8
BRAIN RESEARCH ELSEVIER Brain Research 726 (1996) 181 - 188 Research report In vitro pre-degenerated nerve autografts support CNS axonal regeneration Patrick Decherchi, Patrick Gauthier * Ddpartement de Physiologie et de Neurophysiologie, Laboratoire de Biologie des Rythmes et du Ddveloppement, URA CNRS 1832, Groupe Neurobiologie fonctionnelle des greffes nerveuses, Facult~ des Sciences et des Techniques de Saint-J~r3me, Case 332-351-352, Avenue Escadrille Normandie Niemen, 13397 Marseille Cedex, 20 France Accepted 5 March 1996 Abstract In the present study we compared, in adult rats, the axonal regeneration of central respiratory neurons within autologous fresh (f-; grafted immediately after removal) and pre-degenerated (pd-; grafted after being stored during 3 days in saline at + 8°C) peripheral nerve grafts (PNGs) implanted within the C2 cervical spinal cord. The proximal end of the left peroneal nerve was implanted in the site of projection of medullary respiratory neurons (ventro-lateral quadrant) and the distal part of each nerve graft was left unconnected (blind-ended graft). PNGs were examined 2 to 4 months after grafting. Central neurons regenerating axons within the PNGs were studied by recording spontaneous unit activity from small strands teased from the grafts. In control f-PNGs (n = 9), 248 filaments had spontaneous activities, 58 of these were respiratory-related, i.e. had discharge patterns identical to those of normal respiratory (inspiratory and expiratory) neurons. The presence of regenerated nerve fibers with spontaneous unitary impulse traffic (n = 216) was found in all pd-PNGs (n = 5). Thirty-four had respiratory patterns identical to those found within f-PNGs and corresponded to efferent activity. No statistically significant differences in axonal regrowth were found between f- and pd-PNGs. In conclusion, f- and pd-PNGs were equally capable of promoting axonal regeneration of central neurons. The neural components (Schwann cells and others) required for axonal regeneration of adult central neurons are still effective following 3 days of in vitro peripheral nerve degeneration without special storage conditions (oxygenation, medium inducing ATP synthesis). These results have clinical implications for nerve graft surgery when time is required for typing the tissues of both donor and recipient (post-mortem allografts) or transportation of graft material. Keywords: Axonal regeneration; Respiratory neuron; Peripheral nerve graft; Electrophysiology; Nerve grafting; Rat; Schwann cell; Spinal cord; Wallerian degeneration 1. Introduction Abortive axonal regeneration following central nervous system (CNS) injury in adult mammals [41] can be pre- vented if injured axons are exposed to the glial environ- ment of a peripheral nerve. Many types of mature CNS neurons can express regenerative propensity of their in- jured axons into peripheral nerve grafts (PNGs) [2,10,37,38,42] in which non-neuronal components (par- ticularly Schwann cells) constitute a physical and trophic support for the growth of neurites [8,17,25,35,46,49]. For several years, pre-degenerated PNGs (pd-PNGs) have been used in peripheral nerve system (PNS) reconstruction to bridge extensive injuries. However, the effectiveness of degenerated versus fresh nerve grafts (f-PNGs) remains controversial. Indeed, pd-PNGs may be as effective as * Corresponding author. Fax: (33) 91.28.83.33; e-mail: Neuronet@na- ture.u-3mrs.fr 0006-8993/96/$15.00 Published by Elsevier Science B.V. Pll S0006-8993(96)00331-9 f-PNGs [6,26,45] or more effective [16,41] in promoting more rapid regeneration of peripheral nerve fibers. Pd- PNGs may also reduce the initial time lapse before regen- erating nerve fibers enter the graft without affecting the regeneration rate [4,5,11,12,24,27] or enhance the regener- ation rate and the maturation of the fibers [51 ]. In the PNS, the controversy about the effectiveness of pre-degenerated grafts is further complicated by the possibility of contami- nation of the graft by Schwann cells (SCs) from the host nerve; SCs can migrate from the nerve stump to repopulate the nerve graft [3,19]. This putative contamination of the graft by SCs from the host has been recently avoided in studying axonal regeneration of hippocampal neurons of the adult rat within in situ pre-degenerated autologous PNGs implanted within the CNS [31,32]. To investigate the feasibility of using in vitro degener- ated nerves as grafts to induce axonal regeneration of CNS neurons, we monitored the spontaneous discharge patterns and responses to known physiological inputs of central

In vitro pre-degenerated nerve autografts support CNS axonal regeneration

Embed Size (px)

Citation preview

Page 1: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

BRAIN RESEARCH

E L S E V I E R Brain Research 726 (1996) 181 - 188

Research report

In vitro pre-degenerated nerve autografts support CNS axonal regeneration

Patrick Decherchi, Patrick Gauthier *

Ddpartement de Physiologie et de Neurophysiologie, Laboratoire de Biologie des Rythmes et du Ddveloppement, URA CNRS 1832, Groupe Neurobiologie fonctionnelle des greffes nerveuses, Facult~ des Sciences et des Techniques de Saint-J~r3me, Case 332-351-352, Avenue Escadrille Normandie Niemen,

13397 Marseille Cedex, 20 France

Accepted 5 March 1996

Abstract

In the present study we compared, in adult rats, the axonal regeneration of central respiratory neurons within autologous fresh (f-; grafted immediately after removal) and pre-degenerated (pd-; grafted after being stored during 3 days in saline at + 8°C) peripheral nerve grafts (PNGs) implanted within the C2 cervical spinal cord. The proximal end of the left peroneal nerve was implanted in the site of projection of medullary respiratory neurons (ventro-lateral quadrant) and the distal part of each nerve graft was left unconnected (blind-ended graft). PNGs were examined 2 to 4 months after grafting. Central neurons regenerating axons within the PNGs were studied by recording spontaneous unit activity from small strands teased from the grafts. In control f-PNGs (n = 9), 248 filaments had spontaneous activities, 58 of these were respiratory-related, i.e. had discharge patterns identical to those of normal respiratory (inspiratory and expiratory) neurons. The presence of regenerated nerve fibers with spontaneous unitary impulse traffic (n = 216) was found in all pd-PNGs (n = 5). Thirty-four had respiratory patterns identical to those found within f-PNGs and corresponded to efferent activity. No statistically significant differences in axonal regrowth were found between f- and pd-PNGs. In conclusion, f- and pd-PNGs were equally capable of promoting axonal regeneration of central neurons. The neural components (Schwann cells and others) required for axonal regeneration of adult central neurons are still effective following 3 days of in vitro peripheral nerve degeneration without special storage conditions (oxygenation, medium inducing ATP synthesis). These results have clinical implications for nerve graft surgery when time is required for typing the tissues of both donor and recipient (post-mortem allografts) or transportation of graft material.

Keywords: Axonal regeneration; Respiratory neuron; Peripheral nerve graft; Electrophysiology; Nerve grafting; Rat; Schwann cell; Spinal cord; Wallerian degeneration

1. Introduct ion

Abortive axonal regeneration following central nervous system (CNS) injury in adult mammals [41] can be pre- vented if injured axons are exposed to the glial environ- ment of a peripheral nerve. Many types of mature CNS neurons can express regenerative propensity of their in- ju red axons into per iphera l nerve grafts (PNGs) [2,10,37,38,42] in which non-neuronal components (par- t icularly Schwann cells) constitute a physical and trophic support for the growth of neurites [8,17,25,35,46,49]. For several years, pre-degenerated PNGs (pd-PNGs) have been used in peripheral nerve system (PNS) reconstruction to bridge extensive injuries. However, the effectiveness of degenerated versus fresh nerve grafts (f-PNGs) remains controversial. Indeed, pd-PNGs may be as effective as

* Corresponding author. Fax: (33) 91.28.83.33; e-mail: Neuronet@na- ture.u-3mrs.fr

0006-8993/96/$15.00 Published by Elsevier Science B.V. Pll S 0 0 0 6 - 8 9 9 3 ( 9 6 ) 0 0 3 3 1 - 9

f-PNGs [6,26,45] or more effective [16,41] in promoting more rapid regeneration of peripheral nerve fibers. Pd- PNGs may also reduce the initial time lapse before regen- erating nerve fibers enter the graft without affecting the regeneration rate [4,5,11,12,24,27] or enhance the regener- ation rate and the maturation of the fibers [51 ]. In the PNS, the controversy about the effectiveness of pre-degenerated grafts is further complicated by the possibil i ty of contami- nation of the graft by Schwann cells (SCs) from the host nerve; SCs can migrate from the nerve stump to repopulate the nerve graft [3,19]. This putative contamination of the graft by SCs from the host has been recently avoided in studying axonal regeneration of hippocampal neurons of the adult rat within in situ pre-degenerated autologous PNGs implanted within the CNS [31,32].

To investigate the feasibility of using in vitro degener- ated nerves as grafts to induce axonal regeneration of CNS neurons, we monitored the spontaneous discharge patterns and responses to known physiological inputs of central

Page 2: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

182 P. Decherchi, P. Gauthier/Brain Research 726 (1996) 181-188

respiratory neurons previously shown to regenerate axons in bl ind-ended (i.e. not connected to a target) PNGs in- serted into the brainstem [22,23,28,29] or the spinal cord [15]. Fresh and in vitro pre-degenerated PNGs were im- planted within the descending respiratory pathways of the C2 spinal cord and the discharge characteristics of central respiratory neurons regenerating axons were compared. Spontaneous activities recorded from both type of grafts had normal patterns of periodic efferent respiratory dis- charges originating from central respiratory neurons which retained their afferent connections.

the removed nerve was stored in saline at + 8°C and the animals returned to their quarters after the muscles and skin were sutured. Fol lowing 3 days of in vitro nerve degeneration, animals were reanesthetized (Nembutal ®, i.p. 48 m g / k g ) and the degenerated autologous nerve was then grafted as for the control group. These nerve grafts were called pre-degenerated grafts since the grafting was per- formed after a period of Wallerian degeneration resulting from removal of the nerve.

2.2. Electrophysiological examination

2. Mater ia ls and m ethods

2.1. Animals and surgical procedures

All experimental procedures were carried out on adult female (n = 14) Sprague -Dawley rats (body weight 2 6 0 - 300 g) deeply anesthetized with sodium pentobarbital (Nembutal ®, i.p. 60 m g / k g ) . Atropine (i.p. I m g / k g ) was administered to reduce secretions. All animals were housed in smooth bottomed plastic cages at + 2 2 _+ I°C with a 12 h l i gh t / da rk cycle. For the grafting experiments, auto- grafts were chosen specifically to eliminate possible im- mune responses. As previously described [15] a segment (3-4 cm in length) of the left peroneal branch of the sciatic nerve (pe roneus or f ibularis) was excised and 2 -2 .5 mm of the distal part was grafted to the left spinal cord in the region where medullary respiratory neurons project their axons [18]. In the control group (n = 9), nerve was grafted 15-30 min after removing. In the second group (n = 5),

After a post-grafting period ranging from 2 to 4 months, by which time respiratory neurons regenerate axons into medullary [22,23,28,29] and spinal [15] PNGs, the animals were reanesthetized (Nembutal x, i.p. 48 m g / k g ) . They were prepared for recording of the firing pattern of neu- rons sending regenerated axons into the graft as previously described [15]. The experimental design is illustrated schematically in Fig. 1. Briefly, the phrenic neurogram (Phr. N.) and the tracheal pressure (T.P.) were monitored. The unitary discharge pattern of neurons with regenerating axons (Graft Unit) in the PNGs was recorded using monopolar electrodes. Electrical activities were filtered (0 .1 -3 kHz for the phrenic, 0 .1 -10 kHz for graft unit activity), amplified and monitored on an oscil loscope and a chart recorder. Graft units were classified as either respira- tory (R) or non-respiratory (NR) on the basis of the relationship of their discharge to that of the phrenic nerve. R units which discharged in phase with phrenic nerve activity were inspiratory (I) and expiratory (E) units fired during the silent period between phrenic bursts (Fig. 1).

T.P.

NR graft unit

R graft unit (inspiratory)

R graft unit (expiratory)

Phr. N.

iN II , lilll llll llBlill _ . [11111111l INII$ l[fl~llllll[ IIIIllUN!Ulil lll~illlllllll[ \

~ I

Phrenic nerve

//~ /~ A m Tracheal Pressure + ~ 500., k, . ._. J )

C2

C3

(24

Medulla Oblongata

Cervical Spinal Cord

Fig. 1. On the right side, schematic representation of the experimental set-up and recording of spontaneous activity from a C2 spinal peripheral nerve graft in a paralyzed, ventilated rat. The unitary activity of neurons that have regenerated an axon into the nerve graft is recorded from teased graft filaments and compared with phrenic nerve activity. On the left side, from top to bottom: tracheal pressure (T.P) with upward deflection indicating lung inflation; examples of spontaneous phasic unitary discharges (Graft Unit) recorded within the graft. The non-respiratory unit (NR graft unit, here phasic) was non-related to the respiratory rhythm even when the artificial ventilation is stopped (arrows). The respiratory units (R graft units) discharged in phase (inspiratory) or in opposite phase (expiratory) with the phrenic nerve activity (Phr. N., bottom trace), even when the artificial ventilation is temporarily stopped (arrows), indicating (asterisks) that these units were efferent.

Page 3: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

P. Decherchi, P. Gauthier/Brain Research 726 (1996) 181-188 183

We took into account only efferent R units in which respiratory-related activity persisted when artificial ventila- tion was temporarily stopped in paralyzed (Flaxedil ®, i.v. 10 m g / k g ) animals. The effect of asphyxia on the activity of R graft units was tested by stopping artificial ventilation until peak integrated phrenic nerve activity increased by at least 50%.

2.3. Histological examination

At the end of each experiment, animals were killed with an overdose of pentobarbital. The spinal cord was removed and fixed in 10% formaline for at least 72 h. In order to verify the anatomical position of the grafts, frozen trans- verse sections (thickness 60 Ixm) were counterstained with Cresyl violet, dehydrated and examined by light mi- croscopy. Nerve segments were fixed by 3% glutaralde- hyde in phosphate buffer, stored 12 h in buffered sucrose (pH, 7.4), post-fixed by 2% osmium tetroxide, washed (water and buffered sucrose) and then dehydrated in a graded series of ethyl alcohols. They were then embedded in araldite epoxy resin mixture; semithin (1 I~m) and ultrathin (60 nm) sections were cut and stained with p-phenylenediamine and uranyle acetate (2% in ethanol), respectively. Graft organisation was observed and photo- graphed (Tmax Kodak, 100 ASA) under light (semithin) and electron (ultrathin) microscopy.

2.4. Statistics

Data related to recorded activities were expressed _+ S.E.M. Statistical tests (software graphPAD instat 1.14) were used to determine whether differences were signifi- cant ( P < 0.05) using Non-parametric: Unpaired (Mann- Whitney U two-sample test) or One-sample (Wilcoxon signed rank test), One-tailed or Two-tailed P-test. Non- parametric tests make no restrictive assumptions about the distribution of data. The unpaired test was used to compare the median of the values of the samples while the One- sample test was used to compare each median of the value of the samples with a specified constant (here 100%). The One-tailed test was used when the direction of the differ- ence of the samples was evident in each case and only determines the magnitude of the difference, i.e. whether it was statistically significant. The Two-tailed test was used when the direction of the difference of the samples was not evident.

3. Results

3.1. Graft implantation and organisation

Grafted animals survived without any apparent neuro- logical abnormality except for paralysis of the left hind leg due to the removal of the peroneal nerve. At autopsy, all

the grafts were found to be in overall continuity with the cervical spinal cord. Grafts were anatomically integrated onto the host CNS and cystic necrosis and cavitation of the cord were generally absent in the vicinity of the PNGs. The tip of each PNG reached the ventro-lateral tracts of the cervical spinal cord (Fig. 2A). The overall appearance of the pd-PNGs was similar to that of C2 spinal PNGs previously described [15]. No sign of necrosis of these grafts was seen. They were well revascularized but the vessels were often more dilated than in a f-PNGs. Nerve fibers (myelinated and unmyelinated) were distributed in clusters as is typically observed in regenerating peripheral nerve. At the ultrastructural level, debris resulting from degeneration were not seen, but typical signs of regenera- tion were observed: regenerating axons contained numer- ous neurotubules, neurofilaments, vesicular profiles and mitochondria (Fig. 2B).

3.2. Electrophysiological investigation

3.2.1. Fresh nerve grafts (f-PNGs) All nine spinal f-PNGs were thoroughly investigated

and contained regenerated axons with propagating sponta- neous action potentials. Of a total of 887 teased (T) filaments (98.5 + 12.4 per graft on average), 248 (26.7 _+ 2.8%) had spontaneous respiratory (R; n = 58) and non-re- spiratory (NR; n = 190) unitary activity during both artifi- cial ventilation and temporary arrest of the ventilator in paralyzed animals. The occurrence rate of R and NR units referenced to the total number of T filaments was respec- tively 6.6 + 1.2% (R/T) , and 20.1 + 2.5% (NR/T) (Fig. 3A, Fresh PNGs). Most (45/58) R units were inspiratory (I); the remainder were expiratory (E). All but one (a tonically active unit) of the inspiratory neurons and all the expiratory neurons had phasic discharge patterns. They were classified as 'early' (8I;2E), 'all ' (27I;2E), 'late' (7I;9E) and 'phase-spanning' (2I), as was done in previous studies of respiratory neurons in rat at the medullary [14,43] or the spinal [34] level. All of the 40 R units tested during asphyxia were strongly activated and presented an increased discharge frequency. Most (52/58) R units were unaffected by lung inflation applied during the expiratory phase; they therefore resemble R~ neurons which receive inputs solely from the respiratory central pattern generator [21]. Of the six R units affected by inflation, four (3I- 'late' and 1 I- 'al l ' ) were excited. The discharge that normally occurred during the last part of the phrenic burst could also be elicited by lung inflation applied during the E phase. The other units (2E-'late ') were inhibited during the same test. Their cell bodies therefore receive inputs both from the central respiratory pattern generator and from respira- tory afferents (probably pulmonary slowly adapting recep- tors). These neurons responded like central neurons in the medulla (the so-called R[3) [14,21,43], and within the spinal cord [13]. Spontaneous active NR units (190/248) were not simply R units with low levels of activity due to

Page 4: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

184 P. Decherchi, P. Gauthier / Brain Research 726 (1996) 181-188

insufficient respiratory drive, because asphyxia (which in- (n = 19) or tonically ( n = 120). Some tonic units had creased phrenic activity) did not affect their discharge discharge patterns synchronized with the cardiac rhythm. patterns. They discharged phasically (n = 51), sporadically For each graft, NR units always outnumbered R units; the

Page 5: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

P. Decherchi, P. Gauthier / Brain Research 726 (1996) 181-188 185

5 0 .

4 5 .

40 .

35-

3 0 -

25 .

2 0 .

15,

10.

5 .

0

A Occurence rate (%)

B %ofcontroi(Fresh)

250,

150,

100,

5 0

0 NR/T

i i

I~ ~ R/'I' R NR

~ 91 P

Fresh PNGs Pre-degenerated PNGs (n=9) (n=5)

Fig. 3. A: the diagrams shows the mean occurrence rate of R and NR units per graft for the fresh (f-) and pre-degenerated (pd-) nerve grafts (PNGs). This occurrence is referenced to the number of teased filaments (T). No statistically significant difference was found between f and pd-PNGs. B: the diagrams show the percentage of reinnervation of pd-PNGs referenced to that of fresh PNGs. The mean percentage of R and NR units in pd-PNGs is referenced respectively to the mean number of R and NR units in the f-PNGs. No statistically significant difference was found with regard to fresh PNGs.

mean number of NR units was significantly greater than the number of R units ( P = 0.0053, Unpaired One-tailed test). The origin(s) of these NR activities was not deter- mined but similar discharges have also been reported in PNGs implanted within the brainstem [22,23,28,29,38,42].

3.2.2. Pre-degenerated nerve grafts (pd-PNGs) In all grafts investigated (n = 5), we recorded unitary

spontaneous activity which continued to discharge during temporary arrest of the ventilator in paralyzed animals. On 479 teased (T) filaments (95.8:2 13.4 per graft on average), 216 (38.2 _+ 13.1%) had spontaneous R (n = 34) and NR (n = 182) unitary pattem of discharge. The occurrence rate of R and NR units was respectively 6 _+ 2.4% ( R / T ) and 32.1 --L-_ 10.8% ( N R / T ) (Fig. 3A, Pre-degenerated PNGs). All R units had phasic patterns of discharge and were mainly inspiratory (n = 27). Respiratory discharge patterns were 'early' (19I;1E), 'all ' (4I;4E), and ' late ' (4I;2E). Lung inflation applied during the E phase revealed that one I unit ( I - ' la te ' ) was excited and four E units (3 E- 'a l l ' and 1 E- ' la te ' ) were inhibited. As in the control group, all of the R units (n = 30) recorded during asphyxia were strongly recruited. The NR units (182 /216) had patterns of discharge similar to that found in the control group. These patterns were phasic (n = 33), sporadic (n = 37) or tonic (n = 112). As in the control group, for each graft, the number of NR units was always higher than the number of R units and consequently the mean number of NR units was significantly greater than those of R units ( P = 0.047, Unpaired One-tailed test). Comparisons between pd- and f- spinal PNGs (R, NR, R + NR, T, R / T , N R / T and R +

N R / T ) revealed no statistically significant differences. In the pd-PNGs group, the mean number of R, NR and R + NR units represented respectively 105.5 _ 48.9%, 172.4 + 67.9% and 156.8 + 63.1% of those found in the control group (Fig. 3B). Statistical analysis (One-sample tests) of these data revealed no significant differences from the control group.

4. D i s c u s s i o n

The present study provides the first evidence that nerves allowed to degenerate in vitro and then grafted into the CNS can induce axonal regeneration of functional central neurons. The degenerated/grafted nerves (pd-PNGs) were repopulated from central fibers similarly to that obtained using fresh nerve grafts (f-PNGs) implanted in the same area of the CNS at the level of the descending respiratory pathways in the C2 spinal cord.

4.1. CNS axonal regeneration within fo and pd-PNGs

Facilitation of neurite outgrowth from the hippocampus has been reported to occur within in situ pregenerated PNGs in the initial stage of regrowth at 1 month post-graft- ing whereas no effect could be seen at 2 and 4 months following grafting [31,32]. Unfortunatly, such an early beneficial effect of pd-PNGs could not be determined in the present study since our investigation focused exclu- sively on 2 to 4 months post-grafting. At this time period, pd-PNGs and f-PNGs likely had the same potential to

Fig. 2. A: histological examination of the spinal cord at the level of the graft site (pre-degenerated nerve graft). The graft (G) was found to be firmly and continuously attached to the dorsolateral part of the cervical spinal cord. The graft was anatomically integrated onto the host CNS and the tip reached the ventral tract of the cord. Sections were stained with Cresyl violet. Scale bar: 0.5 mm. B: histology of pre-degenerated nerve graft examined 2.5 months post-transplantation at the level of it distal part. Note the presence of regenerated fibers (myelinated, m and unmyelinated, um) and Schwann cell nuclear profiles (SCn). These features indicate that Schwann cells were also present within the degenerated nerve after cold storage and thus before grafting since the host was CNS structure. Scale bar: 500 nm.

Page 6: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

186 P. Decherchi, P. Gauthier/Brain Research 726 (1996) 181-188

induce outgrowth of central neurites. They presented (1) a similar number of teased-regenerated fibers (T), (2) a similar number of fibers exhibiting spontaneous unitary (R and NR) discharges, and (3) a similar incidence of R, NR, and R + NR units. In addition, in both f- and pd-PNGs, the number of NR units always exceeded the number of R units; consequently, the mean number of NR units was significantly greater than of R units. NR neurons could not be readily determined under our experimental conditions and we do not know if these discharge patterns were 'normal' . However they still generated spontaneous action potentials and some had discharge patterns (e.g. rhythmic bursting or synchronized with the cardiac rhythm) suggest- ing that their cell bodies had maintained at least part of their original function. R units were identified unambigu- ously as emanating from central efferent respiratory neu- rons because, in the absence of respiratory afferent feed- back (arrest of the artificial ventilation in paralyzed ani- mals), they continued to discharge in or out of phase with phrenic nerve activity which indicates central respiratory rhythm. Several lines of evidence indicate that R neurons retained their original function after axonal regeneration within f- and pd-PNGs: (1) they continued to fire sponta- neously during each respiratory cycle with discharge pat- terns similar to those of normal medullary [14,43] and spinal [34] respiratory neuron, (2) they demonstrated the same kind of responses to lung inflation and asphyxia as respiratory neurons in intact animals, indicating retention of inputs from peripheral and central sources. These re- sponses are reminiscent of those obtained using PNGs inserted into the medullary respiratory centers [22,23,28,29] or the spinal cord [15].

4.2. Schwann cell surt,iL, al during in citro nerve degenera- tion

Data from the present experiments indicate that Schwann cells (SCs) survived within nerves stored in vitro for 3 days at + 8°C. Firstly, anatomical data showed that PNGs contained regenerated axons surrounded by SCs. Because SCs were present within grafts, they were also likely present before grafting within the degenerated nerves after cold storage. Secondly, the similarity of graft reinnervation in f- and pd-PNGs suggests that, after in vitro nerve storage, there were no major changes in axonal growth promoting factors and SC properties. Indeed, living SCs in PNGs seem to be a prerequisite for the induction of axonal regeneration of central neurons because PNGs with freeze-killed SCs do not promote axonal growth of CNS neurons [47]. Such a survival of SCs after 3 days of in vitro nerve degeneration is consistent with recent work showing that SCs remain functional after as long as one week of in vitro cold storage of the nerves in a special nutritive medium (U.W Belzer's solution) [30]. One expla- nation of the SCs' survival during cold storage in a normal saline solution would be that, although SCs have a rapid

metabolism like pericysts or endothelial cells, they can tolerate ischemia for several hours even if stored in a medium that does not permit ATP synthesis [36].

4.3. Comparison between fresh, in situ and in uitro degen- erating implanted nerL'es

Within in situ predegenerated and freshly implanted nerves, axons and myelin begin to break up the first day after removal of the nerve and, at about the same time, non-neuronal nerve cells (e.g. Schwann cells, fibroblasts) proliferate [1,44,48,50]. Infiltrating macrophages, present in large numbers 3 to 5 days post-lesion, phagocyte debris of axons and the myelin sheath [48], secrete Interleukine-I (IL-I) implicated in NGF synthesis by SCs [20,25,33,40], product apolipoprotein-E which may aid axonal elongation [9] and may influence proliferation of SCs [7]. The peak of SC division also occurs 3-5 days after the axonal section [48] and the increase in NGF level, which reaches its maximum 3 days post-transection, is thought to be due mainly to an increase in expression by SCs of NGF and NGF receptors [25,49]. This is why we used nerves degen- erated in vitro for 3 days. In fact, the major difference between in vitro and in situ degenerating implanted nerves (predegenerated or fresh) is that in vitro nerve degenera- tion excluded the possibility of an early intervention of infiltrating macrophages during the storage phase and their consequent influence both on SC activity and on removal of debris. Unfortunately, the influence of macrophages on the interactions between early removal of debris and subse- quent reinnervation of the nerve graft still remains unclear. For some authors, premature removal of debris within in situ pre-degenerated peripheral nerves is described as facil- itating regeneration by reducing the initial stage of regen- eration [4,5,11,12,24,27] whereas for others the presence of debris is described as having little effect especially in the initial stage when advancing axons are so thin that they can easily pass through the debris [39]. The fact that f- and pd-PNGs have been found similarly reinnervated suggests that persistent debris on subsequent axonal regeneration have only limited effect and that a lack of infiltrating macrophages at the early stages of in vitro nerve degenera- tion is not essential for subsequent reinnervation of the grafted nerve by CNS axons.

4.4. Conclusion

Nerves stored in vitro for 3 days without oxygenation in a medium (saline) excluding ATP synthesis can be used as grafts to induce axonal regeneration of central neurons. Electrophysiological and morphological analyses indicated no differences in axonal regeneration between fresh and pre-degenerated peripheral nerve grafts, indicating that sufficient numbers of Schwann cells and other components required for axonal regeneration are present in degenerated nerves. These results have implications for the improve-

Page 7: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

P. Decherchi, P. Gauthier / Brain Research 726 (1996) 181-188 187

m e n t o f s taged per iphera l nerve recons t ruc t ion . A wide r

exper imenta l t ime-sca le wou ld al low more t ime for test ing

the h i s tocompat ib i l i ty o f donor and host , or t ranspor ta t ion

o f graft material b e t w e e n source and site o f use.

Acknowledgements

W e are grateful to Dr Steve Iscoe for crit ical c o m m e n t s

and sugges t ions on the manuscr ip t . Exper t technica l assis-

tance was p rov ided by J. Pio, A .M. Lajard, and M. Man-

nevil le . W e also thank Pr J.F. P611issier and E. Pasqua le

(Serv ice d ' A n a t o m i e Pa tho log ique et de Neuropa tho log ie

de la Facult6 de M r d e c i n e de Marse i l le ) for their help in

e lec t ron mic roscopy . This work was suppor ted by C N R S

(Cent re Nat ional de la R e c h e r c h e Scient i f ique) , D R E T

(Direc t ion des R e c h e r c h e s et E tudes Techn iques , grant

9 2 - 0 5 3 ) , I R M E (Inst i tut pour la R e c h e r c h e sur la Moe l l e

Epini~re) and by the Univers i ty o f Aix -Marse i l l e III.

References

[1] Abercrombie, M. and Johnson, M.L., Quantitative histology of Wallerian degeneration. I. Nuclear population in rabbit sciatic nerve, J. Anat., 80 (1946) 37-40.

[2] Aguayo, A.J., Axonal regeneration from injured neurons in adult mammalian nervous system. In C.W. Cotman (Ed.), Synaptic Plas- ticity, Guilford, New York, 1985, pp. 457-484.

[3] Anderson, P.N., Nadim, W. and Turmain, M., Schwann cell migra- tion through freeze killed peripheral nerve grafts without accompa- nying axons, Acta Neuropathol., 82 (1991) 193-199.

[4] B~ihr, M., Eschweiler, G.W. and Wollburg, H., Precrushed sciatic nerve grafts enhance the survival and axonal regrowth of retinal ganglion cells in adult rats, Exp. Neurol., 116 (1992) 13-22.

[5] Bedi, K.S., Winter, J., Berry, M. and Cohen, J., Adult rat dorsal root ganglion neurons extend neurites on pre-degenerated but not normal peripheral nerves in vitro, Eur. J. Neurosci., 4 (1992) 193-200.

[6] Bentley, F.G. and Hill, M., Nerve grafting, Br. J. Surg., 24 (1936) 368-387.

[7] Beuche, W. and Friede, R.L., The role of non-resident cells in Wallerian degeneration, J. Neurocytol., 13 (1984) 767-796.

[8] Blexrud, M.D., Lee, D.A., Windebank, A.J. and Brunden, K.R., Kinetics of production of a novel growth factor after peripheral nerve injury, Z Neurol. Sci., 98 (1990) 287-299.

[9] Boyles, J.K., Zoellner, C.D., Anderson, L.J., Kosis, L.M., Pitas, R.E., Weisgraber, K.H., Hui, D.Y., Mahley, R.W., Gebicke-Harter, P.J. and Ignatius, M.J., A role for Apolipoprotein A-1 and low density lipoprotein receptors in cholesterol transport during regener- ation and remyelination in the rat sciatic nerve, J. Clin. Invest., 83 (1989) 1015-1031.

[10] Bray, G.M., Vidal-Sanz, M. and Aguayo, A.J., Regeneration of axons from the central nervous system of adult rats, In F.J. Seil, E. Herbert and B.M. Carlson (Eds.), Progress in Brain Research, Vol. 71, Elsevier Science Publishers B.V., Biomedical Division, 1987, pp. 373-379.

[11] Danielsen, N., Kerns, J.M., Holmquist, B., Zhao, Q., Lundborg, G. and Kanje, M., Pre-degenerated nerve grafts enhance regeneration by shortening the initial delay period, Brain Res., 666 (1994) 250-254.

[12] Danielsen, N., Kerns, J.M., Holmquist, B., Zhao, Q., Lundborg, G.

and Kanje, M., Predegeneration enhances regeneration into acellular nerve grafts, Brain Res., 681 (1995) 105-108.

[13] Dawkins, M.A., Foreman, R.D. and Farber, J.P., Short latency excitation of upper cervical respiratory neurons by vagal stimulation in the rat, Brain Res., 594 (1992) 319-322.

[14] De Castro, D., Lipski, J. and Kanjhan, R., Electrophysiological study of dorsal respiratory neurons in the medulla oblongata of the rat, Brain Res., 639 (1994) 49-56.

[15] Decherchi, P., Lammari-Barreault, N. and Gauthier, P., Regenera- tion of respiratory pathways within spinal peripheral nerve grafts, Exp. NeuroL, 137 (1996) 1-14.

[16] Duel, A.R., History and development of the surgical treatment of facial palsy, Surg. Gynecol. Obstet., 30 (1933) 382-390.

[17] Edgar, D., Timpl, R. and Thoenen, H., The heparin binding domain of laminin is responsible for its effects on neurite outgrowth and neuronal survival, EMBO J., 3 (1984) 1463-1468.

[18] Ellenberger, H.H. and Feldman, J.L., Monosynaptic transmission of respiratory drive to phrenic motoneurons from brainstem bulbospinal neurons in rats, J. Comp. Neurol., 269 (1988) 47-57.

[19] Fawcett, J.W. and Keyne, R.J., Axonal growth and Schwann cells migration into basal lamina grafts in the rat sciatic nerve, Abstr. Soc. Neurosci., 14 (1988) 164.

[20] Fawcett, J.W. and Keynes, R.J., Peripheral nerve regeneration, Annu. Rev. Neurosci., 13 (1990) 43-60.

[21] Feldman, J.L., Neurophysiology of breathing in mammals. In F.E. Bloom (Ed.), Intrinsic Regulatory System of the Brain, handbook of Physiology, Sect. 1, The Nervous System, Vol IV, American Physio- logical Society, Bethesda, 1986, pp 463-524.

[22] Gauthier, P. and Lammari-Barreault, N., Central respiratory neurons of the adult rat regrow axons preferentially into peripheral nerve autografts implanted within ventral rather than within dorsal parts of the medulla oblongata, Neurosci. Lett., 137 (1992) 33-36.

[23] Gauthier, P. and Rasminsky, M., Activity of medullary respiratory neurons regenerating axons into peripheral nerve grafts in the adult rat, Brain Res., 438 (1988) 225-236.

[24] Gordon, L., Buncke, H., Jewett, D.L., Muldowney, B. and Buncke, G., Predegenerated nerve autografts as compared with fresh nerve autografts in freshly cut and pre-cut motor nerve defects in the rat, J. Hand Surg., 4 (1979) 42-47.

[25] Heumann, R., Korsching, S., Brandtlow, C. and Thoenen, H., Change in nerve growth factor synthesis in non-neuronal cells in response to sciatic nerve transection, ,L Cell. Biol., 104 (1987) 1623-1631.

[26] Huber, G.C., Repair of peripheral nerve injuries, Surg. Gynecol. Obstet., 30 (1920) 464-471.

[27] Kerns, J.M., Danielsen, N., Holmquist, B., Kanje, M. and Lundborg, G., The influence of predegeneration on regeneration through pe- ripheral nerve grafts in the rat, Exp. Neurol., 122 (1993) 28-36.

[28] Lammari-Barreault, N., Rega, P. and Gauthier, P., Axonal regenera- tion from central respiratory neurons of the adult rat into peripheral nerve autografts: effect of graft location within the medulla, Neu- rosci. Lett., 125 (1991) 121-124.

[29] Lammafi-barreault, N., Rega, P. and Gauthier, P., Central respira- tory neuronal activity after axonal regeneration within blind-ended peripheral nerve grafts: time course of recovery and loss of func- tional neurons, Exp. Brain. Res., 98 (1994) 238-244.

[30] Levi, A.D.O., Evans, P.J., MacKinnon, S.E. and Bunge, R.P., Cold storage of peripheral nerves: An in vitro assay of cell viability and function, Glia, 10 (1994) 121-131.

[31] Lewin-Kowalik, J., Sieron, A.L., Krause, M. and Kwiek, S., Prede- generated peripheral nerve grafts facilitate neurite outgrowth from the hippocampus, Brain Res. Bull., 25 (1990) 669-673.

[32] Lewin-Kowalik, J., Sieron, A.L., Krause, M., Barski, J.J. and G6rka, D., Time-dependent regenerative influence of predegenerated nerve grafts on hippocampus, Brain Res. Bull., 29 (1992) 831-835.

[33] Lindholm, D., Heumann, R., Meyer, M. and Thoenen, H., Inter- leukin-1 regulates synthesis of nerve growth factor in non-neuronal cells of the rat sciatic nerve, Nature, 330 (1987) 659-659.

Page 8: In vitro pre-degenerated nerve autografts support CNS axonal regeneration

188 P. Decherchi, P. Gauthier/Brain Research 726 (1996) 181-188

[34] Lipski, J., Duffin, J., Kruszewska, B. and Zhang, X., Upper cervical inspiratory neurons in the rat: an electrophysiological and morpho- logical study, Exp. Brain Res., 95 (1993) 477 487.

[35] Mantborpe, M., Skaper, S.D., Williams, L.R. and Varon, S., Purifi- cation of adult rat sciatic nerve ciliary neurotrophic factor, Brain Res., 367 (1986) 282-286.

[36] Meryman, H.T., Tissue and organ banking: an inevitability, In H.T. Meryman (Eds.), Progress in Clinical and Biological Research, Transplantation: Approaches to Graft Rejection, Vol. 224, Alan R. Liss, Inc, New York, 1986, pp. 337 343.

[37] Morrow, D.R., Campbell, G., Lieberman, A.R. and Anderson, P.N., Differential regenerative growth of CNS axons into tibial and per- oneal nerve grafts in the thalamus of adult rats, Exp. Neurol., 120 (1993) 60-69.

[38] Munz, M., Rasminsky, M., Aguayo, A.J., Vidal-Sanz, M. and Devor, M., Functional activity of rat brainstem neurons regenerating axons along peripheral nerve grafts, Brain Res., 340 (1985) 115-125.

[39] Ochi, M., Wakasa, M., Ikuta, Y. and Kwong, W.H., Nerve regenera- tion in predegenerated basal lamina graft: The effect of duration of predegeneration on axonal extension, Exp. Neurol., 128 (1994) 216- 225.

[40] Perry, V.H. and Brown, M.C., Macrophages and nerve regeneration, Cur. Opin. Neurobiol., 2 (1992) 679-682.

[4t] Ramon y Cajal, S., Degeneration and Regeneration in the Neruous System, Vols. 1 and 2, Hafner, New York, 1928.

[42] Rasminsky, M., Aguayo, A.J., Munz, M. and Vidal-Sanz, M., Electrical activity in axons regenerating along peripheral nerve grafts inserted into the rat brainstem and sensory cortex, In A. Bl~rklund and U. Stenevi (Eds.), Neural Grafting in the Mammalian CNS, Elsevier, Amsterdam, 1985, pp. 421-429.

[43] Saether, K., Hilaire, G. and Monteau, R., Dorsal and ventral respira- tory groups of neurons in the medulla of the rat, Brain Res., 419 (1987) 87-96.

[44] Salonen, V., Aho, H., R~Sytt~i, M. and Peltonen, J., Quantitation of Schwann cells and endoneurial fibroblast-like cells after experimen- tal nerve trauma, Acre Neuropathol., (Berlin), 75 (1988) 331-336.

[45] Sanders, F.K. and Young, J.Z., The degeneration and reinnervation of grafted nerves, J. Anat., 76 (1942) 143-166.

[46] Sendtner, M., St~Sckli, K.A. and Thoene, H., Synthesis and localiza- tion of ciliary neurotrophic factor in the sciatic nerve of the adult rat after lesion and during regeneration, J. Cell Biol., 118 (1992) 139-148.

[47] Smith, G.V. and Stevenson, J.A., Peripheral nerve grafts lacking viable schwann cells fail to support central nervous system axonal regeneration, Exp. Brain. Res., 69 (1988) 299-306.

[48] Sunderland, S., NerL, es Injuries and their Repair, Churchill Living- stone, Edinburgh. 1991.

[49] Taniuchi, M., Clark, H.B., Schweitzer, J.B. and Johnson, E.M. Jr., Expression of nerve growth factor receptors by Schwann cells of axotomized peripheral nerves: Ultrastructural location, suppression by axonal contact, and binding properties, J. Neurosci., 8 (1988) 664-681.

[50] Thomas, P.K., Nerve injury. In R. Bellairs and E.G. Gray (Eds.), Essays on the Nervous System, A Festschrift for Professor J.Z. Young, Clarendon Press, Oxford. 1974, pp. 44-70.

[51] Zhao, Q. and Kern, J.M., Effects of predegeneration on nerve regeneration through silicone Y-chambers, Brain Res., 633 (1994) 97-104.