@ Pergamon Int. J. Devl Neuroscwm. Vol. 12. No.]. pp. 49-56, 1994 Elsevier Science Ltd.

Copyright 0 1994 ISDN Printed in Great Britain. All rights reserved

M36-574894 $6.~+0.~

THYROIDAL STIMULATION OF TUBULIN AND ACTIN IN RAT BRAIN CYTOSKELETON

ARUNANGSU DE, SUMANTRA DAS, SUKANYA CHAUDHURY and PRANAB KUMAR SARKAR*

Division of Neurobiology, Indian Institute of Chemical Biology, 4 Raja S.C. Mullick Road, Jadavpur, Calcutta 700032, India

(Received 19 April 1993; in revised form 263uiy 1993; accepred 2 September 1993)

Abstract-In cultures of neonatal rat brain cells, labeled with 35Smethionine in the presence or absence of triiodothyronine (Ts), the hormone promoted a significant enhancement of labeled tubulin and actin in the insoluble fraction (30,OUO g pellet) of cell homogenate. To identify the specific sub-cellular fraction associated with this induction, organ cultures of 1 day rat cerebra were labelled with ass-methionine in the presence and absence of TJ and the insolubte fraction (30,000 g pellet) was subfractionated into mitochondria, plasma membrane and cytoskeleton. Analysis of the labeled proteins by SDS-PAGE, autoradiography and densitometry revealed a D-induced increase of Xl-&?% for both tubulin and actin, only in the cytoskeleton fraction without any significant effect on the other fractions. Similar results were obtained when plasma membrane or cytoskeleton were isolated directly from labeled cerebrum by conventional methods instead of fractionating from the 3O,O(Xl g pellet. Analysis of relative stimulation of labeled tubulm and actin by Ts in cytoskeleton fraction derived from primary cultures of neuronal (N) and glial (G) cells labeled with s5-methionine show that the stimulatory effect is predominantly on the N cells. Studies on the kinetics of induction of labeled tubulin and actin by Ts in the cytoskeleton fraction prepared from cerebra labeled with s?&methionine for 2,s and 18 hrs revealed no significant difference at 2 hrs; at 8 hrs, an increased incorporation into both tubulin and actin was reproducibly seen in the controls relative to Ts-treated samples. However by 18 hrs, this pattern reversed and an enhanced accumulation of both labeled tubulin and actin was observed under the influence of Ts. The mechanism of this apparently intriguing effect of Ts on the kinetics of association of tubulin and actin with the cytoskeleton has been discussed in the light of the dual effect of the hormone on tubulin viz. enhancing its stability as well as rate of synthesis. The overall results indicate that the thyroid hormones play a major role in the cytoskeletal transport of tubulin and actin from their site of synthesis to that of assembly thus facilitating axodendritic outgrowth and morphological differentiation.

Key words thyroid hormone, tubulin, actin, cytoskeleton, rat brain, neuronal and glial cells.

We have previously demonstratedlo that in primary cultures of neuronal (N) and glial (G) cells from l-day rat brain, labeled for 18 hrs with 14C leucine, triiodothyronine (T3) enhanced the level of both radiolabeled tubulin and actin relative to those in controls. Analysis of the distribution of the total labeled proteins in the soluble (30,000 g supernatant) and insoluble (30,000 g pellet) fractions of the cell sonicate showed that the major proportion (6045%) of the T3-stimulated labeled proteins including tubulin and actin are associated with the insoluble fraction. Since this fraction consists of at least three important subcellular fractions viz. mitochondria, plasma membrane and cytoskeleton and both tubulin and actin have been reported to be integral components of plasma membrane3l24 and cytoskeleton,*3,2t it was of interest to determine if the st~ulation of radiolabeled tubulin and actin by T3 was preferentially associated with any of the above sub~~ular fractions.

We report here, from analysis of the radioactivity profiles of proteins of each of the three subcellular fractions, prepared from organ cultures of rat cerebra labeled with 35S-methionine for 18 hrs, that relative to controls, T3 significantly enhances the accumulation of labeled tubulin and actin only in the cytoskeleton fraction with little or no effect in the mitochondrial or in the plasma membrane fraction. The significance of this stimulation of cytoskeletal proteins by T3 has been discussed.

*Author to whom correspondence should be addressed.

49

A, De L'f ffl.

EXPERIMENTAL PROCEDURES

Materia1.y

Tissue culture plasticwares were obtained from Nunclon, Denmark. Dulbecco’s modified Eagle’s medium deficient in four amino acids viz. I.-leucine, L-lysine, L-glutamine and L-methionine (DMEM-D), pipes, EGTA, ouabain, ATP, AMP, glu~se-6-phosphate and Dowex 1X-8-200 were from Sigma Chemicals, U.S.A. “5S-methionine (Specific activity>2~ ~~mmole) was purchased from Bhaba Atomic Research Centre (BARC), India. All other chemicals were of analyti~l grade, purchased locally. Fetal bovine serum and gentamycin were from Gibco-BRL, U.S.A.

Thyroid hormone depleted fetal calf serum

The endogenous thyroid hormones of fetal calf serum were depleted by successive adsorption (for 5 and 18 hrs) with Dowex 1 X-8-200 according to Samuels et aLz2

Radiolabelling of tissues

Detailed procedures for organ culture of cerebra from 1 day (O-24 hrs) old rat brain as well as separation and identification of N and G cells and their labeling has been described earlier.‘O Briefly, cerebra were freed from blood vessels and connective tissues using the isotonic buffer solution of Ahmed et al.’ Each cerebrum was minced into two pieces and the pooled tissues were cultured (2 ml/cerebra) in DMEM-D supplemented with appropriate amino acids and other nutrients. Since 3sS-methionine was used for labeling, the medium was supplemented with L-glutamine (0.58 g/l), t_-lysine (0.15 g/l) and L-leucine (0.105 g/l) in addition to KC1 (1.08 g/l), glucose (5 g/l), 1% penicillin streptomycin mixture (Gibco) and 5% thyroid hormone depleted fetal calf serum. Tissues were gassed with 5% CC&95% air and labeled in the presence or absence of 5 nM T3 for the indicated times using a gyratory shaker (70 RPM/mm). More than 90% of the cells derived from the cultured tissues after 18 hrs were viable as judged by trypan blue exclusion test. Primary cultures of N and G cells, separated on the basis of their differential attachment to poly-t-lysine coated dishes, were set up as described earlier.‘*

~l~bcelfular fractionation of the 30,000 g pellet

Cultured tissues were washed three times with ice cold PBS, homogenized with a Dounce type homogenizer using PBS containing 1 mM PMSF. The 30,OOOg pellets of the homogenate obtained by centrifugation at 4°C for 15 min, were fractionated into mitochondria, plasma membrane and cytoskeleton following the procedure of Carey and Hirschberg. The pellet was first suspended in sucrose-hepes buffer, pH 7.4, containing 1 mM PMSF by mild homogenization on ice. The suspension was spun at 2000 g for 10 min at 4°C and the supernatants were centrifuged at 11,000 g for 10 min to obtain the mitochondrial fraction. The 2000 g pellets were brought to 47% sucrose using a 60% stock solution, overlayered with 41% and 8.2% sucrose and centrifuged at 100,000 g in a swing bucket rotor for 90 min at 5°C. The plasma membrane fraction (a distinct band at the interface of 8.2% and 41% sucrose) was collected, diluted 5 times with PBS and then pelleted at 10.000 g for 10 min at 4°C. The purity of the plasma membrane was tested by meas~ing the enrichment of specific activities of Na’-Ki-ATPase, 14,23 5’ nucleotidase2 and glucose-6-phos- _ phatase* with respect to those in the total cell homogenate.

Cytoskeleton is operationally defined as the material sedimentable at 12,OOOg (10min) following lysis of cells in triton-X containing buffer. 7J* However, due to the presence of 2 M glycerol in the 0.5% triton-X containing microtubule stabilizing buffer, MSB (0.01 M pipes, pH 6.9,2 M glycerol, 1 mM MgS04, 2 mM EGTA)4 used for suspending the 30,000 g pellet prior to fractionation, the suspension was first sedimented at 100,000 g for 30 min at 4°C and the resulting pellet was resuspended in MSB without the glycerol in the presence of 1 mM PMSF by mild sonication. This suspension was then centrifuged at 12,000 g for 10 min at 4°C to derive the cytoskeleton fraction.

~irect~reparat~on of c~to~keleton and plasma membrane

For the direct preparation of cytoskeleton, labeled cerebra were washed thoroughly with PBS and then lysed with MSB containing 0.2% triton-X and 1 mM PMSF. Lysates were centrifuged at

Thyroidal induction of cytoskeletal tubulin and actin 51

100,000 g for 30 min, suspended in MSB without glycerol and then spun at 12,000 g for 10 min to obtain the cytoskeleton fraction.

The procedure of Carey and Hirschberg was employed for the direct preparation of plasma membrane from labeled tissues. The procedures employed for estimation of protein, determination of incorporation of 35S-methionine into protein, SDS-PAGE and autoradiography have been described earlier.‘O Densitometric scanning of gels was performed using an LKB Ultrascan Laser Densitometer.

RESULTS

Initial studies on the effect of T3 on the stimulation of labeled tubulin and actin in different subcellular fractions of the 30,000 g pellet were performed with organ cultures of cerebra from neonatal rat brain to obtain sufficient and reliable amounts of protein and TCA precipitable radioactivity in each of the three subcellular fractions. Figure 1 depicts the Coomassie blue pattern (A), autoradiography (B) and the densitometric scanning (C) of the solubilized 30,000 g pellets derived from homogenates of control and T3-treated cerebra labeled for 18 hrs with 35S-methionine and resolved by SDS-PAGE. Equal amounts of proteins were loaded from control and Ts-treated samples and no difference with regard to protein profiles could be discerned in the stained gels. Incorporation of 35S-methionine into total TCA precipitable proteins increased by 10-20% in the Ts-treated case. Densitometric scanning of autoradiograms from 3 different gels revealed that T3 significantly (JYO.005) increased the levels of a-tubulin, B-tubulin and actin by 2823, 3225 and 42?6%, respectively. The results of incorporation of 35S-methionine into total protein in three

-Cl

--B

-A

C T3 c T3

(A) (B) cc> Fig. 1. Coomassie blue staining pattern (A), autoradiogram (B), and densitometric scan of B (C) for the SDS-PAGE resolved proteins of the 30,OOOg pellets from 18 hr cultures of l-day rat cerebra labeled with 35S-methionine in the presence or absence (control) of 5 nM T3 and homogenized in microtubule stabilization buffer. Equal amounts of protein (40 ug) were loaded in each lane and the radioactivity applied were 40,000 and 44,008 CPM for control and T3. resnectivelv. T (a. 8) and A renresent the oositions _ \..I

of-marker cr- and 8-tubulin, and actin, run simultaneously.‘In this and all other susequent autoradiograms of 35S-methionine labeled proteins, the intensity of the 8-tubulin band is significantly higher than that of

cx-tubulin due to 50% greater methionine content of P-tubulin relative to a-tubulin.”

c f-3 c -f 3

Fig. 2. Autoradiograms of the SDS-PAGE resolved 3”S-methionine labeled poreins of different subcellular fractions prepared by fractionating the 3O.ooOg peilet as described in “Experimental Procedures”. Labeling conditions are same as described under legends to Fig. 1. Equal amounts of proteins were loaded for control

and T3-treated samples from each subfraction.

different subcellular fractions viz. mitochondria, plasma membrane and cytoskeleton derived from the 30,OOOg pellet following labeling of cerebra for 18 hrs in the presence and absence of Tj showed no differences in 3sS-methioniae incorporation into total proteins between control and TJ-treated sample in the case of mito~ho~dria and plasma membrane, but a significant increase (18-t-4%) by T3 only in the case of cytoskeleton. correspondingly, analysis of these fractions by SDS-PAGE and autoradiography showed a distinct T3-induced st~ulation of tubulin and actin only in the case of cytoskeleton (Fig. 2) without any signifi~nt difference in the case of mitochondria and plasma membrane.

To ensure that the results obtained with the cytoskeleton fraction derived from the 30,OOOg pelfet are not artifacts, we purified plasma membrane and the cytoskeleton fraction directly (avoiding the 30,OOOg pellet step) from labeled tissues and analyzed the effect of T3 on the stimulation of labeled tubulin and actin in such fractions, The purity of the plasma membrane preparations was indicated by at least B-10 fold enrichment in the specific activity of the marker enzymes (Na’-KS‘-ATPase and S-nucleotidase) and the absence of detectable G-(i-Pax activity (data not shown). The autoradiographic patterns of gels loaded with equal amounts of protein from the plasma membrane and cytoskeleton fractions derived from cerebra labeled with ~‘S-methionine for 18 hrs in the presence and absence of Ts are shown in Fig. 3. In agreement with the results obtained from the fractions derived from the 30,~g pellet (Fig. a), signifi~nt stimulation by T3 was seen only in the case of cytosk~leton where densitometric tracing of the autoradiograms demonstrated SO-70% increase in the intensities of the tubulin and the actin bands.

To determine the earliest time point of increase in labeled tubulin and actin in the cytoskeletal fraction in response to ‘I’s, cytoskeleton was prepared by direct tysis of cerebra fabeled with ?S-methionine in the presence and absence of T3 for 2 hrs, 8 hrs and 18 hrs. Analysis of the labeled cytoskeletal proteins at different time points by SDS-PAGE and autoradiography revealed no detectable stimulation of tubulin or actin by T3 at 2 hrs. However, at 8 hrs, even with equal amounts of protein applied on the gels, the corresponding autoradiogram (Fig. 4) displayed increased

Thyroidai induction of cytoskeletal tubulin and actin 53

Plasma mern brane Cytoskeleton

:; IT ,A

Fig. 3. Autoradiograms of the SDS-PAGE resolved proteins of plasma membrane and cytoskeleton prepared by direct fractionation of homogenates of l-day rat brain labeled for 18 hr with 35S-methionine in the presence or absence of T3. The amount of protein loaded in each lane for these two fractions was 100 kg and the corresponding amounts of radioactivities applied for control and Ts-treated samples were

as follows: plasma membrane (20,ooO and 21,000 CPM) and cytoskeleton (20,000 and 25,100 CPM).

intensities of the tubulin and actin bands in the controls relative to those of the T3-treated samples. This pattern, although surprising, was found to be reproducible in at least twoseparate experiments. Nevertheless, by 18 hrs the labeling pattern reversed and the densitometric tracings of the autoradiograms revealed that T3 stimulated the accumulation of both labeled tubulin and actin by more than 50%.

To further investigate the relative effect of T3 on the st~ulation of labeled tubulin and actin in N and G cells, cytoskeleton was prepared from primary cultures of N and G cells labeled for 18 hrs with 35S-methionine. Analysis of the radioactivity profile of these cytoskeletal proteins by SDS-PAGE and autoradiography showed that induction of labeled tubulin and actin was predominantly in the neuronal cells (Fig. 5). However, in addition to tubulin and actin the intensity of several other cytoskeletal proteins also appeared to be enhanced in the N cells. In contrast, the G cells displayed a relatively small but specific stimulation of tubulin and actin.

DISCUSSION

The principal finding from the present exper~ents is that out of the three sub~~ular fractions (mitochondria, plasma membrane and cytoskeleton) derived from organ cultures of l-day rat cerebra labeled for 18 hrs with 35S-methionine in the presence and absence of T3, stimulation of labeled tubulin and actin by T3 was found only in the cytoskeletal fraction. This conclusion was found to be valid irrespective of whether the fractions were isolated from the 30,OOOg pellet wherein the initial stimulation of these proteins were first observed or whether they were isolated directly from labeled cerebra involving conventional procedure6 for subcellular fractionation. Thus the results obtained are unlikely to be due to artifacts generated during the fractionation of the 30,000 g pellet.

54

C t

T3 C T3 C T3 1ur

C T3 C T, C T, L___.--___I I__--1 L.----J

2h 8h 18h 2h 8h 18h

(A) (B> Fig. 4. Kinetics of the effect of T.J on the induction of labeled tubulin and actin in the cytoskeletat fraction. Cposkeleton was prepared by direct fractionation of homogenates of I-day rat brain labeled for 2,X or 1 S hrs with 3sS-methionine in the presence or absence of TJ. The Coomassie blue staining patterns (A) and the autoradiograms (B) of the soluhihzed cy$oskeletal proteins at different time points are shown. The amount of protein loaded in each Iane was 150 kg and the corresponding radioactivities in the control- and Ti-treated samples at 2.8 and 1 S hr were 4fK)O and 4200 CPM, 1 O,StK) and 9000 CPM and 2O.tK10 and 28,OtX)

CPM respectively.

Fig. 5. Effects of Tj on the relative stimulation of cytoskeletal proteins on N and G cells. Primary cultures of N and G cells were labeled with ‘?-methionine for 18 hr in the presence or absence of 5 nM T3. Cells were lysed in microtubule stabilization buffer containing 2 mM PMSF and the 12,000 g, 10 min pellets (cytoskeleton) obtained from the lysates were solubilized in PBS containing 2% SDS by mild sonication and heating in a boiling water bath. The autoradiograms of gels loaded with equal amounts of cytoskeletal proteins from N and G cells are shown. NC,, NT, Gc and CT represent N-control, N-Ts treated, G-control

and G-TX treated. respectively.

Thyroidal induction of cytoskeletal tubulin and actin 55

A close analysis of the nature of stimulation of tubulin and actin by T3 in the cytoskeletal fraction prepared from labeled cerebra reveals several interesting points. First, when equal amounts of labeled cytoskeletal proteins from control and Ts-treated tissues were loaded on SDS-polyacry- lamide gels, there is no selective stimulation of total tubulin or actin since the Coomassie blue stained patterns of these gels were virtually identical (Fig. 2A). Significant stimulation of tubulin and actin could be visualized only in the autoradiograms of the corresponding gels (Fig. 2B), i.e. at the level of newly synthesized labeled proteins. Furthermore, analysis of the relative stimulation in the cytoskeleton fraction prepared fromN and G cells showed that this stimulation is predominantly in the neuronal cytoskeleton (Fig. 5). It should also be pointed out that comparison of the intensity of the protein bands in the cytoskeletal fractions in the autoradiograms from control and T3-treated samples indicated a general stimulation of all cytoskeletal proteins by T3 (Fig. 5).

In this context, it is appropriate to consider the role played by cytoskeleton in the developing brain cells, particularly during the critical period of cell differentiation and synaptogenesis which require a large amount of both tubulin9,16 and actin. 5,20 These proteins are continuously needed for axodendritic outgrowth 20,25 but are synthesized only in the cell body, and the cytoskeleton, apart from maintaining the morphology of the cells, is thought to play a main role in the transport of these proteins from the site of synthesis to that of assembly. 13,21 Neonatal hypothyroidism in rats is known to cause severe impairment of the morphological differentiation of neurons12,15,17,19 and to delay the morphological maturation of glial cells. l5 Since the use of T3-depleted serum in the culture medium employed here mimics that of hypothyroid condition, the ability of exogenous T3 to stimulate the level of these proteins in the cytoskeleton is consistent with the in viva findings.

Another intriguing finding here is the time dependent alteration in the pattern of distribution of labeled proteins in the cytoskeletal fraction from control and T3 treated tissues seen during the kinetic studies. Comparison of the labeled protein profiles between control and T3 treated cytoskeleton at different time points (Fig. 4B) revealed marginal difference at 2 hrs, consistently more labeled proteins in the control relative to the T3 treated sample at 8 hrs and a clear stimulation of most of the labeled proteins by T3 at 18 hrs. The exact reasons leading to this altered distribution of labeled proteins between 8 and 18 hrs are not clear. However, the mechanism of the stimulatory effect of T3 on the induction of tubulin has been investigated by us earlier’ and is known to be largely due to an increased stability of tubulin, and partly due to its enhanced synthesis. It is conceivable that a similar situation exists for the cytoskeletal proteins and that at 8 hrs time point the enhanced accumulation of the labeled proteins in the controls could be primarily due to the stabilization effect of T3 with virtually no increase in the rate of synthesis since incorporation of 35S-methionine into total proteins at 8 hrs was found to be relatively lower than that in the controls. This is in accord with the fact that in neuroblastoma cells in culture, early neurite outgrowth during the first 6 hrs of culture is independent of de mvo protein synthesis23 suggesting the use of a preformed pool of tubulin subunits. By 18 hrs, the initial lag in protein synthesis is over, presumably due to the pool reaching a threshold level, and the synthesis of tubulin and actin is enhanced due to the continuing need for the microtubule and microfilament proteins for morphological differentiation and axodendritic outgrowth 25 induced by T3. The fact that at 18 hrs, despite the enhanced level of newly synthesized cytoskeletal proteins under the influence of T3 (Fig. 5) the amount of total tubulin or actin in the cytoskeleton does not increase with respect to control, suggests that their presence in the cytoskeleton is of a transient nature and the cytoskeleton is involved in the translocation of these proteins from their site of synthesis to their ultimate destination. This is consistent with the recent report that microinjection of tubulin into nerve cells followed by electron microscopy shows the movement of free subunits from the proximal to the distal end of the cytoskeleton.13

Acknowledgements-This work has been partially supported by a grant from the Department of Science and Technology, Government of India. Technical help from Mr B. Sanyamat and Mrs L. Chakraborty and secretarial assistance from Mr D. K. Guin is also gratefully acknowledged.

REFERENCES

1. Ahmed Z., Walker P. S. and Fellous R. E. (1983) Properties of neurons from dissociated fetal brain in serum free culture. J. Neurosci. 3,2448-2462.

2. Aronson N. N.. Jr and ‘l‘ouster 0. ( lY74) Isolation cd rat liver plasma membrane fragments in isotonic sucro\c blcr/l Enzymol. 31,Yl-102.

3. Bhattacharya B. and Wolff J. (1975) Membrane bound tubulin in brain and thyroid tissue. J. hid. Chrm. 250,7639-7646. 4. Black M. M., Keyser P. and Sobel E. (I CM) Interval hctwccn synthesis and assembly ofcytoskeletal proteins in cultured

neurons. J. Neurochem. 6,1(x)5-10 12. 5. Bray D. and Thomas C. (1976) [Jnpolymerized actin in fibroblasts and brain. J. m&t. &c/l. 105,527-544. 6. Carey D. J. and Hirschberg C. B. (lYX0) Kinetics of glycosylation and intracellular transport of sialoglyc~,pr~,tcinh iii

mouse liver. J. hid. Chum. 255,434X-4354. 7. Carraway K. L. and Carraway C‘. A. r’. ( IYXY) Membrane cytoskelcton interactions m aonnal cells. ~i~~~~?/~~? hrtrpht 5

Aciu Y&3,147-171. 8. Chaudhury S.. Chatterjec D. and Sarkar P. K. (lYg5) induction of brain tubulin by tritod~~thyr(~rline-dual ellecis ol

the hormone on the synthesis and turnover of the protein. Brain Res. 339, IYI-lY4. Y. Daniels M. P. (1973) Fine structural changes in neurons and nerve fihrcs associated with colchicine inhibition of nerve

fibre formation in i&+0. J. C’i+l Bid 58, 463.-470. IO. De A.. C’haudhury S. and Sarkar P. K. ( I YYI) Thyroidal stimulation of tubulin and actin in primary cultures of ncmonal

and glial cells in rat brain. Inr. J. dr,vl Ne~rrosc~i. Y, 3X1-390. 1 I. Eipper B. A. (1974) Properties of rat brain tubulin. J. Hiol. C‘lim~. 24Y, 1407-I4lh. 12. Faivre C., Legrand c‘. and Rabic A. (tOX3) Effect of thyroid deficiency and corrective effects of thyroxine on

microtubules and mitochondria in cerebcllar Purkinje cell dendrites of developing rats. Devl Rrain Res. 8,21--30.

13. Hirokawa N. (1991) Molecular architecture and dynamics ot neuronal cytoskeleton. In Neuronal Cytoskcleton (cd. Burgoyne R. E.). pp. 5-74. Wiley Liss Inc.. New York.

14. Jorgensen P. L. and Skou J. C. (lY71) Purification and characterization of Na +-I( ‘-ATPase. The influence of detergents on the activity of Na’-K+-ATPase in preparation from outer medulla of r&bit kidney. Biochinr. hio&rt~ .,lcm 233, 366-380.

15. Legrand J. (IYXO) Effect of thyroid hormones on brain development with particular emphasis on glial cells and n,~elination. In Mrcltirfi~~ci~linrtr~, Api~~(~~~~~r to Brrritz ~~~~l~~~~,~~e~~. (eds DiBenedetta C.. Bahrzas R.. Combos G. and Porallati G.). DD. 27Y-2Y2. Else~ier/N~,rth Holland Biomedical Press. Amsterdam. , .&

Ih. Mogan J. L. and Seeds N. W. ( lY75) Tuhulin constancy during morphological diffcrenti~~ti(~ll ot mouse neurohlastoma cells. J. Cell Viol. 67, 136 145.

17 Nicholson J. L. and Atlaman J. (lY72) Synaptopencsis in the rat ccrehellum. Effect of early hypo and hyperthyroidism. Science 176,5X-532.

18. Niggli V. and Borger M. M. (19X7) Interaction of the cytoskeleton with plasma membrane 100,97-l 21. 19. Nunej J. (ISSS) Thyroid hormones, In Handbook [~~N~llrochemisrrv, Vol. 8 (cd. Lajtha A.), pp. l-23. 20. Peters A.. Palay S. S. and Webster H. De. F. ( IY76) Microtubules and neurofilaments. In The fin~srrrtcncreof~~~c’nr’n’olr,

sytem. pp. 25-26. Harper E. Row Publishers, New York. 21. Rotman A., Helman J. and tinder S. (1982) Association of membrane and cytoplasmic proteins with cytoskeletons in

blood platelets. Biochemisrr~ 21, 17 13-l 719. 12. Samuels H. H., Stanley F. and Casanova J. (1079) Depletion of L-3-5-3’ triiodothyronine and L-thyroxine in euthyroid

calf serum for use in culture studies of the action of thyroid hormone. Endocrinakt,~,v 105, MRS. 2.3. Seeds N. W.. Gilman A. G.. Amano 7‘. and Nirenberg M. W. ( 1970) Regulation of axon f[~rmatjon by clonal lines of a

neuronal tumor. Proc. mm. Acad. Ser. U.S.A. 66, IhO-167. 14. Valenti F. B., Aunis D. and Roussct B. (1’183j Evidences for tubuiin binding sites on cellular membrane: plasma

membrane, mitochondriel membrane and secretory granular membrane. 1. Cd{ Rid 97.209-216. 25. Yamada K. M.. Spooner B. S. and Wells N. K. (lY70) Axon growth, role of microfilaments and microtubules. 66,

1206-1212.