- Email: [email protected]
Glial cell plasticity in sensory ganglia induced by nerve damage
Neuroscience Vol. 114, No. 2, pp. 279^283, 2002 K 2002 IBRO. Published by Elsevier Science Ltd All rights reserved. Printed in Great Britain 0306-4522 / 02 $22.00+0.00
PII: S 0 3 0 6 - 4 5 2 2 ( 0 2 ) 0 0 2 7 9 - 8 www.neuroscience-ibro.com
Letter to Neuroscience GLIAL CELL PLASTICITY IN SENSORY GANGLIA INDUCED BY NERVE DAMAGE M. HANANI,a T. Y. HUANG,a P. S. CHERKAS,a M. LEDDAb and E. PANNESEb a b
Laboratory of Experimental Surgery, Hadassah University Hospital, Mount Scopus, Jerusalem 91240, Israel
Institute of Histology, Embryology and Neurocytology, University of Milan, 14 Via Mangiagalli, I-20133 Milan, Italy Key words: dorsal root ganglia, satellite glial cells, axotomy, gap junctions, dye coupling, ultrastructure.
activity of DRG neurons; for example, injured sensory neurons develop spontaneous electrical activity (Zhang et al., 1997; Amir et al., 1999). The augmented electrical excitability of these neurons may be associated with certain manifestations of chronic pain (Devor and Seltzer, 1999). In spite of the progress in this ¢eld, open questions remain concerning the contribution of DRG changes to neuropathic pain. Individual DRG neurons are surrounded by an envelope consisting of satellite glial cells (SCs; Pannese, 1981). These cells are in close contact with the neurons, but their functions are little understood. There is evidence that SCs undergo morphological and biochemical changes after nerve damage (Woodham et al., 1989; Stephenson and Byers, 1995), but there is no information on their possible contribution to pain mechanisms. It was found that axotomy of retinal ganglion cells (Becker and Cook, 1990) and spinal motor neurons (Chang et al., 2000) can increase electrical coupling among neurons, but information on similar changes in glial cells is not available. The aim of this work was to ¢nd out whether intercellular coupling is altered in DRG of mice following axotomy. The methods we used were intracellular dye injections and serial section electron microscopy.
Numerous studies have been done on the e¡ect of nerve injury on neurons of sensory ganglia but little is known about the contribution of satellite glial cells (SCs) in these ganglia to post-injury events. We investigated cellto-cell coupling and ultrastructure of SCs in mouse dorsal root ganglia after nerve injury (axotomy). Under control conditions SCs were mutually coupled, but mainly to other SCs around a given neuron. After axotomy SCs became extensively coupled to SCs that enveloped other neurons, apparently by gap junctions. Serial section electron microscopy showed that after axotomy SC sheaths enveloping neighboring neurons formed connections with each other. Such connections were absent in control ganglia. The number of gap junctions between SCs increased 6.5-fold after axotomy. We propose that axotomy induces growth of perineuronal SC sheaths, leading to contacts between SCs enveloping adjacent neurons and to formation of new gap junctions between SCs. These changes may be an important mode of glial plasticity and can contribute to neuropathic pain. ( 2002 IBRO. Published by Elsevier Science Ltd. All rights reserved. Mammalian sensory ganglia, and in particular dorsal root ganglia (DRG), have been the focus of intense research because of their importance in the transmission of sensory signals and their contribution to acute and chronic pain states. There is evidence that neuropathic pain resulting from peripheral nerve injury may be partly due to changes taking place in the DRG (LaMotte et al., 1996; Devor and Seltzer, 1999). Morphological changes of DRG neurons after nerve damage have been observed under both light and electron microscopes (Lieberman, 1971). Nerve damage also causes changes in the electrical
EXPERIMENTAL PROCEDURES
Young adult BALB/c mice of either sex, weighing 21^25 g, were used. The animals were anesthetized with pentobarbital (48 mg/kg). The right sciatic and saphenous nerves were exposed at mid-thigh level and a 5^10 mm segment of the nerves was dissected out to prevent regeneration. After 8^14 days the animals were killed by cervical dislocation. The DRG (L4, L5) were removed from both sides. Ganglia from the contralateral side and from unoperated animals were used as controls. In all cases we found that there was no regeneration of the cut nerve. The isolated ganglia were pinned onto a chamber with a Sylgard bottom. The chamber was superfused with Krebs solution at 23^24‡C that was bubbled with 95% O2 /5% CO2 . The chamber was placed on the stage of an upright microscope equipped with £uorescent illumination. Single cells were impaled with glass microelectrodes and injected with Lucifer Yellow (LY) as
*Corresponding author. Tel.: +972-2-5844721; fax: +972-25823515. E-mail address: [email protected] (M. Hanani). Abbreviations : DRG, dorsal root ganglia; LY, Lucifer Yellow; SC, satellite glial cell. 279
NSC 5720 15-8-02
280
M. Hanani et al.
Fig. 1. The morphology of cells in control DRG. (a) A single neuron in mouse DRG injected with LY. (b) SC in mouse DRG injected with LY. In control ganglia, such as this one, SCs were in many cases connected only to other SCs forming an envelope around a given neuron. (c) An example of a case where the SC was not coupled to other SCs. Scale bars = 20 Wm.
described previously (Hanani et al., 1999). The cells were viewed with a 40U water immersion objective, which allowed the observation and photography in real time of the staining of the cells and of the dye spread from the injected cell to adjacent ones. Images of stained cells were obtained using a digital camera (Pixera 120es). After ¢xation of the tissues in 4% phosphatebu¡ered paraformaldehyde (pH 7.4) the cells were imaged using a Bio-Rad confocal microscope. Fourteen days after axotomy four animals were perfused transcardially with a solution containing 2% formaldehyde and 2% glutaraldehyde in 0.1 M sodium cacodylate bu¡er (pH 7.3) under deep pentobarbital anesthesia (48 mg/kg). After ¢xation for 3 h, DRG L4 and L5 were removed from both sides and processed for electron microscopy. Thin sections (about 60 nm) were cut from ¢ve contralateral and ¢ve ipsilateral ganglia. These sections were examined under the electron microscope. The mean density of gap junctions was determined by measuring the number of gap junctions per 100 Wm2 of the section occupied by SCs. The mean length of gap junctions was also determined. The values obtained for the ganglia on each side were compared using a test for variance to establish whether they di¡ered signi¢cantly. The mean values for the contralateral ganglia were compared with those for the ipsilateral ganglia. The statistical comparisons employed the two-tailed Student’s t-test. Finally, a series of sections was cut from each of four ganglia (two contralateral and two ipsilateral to the axotomy). Each series consisted of about 130 consecutive sections. All the sections were
photographed under the electron microscope at a magni¢cation of 4000U and the negatives printed to a ¢nal magni¢cation of 16 000U. Serial sections were used to de¢ne connections between SC sheaths enveloping neighboring neurons.
RESULTS
LY was injected into single DRG neurons. In both control and axotomized ganglia this revealed that DRG neurons were not coupled to other neurons or to SCs (Fig. 1a). Injection of LY into single SCs in control ganglia showed that in many cases SCs were coupled to other SCs that formed the sheath around a given neuron (Fig. 1b). In 2.9% of the cases (6/206) there was also dye coupling among SCs that formed the sheath enveloping a given neuron and SCs surrounding one to four adjacent neurons. Eight to 14 days after axotomy, striking changes were observed in the DRG L4, L5. Injection of a single SC in contralateral DRG resulted in the staining of the SC sheaths surrounding neighboring neurons in only 6.4%
Fig. 2. The e¡ect of axotomy on SCs. (a) A SC injected with LY in an axotomized ganglion. This cell appears to be connected to other SCs enveloping adjacent neurons. (b) Histograms showing the incidence of inter-envelope coupling. (c) A SC in an axotomized ganglion showing a process emerging from the glial envelope, but not coupled to other SCs. The asterisks indicate the injected cells. Scale bars = 20 Wm.
NSC 5720 15-8-02
Glial plasticity in sensory ganglia
281
Fig. 3. Electron microscopic observations. (a) Control DRG; each nerve cell body (N) is enveloped by a SC sheath (sc), which is clearly separated from sheaths encircling the adjacent neurons by connective tissue (ct). Note that the perineuronal SC sheaths have a smooth contour, U10 000. (b) Axotomized DRG; process (arrows) arising from a perineuronal SC sheath (sc) projects into the connective tissue space. v: blood vessel. Scale bars = 2.5 Wm.
of the cases (8/125, eight animals). In contrast, on the ipsilateral side LY injection into single SCs resulted in 21.3% of the cases (37/150, eight animals; P 6 0.0005) in the staining of SCs surrounding up to 15 neighboring neurons (Fig. 2a). In both contralateral and ipsilateral ganglia SC dye coupling was completely blocked (in all
38 cases) by 1 mM octanol, indicating that it was mediated by gap junctions. The results are summarized in Fig. 2b. In several cases SCs in axotomized ganglia had distinct processes that projected into the connective tissue space (Fig. 2c). Such processes were never observed in control ganglia.
Fig. 4. Axotomized DRG. Electron micrographs of two sections from a series, separated by 1.4 Wm. A bridge (outlined) connecting the SC sheaths that envelop the neurons marked N is present in b, but not in a. The area outlined in b is shown at greater enlargement in the inset. Scale bars = 5 Wm.
NSC 5720 15-8-02
282
M. Hanani et al.
glia (0.360 Q 0.040 Wm and 0.361 Q 0.023 Wm, respectively; P 6 0.01).
DISCUSSION
Fig. 5. Axotomized DRG. An electron micrograph showing that the sheaths that envelop two neurons (N) are connected by a bridge (arrows). A gap junction is present in the bridge. Scale bar = 2 Wm. The outlined area is shown at greater enlargement in the inset. The arrowhead points to the gap junction. Scale bar = 0.5 Wm.
Electron microscopy showed that in control ganglia, the SC sheath enveloping a given neuron had a rather smooth outer contour and was distinctly separated from sheaths encircling adjacent neurons (Fig. 3a). In striking contrast, in the ipsilateral ganglia, processes of variable thickness often arose from the SC sheath, projecting into the connective tissue space (Fig. 3b). Examination of the serial sections from control ganglia never revealed any contacts between glial sheaths enveloping di¡erent neurons, whereas in ipsilateral ganglia the SC sheaths belonging to 76 neurons out of 258 (29.46%) were found to be connected by these processes (Fig. 4a, b). An enlargement of the area of contact between SC processes is presented in the inset in Fig. 4b. Fig. 5 is another example of a bridge connecting SC envelopes of two adjacent neurons. The inset in Fig. 5 shows the presence of a gap junction between the newly formed SC processes forming the bridge. Gap junctions were found in SC processes, and in all the bridges that were examined by serial sections; these gap junctions are new by de¢nition, as glial processes were absent in control ganglia. New gap junctions were also found within the perineuronal sheaths. The mean number of gap junctions per 100 Wm2 of the SC sheath sectional area was 6.5 times greater after axotomy (0.20 Q 0.04 in contralateral ganglia vs. 1.35 Q 0.11 in ipsilateral ganglia; P 6 0.01), whereas the mean length of individual gap junctions did not di¡er between ipsilateral and contralateral gan-
The results demonstrate that SCs in mouse DRG undergo profound changes after axotomy. Dye injections showed that coupling among SCs increased by more than seven-fold after axotomy. The electron microscopic observations support this ¢nding as they revealed that axotomy induced the formation of bridges that connect SC sheaths enveloping neighboring neurons. Moreover, the number of gap junctions between SCs increased by 6.5 times after axotomy, which correlates with the increase in the incidence of dye coupling among SCs. In normal ganglia SCs are frequently coupled, but the coupling is restricted in most cases to SCs enveloping a given neuron. In control ganglia we found inter-envelope coupling in a small proportion of cases, which probably re£ects the fact that nerve cell bodies can occur in clusters having SC sheaths in common (Pannese et al., 1991). The number of such clusters is high in young animals, and declines with age. As we used young adult mice the proportion of coupling between SCs around di¡erent neurons is in agreement with the presence of such clusters. The correlation between age and cell coupling is interesting in the present context because it has been shown (Fulton, 1995) that damage induces changes in the nervous system that resemble early stages of development. Thus it was shown that both after axotomy and during development DRG neurons are characterized by an eccentric nucleus, many mitochondria concentrated in the central region of the perikaryon, a profusion of free polysomes and only a few polysomes attached to endoplasmic reticulum membranes (Pannese, 1963). We believe that this is the ¢rst demonstration that nerve injury leads to changes in glial coupling. The involvement of glial cells in various pathological processes in the CNS is well established (Ridet et al., 1997; Aldskogius and Kozlova, 1998). Axotomy of central neurons induces changes in astrocytes, which are referred to as ‘glial activation’ and consist of glial hypertrophy and increased production of molecules such as cytokines and glial ¢brillary acidic protein (Graeber and Kreuzberg, 1986; Rohlmann et al., 1994; Ridet et al., 1997; Pastor et al., 2000). The functional signi¢cance of these observations is not clear, but it has been proposed that glial activation may support the survival of neurons, as activated astrocytes can release growth factors (Ridet et al., 1997; Pastor et al., 2000). Nerve injury causes an increase in the immunoreactivity of the protein connexin 43, which forms the gap junctions among astrocytes (Rohlmann et al., 1994). However, from other studies it appears that the e¡ects of injury on glial connexins are regulated by several processes, and that the overall outcome is still not clear (Nagy and Rash, 2000). The functional implications of the changes observed in glial connexin levels after injury is not known. We demonstrated both morphologically and functionally that axotomy increased glia^glia coupling in the mouse DRG
NSC 5720 15-8-02
Glial plasticity in sensory ganglia
and thus provided direct evidence for glial plasticity after nerve damage. The contribution of these changes to pathological states remains to be explored. Raisman (1991) suggested that glial cells respond to injury as a large, interconnected network. As normally DRG glial cells around di¡erent neurons are rarely coupled, the increased coupling after axotomy may endow these cells with the ability to communicate over long distances. Glial plasticity is part of CNS response to pathological stimuli such as dehydration (Bobak and Salm, 1996) or cortical lesion (Hailer et al., 1999). We propose that augmented glia^glia coupling may be an important mode of glial contribution to plasticity in the nervous system. It will be worthwhile to search for such changes in the CNS under a variety of pathological conditions.
283
The ¢ndings above may have interesting implications for the understanding of chronic pain that results from nerve injury. One prominent feature in neuropathic pain is the ‘spread’ of pain sensation from the locus of the injured nerves to areas where the nerves were apparently intact (Devor and Seltzer, 1999). We showed that axotomy increased SC^SC communication by gap junctions. Gap junctions allow the passage of small ions and of molecules with a molecular weight of up to about 1000 Da. Thus it can be postulated that axotomy enhances the spread of electrical currents and/or metabolites (e.g., second messengers) within the DRG. We propose that this enhanced propagation of signals in the DRG may contribute to the mechanisms of neuropathic pain.
REFERENCES
Aldskogius, H., Kozlova, E.N., 1998. Central neuron-glia and glial-glial interactions following axon injury. Prog. Neurobiol. 55, 1^26. Amir, R., Michaelis, M., Devor, M., 1999. Membrane potential oscillations in dorsal root ganglion neurons: role in normal electrogenesis and neuropathic pain. J. Neurosci. 19, 8589^8596. Becker, D.L., Cook, J.E., 1990. Changes in gold¢sh retinal ganglion cells during axonal regeneration. Proc. R. Soc. London B 241, 73^77. Bobak, J.B., Salm, A.K., 1996. Plasticity of astrocytes of the ventral glial limitans subjacent to the supraoptic nucleus. J. Comp. Neurol. 376, 188^ 197. Chang, Q., Pereda, A., Pinter, M.J., Balice-Gordon, R.J., 2000. Nerve injury induces gap junctional coupling among axotomized adult motor neurons. J. Neurosci. 20, 674^684. Devor, M., Seltzer, R., 1999. In: Wall, P.D., Melzack, R. (Eds.), Textbook of Pain, 4th edn. Churchill Livingstone, London, pp. 129^164. Fulton, B.P., 1995. Gap junctions in the developing nervous system. Perspect. Dev. Neurobiol. 2, 327^334. Graeber, M.B., Kreuzberg, G.W., 1986. Astrocytes increase in glial ¢brillary acidic protein during retrograde changes of facial motor neurons. J. Neurocytol. 15, 363^373. Hailer, N.P., Grampp, A., Nitsch, R., 1999. Proliferation of microglia and astrocytes in the dentate gyrus following entorhinal cortex lesion: a quantitative bromodeoxyuridine labeling study. Eur. J. Neurosci. 11, 3359^3364. Hanani, M., Maudlej, N., Ha«rtig, W., 1999. Morphology and intercellular communication in glial cells of the intramural ganglia from the guineapig urinary bladder. J. Auton. Nerv. Syst. 76, 62^67. LaMotte, R.H., Zhang, J.M., Petersen, M., 1996. Alterations in the functional properties of dorsal root ganglion cells with unmyelinated axons after a chronic nerve constriction in the rat. Prog. Brain Res. 110, 105^111. Lieberman, A.R., 1971. The axon reaction : a review of the principal features of perikaryal responses to axon injury. Int. Rev. Neurobiol. 14, 49^ 124. Nagy, J.I., Rash, J.E., 2000. Connexins and gap junctions of astrocytes and oligodendrocytes in the CNS. Brain Res. Rev. 32, 29^44. Pannese, E., 1963. Investigations on the ultrastructural changes of the spinal ganglion neurons in the course of axon regeneration and cell hypertrophy. I. Changes during axon regeneration. Z. Zellforsch. 60, 711^740. Pannese, E., 1981. The satellite cells of sensory ganglia. Adv. Anat. Embryol. Cell Biol. 65, 1^111. Pannese, E., Ledda, M., Arcidiacono, G., Rigamonti, L., 1991. Clusters of nerve cell bodies enclosed within a common connective tissue envelope in the spinal ganglia of the lizard and rat. Cell Tissue Res. 264, 209^214. Pastor, A.M., Delgado-Garc|¤a, J.M., Mart|¤nez-Guijarro, F.J., Lo¤pez-Garc|¤a, J.M., de la Cruz, R.R., 2000. Response of abducens internuclear neurons to axotomy in the adult cat. J. Comp. Neurol. 427, 370^390. Raisman, G., 1991. Glia, neurons, and plasticity. Ann. NY Acad. Sci. 633, 209^213. Ridet, J.L., Malhotra, S.K., Privat, A., Gage, F.H., 1997. Reactive astrocytes: cellular and molecular cues to biological function. Trends Neurosci. 20, 570^577. Rohlmann, A., Laskawi, R., Hofer, A., Dermietzel, R., Wol¡, J.R., 1994. Astrocytes as rapid sensors of peripheral axotomy in the facial nucleus of rats. NeuroReport 5, 409^412. Stephenson, J.L., Byers, M.R., 1995. GFAP immunoreactivity in trigeminal ganglion satellite cells after tooth injury in rats. Exp. Neurol. 131, 11^ 22. Woodham, P., Anderson, P.N., Nadim, W., Turmaine, M., 1989. Satellite cells surrounding axotomized rat dorsal root ganglion cells increase expression of GFAP-like protein. Neurosci. Lett. 98, 8^12. Zhang, J.M., Donnelly, D.F., Song, X.J., LaMotte, R.H., 1997. Axotomy increases the excitability of dorsal root ganglion cells with unmyelinated axons. J. Neurophysiol. 78, 2790^2794. (Accepted 24 May 2002)
NSC 5720 15-8-02
PII: S 0 3 0 6 - 4 5 2 2 ( 0 2 ) 0 0 2 7 9 - 8 www.neuroscience-ibro.com
Letter to Neuroscience GLIAL CELL PLASTICITY IN SENSORY GANGLIA INDUCED BY NERVE DAMAGE M. HANANI,a T. Y. HUANG,a P. S. CHERKAS,a M. LEDDAb and E. PANNESEb a b
Laboratory of Experimental Surgery, Hadassah University Hospital, Mount Scopus, Jerusalem 91240, Israel
Institute of Histology, Embryology and Neurocytology, University of Milan, 14 Via Mangiagalli, I-20133 Milan, Italy Key words: dorsal root ganglia, satellite glial cells, axotomy, gap junctions, dye coupling, ultrastructure.
activity of DRG neurons; for example, injured sensory neurons develop spontaneous electrical activity (Zhang et al., 1997; Amir et al., 1999). The augmented electrical excitability of these neurons may be associated with certain manifestations of chronic pain (Devor and Seltzer, 1999). In spite of the progress in this ¢eld, open questions remain concerning the contribution of DRG changes to neuropathic pain. Individual DRG neurons are surrounded by an envelope consisting of satellite glial cells (SCs; Pannese, 1981). These cells are in close contact with the neurons, but their functions are little understood. There is evidence that SCs undergo morphological and biochemical changes after nerve damage (Woodham et al., 1989; Stephenson and Byers, 1995), but there is no information on their possible contribution to pain mechanisms. It was found that axotomy of retinal ganglion cells (Becker and Cook, 1990) and spinal motor neurons (Chang et al., 2000) can increase electrical coupling among neurons, but information on similar changes in glial cells is not available. The aim of this work was to ¢nd out whether intercellular coupling is altered in DRG of mice following axotomy. The methods we used were intracellular dye injections and serial section electron microscopy.
Numerous studies have been done on the e¡ect of nerve injury on neurons of sensory ganglia but little is known about the contribution of satellite glial cells (SCs) in these ganglia to post-injury events. We investigated cellto-cell coupling and ultrastructure of SCs in mouse dorsal root ganglia after nerve injury (axotomy). Under control conditions SCs were mutually coupled, but mainly to other SCs around a given neuron. After axotomy SCs became extensively coupled to SCs that enveloped other neurons, apparently by gap junctions. Serial section electron microscopy showed that after axotomy SC sheaths enveloping neighboring neurons formed connections with each other. Such connections were absent in control ganglia. The number of gap junctions between SCs increased 6.5-fold after axotomy. We propose that axotomy induces growth of perineuronal SC sheaths, leading to contacts between SCs enveloping adjacent neurons and to formation of new gap junctions between SCs. These changes may be an important mode of glial plasticity and can contribute to neuropathic pain. ( 2002 IBRO. Published by Elsevier Science Ltd. All rights reserved. Mammalian sensory ganglia, and in particular dorsal root ganglia (DRG), have been the focus of intense research because of their importance in the transmission of sensory signals and their contribution to acute and chronic pain states. There is evidence that neuropathic pain resulting from peripheral nerve injury may be partly due to changes taking place in the DRG (LaMotte et al., 1996; Devor and Seltzer, 1999). Morphological changes of DRG neurons after nerve damage have been observed under both light and electron microscopes (Lieberman, 1971). Nerve damage also causes changes in the electrical
EXPERIMENTAL PROCEDURES
Young adult BALB/c mice of either sex, weighing 21^25 g, were used. The animals were anesthetized with pentobarbital (48 mg/kg). The right sciatic and saphenous nerves were exposed at mid-thigh level and a 5^10 mm segment of the nerves was dissected out to prevent regeneration. After 8^14 days the animals were killed by cervical dislocation. The DRG (L4, L5) were removed from both sides. Ganglia from the contralateral side and from unoperated animals were used as controls. In all cases we found that there was no regeneration of the cut nerve. The isolated ganglia were pinned onto a chamber with a Sylgard bottom. The chamber was superfused with Krebs solution at 23^24‡C that was bubbled with 95% O2 /5% CO2 . The chamber was placed on the stage of an upright microscope equipped with £uorescent illumination. Single cells were impaled with glass microelectrodes and injected with Lucifer Yellow (LY) as
*Corresponding author. Tel.: +972-2-5844721; fax: +972-25823515. E-mail address: [email protected] (M. Hanani). Abbreviations : DRG, dorsal root ganglia; LY, Lucifer Yellow; SC, satellite glial cell. 279
NSC 5720 15-8-02
280
M. Hanani et al.
Fig. 1. The morphology of cells in control DRG. (a) A single neuron in mouse DRG injected with LY. (b) SC in mouse DRG injected with LY. In control ganglia, such as this one, SCs were in many cases connected only to other SCs forming an envelope around a given neuron. (c) An example of a case where the SC was not coupled to other SCs. Scale bars = 20 Wm.
described previously (Hanani et al., 1999). The cells were viewed with a 40U water immersion objective, which allowed the observation and photography in real time of the staining of the cells and of the dye spread from the injected cell to adjacent ones. Images of stained cells were obtained using a digital camera (Pixera 120es). After ¢xation of the tissues in 4% phosphatebu¡ered paraformaldehyde (pH 7.4) the cells were imaged using a Bio-Rad confocal microscope. Fourteen days after axotomy four animals were perfused transcardially with a solution containing 2% formaldehyde and 2% glutaraldehyde in 0.1 M sodium cacodylate bu¡er (pH 7.3) under deep pentobarbital anesthesia (48 mg/kg). After ¢xation for 3 h, DRG L4 and L5 were removed from both sides and processed for electron microscopy. Thin sections (about 60 nm) were cut from ¢ve contralateral and ¢ve ipsilateral ganglia. These sections were examined under the electron microscope. The mean density of gap junctions was determined by measuring the number of gap junctions per 100 Wm2 of the section occupied by SCs. The mean length of gap junctions was also determined. The values obtained for the ganglia on each side were compared using a test for variance to establish whether they di¡ered signi¢cantly. The mean values for the contralateral ganglia were compared with those for the ipsilateral ganglia. The statistical comparisons employed the two-tailed Student’s t-test. Finally, a series of sections was cut from each of four ganglia (two contralateral and two ipsilateral to the axotomy). Each series consisted of about 130 consecutive sections. All the sections were
photographed under the electron microscope at a magni¢cation of 4000U and the negatives printed to a ¢nal magni¢cation of 16 000U. Serial sections were used to de¢ne connections between SC sheaths enveloping neighboring neurons.
RESULTS
LY was injected into single DRG neurons. In both control and axotomized ganglia this revealed that DRG neurons were not coupled to other neurons or to SCs (Fig. 1a). Injection of LY into single SCs in control ganglia showed that in many cases SCs were coupled to other SCs that formed the sheath around a given neuron (Fig. 1b). In 2.9% of the cases (6/206) there was also dye coupling among SCs that formed the sheath enveloping a given neuron and SCs surrounding one to four adjacent neurons. Eight to 14 days after axotomy, striking changes were observed in the DRG L4, L5. Injection of a single SC in contralateral DRG resulted in the staining of the SC sheaths surrounding neighboring neurons in only 6.4%
Fig. 2. The e¡ect of axotomy on SCs. (a) A SC injected with LY in an axotomized ganglion. This cell appears to be connected to other SCs enveloping adjacent neurons. (b) Histograms showing the incidence of inter-envelope coupling. (c) A SC in an axotomized ganglion showing a process emerging from the glial envelope, but not coupled to other SCs. The asterisks indicate the injected cells. Scale bars = 20 Wm.
NSC 5720 15-8-02
Glial plasticity in sensory ganglia
281
Fig. 3. Electron microscopic observations. (a) Control DRG; each nerve cell body (N) is enveloped by a SC sheath (sc), which is clearly separated from sheaths encircling the adjacent neurons by connective tissue (ct). Note that the perineuronal SC sheaths have a smooth contour, U10 000. (b) Axotomized DRG; process (arrows) arising from a perineuronal SC sheath (sc) projects into the connective tissue space. v: blood vessel. Scale bars = 2.5 Wm.
of the cases (8/125, eight animals). In contrast, on the ipsilateral side LY injection into single SCs resulted in 21.3% of the cases (37/150, eight animals; P 6 0.0005) in the staining of SCs surrounding up to 15 neighboring neurons (Fig. 2a). In both contralateral and ipsilateral ganglia SC dye coupling was completely blocked (in all
38 cases) by 1 mM octanol, indicating that it was mediated by gap junctions. The results are summarized in Fig. 2b. In several cases SCs in axotomized ganglia had distinct processes that projected into the connective tissue space (Fig. 2c). Such processes were never observed in control ganglia.
Fig. 4. Axotomized DRG. Electron micrographs of two sections from a series, separated by 1.4 Wm. A bridge (outlined) connecting the SC sheaths that envelop the neurons marked N is present in b, but not in a. The area outlined in b is shown at greater enlargement in the inset. Scale bars = 5 Wm.
NSC 5720 15-8-02
282
M. Hanani et al.
glia (0.360 Q 0.040 Wm and 0.361 Q 0.023 Wm, respectively; P 6 0.01).
DISCUSSION
Fig. 5. Axotomized DRG. An electron micrograph showing that the sheaths that envelop two neurons (N) are connected by a bridge (arrows). A gap junction is present in the bridge. Scale bar = 2 Wm. The outlined area is shown at greater enlargement in the inset. The arrowhead points to the gap junction. Scale bar = 0.5 Wm.
Electron microscopy showed that in control ganglia, the SC sheath enveloping a given neuron had a rather smooth outer contour and was distinctly separated from sheaths encircling adjacent neurons (Fig. 3a). In striking contrast, in the ipsilateral ganglia, processes of variable thickness often arose from the SC sheath, projecting into the connective tissue space (Fig. 3b). Examination of the serial sections from control ganglia never revealed any contacts between glial sheaths enveloping di¡erent neurons, whereas in ipsilateral ganglia the SC sheaths belonging to 76 neurons out of 258 (29.46%) were found to be connected by these processes (Fig. 4a, b). An enlargement of the area of contact between SC processes is presented in the inset in Fig. 4b. Fig. 5 is another example of a bridge connecting SC envelopes of two adjacent neurons. The inset in Fig. 5 shows the presence of a gap junction between the newly formed SC processes forming the bridge. Gap junctions were found in SC processes, and in all the bridges that were examined by serial sections; these gap junctions are new by de¢nition, as glial processes were absent in control ganglia. New gap junctions were also found within the perineuronal sheaths. The mean number of gap junctions per 100 Wm2 of the SC sheath sectional area was 6.5 times greater after axotomy (0.20 Q 0.04 in contralateral ganglia vs. 1.35 Q 0.11 in ipsilateral ganglia; P 6 0.01), whereas the mean length of individual gap junctions did not di¡er between ipsilateral and contralateral gan-
The results demonstrate that SCs in mouse DRG undergo profound changes after axotomy. Dye injections showed that coupling among SCs increased by more than seven-fold after axotomy. The electron microscopic observations support this ¢nding as they revealed that axotomy induced the formation of bridges that connect SC sheaths enveloping neighboring neurons. Moreover, the number of gap junctions between SCs increased by 6.5 times after axotomy, which correlates with the increase in the incidence of dye coupling among SCs. In normal ganglia SCs are frequently coupled, but the coupling is restricted in most cases to SCs enveloping a given neuron. In control ganglia we found inter-envelope coupling in a small proportion of cases, which probably re£ects the fact that nerve cell bodies can occur in clusters having SC sheaths in common (Pannese et al., 1991). The number of such clusters is high in young animals, and declines with age. As we used young adult mice the proportion of coupling between SCs around di¡erent neurons is in agreement with the presence of such clusters. The correlation between age and cell coupling is interesting in the present context because it has been shown (Fulton, 1995) that damage induces changes in the nervous system that resemble early stages of development. Thus it was shown that both after axotomy and during development DRG neurons are characterized by an eccentric nucleus, many mitochondria concentrated in the central region of the perikaryon, a profusion of free polysomes and only a few polysomes attached to endoplasmic reticulum membranes (Pannese, 1963). We believe that this is the ¢rst demonstration that nerve injury leads to changes in glial coupling. The involvement of glial cells in various pathological processes in the CNS is well established (Ridet et al., 1997; Aldskogius and Kozlova, 1998). Axotomy of central neurons induces changes in astrocytes, which are referred to as ‘glial activation’ and consist of glial hypertrophy and increased production of molecules such as cytokines and glial ¢brillary acidic protein (Graeber and Kreuzberg, 1986; Rohlmann et al., 1994; Ridet et al., 1997; Pastor et al., 2000). The functional signi¢cance of these observations is not clear, but it has been proposed that glial activation may support the survival of neurons, as activated astrocytes can release growth factors (Ridet et al., 1997; Pastor et al., 2000). Nerve injury causes an increase in the immunoreactivity of the protein connexin 43, which forms the gap junctions among astrocytes (Rohlmann et al., 1994). However, from other studies it appears that the e¡ects of injury on glial connexins are regulated by several processes, and that the overall outcome is still not clear (Nagy and Rash, 2000). The functional implications of the changes observed in glial connexin levels after injury is not known. We demonstrated both morphologically and functionally that axotomy increased glia^glia coupling in the mouse DRG
NSC 5720 15-8-02
Glial plasticity in sensory ganglia
and thus provided direct evidence for glial plasticity after nerve damage. The contribution of these changes to pathological states remains to be explored. Raisman (1991) suggested that glial cells respond to injury as a large, interconnected network. As normally DRG glial cells around di¡erent neurons are rarely coupled, the increased coupling after axotomy may endow these cells with the ability to communicate over long distances. Glial plasticity is part of CNS response to pathological stimuli such as dehydration (Bobak and Salm, 1996) or cortical lesion (Hailer et al., 1999). We propose that augmented glia^glia coupling may be an important mode of glial contribution to plasticity in the nervous system. It will be worthwhile to search for such changes in the CNS under a variety of pathological conditions.
283
The ¢ndings above may have interesting implications for the understanding of chronic pain that results from nerve injury. One prominent feature in neuropathic pain is the ‘spread’ of pain sensation from the locus of the injured nerves to areas where the nerves were apparently intact (Devor and Seltzer, 1999). We showed that axotomy increased SC^SC communication by gap junctions. Gap junctions allow the passage of small ions and of molecules with a molecular weight of up to about 1000 Da. Thus it can be postulated that axotomy enhances the spread of electrical currents and/or metabolites (e.g., second messengers) within the DRG. We propose that this enhanced propagation of signals in the DRG may contribute to the mechanisms of neuropathic pain.
REFERENCES
Aldskogius, H., Kozlova, E.N., 1998. Central neuron-glia and glial-glial interactions following axon injury. Prog. Neurobiol. 55, 1^26. Amir, R., Michaelis, M., Devor, M., 1999. Membrane potential oscillations in dorsal root ganglion neurons: role in normal electrogenesis and neuropathic pain. J. Neurosci. 19, 8589^8596. Becker, D.L., Cook, J.E., 1990. Changes in gold¢sh retinal ganglion cells during axonal regeneration. Proc. R. Soc. London B 241, 73^77. Bobak, J.B., Salm, A.K., 1996. Plasticity of astrocytes of the ventral glial limitans subjacent to the supraoptic nucleus. J. Comp. Neurol. 376, 188^ 197. Chang, Q., Pereda, A., Pinter, M.J., Balice-Gordon, R.J., 2000. Nerve injury induces gap junctional coupling among axotomized adult motor neurons. J. Neurosci. 20, 674^684. Devor, M., Seltzer, R., 1999. In: Wall, P.D., Melzack, R. (Eds.), Textbook of Pain, 4th edn. Churchill Livingstone, London, pp. 129^164. Fulton, B.P., 1995. Gap junctions in the developing nervous system. Perspect. Dev. Neurobiol. 2, 327^334. Graeber, M.B., Kreuzberg, G.W., 1986. Astrocytes increase in glial ¢brillary acidic protein during retrograde changes of facial motor neurons. J. Neurocytol. 15, 363^373. Hailer, N.P., Grampp, A., Nitsch, R., 1999. Proliferation of microglia and astrocytes in the dentate gyrus following entorhinal cortex lesion: a quantitative bromodeoxyuridine labeling study. Eur. J. Neurosci. 11, 3359^3364. Hanani, M., Maudlej, N., Ha«rtig, W., 1999. Morphology and intercellular communication in glial cells of the intramural ganglia from the guineapig urinary bladder. J. Auton. Nerv. Syst. 76, 62^67. LaMotte, R.H., Zhang, J.M., Petersen, M., 1996. Alterations in the functional properties of dorsal root ganglion cells with unmyelinated axons after a chronic nerve constriction in the rat. Prog. Brain Res. 110, 105^111. Lieberman, A.R., 1971. The axon reaction : a review of the principal features of perikaryal responses to axon injury. Int. Rev. Neurobiol. 14, 49^ 124. Nagy, J.I., Rash, J.E., 2000. Connexins and gap junctions of astrocytes and oligodendrocytes in the CNS. Brain Res. Rev. 32, 29^44. Pannese, E., 1963. Investigations on the ultrastructural changes of the spinal ganglion neurons in the course of axon regeneration and cell hypertrophy. I. Changes during axon regeneration. Z. Zellforsch. 60, 711^740. Pannese, E., 1981. The satellite cells of sensory ganglia. Adv. Anat. Embryol. Cell Biol. 65, 1^111. Pannese, E., Ledda, M., Arcidiacono, G., Rigamonti, L., 1991. Clusters of nerve cell bodies enclosed within a common connective tissue envelope in the spinal ganglia of the lizard and rat. Cell Tissue Res. 264, 209^214. Pastor, A.M., Delgado-Garc|¤a, J.M., Mart|¤nez-Guijarro, F.J., Lo¤pez-Garc|¤a, J.M., de la Cruz, R.R., 2000. Response of abducens internuclear neurons to axotomy in the adult cat. J. Comp. Neurol. 427, 370^390. Raisman, G., 1991. Glia, neurons, and plasticity. Ann. NY Acad. Sci. 633, 209^213. Ridet, J.L., Malhotra, S.K., Privat, A., Gage, F.H., 1997. Reactive astrocytes: cellular and molecular cues to biological function. Trends Neurosci. 20, 570^577. Rohlmann, A., Laskawi, R., Hofer, A., Dermietzel, R., Wol¡, J.R., 1994. Astrocytes as rapid sensors of peripheral axotomy in the facial nucleus of rats. NeuroReport 5, 409^412. Stephenson, J.L., Byers, M.R., 1995. GFAP immunoreactivity in trigeminal ganglion satellite cells after tooth injury in rats. Exp. Neurol. 131, 11^ 22. Woodham, P., Anderson, P.N., Nadim, W., Turmaine, M., 1989. Satellite cells surrounding axotomized rat dorsal root ganglion cells increase expression of GFAP-like protein. Neurosci. Lett. 98, 8^12. Zhang, J.M., Donnelly, D.F., Song, X.J., LaMotte, R.H., 1997. Axotomy increases the excitability of dorsal root ganglion cells with unmyelinated axons. J. Neurophysiol. 78, 2790^2794. (Accepted 24 May 2002)
NSC 5720 15-8-02