Protein synthesis and axonal transport in goldfish retinal ganglion cells during regeneration accelerated by a conditioning lesion

Brain Research, 251 (1982) 25-37 Elsevier Biomedical Press

25

Protein Synthesis and Axonal Transport in Goldfish Retinal Ganglion Cells During Regeneration Accelerated by a Conditioning Lesion IRVINE G, McQUARRIE* and BERNICE GRAFSTEIN Department of Physiology, Cornell University Medical College, New York, N Y 10021 (U.S.A)

(Accepted April 22nd, 1982) Key words: axonal growth - - retinal ganglion cell - - goldfish - - axonal transport - - optic nerve regeneration - - conditioning lesion effect

Axonal outgrowth in goldfish retinal ganglion cells following a testing lesion of the optic axons is accelerated by a prior conditioning lesion. Changes in protein synthesis and axonal transport were examined during the accelerated regeneration. The conditioning lesion was an optic tract cut made 2 weeks prior to the testing lesion, which consisted of a tract cut at the chiasma, so that nerves subjected to either a conditioning lesion ('conditioned nerves') or a sham operation ('sham-conditioned nerves') could be examined in the same animal. In the retinal ganglion cells of conditioned nerves, the incorporation of [SH]proline into protein began to increase between 1 and 8 days after the testing lesion. The amount of fast-transported labeled protein was elevated to about 8 × normal by 1 day after the testing lesion but had decreased to about 3-5 × normal at 8 and 22 days. The 8 and 22 day values were not significantly different from those in sham-conditioned nerves or nerves that had received a testing lesion alone. For slow protein transport, the instantaneous amount transported was 15-16 × normal in the conditioned nerves at 1 and 8 days after the testing lesion, and the velocity of slow transport, which was already elevated above normal by I day after the testing lesion, was elevated still further by 8 days - - to a value in excess of 1.5 ram/day (compared to 0.2-0.4 mm/day in normal animals). We believe that the enhanced outgrowth resulting from the conditioning lesion is due to a transient increase in the amount of fast transport (possibly responsible for a decreased delay in the initiation of sprouting), and a sustained increase in the amount and velocity of slow transport (which may account for an increased rate of elongation). INTRODUCTION

tioning interval is 2 days and the o p t i m u m conditioning interval is 2 weeks8,9,10.

In m a n y kinds o f axons, axonal regeneration is faster if the axons have been previously injured. Thus, if axonal outgrowth is measured after a standard testing lesion that has been preceded by a conditioning lesion, the initial delay required for axons to transverse the lesion site is reduced, and, in most instances, the rate at which newly-formed axons advance is more rapid, c o m p a r e d to values obtained after a testing lesion alonel,9,10,17,24,27, 3133,35,37,3s,45,50. The most marked conditioning lesi-

The newly-formed portion o f the regenerating axon presumably receives proteins either from the nerve cell b o d y via axonal transport, or from the mobilization o f proteins within the surviving portion o f the axon lying proximal to the lesion12,16,17, 19,21. There have been a n u m b e r o f studies dealing with the axonal transport changes that a c c o m p a n y axonal regeneration after a single lesion o f goldfish optic axons13,1s, 36. These studies have shown that during axonal regeneration b o t h the slow and fast components o f axonal transport exhibit increases in the velocity and a m o u n t o f protein transport, beginning within 7-10 days after a x o t o m y and reaching peak values o f 2-5 times normal at 2-3 weeks. These changes are accompanied by an increase in amino

on effect has been seen in the goldfish optic axons, in which the initial delay before the axons traverse the lesion site is reduced by half, and the outgrowth rate is doubled 35. In b o t h the rat sciatic nerve and the goldfish optic nerve, the m i n i m u m effective condi-

*To whom all correspondence should be addressed at: Veterans Administration Medical Center, Medical Research Service, 10701 East Boulevard, Cleveland, OH 44106, U.S.A. 0006-8993/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

26 acid incorporation by the retinal ganglion cell bodiesa6, 42. Beginning even earlier, at 3 days after axotomy, there is an increase of RNA synthesis, reflected in the incidence and size of nucleoli detectable by light microscopy 42, and in the incorporation of labeled nucleosides 40. While it is clear from these studies that axonal outgrowth, normally beginning at 2~, days after axotomy a3,26,35, is associated with an increase in protein synthesis and axonal transport, it is not known how this program is modified during the accelerated axonal outgrowth that is caused by using a conditioning lesion. We have carried out this study in order to identify the axonal transport changes that are specifically linked to accelerated outgrowth, and have found that the amount of protein carried by both fast and slow transport is elevated above that seen during normal regeneration. Brief summaries of some of our findings have appeared previously 9,17,34. MATERIALS AND METHODS Goldfish (Carassius auratus), 9-11 cm in body length (13-15 cm from nose to tip of tail), were obtained in the fall and winter months from Ozark Fisheries (Stoutland, MO) and kept at 19-23 °C. Surgical procedures were carried out with the goldfish anesthetized by immersion in 3 ~ (w/v) urethane until respirations had ceased. For the initial lesion, a flap in the skull was opened, the intracranial adipose tissue and cerebral hemispheres were removed by aspiration, and the left optic tract was cut with iridectomy scissors at its bifurcation near the optic tectum (5.5 ± 0.3 mm from the choroid-sclera junction36). The cranial flap was replaced and fastened with a single 10 mm Michel clip. Some fish received no further lesions: in these fish, the nerve that had been transected would have been subjected to a 'testing lesion' alone, and the contralateral nerve was considered to be the 'sham-operated control', whereas, nerves in unoperated fish were considered to be 'normal'. Some fish were re-operated 2 weeks later: the cranial flap was reopened under urethane anesthesia and the optic chiasm was cut in the coronal plane with iridectomy scissors. Thus, in these animals both nerves would have been transected at the second operation (the 'testing' lesion), but the testing lesion in one nerve would have been preceded 2

weeks earlier by a conditioning axotomy ('conditioned nerve') and in the other nerve by the same operation without axotomy Csham-conditioned nerve'). All operated goldfish received bilateral intravitreal isotope injections at either 2 or 72 h before being killed at 2-36 days after the initial operation; unoperated goldfish were similarly injected. For the injection solution, L-2,3[aH]proline (45.7 Ci/mM; New England Nuclear, Boston, MA.) was dried to a residue and reconstituted with 0.15 M NaCI to 0.65-0.85 ~Ci//~l. Two /~1 of this solution were injected using a 10 #1 syringe (Hamilton Co., Reno, NV.). For tissue retrieval, goldfish were killed by decapitation; the intracranial contents and optic nerves were immediately exposed and the head was placed in either Bouin's fixative or a trichloroacetic acid (TCA) fixative (90 ml 5 0 ~ ethanol, 10 ml 3 7 ~ formaldehyde, 5 g TCA), both of which are effective in removing unincorporated amino acid, so that the remaining radioactivity can be assumed to be in protein 6. After 48-72 h, the tissues were decolorized in 70~o ethanol. The optic tecta were dried at 37 °C and weighed, then solubilized in Soluene-100 (Packard Instrument Co., Downer's Grove, IL.); radioactivity was measured in a liquid scintillation spectrometer and expressed as cpm/mg. The optic nerves (with contiguous optic tracts) were removed and embedded in paraffin. At each stage of dehydration, the nerves were individually straightened by applying gentle pressure with angled jeweler's forceps. The nerves were either cut in 15 /zm serial longitudinal sections for axonal staining by a reduced silver method 5 or serially cross-sectioned into 50 #m segments, consecutive groups of 5 segments being collected into vials for liquid scintillation spectrometry is. The velocity and amount of fast transport were determined from nerves taken at 1 and 2 h after injection of 3H-proline 38. The velocity was determined from the rate of advance of the wave-front of protein-bound radioactivity; the amount was determined by measuring the 2 h value for radioactivity at 1.5 mm behind the eye, a point far enough from the eye to elinfinate radioactivity leaking from the eye16 and far enough from the testing lesion to eliminate radioactivity from transported proteins accumulating at the lesion site a6. Slow transport determi-

27 nations were made on nerves taken at 3 days after injection of aH-proline. The transported material entering the nerve generates an exponentially rising wavefront of radioactivity that can be described by a straight line on a semi-logarithmic plot is. The slope of this line was calculated from the regression function of logl0 dpm on distance for the nerve segments lying 0.5-2.5 mm from the eye. The slope of the line is initially negative but becomes increasingly positive as the peak of the wave of slowly transported protein enters the nerve. Thus, the slope can be taken as an index of the transport velocity 14 since, at a fixed time after the injection, a faster velocity of transport would express itself as a less negative or even a positive slope. In the present study, the slope has been expressed as the angle of the calculated regression line to the horizontal plane (arctangent) - - a method similar in principle to that originally used by Grafstein and Murray is but employing different units of slope. The amount of labeled protein in the slow component was measured by determining the amount of protein-bound radioactivity in the initial 3.25 mm of optic nerve at 3 days after isotope injection. This limited length was taken in order to insure that the accumulation of labeled fast-transported protein at the lesion site was excluded 14. (The location of this accumulation in nerves taken 1 day after transection of the chiasma indicated that the mean distance between the choroid-sclela junction and the lesion was 4.5 ± 0.2 mm (n = 12).) The eyes were paraffin-embedded and sectioned at 8 /~m for light microscopic radioautography aa. In each retina, cell area measurements and silver grain counts were made at 400 × magnification in 7-10 retinal ganglion cells on each side of the optic nerve head; 3 4 eyes (55-70 cells) were examined at each time interval post-lesion, and eyes from normal goldfish were similarly examined. Only cell profiles located adjacent to the optic nerve layer and containing large nuclei surrounded by cytoplasm were selected for measurement. Because the amounts of incorporated radioactivity varied considerably within each retina, it was necessary to correct for this variation by relating the grain density in each retinal ganglion cell to overall grain density in that part of the retina. Accordingly, the grain density (grains per /am 2) for each retinal ganglion cell was expressed as

a multiple of the grain density measured in a 155 /~m2 area of the subjacent outer plexiform layer. The values for [3H]proline incorporation per cell were obtained by multiplying each mean value for the corrected grain density by the corresponding mean cell area 36. A second assay of the cellular response to axotomy consisted of determining the frequency with which nucleoli were seen in retinal ganglion cells42. Paraffin-embedded eyes were serially sectioned at 15/~m and stained with toluidine blue. Alternate sections were examined under oil immersion at 1000 ×, and the nucleoli were counted in every fifth retinal ganglion cell having a large nucleus and circumferential cytoplasm. For each retina, cells in one section were examined beginning at the temporal rim, in one beginning at the nasal rim, and in one beginning at the optic nerve head region. Nucleoli were counted in 25 cells of each section; thus, nucleoli were counted in 75 cells contained in a 90 #m strip of retina passing through the optic nerve head. RESULTS

Changes in retinal ganglion cell size and protein synthesis We have previously shown ae that during regeneration under normal conditions (i.e. tbllowing a testing lesion alone), the labeled amino acid incorporation into proteins in the axotomized retinal ganglion cells increases to reach a maximum of about 5 × normal (unoperated) at 15 days after the testing lesion (Fig. 1C, open circles). If the testing lesion had been preceded by a conditioning lesion 14 days earlier (Fig. 1, black squares), we found that at 1 day after the testing lesion the ganglion cell size and labeled amino acid incorporation were identical to the values observed at the corresponding time following a testing lesion alone, i.e., 15 days (Fig. 1). Thus, at 1 day after the testing lesion, the incorporation into the ganglion cells of the conditioned neive was unchanged from what would have been expected as a result of the conditioning lesion itself. Seven days later, i.e., at 8 days following the testing lesion in the conditioned nerves, the ganglion cell size (Fig. 1A) was only 15 ~o greater but the amino acid incorporation (Fig. 1C) was 60~o greater i.e. about 7.5 × normal. At 22 days following the testing lesion, the

28

160 -

/•

• testing lesion after conditioningtes,on D testing lesion after sham operation I-- "~ 0 testing lesion • - " ' " ' " " . r l - ) - ~ " ~ : ' O ~ n......... • ¢ sham operation .~./0 ~...~eroted

120-

CeLl Area

.-

% 80-

.

............

.

40-

3'0 80-

B

4'0

Corrected Grain Density

~_~ 60~E

~s~

3'5

..........

../s

4o-

,[j]._ _ - -

.o "6 o~ 20- oa ...............=~ ..........®............... ®.................................® 0

I000

5

I0

15

2'0

2'5

30

3'5

4'0

-

C

incorporation per Ceil

/a

800 J ...............

; ./-/..-.•

x ~ 600 ..,=

•/

"6 400 .c

7 ~ o

/

c

~, ~

//o

200 e

o

8

{

]

~

~ ......

;

@

..........

,'o

¢

...............

,'s

Fo

@

o

.................................

2's

~o

•

~5

4'0

TIME AFTER TESTING LESION (days)

Fig. 1 Time course of changes in retinal ganglion cells. A: cross-sectional cell area. B: density of silver grains (expressed as a multiple of background grain density) in autoradiograms at 2 h after injection of [3H]proline. C: [SH]proline incorporation into protein (cell area x grain density), expressed as percent of mean value for normal (unoperated) animals. The symbols indicate values for cells of nerves subjected to the following conditions: O, testing lesion alone; 03, sham-operation alone, i.e., contralateral to a testing lesion alone; I , conditioned, i.e., testing lesion preceded by a conditioning lesion; [3, sham-conditioned, i.e., testing lesion preceded by a contralateral conditioning lesion; and 0 , unoperated normal. Each point represents the mean value for a total of 55-70 cells from 3-4 eyes. Vertical lines in A and B denote S.E.M. ; absence of these indicates an S.E.M. that is less than the radius of the marking symbol. Values for unoperated normals, sham-operated controls, and testing-lesion alone series are reproduced from an earlier reporta% picture remained essentially unchanged in these neurons, whereas at the corresponding time after a testing lesion alone there was already some decline in both cell size and amino acid incorporation (Fig.

1). In the retinal ganglion cells o f sham-operated control nerves, i.e. nerves contralateral to those that had been subjected to a testing lesion alone, we have previously 36 found no significant changes in amino acid incorporation (Fig. 1, crossed circles). A t 1 day

after a testing lesion in sham-conditioned nerves (Fig. 1, open squares), the ganglion cell area was already increased by 30% but amino acid incorporation remained normal because o f a 20 % decrease in grain density. However, at both 8 days and 22 days after the testing lesion, the cell area was somewhat higher than the corresponding values seen after a testing lesion alone, and the labeled amino acid incorporation was significantly higher - - 5 × normal at 8 days, and 8 x normal at 22 days. In fact,

29

NUCLEOLI PER CELL • testing lesion 14 days after conditioning lesion [] testing lesion 14 days after sham operation

0 15 days after testing lesion D~ UJ 13_

1.5-

•

_J * (6l r-]~

O W UJ C)

(~

1.0 Z LL O n,uJ

133

15 days after sham operation

•-(4~"-- -.

(7)

_j (.9 Z .~ ._I .~

/

(5)

•

(4)

/ / /

0.5

/

=~ E:) z -Z Z

l~J

I-UJ n,-

(3)

0

I

I

control

I

I

I

I

I

I

0

5

I0

15

20

25

TIME AFTER TESTING LESION (days) Fig. 2. Incidence of nucleoli in retinal ganglion cells after various operations (symbols as in Fig. 1). The values at 15 days after a testing lesion or sham operation alone are considered to be 'control' values because they represent the conditions that existed at the time that the testing lesion was made in the conditioned and sham-conditioned nerves, respectively. The sham-operated control value also represents the unoperated normal value 42. Numbers in parentheses denote the number of retinas examined. Asterisks denote significant differences (P < 0.05 by Student's t-test) between each experimental point and its appropriate control, the control for the cells of the conditioned nerves being the value at 15 days after a testing lesion alone, and the control for the cells of the sham-conditioned nerves being the value at 15 days after a sham operation.

the incorporation in the ganglion cells of the shamconditioned nerves at 22 days was at least as great as that observed at the same time in the ganglion cells of the conditioned nerves.

trol at 1 day after the testing lesion, but at 8 days and 22 days after the testing lesion the frequency was similar to that for the conditioned nerves.

Changes in axonal transport Changes in nucleolar incidence The nucleolar incidence provides an indication of nucleolar size and hence of the status of ribosomal R N A synthesis in the retinal ganglion cells4L During regeneration following a testing lesion alone (Fig. 2, open circles) the nucleolar incidence rose from a sham-operated control value of about 0.3 to approximately 1.0, confirming an earlier study 42. In the retinal ganglion cells of the conditioned nerves the nucleolar incidence remained elevated at about 1.0 until at least 22 days after the testing lesion (Fig. 2, black squares). The cells of the sham-conditioned nerves showed a nucleolar frequency equal to con-

(a) Amount of protein conveyed by fast transport. The amount of labeled protein in the fast component of axonal transport was measured by determining the amount of protein-bound radioactivity seen at 2 h after isotope injection at a point 1.5 mm behind the eye36. After a testing lesion alone (Fig. 3, open circles), the amount reached a peak value of 5 × normal (unoperated) at 15 days. In the conditioned nerves (Fig. 3, black squares), the amount had increased to 8 × normal by 1 day after the testing lesion, a value 70 ~o greater than the value for the corresponding time (15 days) after a testing lesion alone, although the difference was not significant

30

PROLINE injected 2 hours before sacrifice W > IX: W Z

A

>, m

• testing lesion after conditioning lesion [] testing lesion after sham operation O testing lesion sham operation • unoperated

2,000-

"o roD

I.a a_ o E

z

t

1,500

',..\,.

E

--

tt3

z

--

w F--

a

o

E

a:: =~ n 0

I000

500

iz3

w

dO

"

w

:~

<

E

._1

o.

0

.

I

I

I

I

I

I

I

i

0

5

I0

15

20

25

30

35

TIME A F T E R

TESTING

i

40

LESION (days)

Fig. 3. Changes in amount of labeled protein carried by fast axonal transport after various operations (symbols as in Fig. 1). The mean amount of labeled protein was measured in optic nerves (n = 3-9) at 1.5 mm behind the eye 2 h after intravitreal injection of [SH]proline3n. Values for unoperated normals, sham-operated controls, and testing-lesion alone series are reproduced from an earlier reportz6. (0.05 < P < 0.1). At 8 and 22 days after a testing lesion in the conditioned nerves, the amounts of fast-transported protein were approximately the same as at the corresponding times after a testing lesion alone. Thus, in the conditioned nerves the increase in the amount of fast-transported protein evoked by the testing lesion was only transiently greater than that occurring after a testing lesion alone. In the sham-conditioned nerves (Fig. 3, open squares), the amount of fast-transported protein was unchanged from normal at 1 day after the testing lesion and then increased to 3 × normal at 8 days and 6 × normal at 22 days. This response is not significantly different from that seen following a testing lesion alone. (b) Amount of protein conveyed by slow transport. The characteristics of slow transport were determined from the distribution of radioactively-labeled protein in the nerve at 3 days after injection of [3H]proline. Typical radioactivity distributions over the initial 3.25 m m of the nerve are shown in Fig. 4. In Fig. 4A, the distribution at 15 days after a testing

lesion alone is compared with that seen in the contralateral sham-operated control nerve. Fig. 4B shows the distribution at 8 days after a testing lesion in a conditioned nerve, compared with that seen in the contralateral sham-conditioned nerve. The instantaneous amount of labeled protein in the slow component of transport was determined by measuring the total amount of radioactivity in the initial 3.25 m m of optic nerve (see Materials and Methods). At 15 days after a testing lesion alone (Fig. 5A), this amount was 7 × as much as in the sham-operated control nerves. In the conditioned nerves, the amount of slow transport at 1 and 8 days after the testing lesion was more than twice as great as the maximum after a testing lesion alone (P -< 0.05). In the sham-conditioned nerves, the amount had not increased significantly above control levels by 1 day after the testing lesion but had increased by 8 days to nearly the amount seen at the same time in the conditioned nerves. (c) Velocity of slow transport. Because of the short distance of 7-8 m m between the eye and the optic tectum zS, only a part of the wave of labeled

31 10 6

B

A 0 15 days after testing lesion

14 day Conditioning Interval 8 days after testing lesion n 8 days after testing lesion

E) 15 clays after

(sham conditioning lesion)

No Conditioning Lesion IO s

E tO OJ

o

sham operation 10 4 I

E

I

I °

I

10 3

I testing

t e st ing

% @ I0 =

lesion site

lesion site

I

I

I

I

|

I

I

I

l

1

/

0

I

2

3

4

5

0

I

2

I

I

I

5

4

5

DISTANCE FROM EYE (ram) Fig. 4. Typical examples of distribution of labeled protein carried by slow transport at 3 days after intra-vitreal injection of [SH]proline (A) in goldfish with testing lesion in one nerve ((3) and sham operation in the other (G), (B) in goldfish with testing lesions in previously conditioned nerve on one side ( i ) and sham-conditioned nerve on the other side ([]). Only the values for the first 3.25 mm behind the eye are shown.

slow component proteins is present in the optic nerve and tract at any one time ~s. An absolute value for the slow transport velocity is therefore difficult to obtain. However, a change in velocity can be detected at injection-sacrifice intervals as short as 3 days by examining the distribution of slow-transported proteins as they enter the optic nerve 2,leAS,Is. For a distance of a few mm from the eye, there is an exponential distribution of radioactivity: the regression function of log10 dpm on distance is approximately linear. The slope at 15 days after a testing lesion alone was less negative than that seen at 15 days after a sham operation (Fig. 5B), indicating an acceleration of slow transport during normal axonal regeneration - - as has been previously reported TM. For the conditioned nerves, the angle at 1 day after the testing lesion was approximately equal to that seen at 15 days after a testing lesion alone but the angle at 8 days showed a positive value, indica-

ting a significant further acceleration of slow transport by that time. (An angle of 0 ° would have denoted that the peak of the labeled protein wave had advanced along the nerve to the middle of the sampling zone, i.e., to a point 1.5 mm behind the eye. The presence of a positive angle in the conditioned nerves indicates that the peak had advanced even farther. Since 3 mm may be allowed for the intraocular length of an average optic axonlS, 51, that peak had advanced more than 4.5 mm in 3 days. This implies a transport rate in excess of 1.5 mm/day for the conditioned nerves, compared to values of 0.2-0.4 mm/day for normal nervesla,39.) In sham-conditioned nerves, the angle at 1 day after the testing lesion was approximately equal to that seen at the same time in the conditioned nerves. The value at 8 days, however, remained negative and was significantly less than in the conditioned nerves (P < 0.Ol), although still significantly greater than

32 hi

>

~ 2oo,ooo-

A

E

(D tO

F- c4

o_ r,5

O

-6

~

z Z w

Axonal Transport Amount (3 days after

160,000 ,

~"~"~'- I~151

120,000

.c_

F- _ a:..a a_ E c~ ~

/

_____RO,~OtqP
(4)

• O- (4) f--1

m 2 d

(5)

/ /

I-I testing lesion 14 days after sham operation 0

15° -

[~/(4)

~

B

15 days after testing lesion

(9 15 days after sham operation

' 5

control

tt3

testing lesion 14 days after conditioning lesion

/

p
g

*

•

/

40,O00-

la,J

~H-proline injection)

(41

, I0

Advance of Slow Transport Wave (3 days after 3H-proline injection)

i0 o 50 -

I(.5) /' ~]

,~ E

./p(O.Ol

00-

Z o Q- ....

a_ L5 o ..~

(41/" -5 o _

-I0°-

(4}

{

{

0 _15 • _ 1141

control

I

I

I

I

5 I0 TIME AFTER TESTING LESION (days)

o

Fig. 5. Slow axonal transport in nerves after various operations (symbols as in Figs. 1 and 2). A: amount of transported labeled protein in the initial 3.25 mm of nerve at 3 days after intra-vitreal injection of pH]proline. B: advance of slow transport wave by 3 days after intravitreal pH]proline injections, as determined from distribution of labeled protein in the 2 mm length of nerve lying 0.5 -- 2.5 mm behind the eye (cf. Fig. 4 and text). The value in the sham-operated control nerves is presumably not significantly different from that in normal unoperated nervesTM.

at 15 days after a testing lesion alone. Thus, during regeneration in the sham-conditioned nerves the velocity of slow transport was increased more than during normal regeneration but not as much as during regeneration in the conditioned nerves.

Distance of axonal outgrowth We have previously shown aS, that the outgrowth distance for regenerating optic axons following an optic nerve crush in goldfish could be determined either by direct histological examination to measure the length of the leading fascicle of axons, or by a method based on axonal transport. The site where labeled proteins accumulate proximal to the testing lesion represents the site of sprout formation, and the distance from that peak to the most distal 0.25 mm nerve segment having more that 10% of

peak radioactivity is approximately equal to the length of the longest fascicle detectable by light microscopyaS. When these methods were applied in the present study at 8 days after the testing lesion, the mean outgrowth distance was found to be 0.85 mm on the side of the conditioning lesion compared to 0.53 mm on the side of the sham operation, a statistically significant difference (Table I). We are not able to calculate the axonal outgrowth rates from these values, since we do not know the value for the initial delay, but it is clear that the outgrowth rates would be less than we found in our previous study of regeneration following optic nerve crush aS. For example, if we assume an initial delay of 5 days in the sham-conditioned nerves 1s,25, the outgrowth rate would be 0.2 mm/day, compared to a normal value of about 0.4 mm/day after optic nerve crush ~5.

33 TABLE I Axonal outgrowth distances at 8 days after bilateral testing lesions (optic chiasma cut) made at 14 days following a unilateral conditioning lesion (optic tract cut)

Each value is the mean 4- S.E.M. Method o f outgrowth measurement

Fast axonal transport (injection-sacrifice interval = 2 h) Slow axonal transport (injection-sacrifice interval = 3 days) Histological Combined methods

Outgrowth distance (mm) Sham-conditioned nerves

Conditioned nerves

0.50 4- 0.06 (n = 6)

0.79 4- 0.12 (n = 6)*

0.63 4- 0.07 ( n = 4 ) 0.40 4- 0.05 (n = 2)

1.06 4- 0.24 (n = 4) 0.67 4- 0.07 (n = 3)

0.53 4- 0.05 (n = 12)

0.85 4- 0.10 (n = 13)*

* P < 0.05 (Student's t-test)

The lower rate in the present study may be due to the greater surgical trauma of a craniotomy, the reaction of nerve cell bodies to the loss of axons by a cut as opposed to a crush 48, or the absence of a connective tissue bridge for axons to follow while crossing the injury zone. DISCUSSION

(a) Effects o f a single testing lesion o f the goldfish optic axons We have previously shown that, at 2-36 days after cutting the optic tract, changes in the amount of labeled fast-transported protein in the regenerating optic nerve proceeded in tandem with changes in the incorporation of labeled amino acid into the retinal ganglion cells: both had increased 2 x by 10 days post-lesion, reached a peak of 5 x normal at 15 days post-lesion, and by 36 days had decreased again to a level of 2 × normalae. In the present study we have shown that at 15 days post-lesion the instanteous amount of labeled slow-transported protein was 7 x that in contralateral sham-operated controls, and the slow transport velocity was faster than in the controls. These changes in protein synthesis and axonal transport are consistent with previously reported findings in the goldfish optic system3,1x, 18, 18,20,41,42,52.

(b) E.ffects o f a testing lesion preceded by a sham conditioning lesion At 1 day after a testing lesion in sham-conditioned

nerves, the ganglion cell area was increased 30 70 over normal and the velocity of slow axonal transport was elevated like that in conditioned nerves, but there were no changes in amino acid incorporation, nucleolar incidence, or in the amount of protein carried by either fast or slow transport. However, by 8 days after a testing lesion a number of changes had occurred: cell area had doubled, amino acid incorporation had increased 5 x , and the amount of fast-transported labeled protein had increased 3 x , compared to normal. The fast transport amount was in line with that seen after a testing lesion alone, but the other values were greater than the corresponding values after a testing lesion alone. Instantaneous slow transport amount and slow transport velocity had both increased to significantly greater values than those seen at 15 days after a testing lesion alone. At 22 days the cell area and incorporation values for the sham-conditioned nerves were significantly higher than the maximum values seen after a testing lesion alone. It is evident, therefore, that the sham operation had some effects that eventually influenced the retinal ganglion cells' response to axotomy. In the sham-conditioned cells at 1 day after the testing lesion, the low values for incorporation, incidence of nucleoli, and the amounts of fast and slow transport show that, in spite of the elevated velocity of slow transport, RNA and protein metabolism in these cells was not obviously perturbed at the time of the testing lesion: this confirms that they had not been inadvertently axotomized.

34 in a previous study, we showed that the neurons subjected to the sham operation alone showed a small increase in retinal ganglion cell area at 10 and 15 days post-lesion, without any change in labeled amino acid incorporation; a decrease in the amount of labeled fast-transported proteins at 2 and 10 days post-lesion; and a doubling of the velocity of fast transport at 10 days post-lesion. We have suggested 36 that under the conditions of our experiments these changes may be attributable to removal of the cerebral hemispheres at the time of the optic tract cut, resulting in axotomy of the second-order visual pathway neurons that project from thalamic nuclei (anterior, dorsal posterior) to the cerebral hemispheres7,46: since axotomy of motor neurons or postganglionic sympathetic neurons causes retraction of pre-synaptic axon terminals43, 47, the optic axons may similarly retract from their thalamic target neurons following axotomy of these neurons due to removal of the cerebral hemispheres. However, other explanations have not been ruled out, since some sham operation effects have also been observed even when the hemispheres have not been disturbed. For example, the retinal ganglion cells of sham-operated nerves show changes in the perikaryal routing of newly synthesized proteins 5z,53. The observations of Landreth and Agranoff 25 on the outgrowth of neurites from goldfish retinal explants in vitro show a slightly increased capacity for outgrowth in the retinas opposite to those whose nerves had been crushed 2-7 days earlier; also, ornithine decarboxylase activity is increased in the retinas on the side opposite to the nerve crushes and in nonneural tissues 22, indicating that a systemic mechanism is involved. (Recent observations that nerve damage may lead to the appearance of cilculating antibodies to nerve cell constituents 44 suggest one possible mechanism.) It is not yet clear whether the exaggerated response that a testing lesion elicits in the ganglion cells and axons of sham-conditioned nerves is associated with enhanced axonal outgrowth. That this response is insufficient to produce a full conditioning lesion effect is clear from the fact that these neurons showed slower axonal outgrowth than those that had received a bona fide conditioning lesion.

(c) Effects of a testing lesion preceded by a conditioning lesion In comparison to regeneration after a testing lesion alone, regeneration in conditioned nerves was characterized by (a) a higher level of amino acid incorporation into retinal ganglion cells throughout regeneration; (b) a transiently larger amount of fasttransported protein; (c) a larger instantaneous amount of slow-transported protein; and (d) a faster velocity of slow transport. These differences may be the basis for the enhanced regeneration resulting from the conditioning lesion. The increased amount of fast transport, which is evident only at 1 day following the testing lesion and not at 8 days or later, may be responsible for an accelerated initiation of sprouting, revealed as a reduction in the initial delay from 4.5 to 2.5 days 35. This would be consistent with our previous finding that regeneration is accelerated with conditioning intervals as short as 2 daysS, 9. However, we do not believe that the change in fast transport was of adequate duration to account for the maintained acceleration of outgrowth that was evident in our previous study of the conditioning lesion effect on goldfish optic axons (ref. 35, Figs. 1 and 5). It seems more likely that the change in outgrowth rate is related to the increase in slow transport, which lasts for at least 8 days after the testing lesion. Comparison of the effects of a testing lesion in the conditioned and sham-conditioned nerves reveals that the elevated amounts of fast and slow axonal transport seen at 1 day after the testing lesion in the conditioned nerves were not replicated in the shamconditioned nerves; also, at 8 days after the testing lesion, although the conditioned and sham-conditioned nerves were equal with respect to the amounts of fast and slow transport constituents they contained, the higher velocity of slow transport in the conditioned nerves would result in an increased amount of slow-transported material being delivered in unit time. Thus, there are differences in both fast and slow transport that would account for the faster regeneration in the conditioned nerves in relation to the sham-conditioned nerves as well as in relation to the nerves that had received a testing lesion alone. The idea of a linkage between slow transport and regeneration rate was first advanced by Weiss and

35 Hiscoe49 in their initial study of slow transport (which they called 'axoplasmic flow'), in which they remarked on the similarity between the rate of axoplasmic movement and the rate of axonal outgrowth. Subsequent studies have tended to bear out this relationship (reviewed by Grafstein and Forman16). It has been observed29,3° that a particularly close correspondence exists between the rate of outgrowth and the velocity of component SCb of slow transport, which comprises the axonal microtrabecular network that includes microfilaments4 and possibly the axoplasm 2s. In the goldfish, the peak of SCb has been found to advance at a velocity of 0.2-0.4 mm/day39, which is close to the normal rates of slow transport and axonal outgrowth previously reported in this systemS,iS,35. However, some discrepancy arises from the fact that the velocity of slow transport in the goldfish rises 2-3-fold during the course of regeneration is, although the rate of outgrowth apparently remains linear (ref. 35, Fig. 1 and Table IV). A discrepancy between the velocity of SCb and outgrowth rate may also be seen in a recent study of age-related changes in rat sensory neurons, which showed that in young animals the maximum velocity of SCb (as indicated by the transport of actin) may exceed the rate of axonal regeneration by several-fold 23. Thus, although our present findings do reiterate the importance of slow transport in regulating the outgrowth rate, it is unlikely that the velocity of SCb is the direct determinant of the rate of outgrowth. As we have previously suggested16A7, slow transport may control regeneration by providing to the axon tip a pool of eytoskeletal constituents that can be drawn upon for the growth of the axon. Conceivably, under various circumstances the velocity of one or another of the slow-transport constituents might become the limiting factor for outgrowth, if, for example, a particula~ equilibrium ratio among various constituents were necessary for outgrowth. Even so, there would not necessarily be a direct relationship between the transport velocity of

the outgrowth-limiting consUtuent and the axonal outgrowth rate if the rate of build-up of that constituent in the pool were a function not only of its transport velocity, but also of the instantaneous amount transported and the rate of metabolic turnover. The increased amounts of slow- and fast-transported protein at 1 day in conditioned nerves compared to that at 15 days in the 'testing lesion alone' group occurred without a corresponding increase in amino acid incorporation. This suggests that in the conditioned nerves larger fractions of newly-synthesized protein were shunted to the axons. Possibly, the resulting depletion of cell body materials may act as a stimulus for the subsequent increase in protein synthesis attested to by the increased amino acid incorporation seen at 8 and 22 days after the testing lesion in conditioned nerves. We have also found that at 1 day following a testing lesion alone the amount of axonally transported protein appearing in the optic nerve is increased without any concomitant increase in amino acid incorporation (Grafstein and Alpert, unpublished results). These findings lead us to speculate that the changes in protein synthesis that accompany regeneration in many neurons, including goldfish retinal ganglion cellsa6' may be initiated via a process that begins with regulation of the amounts of these proteins that are diverted into axonal transport. ACKNOWLEDGEMENTS This study was supported by USPHS Grant NS09015 to B.G. USPHS Grant NS-14967 B.G, and I.G.M. and a grant from Paralyzed Veterans of America to I.G.M.I.G.M. was also supported in part by the Medical Research Service of the U.S. Veterans Administration. We particularly thank Roberta Alpert for technical assistance, Marguarita Schmid for executing the illustrations, and Maureen McEntee for typing the manuscript.

REFERENCES 1 Agranoff, B. W., Field, P. and Gaze, R. M., Neurite outgrowth from explanted Xenopus retina: an effect of prior optic nerve section, Brain Research, 113 (1976) 225-234.

2 Alpert, R. M., Grafstein, B. and Edwards, D. L., Slow axonal transport in goldfish optic axons, Neurosci. Abstr., 6 (1980) 83. 3 Benowitz, L. I., Shashoua, V. and Yoon, M. G., Specific changes in rapidly transported proteins during

36 regeneration of the goldfish optic nerve, J. Neurosci., 1 (1981) 300--307. 4 Black, M. M. and Lasek, R. J., Slow components of axonal transport: two cytoskeletal networks, J. Cell BioL, 86 (1980) 616-623. 5 Bodian, D., A new method for staining nerve fibers and nerve endings in mounted paraffin sections, Anat. Rec., 65 (1936) 8%97. 6 Droz, B. and Warshawsky, H., Reliability of the radioautographic technique for the detection of newly synthesized protein, J. Histochem. Cytoehem., 11 (1963) 426-435. 7 Echteler, S. M. and Saidel, W. M., Forebrain connections in the goldfish support telencephalic homologies with land vertebrates, Science, 212 (1981) 638-685. 8 Edwards, D. L., Alpert, R. M. and Grafstein, B., Recovery of vision in regeneration of goldfish optic axons: enhancement of axonal outgrowth by a conditioning lesion, Exp. NeuroL, 72 (1981) 672-686. 9 Forman, D. S., McQuarrie, I. G., Grafstein, B. and Edwards, D. L., Effect of a conditioning lesion on axonal regeneration and recovery of function. In H. Flohr and W. Precht (Eds.), Lesion-induced Neuronal Plasticity in Sensorimotor Systems, Springer-Verlag, Berlin, 1981, pp. 103-113. 10 Forman, D. S., McQuarrie, I. G., Labore, F. W., Wood, D. K., Stone, L. S., Braddock, C. H. and Fuchs, D. A., Time course of the conditioning lesion effect on axonal regeneration, Brain Research, 182 (1980) 180185. 11 Giulian, D., Des Ruisseaux, H. and Cowburn, D., B;osynthesis and intra-axonal transport of proteins during neuronal regeneration J. bioL Chem., 255 (1980) 64946501. 12 Goldowitz, D. and Cotman, C. W., Axonal transport and axon sprouting in the adult rat dentate gyrus: an autoradiographic study, Neuroscience, 5 (1980) 21632174. 13 Grafstein, B., Role of slow axonal transport in nerve regeneration, Acta neuropath., Suppl. V (1971) 144-152. 14 Grafstein, B. and Alpert, R. M., Properties of slow axonal transport: studies in goldfish optic axons. In D. Weiss (Ed.), Axoplasmic Transport, Raven Press, New York, 1982, pp. 226-231. 15 Grafstein, B., Edwards, D. L. and Alpert, R. M., Axonal transport and neuronal activity. In H. G. Schweiger (Ed.) International Cell Biology 1980-1981, SpringerVerlag, Berlin, 1981, pp. 728-736. 16 Grafstein, B. and Forman, D. S., Intracellular transport in neurons, PhysioL Rev., 60 (1980) 1167-1283. 17 Grafstein, B. and McQuarrie, I. G., The role of the nerve cell body in axonal regeneration. In C. W. Cotman (Ed.) Neuronal Plasticity, Raven Press, New York, 1978, pp. 155-195. 18 Grafstein, B. and Murray, M., Transport of protein in goldfish optic nerve during regeneration, Exp. Neurol., 25 (1969) 494-508. 19 Griffin, J. W., Price, D. L., Drachman, D. B. and Morris, J., Incorporation of axonally transported glycoproteins into axolemma during nerve regeneration, J. CellBioL, 88 (1981) 205-214. 20 Heacock, A. M. and Agranoff, B. W., Enhanced labeling of a retinal protein during regeneration of optic nerve in goldfish, Proc. nat. Acad. Sci. U.S.A., 73

(1976) 828-832. 21 Hoffman, P. N. and Lasek, R. J., Axonal transport of the cytoskeleton in regenerating motor neurons: constancy and change, Brain Research, 202 (1980) 317-333. 22 Kohsaka, S., Schwartz, M. and Agranoff, B. W., Increased activity of ornithine decarboxylase in goldfish following optic nerve crush, Brain Research, 227 (1981) 391-401. 23 Komiya, Y., Growth, aging and regeneration of axon and slow axonal flow. In A. D. Kidman, J. K. Tomkins and R. A. Westerman (Eds.), New Approaches to Nerve and Muscle Disorders. Basic and Applied Contributions, Excerpta Medica, Amsterdam, 1981, pp. 173-182. 24 Landreth, G. E. and Agranoff, B. W., Explant culture of adult goldfish retina: effect of prior optic nerve crush, Brain Research, 118 (1976) 299-303. 25 Landreth, G. E. and Agranoff, B. W., Explant culture of adult goldfish retina: a model for the study of CNS regeneration, Brain Research, 161 (1979) 39-53. 26 Lanners, H. N. and Grafstein, B., Early stages of axonal regeneration in the goldfish optic tract: an electron microscopic study, J. Neurocytol., 9 (1980) 733-751. 27 Lanners, H. N. and Grafstein, B., Effect of a conditioning lesion on regeneration of goldfish optic axons: ultrastructural evidence of enhanced outgrowth and pinocytosis, Brain Research, 196 (1980) 547-553. 28 Lasek, R. J., Cytoskeletons and cell motility in the nervous system. In G. S. Siegel, R. W. Albers, B. W. Agranoff and R. Katzman (Eds.), Basic Neurochemistry, 3rd edn. Little Brown, Boston, 1981 pp. 403-412. 29 Lasek, R. J. and Hoffman, P. N., The neuronal cytoskeleton, axonal transport and axonal growth. In R. Goldman, T. Pollard and J. Rosenbaum (Eds.), Cell Motility, Book C, Mierotubules and Related Proteins, Cold Spring Harbor, 1976, pp. 1021-1051. 30 Lasek, R. J., McQuarrie, I. G. and Wujek, J. R., The central nervous system regeneration problem: neuron and environment. In A. Gorio, H. Millesi and S. Mingrino (Eds.), Posttraumatic Peripheral Nerve Regeneration. Experimental Basis and Clinical Implications, Raven Press, New York, 1981, pp. 59-70. 31 McQuarrie, I. G., The effect of a conditioning lesion on the regeneration of motor axons, Brain Research, 152 (1978) 597-602. 32 McQuarrie, I. G., Accelerated axonal sprouting after nerve transection, Brain Research, 167 (1979) 185-188. 33 McQuarrie, I. G. and Grafstein, B., Axon outgrowth enhanced by a previous nerve injury, Arch. NeuroL, 29 (1973) 53-55. 34 McQuarrie, 1. G. and Grafstein, B., Protein synthesis and fast axonal transport in regenerating goldfish retinal ganglion ceils: effect of a conditioning lesion, Neurosci. Abstr., 4 (1978) 533. 35 McQuarrie, I. G. and Grafstein, B., Effect of a conditioning lesion on optic nerve regeneration in goldfish, Brain Research, 216 (1981) 253-264. 36 McQuarrie, I. G. and Grafstein, B., Protein synthesis and fast axonal transport in regenerating goldfish retinal ganglion cells, Brain Research, 235 (1982) 213-223. 37 McQuarrie, I. G., Grafstein, B., Dreyfus, C. F. and Gershon, M. D., Regeneration of adrenergic axons in rat sciatic nerve: effect of a conditioning lesion, Brain Research, 141 (1978) 21-34. 38 McQuarrie, I. G., Grafstein, B. and Gershon, M. D.,

37

39

40 41 42 43 44

45

46

Axonal regeneration in the rat sciatic nerve: effect of a conditioning lesion and of dbcAMP, Brain Research, 132 (1977) 443-453. McQuarrie, I. G. and Lasek, R. J., Axonal transport of labeled neurofilament proteins in goldfish optic axons, J. Cell BioL, 91 (1981) 234a. Murray, M., 3H-uridine incorporation by regenerating retinal ganglion cells of goldfish, Exp. NeuroL, 39 (1973) 489-497. Murray, M. and Forman, D. S., Fine structural changes in goldfish retinal ganglion cells during axonal regeneration, Brain Research, 32 (1971) 287-298. Murray, M. and Grafstein, B., Changes in the morphology and amino acid incorporation of regenerating goldfish optic neurons, Exp. NeuroL, 23 (1969) 544-560. Purves, D., Functional and structural changes in mammalian sympathetic neurones following interruption of their axons, J. Physiol. (Lond.), 252 (1975) 429-463. Schwartz, M. and Eshhar, N., Production of antibodies against autoantigens induced by peripheral nerve injury, in mice. Proc. Int. Symp. on Nerv. Syst. Regen., Catania, Italy, in press. Sebille, A. and Bondoux-Jahan, M., Effects of electric stimulation and previous nerve injury on motor function recovery in rats, Brain Research, 193 (1980) 562-565. Sharma, S. C., The retinal projections in the goldfish:

47 48

49 50

51

52

53

an experimental study, Brain Research, 39 (1972) 213223. Sumner, B. E. H., A quantitative analysis of the response of presynaptic boutons to postsynaptic motor neuron axotomy, Exp. Neurol., 46 (1975) 605-615. Torvik, A. and Skj/~rten, F., Electron microscopic observations on nerve cell regeneration and degeneration after axon lesions. I. Changes in the nerve cell cytoplasm. Acta Neuropathol., 17 (1971) 248-264. Weiss, P. and Hiscoe, H. B., Experiments on the mechanism of nerve growth, J. Exp. Zool., 107 (1948) 315-395. Wells, M. R. and Bernstein, J. J., Amino acid incorporation into rat spinal cord and brain after simultaneous and interval sciatic nerve lesions, Brain Research, 139 (1978) 249-262. Whitnall, M. H., Currie, J. R. and Grafstein, B., Bidirectional axonal transport of glycoproteins in goldfish optic nerve, Exp. Neurol., 75 (1982) 191-207. Whitnall, M. H. and Grafstein, B., The relationship between extracellular amino acids and protein synthsis is altered during axonal regeneration, Brain Research, 220 (1981) 362-366. Whitnall, M. H. and Grafstein, B., Perikaryal routing of newly synthesized proteins in regenerating neurons: quantitative electron microscopic autoradiography, Brain Research, 239 (1982) 41-56.