- Email: [email protected]
An immunocytochemical analysis of growing axons in a silicone chamber prefilled with artificial sponge matrix
Acta histochem. (Jena) 98, 123-130 (1996) Gustav Fischer Verlag Jena . Stuttgart . New York
Atla hislork.ita
An immunocytochemical analysis of growing axons in a silicone chamber prefilled with artificial sponge matrix Petr Dubovy' r' and J. Bednarova' I Department of Anatomy, School of Medicine, Masaryk University Brno, Komenskeho nam, 2, CZ-66243 Brno, Czech Republic and 2 Department of Neuroscience, Division of Neuroanatomy and Neuronal Plasticity, Karolinska Institute, S-17177 Stockholm, Sweden
Accepted 14 December 1995
Summary We have used antibodies against growth associated protein (GAP-43), phosphorylated neurofilament protein of 200 kDa molecular weight (RT-97) and substance P (SP) to analyze regrowing axons and their features in a silicone chamber filled with resorbable sponge matrix within the first two weeks after sciatic nerve transection in the rat. Growing axons identified with the GAP-43 antibody extended over a distance of about 7 mm from the proximal stump at 7 days and grew over a 10 mm gap within 14 days. This is a markedly longer distance than in the case of the standard chamber model without artificial sponge matrix. The regrowing axons were labelled with RT-97 already on the 7th day up to a distance of 5 mm and they made up about 750/0 of all axons in the first segments. The number of RT-97-positive axons did not increase significantly over the next 7 days, although they could be identified over a longer distance. Some of the growing axons expressed SP-like immunoreactivity (LI) 14 days, but not 7 days after chamber application and constituted about 30% of all growing axons in the first segment. The SPLI fibres also appeared to grow from the distal stump since they were found in larger numbers in the distal segments than in central ones. Those fibres accompanying blood vessels are probably sympathetic. Our findings demonstrate that axons are able to bridge a 10 mm gap within 14 days under appropriate substrate conditions, which are provided by the resorbable fibrin sponge.
Key words: peripheral nerve neurofilaments - substance P
transection -
silicone chamber -
GAP-43 -
Introduction The idea of attempting to create a conduit for regenerating axons across a gap between nerve stumps started in the 1940s (Weiss, 1944). Since then, one of the most useful designs of such a conduit was introduced by Lundborg and co-workers for the study of Correspondence to: P. Dubovy
124
P. Dubovy and J. Bednarova
cellular and molecular mechanisms of peripheral nerve regeneration (Lundborg et al., 1982a, b). In this experimental model, the proximal and distal stumps of the transected sciatic nerve are sutured into the ends of a silicone tube which form a chamber with a structurally organized preformed space where the nerve microenvironment is protected from external influences. These features make chambers from unpermeable material suitable for analyzing cellular events and molecular processes during nerve regeneration (Williams et al., 1983; Dubovy and Svizenska, 1992, 1994; Zhao et al., 1993). During the first week after implantation, the standard chamber is filled with nerve stump-derived fluid, in which a longitudinally oriented fibrin matrix is formed. The fibrin matrix formation is critical for the migration of fibroblasts, Schwann and endothelial cells as well as for the regrowth ofaxons through the chamber (Williams et al., 1985). A similarly suitable matrix for the enhanced migration of these cells and the subsequent regrowth ofaxons in the chamber can be provided by exogenous fibrin precursors or an artificial fibrin sponge (Williams et al., 1987; Dubovy and Svizenska, 1992). The spatial-temporal progress of nerve regeneration across a 10 mm gap inside the silicone chamber has been examined previously by means of classic light and electron microscopy (Williams et al., 1983, 1988). However, little is known about the immunocytochemical characteristics of regrowing axons in this chamber model. In the present study we have characterized immunocytochemically the growth of axons and their features in a silicone chamber model within the first two weeks after sciatic nerve transection by means of antibodies against growth associated protein (GAP-43), phosphorylated neurofilament protein (RT-97) and substance P. Material and methods Ten female Sprague-Dawley rats (200 - 250 g) were used for the experiments. All surgical procedures were performed under deep anaesthesia by intramuscular injection of a mixture (0.2 mill 00 g) containing xylazinum (4 mg/ml) and ketaminum (40 mg/ml). The right sciatic nerve was exposed and mobilized, stripped of epineurium in a 4 mm segment, and transected at mid-thigh level proximal to the tibial and peroneal bifurcation. A 2 mm nerve segment was removed and the proximal and distal stumps were introduced into each end of a 14 mm long silicone tube. The stumps were fixed with single transepineurial 10-0 sutures to the chamber wall, leaving a 10 mm distance between them. The chambers were prepared from silicone tubes with an internal diameter of 1.7 mm and an outer diameter of 3 mm. They were sterilized, filled with artificial fibrin sponge tissue (Gelaspon, Ankerwerk, FRG) and soaked with sterilized phosphate-buffered saline (PBS). After chamber implantation, the muscles and overlying skin were closed, and the animals were allowed to recover from the anaesthesia and to survive for a period of 7 (5 animals) and 14 days (5 animals). All animals were fed ad libitum and housed on soft material to minimize the possibility of trauma to their denervated hind limbs. At the time of sacrifice the animals were deeply anaesthetized with chloral hydrate and perfused via the ascending aorta with warm (37°C) phosphate-buffered saline (PBS) at pH 7.4. This was followed by cold (4°C) 4070 paraformaldehyde (w/v) in PBS (pH 7.4) containing 0.4% picric acid (w/v) (Zamboni and de Martino, 1967). The content of each silicone tube was removed from the chamber, stored at 4°C overnight in Zamboni's fixative solution, and then rinsed in PBS containing 10% sucrose. The nerve outgrowths were divided in 10 proximo-distal segments, each measuring about 1 mm. Cryostat sections (10 urn thick) were cut transversely through each segment and processed for indirect immunofluorescence. Twenty subsequent sections from each segment were incubated simultaneously with mouse monoclonal antibody to neurofilament protein (RT-97, dilution 1: 250, generously supplied by Dr. J. Wood, London, U. K.; Wood and Anderton, 1981) and rabbit polyclonal antibody against GAP-43 (dilution 1: I 000, a generous gift of Prof. W. H. Gispen, Utrecht, The Netherlands). Another twenty subsequent sections from each segment were incubated with polyclonal antibody raised against substance P (dilution 1: 1000, generously provided by Dr. P. Petrusz, North Carolina, U.S.A.).
Growing axons in silicone chambers
125
Goat anti-mouse and goat anti-rabbit secondary antibodies conjugated with fluorescein isothiocyanate (FITC) or rhodamine (TRITe) (dilution 1: 3D, Boehringer Mannheim, Indianapolis, U.S.A.) were used for localization of the primary immune complex. The sections were coverslipped in glycerol containing 0.1 % p-phenylenediamine to reduce fading (Johnson and Nogueira Araujo, 1981) and viewed in a Nikon fluorescence microscope using appropriate filters for FITC or TRITC epifluorescence. Immunohistochemical controls were performed by incubation without the primary antibodies or with appropriate preimmune sera and yielded negative data. The total number of GAP-43-like immunoreactive (GAP-43-LI), substance-P-like immunoreactive (SP-LI), and RT-97 positive immunolabelled fibers was counted in the entire cross-sectioned area of 10 sections using an image analysis system (Leitz-ASBA) at the same magnification. Average values and standard deviations were calculated for individual segments.
Results All chambers contained a delicate bridge of nerve tissue connecting the proximal and distal stumps. Immunocytochemical staining of transversal sections revealed that regenerating axons were closely packed in the central area of the nerve tissue where labelled axons appeared as spots. In addition to these spots some enlargements probably corresponding to growth cones were observed in sections stained for GAP-43. Regenerating axons immunolabelled with the GAP-43 antibody were found in the 7 proximal and in all 10 segments after 7 and 14 days, respectively, implying a distance of axonal elongation of about 7 and 10 mm from the proximal stump. Seven days after chamber application, the RT-97 positive axons were present in 5 proximal segments while following 14 days the RT-97 positive axons were observed in up to seven proximal segments (Fig. 1). Some of the regenerating axons identified by GAP-43 were also immunoreactive for RT-97 (cf. Figs. 2A, B and Figs. 2e, D).
•
200
= 0 1;1)
~
<
7Day RT-97
D
GAP-43
8
9
1 0
eO
J~
..c S Z
=
100
50
o
2
3
4
5
6
7
10
egment Fig. 1. Number of GAP-43 and RT-97 immunoreactive axons seven days after chamber application.
126
P. Dubovy and J. Bednarova
Figs. 2 A, B. A transversal section trough the sixth segment of nerve outgrowth immunostained simultaneously for GAP-43 and RT-97. Axon profiles exhibiting immunopositivity for GAP-43 (Fig. 1A) are not immunolabelled by RT-97 (Fig. 1B). The segment corresponds to a distance of about 6 mm from the end of the proximal stump, 7 days (0) after chamber application. Scale bar: 300 um. Figs. 2 C, O. Part of a transversal section through the second segment of nerve outgrowth (about 2 mm from proximal stump) 14days (0) after chamber implantation. The section was immunostained simultaneously for GAP-43 (Fig. 1 C) and RT-97 (Fig. 10). Some axon profiles display immunolabelling for both GAP-43 and RT-97 (arrows). Scale bar: 300 urn.
Growing axons in silicone chambers
127
14Da 600 •
D
o 400
RT-97 GAP-43
I- I
300 200 100
o
2
3
8
4
9
10
egment Fig. 3. Number of GAP-43 and RT-97 expressing axons 14 days after chamber application.
When comparing the number ofaxons 7 and 14 days after chamber application, a marked increase in the number of GAP-43 immunoreative axons occurred in the first proximal segment (from 180 to 566 in average) in contrast to the absence of a significant increase of RT-97 expressing axons (from 60 to 62) (cf, Figs. 1 and 3). No SP-LI axonal profiles were identified in the nerve tissue on the 7th day after chamber application to the transected sciatic nerve. However, SP-LI axonal profiles were found in all segments of the nerve tissue after 14 days (Figs. 4A, B). The number of SPLI axons was highest in the proximal segments, and more distal segments contained a higher amount of SP-LI dots than the central ones (Fig. 5). Most of the SP-LI axons in the distal segments ran close to blood vessels (Fig. 4B). The proportion of SP-LI axons to all growing axons was about 40010 and 23% in the most proximal and most distal segment, respectively.
Discussion
The first events observed in the standard silicone chamber with a 10 mm gap is the accumulation of fluid and subsequent formation of a fibrin matrix during the first week. In this model, the axons are found in a 1 mm proximal segment of the chamber 7 days after operation and bridge the gap between the nerve stumps by 3 weeks (Williams et aI., 1983, 1987). In our chamber model we have used a resorbable artificial fibrin sponge (Gelaspon) to improve the substrate conditions for the migration of Schwann cells (Dubovy and
128
P. Dubovy and J. Bednafova
Fig. 4. Transversal sections through the nerve outgrowth immunostained for substance P (SP), 14 days (D) after chamber implantation. The section through the first proximal segment of the nerve outgrowth contains many immunopositive axonal profiles (Fig. 4A). A detail of some individual SP-immunopositive axonal profiles (arrows) found in the section through the ninth segment is shown in Fig. 4B. Scale bar: 350 urn for Fig. 4A; 100 urn for Fig. 4B.
14Da
2
3
467
8
9
10
egment Fig. 5. Number of SP-immunoreactive axonal profiles 14 days after chamber application .
Growing axons in silicone chambers
129
Svizenska, 1992), which are critical components for the regenerating axons (Hall, 1986; Feneley et al., 1991). In our experiments growing axons identified with the GAP-43 antibody were observed at a distance of about 7 mm from the proximal stump on the 7th day after chamber application. This is a conspicuously longer distance for axonal elongation than that in standard PBS-prefilled chambers without the gelatine sponge (Williams et al., 1983; for review see Danielsen, 1990). Within the second week, axons bridged the whole 10 mm gap distance of the chamber. In addition, the number of growing axons in the proximal segments was significantly increased. The present results thus indicate that the artificial sponge matrix significantly improves the conditions for growing axons in the chamber model. The monoclonal antibody RT-97 binds predominantly to the phosphorylated 200 kDa neurofilament protein subunit (Wood and Anderton, 1981; Perry et al., 1991). The RT-97 immunoreactive fibres correspond mainly to large A a/fJ primary sensory neurons with rapid conduction velocities (Lawson and Waddell, 1991). Moreover, the expression of phosphorylated neurofilament proteins correlates with the maturation of regenerating axons (Hoffmann et al., 1985). The number of RT-97 positive fibres did not increase significantly from 7 to 14 days after chamber implantation. However, RT-97 expressing fibres were found over a longer distance after 14 than after 7 days. These findings probably indicate that the growing axons are related to the large-sized primary sensory neurons as well as in their increased maturation progress during the second week. Substance P is an undecapeptide which has been demonstrated immunohistochemically in primary sensory neurons (Hokfelt et al., 1975) as well as in sympathetic ganglion neurons (Gamse et al., 1981). In our material we were not able to distinguish between sensory and autonomic SP-LI fibres. However, since most of the SP-LI fibres were found close to blood vessels, especially those in the distal segments, the majority of the SP-LI axons are likely to be sympathetic than sensory (Furness et al., 1982; Barja et al., 1983). On the other hand, at least some of these SP-LI axons may have sensory function (Dalsgaard, 1988). The SP-LI fibres were observed throughout the entire proximo-distal length of the chamber 14 days after entubulization. The presence of SP-LI axons in the nerve tissue only 14 but not 7 days after implantation may be related to a recovery of SP synthesis and the subsequent anterograde SP transport into the regenerating axons (Bisby and Keen, 1985). The observation of a smaller number ofaxons in the central segments compared to the proximal as well as distal segments indicates SP-LI fibres growth into the chamber not only from the proximal but also from the distal stump. However, since SP-LI axons contributed only about 23070 of all growing axons in the most distal segment, ingrowth from the distal stump is only a minor explanation for the presence ofaxons in this segments after two weeks. It is therefore concluded that axons growing from proximal stump under appropriate substrate conditions are able to bridge a gap of about 10 mm within two weeks.
Acklowledgements This work was supported by a grant 309/93/0124 of GA CR, in part by the Swedish Medical Research Council, project 5420, and by a visiting scientist fellowship from the Karolinska Institutet. We gratefully acknowledge Prof. H. Aldskogius for his critical reading of the manuscript.
130
P. Dubovy and J. Bednarova
References Barja F, Mathison R, and Huggel H (1983) Substance P-containing nerve fibres in large peripheral blood vessels of the rat. Cell Tissue Res 229: 411 - 422 Bisby MA, and Keen P (1985) Axonal transport of substance P-Iike immunoreactivity in regenerating rat sciatic nerve. Brain Res 361: 396- 399 Dalsgaard CJ (1988) The sensory system. In: Bjorklund A, Hokfelt T, and Owman C (Eds) Handbook of chemical neuroanatomy, Vol. 6: The peripheral nervous system. Elsevier Science Publisher, pp 599-635 Danielsen N (1990) Regeneration of the rat sciatic nerve in the silicone chamber model. Restor Neurol Neurosci 1: 253 - 259 Dubovy P, and Svizenska I (1992) Migration of Schwarm cells from the distal stump of the sciatic nerve 1 week after transection - The effects of insulin and cytosine arabino side. Glia 6: 281-288 Dubovy P, and Svizenska I (1994) Denervated skeletal muscle stimulates migration of Schwann cells from the distal stump of transected peripheral nerve: An in vivo study. Glia 12: 99 - 107 Feneley MR, Fawcett JW, and KeynesRJ (1991) The role of Schwarm cells in the regeneration of peripheral nerve axons through muscle basal lamina grafts. Exp, Neurol114: 275- 285 Furness JB, Papka RE, Della NO, Costa M, and Eskay RL (1982) Substance P-like immunoreactivity in nerves associated with the cardiovascular system of guinea-pigs. Neuroscience 7: 447 - 459 Gamse R, Wax A, Zigmond RE, and Leeman SE (1981) Immunoreactive substance P in sympathetic ganglia: distribution and sensitivity towards capsaicin. Neuroscience 6: 437 -441 Hall SM (1986) The effect of inhibiting Schwann cell mitosis on the re-innervation of acellular autografts in the peripheral nervous system of the mouse. Neuropathol Appl Neurobiol12: 401-414 Hoffman PN, Thompson OW, Griffin JW, and Price DL (1985) Changes in neurofilament transport coincide temporall y with alterations in the caliber ofaxons in regenerating moto r fibers. J Cell Bioi 101: 1332-1340 Hokfelt T, Kellerth JO, Nilsson G, and Pernow B (1975) Substance P: localization in the central nervous system and in some primary sensory neuron s. Science 190: 889- 890 Johnson D, and deC Nogueira Araujo G (1981) A simple method of reducing the fading of immunofluorescence during microscopy. J Immunol Meth 43: 349- 350 Lawson SN, and Waddell PJ (1991) Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons. J Physiol 435: 41-63 Lundborg G, Longo FM, and Varon S (1982a) Nerve regeneration model and trophic factors in vivo. Brain Res 232: 157-161 Lundborg G, Dahlin LB, Danielsen N, Gelberrnan RH, Longo FM, Powell HC, and Varon S (1982b) Nerve regeneration in silicone chambers : Influence of gap length and of distal stump components. Exp Neurol 76: 361-375 Perry MJ, Lawson SN, and Robertson J (1991) Neurofilament immunoreactivity in populations of rat primary afferent neurons : a quantitative study of phosphorylated and non-phosphorylated subunits. J Neurocytol 20: 746 -758 WeissP (1944) The technology of nerve regeneration: sutureless tubulation or related methods of nerve repair. Neurosurgery 1: 400-450 Williams LR, Longo FM, Powell HC, Lundborg G, and Varon S (1983) Spatial-temporal progress of peripheral nerve regeneration within a silicone chamber: parameters for a bioassay. J Comp Neurol 218: 460-470 Williams LR, Danielsen N, Muller H, and Varon S (1987) Exogenous matrix precursors promote functional nerve regeneration across a 15 mm gap within a silicone chamber in the rat. J Comp Neurol 264: 284 - 290 Williams LR, Danielsen N, Muller H, and Varon S (1988) Influence of the acellular fibrin matri x on nerve regeneration success within the silicone chamber model. Neurol Neurobiol 38: 111 -122 Williams LR, and Varon S (1985) Modifica tion of fibrin matri x formation in situ enhances nerve regeneration in silicone chambers. J Comp Neurol 231: 209 - 220 Wood JW, and Anderton BH (1988) Monoclonal antibodies to mammalian neurofilaments. Biosci Reports 1: 263 - 268 Zamboni L, and de Martino C (1967) Buffered picric acid-formaldehyde: a new rapid fixative for electron microscopy. J Cell Bioi 35: 148A Zhao Q, Dahlin LB, Kanje M, and Lundborg G (1993) Repair of the transected rat sciatic nerve - Matrix formation within implanted silicone tubes. Restor Neurol Neurosci 5: 197- 204
Atla hislork.ita
An immunocytochemical analysis of growing axons in a silicone chamber prefilled with artificial sponge matrix Petr Dubovy' r' and J. Bednarova' I Department of Anatomy, School of Medicine, Masaryk University Brno, Komenskeho nam, 2, CZ-66243 Brno, Czech Republic and 2 Department of Neuroscience, Division of Neuroanatomy and Neuronal Plasticity, Karolinska Institute, S-17177 Stockholm, Sweden
Accepted 14 December 1995
Summary We have used antibodies against growth associated protein (GAP-43), phosphorylated neurofilament protein of 200 kDa molecular weight (RT-97) and substance P (SP) to analyze regrowing axons and their features in a silicone chamber filled with resorbable sponge matrix within the first two weeks after sciatic nerve transection in the rat. Growing axons identified with the GAP-43 antibody extended over a distance of about 7 mm from the proximal stump at 7 days and grew over a 10 mm gap within 14 days. This is a markedly longer distance than in the case of the standard chamber model without artificial sponge matrix. The regrowing axons were labelled with RT-97 already on the 7th day up to a distance of 5 mm and they made up about 750/0 of all axons in the first segments. The number of RT-97-positive axons did not increase significantly over the next 7 days, although they could be identified over a longer distance. Some of the growing axons expressed SP-like immunoreactivity (LI) 14 days, but not 7 days after chamber application and constituted about 30% of all growing axons in the first segment. The SPLI fibres also appeared to grow from the distal stump since they were found in larger numbers in the distal segments than in central ones. Those fibres accompanying blood vessels are probably sympathetic. Our findings demonstrate that axons are able to bridge a 10 mm gap within 14 days under appropriate substrate conditions, which are provided by the resorbable fibrin sponge.
Key words: peripheral nerve neurofilaments - substance P
transection -
silicone chamber -
GAP-43 -
Introduction The idea of attempting to create a conduit for regenerating axons across a gap between nerve stumps started in the 1940s (Weiss, 1944). Since then, one of the most useful designs of such a conduit was introduced by Lundborg and co-workers for the study of Correspondence to: P. Dubovy
124
P. Dubovy and J. Bednarova
cellular and molecular mechanisms of peripheral nerve regeneration (Lundborg et al., 1982a, b). In this experimental model, the proximal and distal stumps of the transected sciatic nerve are sutured into the ends of a silicone tube which form a chamber with a structurally organized preformed space where the nerve microenvironment is protected from external influences. These features make chambers from unpermeable material suitable for analyzing cellular events and molecular processes during nerve regeneration (Williams et al., 1983; Dubovy and Svizenska, 1992, 1994; Zhao et al., 1993). During the first week after implantation, the standard chamber is filled with nerve stump-derived fluid, in which a longitudinally oriented fibrin matrix is formed. The fibrin matrix formation is critical for the migration of fibroblasts, Schwann and endothelial cells as well as for the regrowth ofaxons through the chamber (Williams et al., 1985). A similarly suitable matrix for the enhanced migration of these cells and the subsequent regrowth ofaxons in the chamber can be provided by exogenous fibrin precursors or an artificial fibrin sponge (Williams et al., 1987; Dubovy and Svizenska, 1992). The spatial-temporal progress of nerve regeneration across a 10 mm gap inside the silicone chamber has been examined previously by means of classic light and electron microscopy (Williams et al., 1983, 1988). However, little is known about the immunocytochemical characteristics of regrowing axons in this chamber model. In the present study we have characterized immunocytochemically the growth of axons and their features in a silicone chamber model within the first two weeks after sciatic nerve transection by means of antibodies against growth associated protein (GAP-43), phosphorylated neurofilament protein (RT-97) and substance P. Material and methods Ten female Sprague-Dawley rats (200 - 250 g) were used for the experiments. All surgical procedures were performed under deep anaesthesia by intramuscular injection of a mixture (0.2 mill 00 g) containing xylazinum (4 mg/ml) and ketaminum (40 mg/ml). The right sciatic nerve was exposed and mobilized, stripped of epineurium in a 4 mm segment, and transected at mid-thigh level proximal to the tibial and peroneal bifurcation. A 2 mm nerve segment was removed and the proximal and distal stumps were introduced into each end of a 14 mm long silicone tube. The stumps were fixed with single transepineurial 10-0 sutures to the chamber wall, leaving a 10 mm distance between them. The chambers were prepared from silicone tubes with an internal diameter of 1.7 mm and an outer diameter of 3 mm. They were sterilized, filled with artificial fibrin sponge tissue (Gelaspon, Ankerwerk, FRG) and soaked with sterilized phosphate-buffered saline (PBS). After chamber implantation, the muscles and overlying skin were closed, and the animals were allowed to recover from the anaesthesia and to survive for a period of 7 (5 animals) and 14 days (5 animals). All animals were fed ad libitum and housed on soft material to minimize the possibility of trauma to their denervated hind limbs. At the time of sacrifice the animals were deeply anaesthetized with chloral hydrate and perfused via the ascending aorta with warm (37°C) phosphate-buffered saline (PBS) at pH 7.4. This was followed by cold (4°C) 4070 paraformaldehyde (w/v) in PBS (pH 7.4) containing 0.4% picric acid (w/v) (Zamboni and de Martino, 1967). The content of each silicone tube was removed from the chamber, stored at 4°C overnight in Zamboni's fixative solution, and then rinsed in PBS containing 10% sucrose. The nerve outgrowths were divided in 10 proximo-distal segments, each measuring about 1 mm. Cryostat sections (10 urn thick) were cut transversely through each segment and processed for indirect immunofluorescence. Twenty subsequent sections from each segment were incubated simultaneously with mouse monoclonal antibody to neurofilament protein (RT-97, dilution 1: 250, generously supplied by Dr. J. Wood, London, U. K.; Wood and Anderton, 1981) and rabbit polyclonal antibody against GAP-43 (dilution 1: I 000, a generous gift of Prof. W. H. Gispen, Utrecht, The Netherlands). Another twenty subsequent sections from each segment were incubated with polyclonal antibody raised against substance P (dilution 1: 1000, generously provided by Dr. P. Petrusz, North Carolina, U.S.A.).
Growing axons in silicone chambers
125
Goat anti-mouse and goat anti-rabbit secondary antibodies conjugated with fluorescein isothiocyanate (FITC) or rhodamine (TRITe) (dilution 1: 3D, Boehringer Mannheim, Indianapolis, U.S.A.) were used for localization of the primary immune complex. The sections were coverslipped in glycerol containing 0.1 % p-phenylenediamine to reduce fading (Johnson and Nogueira Araujo, 1981) and viewed in a Nikon fluorescence microscope using appropriate filters for FITC or TRITC epifluorescence. Immunohistochemical controls were performed by incubation without the primary antibodies or with appropriate preimmune sera and yielded negative data. The total number of GAP-43-like immunoreactive (GAP-43-LI), substance-P-like immunoreactive (SP-LI), and RT-97 positive immunolabelled fibers was counted in the entire cross-sectioned area of 10 sections using an image analysis system (Leitz-ASBA) at the same magnification. Average values and standard deviations were calculated for individual segments.
Results All chambers contained a delicate bridge of nerve tissue connecting the proximal and distal stumps. Immunocytochemical staining of transversal sections revealed that regenerating axons were closely packed in the central area of the nerve tissue where labelled axons appeared as spots. In addition to these spots some enlargements probably corresponding to growth cones were observed in sections stained for GAP-43. Regenerating axons immunolabelled with the GAP-43 antibody were found in the 7 proximal and in all 10 segments after 7 and 14 days, respectively, implying a distance of axonal elongation of about 7 and 10 mm from the proximal stump. Seven days after chamber application, the RT-97 positive axons were present in 5 proximal segments while following 14 days the RT-97 positive axons were observed in up to seven proximal segments (Fig. 1). Some of the regenerating axons identified by GAP-43 were also immunoreactive for RT-97 (cf. Figs. 2A, B and Figs. 2e, D).
•
200
= 0 1;1)
~
<
7Day RT-97
D
GAP-43
8
9
1 0
eO
J~
..c S Z
=
100
50
o
2
3
4
5
6
7
10
egment Fig. 1. Number of GAP-43 and RT-97 immunoreactive axons seven days after chamber application.
126
P. Dubovy and J. Bednarova
Figs. 2 A, B. A transversal section trough the sixth segment of nerve outgrowth immunostained simultaneously for GAP-43 and RT-97. Axon profiles exhibiting immunopositivity for GAP-43 (Fig. 1A) are not immunolabelled by RT-97 (Fig. 1B). The segment corresponds to a distance of about 6 mm from the end of the proximal stump, 7 days (0) after chamber application. Scale bar: 300 um. Figs. 2 C, O. Part of a transversal section through the second segment of nerve outgrowth (about 2 mm from proximal stump) 14days (0) after chamber implantation. The section was immunostained simultaneously for GAP-43 (Fig. 1 C) and RT-97 (Fig. 10). Some axon profiles display immunolabelling for both GAP-43 and RT-97 (arrows). Scale bar: 300 urn.
Growing axons in silicone chambers
127
14Da 600 •
D
o 400
RT-97 GAP-43
I- I
300 200 100
o
2
3
8
4
9
10
egment Fig. 3. Number of GAP-43 and RT-97 expressing axons 14 days after chamber application.
When comparing the number ofaxons 7 and 14 days after chamber application, a marked increase in the number of GAP-43 immunoreative axons occurred in the first proximal segment (from 180 to 566 in average) in contrast to the absence of a significant increase of RT-97 expressing axons (from 60 to 62) (cf, Figs. 1 and 3). No SP-LI axonal profiles were identified in the nerve tissue on the 7th day after chamber application to the transected sciatic nerve. However, SP-LI axonal profiles were found in all segments of the nerve tissue after 14 days (Figs. 4A, B). The number of SPLI axons was highest in the proximal segments, and more distal segments contained a higher amount of SP-LI dots than the central ones (Fig. 5). Most of the SP-LI axons in the distal segments ran close to blood vessels (Fig. 4B). The proportion of SP-LI axons to all growing axons was about 40010 and 23% in the most proximal and most distal segment, respectively.
Discussion
The first events observed in the standard silicone chamber with a 10 mm gap is the accumulation of fluid and subsequent formation of a fibrin matrix during the first week. In this model, the axons are found in a 1 mm proximal segment of the chamber 7 days after operation and bridge the gap between the nerve stumps by 3 weeks (Williams et aI., 1983, 1987). In our chamber model we have used a resorbable artificial fibrin sponge (Gelaspon) to improve the substrate conditions for the migration of Schwann cells (Dubovy and
128
P. Dubovy and J. Bednafova
Fig. 4. Transversal sections through the nerve outgrowth immunostained for substance P (SP), 14 days (D) after chamber implantation. The section through the first proximal segment of the nerve outgrowth contains many immunopositive axonal profiles (Fig. 4A). A detail of some individual SP-immunopositive axonal profiles (arrows) found in the section through the ninth segment is shown in Fig. 4B. Scale bar: 350 urn for Fig. 4A; 100 urn for Fig. 4B.
14Da
2
3
467
8
9
10
egment Fig. 5. Number of SP-immunoreactive axonal profiles 14 days after chamber application .
Growing axons in silicone chambers
129
Svizenska, 1992), which are critical components for the regenerating axons (Hall, 1986; Feneley et al., 1991). In our experiments growing axons identified with the GAP-43 antibody were observed at a distance of about 7 mm from the proximal stump on the 7th day after chamber application. This is a conspicuously longer distance for axonal elongation than that in standard PBS-prefilled chambers without the gelatine sponge (Williams et al., 1983; for review see Danielsen, 1990). Within the second week, axons bridged the whole 10 mm gap distance of the chamber. In addition, the number of growing axons in the proximal segments was significantly increased. The present results thus indicate that the artificial sponge matrix significantly improves the conditions for growing axons in the chamber model. The monoclonal antibody RT-97 binds predominantly to the phosphorylated 200 kDa neurofilament protein subunit (Wood and Anderton, 1981; Perry et al., 1991). The RT-97 immunoreactive fibres correspond mainly to large A a/fJ primary sensory neurons with rapid conduction velocities (Lawson and Waddell, 1991). Moreover, the expression of phosphorylated neurofilament proteins correlates with the maturation of regenerating axons (Hoffmann et al., 1985). The number of RT-97 positive fibres did not increase significantly from 7 to 14 days after chamber implantation. However, RT-97 expressing fibres were found over a longer distance after 14 than after 7 days. These findings probably indicate that the growing axons are related to the large-sized primary sensory neurons as well as in their increased maturation progress during the second week. Substance P is an undecapeptide which has been demonstrated immunohistochemically in primary sensory neurons (Hokfelt et al., 1975) as well as in sympathetic ganglion neurons (Gamse et al., 1981). In our material we were not able to distinguish between sensory and autonomic SP-LI fibres. However, since most of the SP-LI fibres were found close to blood vessels, especially those in the distal segments, the majority of the SP-LI axons are likely to be sympathetic than sensory (Furness et al., 1982; Barja et al., 1983). On the other hand, at least some of these SP-LI axons may have sensory function (Dalsgaard, 1988). The SP-LI fibres were observed throughout the entire proximo-distal length of the chamber 14 days after entubulization. The presence of SP-LI axons in the nerve tissue only 14 but not 7 days after implantation may be related to a recovery of SP synthesis and the subsequent anterograde SP transport into the regenerating axons (Bisby and Keen, 1985). The observation of a smaller number ofaxons in the central segments compared to the proximal as well as distal segments indicates SP-LI fibres growth into the chamber not only from the proximal but also from the distal stump. However, since SP-LI axons contributed only about 23070 of all growing axons in the most distal segment, ingrowth from the distal stump is only a minor explanation for the presence ofaxons in this segments after two weeks. It is therefore concluded that axons growing from proximal stump under appropriate substrate conditions are able to bridge a gap of about 10 mm within two weeks.
Acklowledgements This work was supported by a grant 309/93/0124 of GA CR, in part by the Swedish Medical Research Council, project 5420, and by a visiting scientist fellowship from the Karolinska Institutet. We gratefully acknowledge Prof. H. Aldskogius for his critical reading of the manuscript.
130
P. Dubovy and J. Bednarova
References Barja F, Mathison R, and Huggel H (1983) Substance P-containing nerve fibres in large peripheral blood vessels of the rat. Cell Tissue Res 229: 411 - 422 Bisby MA, and Keen P (1985) Axonal transport of substance P-Iike immunoreactivity in regenerating rat sciatic nerve. Brain Res 361: 396- 399 Dalsgaard CJ (1988) The sensory system. In: Bjorklund A, Hokfelt T, and Owman C (Eds) Handbook of chemical neuroanatomy, Vol. 6: The peripheral nervous system. Elsevier Science Publisher, pp 599-635 Danielsen N (1990) Regeneration of the rat sciatic nerve in the silicone chamber model. Restor Neurol Neurosci 1: 253 - 259 Dubovy P, and Svizenska I (1992) Migration of Schwarm cells from the distal stump of the sciatic nerve 1 week after transection - The effects of insulin and cytosine arabino side. Glia 6: 281-288 Dubovy P, and Svizenska I (1994) Denervated skeletal muscle stimulates migration of Schwann cells from the distal stump of transected peripheral nerve: An in vivo study. Glia 12: 99 - 107 Feneley MR, Fawcett JW, and KeynesRJ (1991) The role of Schwarm cells in the regeneration of peripheral nerve axons through muscle basal lamina grafts. Exp, Neurol114: 275- 285 Furness JB, Papka RE, Della NO, Costa M, and Eskay RL (1982) Substance P-like immunoreactivity in nerves associated with the cardiovascular system of guinea-pigs. Neuroscience 7: 447 - 459 Gamse R, Wax A, Zigmond RE, and Leeman SE (1981) Immunoreactive substance P in sympathetic ganglia: distribution and sensitivity towards capsaicin. Neuroscience 6: 437 -441 Hall SM (1986) The effect of inhibiting Schwann cell mitosis on the re-innervation of acellular autografts in the peripheral nervous system of the mouse. Neuropathol Appl Neurobiol12: 401-414 Hoffman PN, Thompson OW, Griffin JW, and Price DL (1985) Changes in neurofilament transport coincide temporall y with alterations in the caliber ofaxons in regenerating moto r fibers. J Cell Bioi 101: 1332-1340 Hokfelt T, Kellerth JO, Nilsson G, and Pernow B (1975) Substance P: localization in the central nervous system and in some primary sensory neuron s. Science 190: 889- 890 Johnson D, and deC Nogueira Araujo G (1981) A simple method of reducing the fading of immunofluorescence during microscopy. J Immunol Meth 43: 349- 350 Lawson SN, and Waddell PJ (1991) Soma neurofilament immunoreactivity is related to cell size and fibre conduction velocity in rat primary sensory neurons. J Physiol 435: 41-63 Lundborg G, Longo FM, and Varon S (1982a) Nerve regeneration model and trophic factors in vivo. Brain Res 232: 157-161 Lundborg G, Dahlin LB, Danielsen N, Gelberrnan RH, Longo FM, Powell HC, and Varon S (1982b) Nerve regeneration in silicone chambers : Influence of gap length and of distal stump components. Exp Neurol 76: 361-375 Perry MJ, Lawson SN, and Robertson J (1991) Neurofilament immunoreactivity in populations of rat primary afferent neurons : a quantitative study of phosphorylated and non-phosphorylated subunits. J Neurocytol 20: 746 -758 WeissP (1944) The technology of nerve regeneration: sutureless tubulation or related methods of nerve repair. Neurosurgery 1: 400-450 Williams LR, Longo FM, Powell HC, Lundborg G, and Varon S (1983) Spatial-temporal progress of peripheral nerve regeneration within a silicone chamber: parameters for a bioassay. J Comp Neurol 218: 460-470 Williams LR, Danielsen N, Muller H, and Varon S (1987) Exogenous matrix precursors promote functional nerve regeneration across a 15 mm gap within a silicone chamber in the rat. J Comp Neurol 264: 284 - 290 Williams LR, Danielsen N, Muller H, and Varon S (1988) Influence of the acellular fibrin matri x on nerve regeneration success within the silicone chamber model. Neurol Neurobiol 38: 111 -122 Williams LR, and Varon S (1985) Modifica tion of fibrin matri x formation in situ enhances nerve regeneration in silicone chambers. J Comp Neurol 231: 209 - 220 Wood JW, and Anderton BH (1988) Monoclonal antibodies to mammalian neurofilaments. Biosci Reports 1: 263 - 268 Zamboni L, and de Martino C (1967) Buffered picric acid-formaldehyde: a new rapid fixative for electron microscopy. J Cell Bioi 35: 148A Zhao Q, Dahlin LB, Kanje M, and Lundborg G (1993) Repair of the transected rat sciatic nerve - Matrix formation within implanted silicone tubes. Restor Neurol Neurosci 5: 197- 204