were centrifuged at 200 000 X g (50 000 rev./min in a Beckman 50 Ti rotor) for 60 min at 0--4°C prior to the electro-optic measurements. Aliquots for sodium dodecyl sulfate polyacrylamide gel electrophoresis [10] were taken from the spectrin preparations just prior to the electro-optic measurement. We ignored data from measurements on preparations revealing proteolytic breakdown or other irregularities by gel electrophoresis. Preparations of human spectrin heterotetramers. Human erythrocyte ghosts were prepared as for the preparation of spectrin heterodimers. The ghost suspension was then dialyzed at 2--4°C for 40--48 h against three times 1 1 of 0.1 mM ethylenediaminetetraacetic acid (EDTA) and 0.05 mM dithiothreitol titrated to pH 7.5 with 0.3 M NaOH. The suspension was centrifuged at 200 000 X g (50 000 rev./min in a Beckman 50 Ti rotor) for 60 min at 0--4°C, the pellet discarded and the supernatant was centrifuged again at 200 000 X g for 60 min. A solution of 1.1 M NaC1, 1.1 mM EDTA, and 0.55 mM dithiothreitol (pH 7.6) was added under gentle stirring until a salt concentration of 100 mM NaC1 was obtained, and 15--20 ml of the solution was then applied to a column (90 X 1.6 cm) of Sepharose CL-4B beads equilibrated with 100 mM NaC1, 0.1 mM EDTA and 0.05 mM dithiothreitol (pH 7.6) at 2--4°C. The fract i o n s off the column were analyzed as described earlier [10]. Not all the preparations exhibited elution profiles with as well-resolved heterotetramer peak as indicated earlier [ 10]. The fractions containing heterotetramers were collected, and then dialyzed as for the heterodimers prior to the electro-optic measurements. Electric birefringence apparatus. Except for an added system to control the Kerr cell temperature the apparatus was as described previously [1]. The Kerr cell was placed in a copper block, where temperature was controlled by a

76 Haake NK 22 Thermostat. Some of the thermostat bath liquid was passed through channels in the copper block and in our experiments this system kept the sample temperature at 5 + 1°C. The electro-optic measurements were done using a photomultiplier anode voltage of 2000 V and the quadratic detection mode. We used an anode resistor of 10 k ~ for the birefringence steady-state study and 1 k~2 for the transient study. The performance of the detection system was tested as described earlier [ 1 ] by measuring the apparent birefringence relaxation time of propylene carbonate {Merck). This test yielded an apparent relaxation time of 0.10 +0.02 ps. In all cases we used rectangular, 100 ps long electrical pulses across the Kerr cell. The employed electrical field strength ranged from 0.05 to 1.5 kV/mm. Results

Human spectrin heterotetramer steady-state birefringence, 5, at 5°C vs. the electric field strength, E 2, and vs. E is shown in Fig. 1A and B, respectively. n v 4

3 x

,x

1

0

0.1

0.2 0.3 E2 ( k V / m m ) 2

05

0.4

!

10

v

B

v

o, 6 0 × 4

IB

o

0.2

0.4

0.6

0.8 1.0 E ( k V / m m)

1.2

,0.

114

~'.6

F i g . 1 . T h e s t e a d y - s t a t e b i r e f r i n g e n c e , ~. vs. the s q u a r e of the electric field s t r e n g t h , E 2 (A) a n d vs. t h e e l e c t r i c f i e l d s t r e n g t h , E (B) for: v , 0.29 rag/m1; ~, 0 . 1 2 m g / m l ; Q, 0.06 m g / m l , a n d o, 0.0 m g / r a l of s p e c o

trin h e t e r o t e t r a m e r s in 1 raM NaC1, 0,1 m M E D T A , a n d 0 . 0 5 raM d i t h i o t h r e i t o l at 5 C a n d p H 7.0.

77

0

,

,

,

0

•

8 -0,4 r-'m~-041 ~-08

~ -0

>

£

£

o -1

-161 0

I

~ 20

I ,40

a 60

80

100

-12 -1E

,

2

t(#s)

,

.

,4 t (#s)

.

,

6

,

.

8

Fig. 2. A. L o g a r i t h m i c p l o t o f the build-up of the p h o t o d e t e c t o r electric signal, V ( t ) , e m p l o y i n g the q u a d r a t i c d e t e c t i o n m o d e and 0.29 m g / m l s p e c t r i n h e t e r o t e t r a m e r s in 1 m M NaC1, 0.1 m M E D T A , a n d 0 . 0 5 m M d i t h i o t h r e i t o l at 5°C a n d p H 7.0. V(¢¢), the s a t u r a t i o n value of V ( t ) . Electric field s t r e n g h t (E) is 0.8 k V / m m . B. L o g a r i t h m i c p l o t of the d e c a y o f t h e p h o t o d e t e c t o r signal, V ( t ) , e m p l o y i n g the q u a d r a t i c d e t e c t i o n m o d e a n d 0 . 1 4 m g / m l s p e c t r i n h e t e r o t e t r a m e r s in 0.5 m M NaC1, 0.1 m M E D T A and 0 . 0 5 m M d i t h i o t h r e i t o l at 5°C a n d p H 7.0. E = 0.5 k V / m m .

Human spectrin heterodimers show qualitatively the same steady-state birefringence as heterotetramers. Representative examples of heterotetramer birefringence build-up and decay (logarithmic plots) are depicted in Fig. 2A and B, respectively. Fitting the plot of the decay with two straight lines gives the two relaxation times Tdl and Td2. The electro-optic properties of human spectrin heterotetramers and heterodimers are summarized in Tables I and II, respectively. Note that by employing the quadratic detection mode the relaxation times detected by the optical system are half the birefringence relaxation times of the sample, whereas the

TABLE I T H E E L E C T R O - O P T I C P R O P E R T I E S OF H U M A N S P E C T R I N H E T E R O T E T R A M E R S I N 0.1 m M E D T A , 0 . 0 5 m M D I T H I O T H R E I T O L , A N D T H E G I V E N NaCl C O N C E N T R A T I O N S A T 5°C A N D p H 6.8--7.3 T h e specific K e r r c o n s t a n t , Bsp , is o b t a i n e d f r o m t h e linear p a r t of t h e p l o t s o f 5 vs. E 2 a n d the specific l i n e a r c o e f f i c i e n t , 5 / ( E • c ) , is o b t a i n e d f r o m the linear p a r t of t h e plots of 6 vs. E. T h e build u p t i m e is Tbu, a n d Vdl a n d r d 2 are t h e t w o r e l a x a t i o n t i m e s o b s e r v e d in p l o t s like Fig. 2B. T h e n u m b e r s are the a v e r a g e f o r s p e c t r i n h e t e r o t e t r a m e r c o n c e n t r a t i o n s , c, r a n g i n g f r o m 0 . 0 5 t o 0 . 5 5 m g / m l a n d for electric field s t r e n g t h s 0.3, 0.5 a n d 0.8 k V / m m across t h e K e r r cell. n, n u m b e r of s e p a r a t e p l o t s f r o m w h i c h the values are a v e r a g e d . Salt (raM NaC1) 0 0.5

Bs p (X101 I ) m 4 • V -2 • kg-1

6/(E • c)(×108) m 4 • V-1 . kg-1

r b u (gs)

r d l (#s)

13.2 ± 3.0

29 ± 6

22 ± 7

3.9 ± 0.8

(n = 11)

(n = 16)

(n = 32)

(n = 59)

27 ± 3

21 ± 6

4.2 ± 1.0

(n = 9)

(n = 18)

(n = 37)

8.2 ± 2.0 (n = 9)

1.0

Mean

11.2

± 6.0

r d 2 (ps)

9.8 ± 3.0 ( n = 57)

9.4 ± 3.0 ( n = 44)

24 ± 5

1 6 -+ 5

4.5 ± 0.8

10.2

(n = 11)

(n = 15)

(n = 29)

(n = 83)

( n = 93)

± 3.0

11 ± 4 (n = 31)

27 ± 5 (n = 40)

20 ± 7 (n = 79)

4.2 ± 1.0 (n = 179)

9.9 ± 3.0 (n = 194)

78 TABLE II THE ELECTRO-OPTIC PROPERTIES OF HUMAN SPECTRIN HETERODIMERS IN 0.1 mM EDTA, 0,05 mM DITHIOTHREITOL,AND THE GIVENNaCl CONCENTRATIONS AT 5°C AND pH 6,8--7.3 The numbers are the averagefor spectrin heterodimer concentrations 0.10--0.65 mg/ml and electric field strengths 0.5 and 0.75 kV/mm. Salt (mM NaCl) 0

0.5

1.0

Mean

B s p" ( × 1 0 1 1 ) (m 4 • V -2 • kg -1)

5/(E.

c)(×108) ( m 4 • V -1 . k g - I )

~'bu ( p s )

V d l (/as)

V d 2 C/as)

18 ± 4

28 ± 5

18 ± 5

4.0 ± 0.8

8.0 ± 1.5

(n = 5)

(n = 5 )

(n = 1 8 )

(n = 3 5 )

(n = 3 3 )

12 ± 3

25 + 3

14 + 4

4.0 ± 0.8

7 . 2 -+ 1 . 2

(n = 3)

(n = 3 )

(n = 7)

(n = 1 6 )

(n = 1 5 )

15 ± 5

2 6 -+ 3

13 ± 5

4 . 2 -+ 0 . 8

7.0 ± 1.0

(n = 5)

(n = 5)

(n = 1 6 )

(n = 3 8 )

(n = 3 8 )

15 ± 5

27 ± 4

15 ± 5

4.1 ± 0.8

7.4 ± 1.2

(n = 1 3 )

(n = 1 3 )

(n = 4 1 )

(n = 8 9 )

(n = 8 6 )

build-up times are essentially detected directly. A more detailed analysis of the birefringence relaxation revealed the presence of a third relaxation time of 20--30 ps in some of the samples (not indicated in Fig. 2B). Centrifugation of spectrin heterotetramer solutions at 200 000 X g for 60 min after the final dialysis and before the electro-optic measurements resulted in only 0--15% reduction in the relaxation times. The birefringence relaxation times were not found to depend on the applied electric field strength when varying this from 0.3 to 0.8 kV/mm. The build-up time constants, however, decreased by 15--30% upon a similar increase in E. We observed no significant dependence upon spectrin concentration in any of the time constants. Discussion

Fig. 1A and B show that the spectrin heterotetramer steady-state birefringence, 6, follows the Kerr law [11,12] for E ~< 0.2 kV/mm, however, when 0.2 k V / m m ~< E ~< 1.5 k V / m m 5 is approximately proportional to E. For the heterodimer we find qualitatively the same steady-state birefringence behaviour. Many macromolecules in solution show similar deviations from the Kerr law [13,14]. In spite of this, 6 is only occasionally [15] presented as function of E instead of E 2, and in our previous study [1] we followed this common practice. However, if the values of 6 we obtained in this study [1] are plotted as a function of E instead of E 2, it becomes clear that for 0.1 k V / m m ~< E ~< 1.2 k V / m m 5 is a linear function of E. For spectrin heterodimers at 22°C the average specific coefficient, 5 / ( E • c), is thus calculated to be (33 + 5) • 10 -8 m 4 • V-' • kg -'. The conclusion is that for both spectrin tetramer and spectrin dimer solutions the linear specific coefficient, 6 / ( E . c), is a more relevant parameter for describing birefringence saturation behaviour when E ~> 0.2 kV/ mm, than the c o m m o n l y used Kerr constant Bsp. Assuming that a change in the temperature does n o t change the structure of

79 the molecules, the temperature dependence of the birefringence relaxation of molecules in solution is mainly accounted for by the temperature dependence of the solvent viscosity [8,9]. The viscosity of water at 5°C is approximately twice the viscosity at 22°C. It is therefore expected that the birefringence relaxation times of molecules with temperature-independent geometry and flexibility will be nearly doubled by reduction of the temperature from 22°C to 5°C. The observed change in the spectrin heterodimer birefringence relaxation time Td 1 by a factor 2.0--2.3 when the temperature is decreased from 22°C to 5°C therefore indicates that the structure of the spectrin heterodimers is roughly the same at these temperatures. Our observation that spectrin solutions exhibit several birefringence relaxation times may indicate that the solutions were not fully monodisperse. However, if the spectrin molecules are flexible, the presence of several relaxation times may also be an inherent property of the spectrin molecules themselves. The presence of some high molecular weight aggregates cannot, however, be fully ruled out. This is particularly true in the spectrin heterotetramer solution since the spectrin aggregate and the heterotetramer peaks in the elution profile from the gel filtration column were not always well resolved. However, since additional centrifugation prior to the electro-optic measurements of spectrin heterotetramers only had a minor effect on the observed birefringence relaxation times, the lack of spectrin monodispersity does n o t appear to be a serious methodological problem. The possibility that the two or more decay constants are inherent of the experimental set-up is ruled out by the fact that the apparent birefringence relaxation of propylene carbonate yielded a simple exponential decay. Tables I and II show that the electro-optic properties of human heterotetramer and heterodimer solution are nearly the same. Of particular interest is the observation that the birefringence relaxation times are roughly the same for heterotetramers and heterodimers. The birefringence relaxation time for rigid rods and rigid prolate ellipsoids can be calculated using the equations of Broersma [8] and Perrin [9], respectively. These equations show that if rigid rod-like molecules double their lengths, the birefringence relaxation time will have a 6-fold increase. The near independence of Bsp on salt concentration may indicate that permanent dipole m o m e n t is largely responsible for the electrically induced birefringence. This is substantiated by the difference between rise and decay birefringence kinetics [11,12]. Our observation that Bsp for spectrin heterodimers and heterotetramers is approximately the same does n o t in any way exclude the possibility that heterotetramers are formed by end-to-end association of heterodimers provided heterodimers and heterotetramers are flexible molecules. The electron microscopic studies by Shotton et al. [4,5] indicate that the spectrin heterotetramers are formed by an end-to-end association of two heterodimers. The work by Shotton et al. [4,5] as well as previous studies in this laboratory [ 1,6,7,10,16] further indicate that spectrin heterodimers are highly elongated molecules. The electro-optic measurements presented here unambiguously rules out the possibility that human spectrin heterotetramers are formed by making a stiff joint between end-t0-end associated rigid heterodimers. The birefringence relaxation time of molecules consisting of two stiff

80 rod~shaped subunits linked by a flexible joint is about three times the birefringence relaxation time of the individual unlinked rigid molecules [8,9,17]. We can therefore also rule o u t the possibility that spectrin heterotetramers consist of e n d - t o ~ n d associated rigid heterodimers joined by a flexible link. The conclusion from this electro~ptic study and our earlier electro-optic and light-scattering studies [1,10] therefore is that the human spectrin heterotetramers as well as heterodimers are molecules with a high degree of flexibility.

Acknowledgements The authors gratefully acknowledge the generous help of Professors K.B. Eik-Nes and P.C. Hemmer. They also are greatly indebted to those who volunteered to donate blood to this project.

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Mikkelsen, A. and Elgsaeter, A. (1978) Biochim. Biophys. Acta 5 3 6 , 2 4 5 - - 2 5 1 Ralston, G., Dunbar, J. and White, M. (1977) Biochim. Biophys. Acta 491, 345--348 UngewickeU, E. and Gratzer, W. (1978) Eur. J. Biochem. 8 8 , 3 7 9 - - 3 8 5 Sh otton, D., Burke, B, and Branton, D. (1978) Biochim. Biophys. Acta 5 3 6 , 3 1 3 - - 3 1 7 Shotton, D.M., Burke, B.E. and Branton, D. (1979) J. Mol. Biol. 131, 303--329 B~e, A., Elgsaeter, A., Oftedal, G. and Strand, K.A. (1979) Acta Chem. Scand. A 3 3 , 2 4 5 - - 2 4 9 Stokke, B.T. and Elgsaeter, A, (1981) Biochim. Biophys, Acta 6 4 0 , 6 4 0 - - 6 4 5 Broersma, S. (1960) J. Chem. Phys. 32, 1626--1631 Perrin, F. (1934) J. Phys. Radium 5 , 4 9 7 - - 5 1 1 Elgsaeter, A. (1978) Biochim. Biophys, Acta 536, 235--244 O'Konski, C.T. and Krause, S. (1976) in Molecular Electro-optics (O'Konski, C,T., ed.), part 1, Chapt. 3, pp. 63--120, M. Dekker, New York Yoshioka, K. and Watanabe, H. (1969) in Physical Principles and Techniques of protein Chemistry (Leach, S.J., ed,), Part A, pp. 335--367, Academic Press, New York Yoshioka, K. and O'Konski, C,T. (1968) J. Polym. Sci. Part A-2, 6 , 4 2 1 - - 4 3 1 Yoshioka, K. and O'Konski, C.T. (1966) Biopolymers 4 , 4 9 9 - - 5 0 7 Nakayama, H. and Yoshioka, K. (1965) J. Polym. Sci., Part A, 3 , 8 1 3 - - 8 2 5 Mikkelsen, A. and Elgsaeter, A. ( ] 9 7 8 ) Acta Chem. Scand. A32, 1017--1020 Yu, H, and Sto ckmayer, W.H. (1967) J. Chem. Phys. 47, 1369--1373