Axonal transport of actin and regeneration rate in non-myelinated sensory nerve fibres

255

Brain Re~earch. 333 (1985) 255-26(I Elsevier BRE 1(1708

Axonal Transport of Actin and Regeneration Rate in Non-Myelinated Sensory Nerve Fibres W. G. McLEAN

Department of Pharmacology and Therapeutics, UniversiO, of Liverpool, l,iverpool L69 3 BX ( U. K. (Accepted August 10th, 1984)

Key words: axonal transport - - regeneration - - actin - - vagus nerve - - cytoskeleton

lhe relationship was examined between the rate of regeneration and rate of axona[ transport of actin in the sensory fibres of the rabbit vagus nerve. Regeneration rate. determined as the distance moved by [~S]methionine-radiolabelled. fast-transported proteins beyond a crush, was about 3 mm,day. The rate of transport of actin, identified by two-dimensional polyacrylamide gel electrophoresis with fluorography, and DNase affinity chromatography, was 25-30 mm/day. No slower rate of actin transport comparable with regeneration rate, could be found in either control or regenerating nerves. While the provision of actin, by slow axonal transport, to the axonal growth cone may be essential for nerve regeneration, the regeneration rate is not directly controlled by the rate of actin transport.

INTROI)U('TION At least three distinct groups of proteins are transported intra-axonally in the sensory fibres of the rabbit vagus nerve 7,~. When radiolabelled with [-~sS]methionine, two low molecular weight proteins can be detected as the major proteins of the fastest moving group, transported at a rate of 15-17 mm/h. A prorein of molecular weight 43,000 is the major radiolabelled protein of the next slower transported group, at 25-30 mm/day, while the slowest group so far detected, at 12-15 mm/day, consists almost entirely of a pair of proteins of molecular weight 54,000 and 56,000. It has been suggested that the major proteins of the slowest groups were actin and tubulin, respectively, bv analogy with transported proteins in other nerves 1.4.11. The rate at which axons regenerate after being damaged is comparable in most nerves with the rate of axonal transport of actin, viz. from 2 - 6 ram/day. Wujek and Lasek 12 have suggested that the correlation between regeneration and actin transport reflects a direct involvement of the axonal transport of actin, along with other slow c o m p o n e n t b (SCb) proteins, in the locomotion of the axonal growth cone. In

view of the apparent discrepancy between the rate of transport of actin in the sensory fibres of the rabbit vagus nerve and other nerves, l have examined the rate of regeneration in those fibres and identified more soundly the rate of axonal transport of actin. MATERIAI.S AND METtlODS

Animals Male albino rabbits weighing around 2 kg were used. Measurements o f nerve regeneration and axonal transport Rabbits were sedated with diazepam (Valium), 2.5 mg/kg i.m., and neuroleptanalgesia was induced with Hypnorm, i.e. a combination of fentanyl, 60 ,ug/kg, and fluanisone, 3.5 mg/kg i.m. One vagus nerve was surgically exposed and crushed with 4/0 silk thread pulled tightly against a glass rod. The site of the crush was marked. The wound was sutured and treated with bupivacaine (Marcain), 2 mg i.d., prior to recovery of the animal. At various times from 5 days to 17 days later, the rabbits were reanaesthetized. Both nodose ganglia were exposed and four in-

Corre,spondence: W. (-i. Mcl,ean. Department of Pharmacology and Therapeutics, University of Liverpool. I,iverpool 1,69 3BX, LJ.K. 0(H)6-8993;85;$(13.3(1© 1985 Elsevier Science Publishers BV, (Biomedical Division)

25~ jections, each of 5 ,ul (40 .uCi) I..-[35S]methionine ( > 1000 Ci/mmol; Amersham) were made into each ganglion through a glass micropipettc, tip diameter 15 um, under hydrostatic pressure, The animals were then allowed to survive for 5 h under sedation, or for a further 3 days, 4 days or 12 days (post-operative procedure as previously). At appropriate times, rabbits were killed and the vagus nerve and nodose ganglia from each side removed and frozen. All experiments were carried out at least four times.

One-dimensional electrophorests In some experiments, whole vagus nerves wcrc cut into 5 mm pieces. Each piece was homogenized in ¢~(I ul of2% w/v SDS and l(l~ wlv glycerol io water l-ire ~1 aliquots were removed fit that stage lor liquid scintillation counting. After furthcr treatment as tiescribed prcviously~, the remainder o! the samples were applied consecutively to a 5-1~% gradient SDS-polyacrylamide slab gel and electrophoresed for 16-18 h.

DNase affinity chromatography Two-dimensional electrophoresis One cm pieces of nerves which had been radiolabelled 4 days previously were subjected to two-dimensional polyacrylamide gel electrophoresis by the method of O'Farrel 9, with slight modificationsS. They were first homogenized in a glass/glass homogenizer in 14ui of an aqueous solution containing 1% w/v sodium dodecyl sulphate (SDS), 10 M urea and 10% v/v 2-mercaptoethanol. Nonidet P-40 and Biolyte ampholines (Bio-rad; 4:1 o f p H 5-7: pH 3-10 by volume) were each added to final concentrations of 2% v/v, along with 4 mg urea (in order to maintain 10 M urea concentration). The homogenizer was rinsed with 20/~1 of a solution containing 10 M urea, 5% v/v 2-mercaptoethanol, 2% v/v Nonidet P-40 and 2e~ ampholines, and the wash added to the sample. Samples were then centrifuged at 85.000 g for 1 h at 2O ~C. Polyacrylamide rod gels containing 10% acrylamide-N,N'-methylene bis-acrylamide (18:1 by weight), 10 M urea, 3% v/v Nonidet P-40 and 8% v/v ampholines (in the above ratio) were prepared, and 20ktl of the supernatant of each nerve sample applied to each gel. Gels were electrofocussed ~ at 400 V for 12 h and finally at 800 V for 1 h, with cooling. A mixture of standard proteins of known isoelectric points was run simultaneously. Rod gels were removed from their tubes and equilibrated for 15 rain in a solution containing 8% w/v glycerol, l % v/v 2-mercaptoethanol, 1% w/v SDS in 0.5 M Tris buffer pH 6.8 (in later experiments, a second equilibration of 5 min was added, without 2-mercaptoethanol). Equilibrated gels were then electrophoresed for 16-18 h in the second dimension on 5- 18% gradient SDS-polyacrylamide slab gels.

Five mg deoxyribonuclease (DNase type i from bovine pancreas; Sigma) was incubated with activated Sepharose 4B (Pharmacia) in 0.1 M sodium bicarbonate at 4 °C for 12 h. The resulting D N a s e Sepharose was washed repeatedly with water and packed into 1 ml polythene columns. They were first equilibrated with 25 ml 0.5 M sodium acetate buffer pH 6.5, containing 4 M guanidine-HCI find 30% v/v glycerol; then with 250 ml 10 mM Tris buffer pH 7 5. containing 5 mM CaCI 2 (buffer A). 100 mm of vagus nerve (from 30 mm to 80 mm from the nodose ganglion, pooled from two nerves), radiolabelled 48 h previously, were homogenized in l ml 10 mM Tris buffer pH 7.5, containing 1 mM CaCI 2. Samples were centrifuged at 85,000 g for 20 rain and the supernatant applied to the DNaseSepharose column. Columns were washed with 5 mi buffer A, followed by 2 ml of a buffer containing 0.5 M sodium acetate, 0.75 M guanidine-HCI and 30% w/v glycerol, pH 6.5. This was followed with 2 ml of 3 M guanidine-HCl, 1.0 M sodium acetate and 30c~ w/v glycerol, pH 6.5. The eluate from the first wash and the last two together were dialyzed against water, freeze-dried and prepared for one-dimensional etectrophoresis as described above.

Fluorography The resulting gels from one- or two-dimensional electrophoresis were impregnated in 2,5-diphenyioxazole (PPO) in dimethyl sulphoxide (DMSO), dried and kept in contact with pre-flashed Fuji RX autoradiography film at -70 °C for up to 14 days, as described previously8. All slab gels also contained 0.2 ~Ci ~4C methylated protein standards (Amersham) for calibration of molecular weights. All reagents used were electrophoresis grade,

257 ewhere available, except for reagent grade D M S O .

°~

RESULTS

Rate of nerve regeneration The rate of regeneration of the sensory vagus nerve fibres was calculated from those e x p e r i m e n t s in which the axonally t r a n s p o r t e d proteins of previously crushed nerves were radiolabelled and the radioactivity measured 5 h later. In that way, the farthest distance from the site of the crush over which the radioactive proteins could be t r a n s p o r t e d was taken as a measure of the m a x i m u m extension of the growing axon tips 2.3. Fig. 1 shows that the regeneration rate in the sensory fibres, as m e a s u r e d in that way, was 2.8 + 0.5 (S.D.) m m / d a y with an a p p a r e n t delay of only 0.8 days.

Axonal transport of actin The presence of actin in the 2 5 - 3 0 mm/day phase of axonal transport was confirmed by two-dimensional electrophoresis of a nerve containing radiolabelled proteins at a point 4 0 - 5 0 mm from the ganglion, 48 h after radiolabelling. O n e heavily radiolabelled protein with pI of 5 . 5 - 5 . 6 and a p p a r e n t molecular weight of 43,000 was present at that time (Fig. 2). This was identical in position to rabbit muscle actin e l e c t r o p h o r e s e d on separate gels at the same time (not shown).

'E 50[

Z_

4o

•

4a

_j= P

e-

4 Q

9.5

I~

3.5

Fig. 2. Fluorogram of a two-dimensional separation of 35S-radiolabelled, slowly transported proteins in the sensory fibres of rabbit vagus nerve 40-50 mm from the nodose ganglion, 48 h after injection of [35S]methionine. The spot marked with an arrow co-migrated with rabbit muscle actin.

R a d i o l a b e l l e d proteins from the 2 5 - 3 0 ram/day phase, again collected 48 h after radiolabelling, were subjected to DNase affinity chromatography5 prior to electrophoresis in one-dimension. Fig. 3 shows a densitometry scan of the fluorogram of the eluates from the DNase affinity column and shows the presence of a radiolabelled protein of molecular weight around 43,000 with an affinity for DNase. Normal vagus nerves therefore transport actin at a rate of 2 5 30 mm/day. It was still considered possible, however, either that actin transport was significantly slowed in regenerating nerves or that a significant amount of actin, in addition to that transported at 2 5 - 3 0 mm/day, was

r

soi. j

5"5 2°I --~ I0~

t..Z_

A

t

5 10 days after crush

i

"15

Fig. 1. The length of regenerated axons in the rabbit sensory vagus nerve at various times after the nerve was crushed. Regenerated fibre length was measured as the distance from the crush site to which 35S-radiolabelled proteins were axonally transported, 5 h after injection of [35S]methionine into the nodose ganglion. Regression coefficient = 0.986.

g2'.s 8'g

4'6

3o

apparent molecular weight (Kdaltons) Fig. 3. Densitometer trace of a fluorogram of the second eluate from a DNase affinity column, containing 3sS-radiolabellcd proteins of rabbit vagus nerve radiolabelled 48 h previously. One radioactive protein of molecular weight around 43,000 was retained.

25~; transported at a rate closer to the regeneration rate of 3 mm/day in normal or regenerating nerves. 72 h

92

after radiolabelling, vagus nerves contain the major

69

bulk of the radiolabelled 25-30 mm/day phase about 60-65 mm from the ganglionS, while the slower

46

12-15 mm/day phase phase has not 3'el reached that

-430

point. Fig. 4 shows densitometer scans of fluorograms of radioactive proteins at that point in both normal and regenerating nerves. Actin and other

~14

radiolabelled at that point in both control and regenerating nerves 17 days after a crush. Results were

i 92

identical in nerves studied 5 days after a crush except

69

that the growing axons had not extendcd as far. Fig. 5 shows overexposed fluorograms of regenerat-

46

ing and control nerves 12 days after radiol:tbelling. In the case of the regenerating nerves, the growing axons would not at that time have reached their target organs, which are farther than 130 mm from the cell of molecular weight 54,000-56,00{). which co-migrated with pig brain tubulin a. and a much less heavily radiolabelled protein of molecular weight identical to actin. There was no evidence for increased intensity of labelling of actin around 40 mm from the ganglion as would be anticipated if a significant bulk of actin was being transported it! 3 ram,day in either control or regenerating nerves.

"o .x,, "" ¢-

proteins in the same transport phase are sccn to bc

bodies. The tnain radiolabclled proteins present are

,.~

3O '

~14 , __

i 0

i

i 20

i

i ~ J 40

mm f r o m n o d o s e

--¢ o E

E

(9

Q-

, __1.... C.] 60 80 ganglion

Fig. 5. Fluorograms of .~SS-radiolabclled proteins m consecutive 5 mm pieces of rabbit vagu~ nerve and m~dosc ganglion (far left). 12 days after radiolabelling. In (h). the nerve had been crushed (at arrow) 14 days prior to radiolat~elline and allowed to ~cgencratc. The conlralatcral uncru'~hcd ner~,c i,, shown in (a).

DISCUSSION

I

9ds 6'9 ,'6

The relatively fast rates of the two phases of axonal transport of cytoskeletal proteins in scnsorv fibres in

3b

apparent molecular weight (Kdaltons) Fig. 4. Densitometer trace of a tluorogram of ,~sS-radio]abelled proteins in rabbit vagus nerve 60-65 mm from the nodosc ganglion 72 h after radiolabelling. The position of actin is marked with an arrow. The upper trace shows a control nerve; the lower one a nerve which had been crushed 40 mm from the ganglion 17 days previously.

rabbit vagus nerve have been discussed b e f o r e < "l'hey are slightly faster than. but comparable with, the apparent rates of transport of proteins in rat and guinea pig vagal motor fibres "~,'4. But they are an order of magnitude faster than the rates of transport of equivalent groups of proteins in the rat sciatic nerve 4 or the guinea pig optic nervc~J. It was theretore clear that a discrepancy existed between the rate of transport of aetin and that of nerve regeneration in the vagus nerve, based on the assumptions that actin was indeed a component of the 2 5 - 3 0 mm/day phase of slow transport, and that vagal sensory fibres regenerated at a rate comparable with motor, or parasympathetic, fibres in the same nerve 3. The present work

259 was designed to confirm those assumptions. Actin was identified as the m a j o r 35S-radiolabelled protein of the 2 5 - 3 0 mm/day phase of transport, by two-dimensional electrophoresis and by DNase affinity c h r o m a t o g r a p h y (Fig. 3). The two-dimensional analysis (Fig. 2) d e m o n s t r a t e d at least 40 radiolabelled proteins in that transport phase. Those proteins remained radiolabelled in the nerve for several days more. The only other radiolabelled proteins which have so far been identified as being transported at a significantly slower rate were the a- and fl-subunits of tubulin s. Nerves removed from rabbits 12 days after radiolabelling (Fig. 5) contained radioactivity in the form of tubulin and actin only; no other 'wave" of transported proteins, or additional transport of radiolabelled actin at a rate less than 2 5 - 3 0 ram/day could be detected. This was the case in both control nerves and nerves crushed 14 days previously. The data also confirmed that the rate of regeneration of sensory fibres of the rabbit vagus nerve was identical to that of m o t o r fibres with the apparent difference that the onset of regeneration was more rapid~: the early onset r e p o r t e d here rests, of course. on the assumption that regeneration proceeds linearly with time at the early stage - - an assumption which may not be justified. The other possible reason for the discrepancy may have a purely experimental origin, viz. that the site of the lesion in the sensory fibres was considerably nearer the nerve cell bodies in the nodose ganglion than in the experiments with motor fibres which derived from the dorsal m o t o r nucle-

p o r t e d proteins were, however, observed. A report of those changes will be the subject of a separate work. It has been shown that the actin-containing phase of slow transport (SCb) reflects the active translocation of elements of the axonal cytoskeleton, including a number of proteins which may comprise the internal structure of the growth cone of regenerating fibres~. This and other evidence t,.~s clearly point to an important role of actin and other SCb proteins in nerve regeneration. The supply of actin by slow axonal transort to the tips of regenerating fibres could therefore limit the rate of regeneration. Wujek and Lasek12 have provided evidence to suggest that the relationship between SCb transport and regeneration rate is more fundamental than that. They have shown a similarity between the rates of axonal regeneration and SCb transport in two branches of a single axon. This would suggest that a direct relationship exists between the rate of SCb transport and the rate of axonal regeneration and that the mechanism which transports SCb proteins is also involved in the movement of the growth cone. In the work reported here I have shown that that similarity does not exist in the sensory vagus nerve. Unless the movement of the cytoskeleton in the vagus nerve is less well reflected in the axonal transport of cytoskeletal proteins than in other nerves, one must conclude that the rate of nerve regeneration is not d e t e r m i n e d directly by the rate of movement of the cytoskeleton but depends on other as yet undiscovered factors.

llS.

Radiolabellcd actin was transported normally during regeneration (Fig. 4). This dispelled the possibility that the rate of actin transport might be significantly reduced in regenerating nerves. Quantitative changes in the amounts of some fast and slowly trans-

REFERENCES

1 Black. M. M. and l,asck, R. J., Axonal transport of actin: slow component b is the principal source of actin for the axon, Brain Research. 171 (1979) 401-413. 2 Forman, D. S. anti Berenberg, R. A., Regeneration of motor axons in thc rat sciatic nerve studied by labelling with axonally transported radioactive proteins, Brain Research. 15(3(1978) 213-225. 3 Frizell, M. and Sjc:,strand, J.. l"he axonal transport of slowly

ACKNOWI.EDGEMENTS

This work was s u p p o r t e d by a Medical Research Council project grant. Technical assistance was proviced by Ms. A. M c K a y and Mr. D. Trafford.

migrating [311]leucine labcllcd proteins and the regeneration rate in regenerating hypoglossal and vagus nervcs of the rabbit. Brain Research. 81 (1974) 267-283. Hoffman. P. N. and Lasek, R. J., The slow component of axonal transport. Identification of major structural components of the axon and their generality among mammalian neurons, J. ('ell Biol., 66 (1975) 351-366. l.azarides, E. and Lindbcrg, V., Actin is the naturally occurring inhibitor of deoxyribonuclcase 1.. t'roc, nat. Acad. Sci. U.S.A.. 71 (1974) 2268-2272.

2 (~( J Lctourneau, P. C., lmmunocytochemical evidence for c,alocalization in neurite growth cones of actin and myosin and their relationship to cell-substratum adhesions, l)eYelop. Biol., 85 (19811 113-122. 7 McLean, W. G., Frizell, M. and Sjostrand, J.. Slow axonal transport of labelled proteins in sensory fibres of rabbit vagus nerve, J. Neurochern., 26 (1"976) 1213- 1216. 8 McLean, W, G., McKay, A. L. and Sj6strand, J.. Electrophoretic analysis of axonally transported proteins in rabbit vagus nerve, J. Neurobiol., 14 (1983) 227-23~. 9 0 ' F a r r e l , P. tt., High resolution two-dimensional electrophoresis of proteins. J. biol. Chem., 250 (1975) 4007-4021. 10 Tashiro, T.. Kasai, H. and Kurokawa, M., A calmodulin related polypeptide rapidly migrates within the mammalian ncrve, Biomed. Res., 1 (19801 202-299

11 Willard, M., Wiscman, M.. l.cvillc..I, rind ,~kcllc, |'.. ,\x.tl nal transport ofactin in retinal ganglion cclh, ; (",,1!B,,/ .~ (1979) 581 --591. 12 Wujck, J R. and Lasek, R .I.. c. orrclation t)l axonal IcgcTIoration and slow component b in two branchc:, ,~f a ~inglc axon, J. Neurosci., 3 (19831 243-251. 13 Yamada. K. M., Spooner, B. ,'q. and Wcssells. N. K., Axon growth: role of microfilamcnts and microtubulc~,./'roe t~at. Acad. Sci. U.S.A., 66 (1970) 121)6--1212. 14 Yokokoyama, K., Tsukita, S., lshikawa, H. and Kurokawa, M., Early changes in the neuronal cytoskeleton caused by fl, fl'-iminodipropio-nitrile: selective impairment of ncurofilament polypeptides, Biomed. Res.. 1 (1980) 537--547