- Email: [email protected]
Flat, whole-mount nerve preparations: A useful tool for studying the process of regenerating axon outgrowth
Journal of Neuroscience Methods, 9 (1983) 205-216
205
Elsevier
Research Papers
Flat, whole-mount nerve preparations" a useful tool for studying the process of regenerating axon outgrowth Yuri G e i n i s m a n a n d Michael T. Shipley Department of Cell Biology and Anatomy, Northwestern University Medical School, Chicago, IL 60611 and Department of Anatomy and Cell Biology, University of Cincinnati, College of Medicine, Cincinnati, OH 45267 (U. S.A.) (Received April l l t h , 1983) (Revised July 18th, 1983) (Accepted August 2nd, 1983)
Key words: nerve regeneration--flat, whole-mount nerve preparations A method, which is based on the use of fiat, whole-mount nerve preparations, has been developed for studying the process of regenerating axon outgrowth, employing the rat sciatic nerve as a model. At various intervals after a nerve crush, animals are perfused with aldehyde fixatives, the nerve dissected out, and its epineurium removed. Next the nerve is flattened between two glass slides, removed and reacted (floating), then whole-mounted on a micro slide and cover-slipped. Regenerating axons have been labeled by means of the horseradish peroxidase tracing technique, a histochemical technique for acetylcholinesterase, or an indirect immunocytochemical technique utilizing antibodies against tubulin. With all these techniques, individual outgrowing axons and their bundles can be clearly visualized. Regenerating axons labeled by horseradish peroxidase are readily traced along their entire undulating courses from the distal margin of the crush zone to axonal tips, which mark the leading edge of several waves of outgrowing axons. It appears that such flat, whole-amount nerve preparations can be useful for obtaining: (1) accurate estimates of the rate of regenerating axon elongation, (2) values characterizing the duration of the initial delay of axonal outgrowth, and (3) information concerning the nature of axonal subpopulations that elongate at different rates.
Introduction
The process of axonal outgrowth distal to the site of an axotomizing injury is an essential part of neuronal regeneration, and the rate of elongation of regenerating axons appears to reflect the varying potential of neurons to regenerate after axonal lesion (Grafstein and McQuarrie, 1978; Lasek et al., 1981). Obtaining accurate estimates of the rate of regenerating axon elongation under various experimental conditions is, therefore, an essential step in the search for factors that limit or promote neuronal regeneration. A number of techniques are available for measuring the elongation rate of regenerating axons: light microscopy of silver impregnated preparations and electron 0165-0270/83/$03.00 © 1983 Elsevier Science Publishers B.V.
206
microscopy for locating the advancing tips of outgrowing axons (Orgel, 1980): fast axonal transport of radioactive proteins for labeling the advancing growing cones (Black and Lasek, 1976; Griffin et al., 1976; Bisby, 1978; Forman and Berenberg, 1978); and the pinch test for locating the advancing tips of the fastest growing sensory axons (Young and Medwar, 1940; Berenberg et al., 1977; McQuarrie et al., 1977). However, being mainly useful for the detection of the 'front' of regenerating axons, these techniques have serious limitations. None of them allow the visualization and measurement of the length of outgrowing axons along their entire courses. In light microscopic and, even more so, in electron microscopic preparations, different segments of a particular regenerating axon are usually found in different sections, and the entire axon course should be reconstructed from a substantial number of sections. This task is extremely difficult, if not unfeasible. The axonal transport and pinch test techniques are also clearly unsuitable for such a purpose. Related to this disadvantage is another limitation of the available techniques: they do not permit the differential direct measurement of outgrowth distances for subpopulations of regenerating axons that elongate at fast. intermediate or slow rates. Nevertheless, the axonal transport technique offers the opportunity to indirectly infer the elongation rate of various axonal subpopulations from the distribution of radioactivity in regenerating nerves, and it has been strongly suggested by such analyses that a selective change in the outgrowth rate can involve only one of several axonal subpopulations. For instance, the 'conditioning lesion' effect (an acceleration of axonal outgrowth following a previous axotomizing lesion) appears to be characteristic of the subpopulation of regenerating motor axons in the rat sciatic nerve that elongate at intermediate rates, whereas the outgrowth rate of faster and slower regenerating axons remains unchanged (McQuarrie, 1981). A slowing of the elongating rate of regenerating motor axons in the rat sciatic nerve during aging appears to be typical of the axonal subpopulations that grow at intermediate and slow rates, while the fastest growing axons do not change their rate of elongation in aged animals (Pestronk et al., 1980). It is obvious, therefore, that an adequate characterization of the process of regenerating axon outgrowth cannot be achieved without differential measurements of the elongation rate for various axonal subpopulations and an analysis of the nature of axons comprising these subpopulations. To circumvent the limitations of the techniques available for studying the process of regenerating axon outgrowth, we have developed a method which is based on the use of flat, whole-mount nerve preparations. In such preparations, axons belonging to various subpopulations of regenerating fibers can be readily visualized along their entire undulating course when labeled by horseradish peroxidase (HRP), and the actual distance of regenerating axon outgrowth can be accurately measured over time after axotomy. Moreover, histochemical and immunocytochemical techniques can also be applied to such preparations, providing resolution at the level of individual regenerating axons. This may be valuable for obtaining information concerning the nature of regenerating axon subpopulations that elongate at different rates.
207 Materials and Methods
Male Holtzman rats (70-80 days of age, 300-350 g body weight) were used. All surgical procedures were done under chloral hydrate anesthesia (300 m g / k g i.p.). The right sciatic nerve (SN) was exposed in the gluteal region and crushed at the point where it crosses the fused tendon of the obturator internus and gemelli muscles. Two 45 s crushes were made at a 60 s interval with No. 5 jeweler's forceps sheathed by polyethylene tubing. At various time intervals after the crush, animals were sacrificed by intracardiac perfusion, the right SN dissected out, and the epineurium covering the SN trunk and its tibial and peroneal divisions removed. The SN divisions were flattened between two glass slides with the help of a rubber roller, after which the slides were pressed together with a flat tubing clamp. The slide 'sandwich' containinj~, the nerve was left in a cold postfixation buffer perfusate for 2 - 4 h. The flat whole-nerve preparations obtained in this manner were then removed from the 'sandwich' and reacted, while floating, in appropriate media. For the HRP tracing technique, the enzyme was injected into the region of the cell bodies of spinal motoneurons that innervate the musculature of the right hind limb, following laminectomy of T13, L1 and L2 vertebrae. To determine the rostro-caudal extent of the motoneuronal pool innervating the hind limb musculature, the procedure developed by Forman and Berenberg (1978) was used. Injections of 0.1-0.2 ~liter saline containing 10 mM sodium glutamate were placed into the ventral cell column on the left side of the spinal cord, and contractions of muscles of the left hind limb indicated the rostro-caudal level of the spinal cord at which HRP injections had to be performed. At this identified spinal cord level, 6 injections of H R P solution (40% Sigma type VI HRP in saline containing 2% dimethyl sulfoxide) were made at 1 mm intervals. HRP injections were performed with a glass micropipette attached to a micromanipulator. To place HRP into the ventral horn region, the micropipette tip (60/am diameter) was positioned 1 mm lateral to the midline and 1.6 mm below the dura (Lasek, 1968; Griffin et al., 1976; Forman and Berenberg, 1978). The HRP solution was injected at a rate of 0.1 /~liter/45 s by means of a 2 #liter Hamilton syringe whose needle was coupled by polyethylene tubing to the micropipette (the system being filled with mineral oil) and whose plunger was driven by a Sage 341 syringe pump. Although injections of 0.05-0.5 /xliter of H R P solution appeared to provide quite similar patterns of labeling of SN axons, the relatively large amount of 0.3/~liter per injection was chosen to ensure the labeling of the majority of motoneurons. Animals were perfused 16-24 h after HRP injections. Perfusion and HRP reaction procedures were essentially those of Mesulam's tetramethylbenzidine method as described by Warr et al. (1981), with the following exceptions: postfixation perfusate was 0.1 M phosphate buffer, pH 7.4, and 0.75% gelatin was added to the preincubation and incubation solutions. Following the reaction, flat nerves were whole-mounted on glass slides subbed with gelatin-chrome alum, left to air dry for 24 h, counterstained with neutral red (2.5 min), rinsed in distilled water (15 s), dehydrated in 70%, 95% and two changes of 100% ethanol (15 s each), cleared in methyl salicylate (two changes, 1 and 5 min), washed in xylene and coverslipped with Permount. In such preparations, the
208 location of the distal margin of the crush zone and tips of the most advanced axons was marked on the coverslip under a microscope and measured with a caliper. To demonstrate acetylcholinesterase (ACHE) activity, animals were perfused with 50 ml of saline at room temperature, followed by ! liter of fixative (1% paraformaldehyde-l.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4) for 30 min at room temperature and 1 liter of the same buffer for 30 min at 4°C. Flat nerves were reacted for ACHE, using the Koelle copper thiocholine method as modified by Geneser-Jensen and Blackstad (1971). The nerves were then whole mounted on subbed glass slides, left to air dry for 24 h, dehydrated in 50, 70, 96 and 100% ethanol (2 min each), cleared in xylene (two changes, 5 min each) and coverslipped with Permount. For localizing tubulin, rats were perfused with 50 ml phosphate buffer -saline (PBS) solution (0.9% NaCI in 0.1 M phosphate buffer, pH 7.4) at room temperature, followed by 1 liter of fixative (4% paraformaldehyde -0.5% glutaraldehyde in PBS) for 30 min at room temperature and 1 liter of PBS for 30 min at 4°C. Flat nerves were reacted by means of a two-step ('sandwich'), indirect immunocytochemical technique involving covalent bonding (Pickel, 1981). A primary antiserum produced in a rabbit against tubulin was used at 1:100 dilution. Preparations were stained with 4-chlor-l-naphthol, whole-mounted and coverslipped with 90% glycerol/PBS.
Results Using the HRP tracing technique, it was possible to demonstrate the dynamics of regenerating axon outgrowth. At 16 h after the crush, virtually all HRP labeled axons were seen proximal to the crush site (Fig. 1). By 2 days after the crush, many labeled axons were observed crossing the crush zone and passing peripheral to it (Fig. 2A), the distance of regenerating axon elongation being progressively increased at 4 (Fig. 2B) and 6 (Fig. 2C) days post-lesion. The crush zone was well-defined in these preparations, especially under lower magnifications, due to an increased background staining by neutral rad. Being relatively moderate in the portion of the SN central to the crush, the intensity of regenerating axon labeling by HRP was markedly increased, starting just proximal to the crush zone and continuing all the distance down to the tips of outgrowing axons. Individual regenerating axons, as well as axonal bundles could be easily traced (while changing focal planes) along their entire undulating course from the distal border of the crush zone to the axonal tips. Axonal tips marked the leading edge of several waves of outgrowing axons most noticeable in Fig. 2B. The leading edge could be readily distinguished, since many axonal tips had larger dimensions than their parent axons and formed elongated bulbs filled with the H R P reaction product (Fig. 3A, B). Examination of flat nerve preparations at higher magnifications, such as used in Fig. 3, showed that a limited number of HRP labeled axons of apparently smaller size extended distally to the prominent waves of outgrowing axons which were detected at lower magnifications and demonstrated in Fig. 2. These fine, most advanced axons (fig. 3A) were
209
Fig. 1. Flat, whole-mount preparation of the tibial division of the rat sciatic nerve, demonstrating the pattern of axonal labeling at 16 h after the crush by the HRP tracing technique. The crush zone is indicated by the arrowhead, the nerve segment proximal to the crush is on the left of the arrowhead, x 50.
designated as the fastest growing fibers. Peripheral to the fastest growing fibers, labeled axons were not seen in any preparation examined, indicating that the crush paradigm used completely axotomized the SN. With the defined subpopulation of the fastest growing fibers, it became possible to estimate the m a x i m u m outgrowth distance and m a x i m u m elongation rate of regenerating axons in the rat SN. For this purpose, rats were sacrificed 4 and 6 days after the SN crush, 7 animals at each post-lesion time. The distance between the peripheral margin of the crush zone and the most advanced tips of the fastest growing axons labeled with H R P was measured along the tibial division of the SN. This m a x i m u m axon outgrowth distance (mean _ S.E.M.) was f o u n d to be 8.2 _+ 0.2 m m (range of 7.5-9.2 mm) at 4 days and 16.2 _+ 0.4 m m (range of 14.4-17.7 mm) at
210
~iiI¸¸Ii
Fig. 2. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating the dynamics of outgrowth of regenerating H R P labeled axons at 2 days (A), 4 days (B) and 6 days (C) after the crush. The distal margin of the crush zone is indicated by arrowheads. Calibration bar = 1 mm.
6 days after the crush. Thus, the most rapidly advancing axons grew 8 mm in 2 days with the maximum outgrowth rate of 4 m m / d a y . Additional experiments were performed to verify whether or not the length of the SN was changed due to the preparative procedures used. The SN was exposed along its course, and two 11-0 surgical sutures were passed through the SN perpendicular to its long axis, separated by an interval of 12-18 mm. The distance between the sutures turned out to be virtually the same when measured in living animals, immediately following the perfusion in situ, after the SN dissection and epineurium removal, following the SN. flattening, or after the H R P reaction, dehydration, mounting and coverslipping. The results of SN length measurements in living animals as compared to all other conditions did not vary more than + 0.5 mm. Using the histochemical technique for ACHE, the AChE activity at 3 h after the crush was found in axonal segments located in the SN portions proximal and distal to the crush, with an increased reaction intensity near the crush site (Fig. 4A). Blood vessels crossed the crush zone at this post-lesion time, but single structures resem-
211
Fig. 3. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating HRP labeled axonal tips as they advanced peripherally (to the right) at various distances by 4 days after the crush (A), many of the axonal tips being enlarged compared to parent axons and filled with the HRP reaction product (B). Dark-field illumination in polarized light, x 190 (A); × 1800 (B).
212
Fig. 4. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating the patterns of axonal labeling at 3 h (A) and 2 days (B) after the crush by means of histochemical technique for acetylchohnesterase. The distal margin of the crush zone is indicated by arrowheads. × 36.
213
Fig. 5, Flat, whole-mount preparation of the tibial division of the rat sciatic nerve, demonstrating the pattern of axonal labeling at 7 days after the crush by means of indirect immunocytochemicaltechnique for localizing tubulin. The distal margin of the crush zone is indicated by the arrowhead. × 16.5.
bling AChE-positive axons were only occasionally observed within this zone in some preparations. By 2 days after the crush, moderate AChE activity was present in a large number of axons along the SN portion proximal to the crush; approaching the crush site, regenerating axons gave a strong positive reaction for ACHE, and they could be traced more peripherally, crossing the crush zone, passing distal to it and ending as enlarged tips (Fig. 4B). Distal to the most advanced axonal tips, no AChE activity was present in axons. Although individual axons were well resolved by this technique, their entire course was difficult to follow due to a fading of staining intensity in the proximity of axonal tips. When SN preparations were incubated without acetylthiocholine, AChE activity was entirely absent. Using the immunocytochemical indirect 'sandwich' technique, distinct and reproducible patterns of immunoperoxidase staining with antibodies against tubulin were demonstrated. Individual regenerating axons and their bundles were intensely stained against the background in the SN portions proximal and distal to the crush (Fig. 5). In the crush zone, however, the axons were not stained, even by 7 days after the crush (Fig. 5). Enlarged axonal tips were never observed, and it was difficult to trace the entire course of individual axons in immunostained nerve preparations. The immunoperoxidase staining of regenerating axons with the antiserum against tubulin appeared to be specific, since no such staining was detected in control SN preparation treated with the preimmune serum.
214
Discussion The method of fiat, whole-mount nerve preparations can be useful for obtaining accurate estimates of the elongation rate of regenerating axons, since it allows the axon outgrowth distance to be precisely assessed by means of several alternative approaches: First, the method developed by us offers a unique opportunity to measure the absolute length of outgrowing individual axons labeled by HRP. This can provide an actual measure of the outgrowth distance for regenerating axons. With the available experimental techniques, the extent of outgrowing axon elongation is estimated as the distance between two points along a nerve: the site of an axotomizing lesion and the region of advancing axonal tips. An undulating course of axons within a nerve is not taken into account by such two-point measurements, which may result in a systematic underestimation of outgrowth distances covered by regenerating axons. Second, traditional two-point measurements of axon outgrowth distances can also be reliably performed in flat, whole-amount nerve preparations, using the H R P tracing technique. Compared to other current techniques (fast axonal transport and pinch test techniques), the described method has the advantage of an unequivocal precision in locating both the distal margin of the zone of axotomy and advancing axonal tips. The accuracy of two-point measurements in flat, whole-mount nerve preparations appears to be unaffected by the preparative procedures, which cause no appreciable distortion of linear nerve dimensions. Finally, a direct differential analysis of outgrowth distances for subpopulations of regenerating neurons that elongate at fast, intermediate or slow rates can now be carried out in flat, wholemount nerve preparations, using either length or two-point measurements. It could be argued, however, that H R P might be transported to the terminal, newly formed parts of regenerating axons in an amount insufficient to render them detectable at the light microscopic level. Since the technique of fast axonal transport of radioactive proteins labels the advancing cones of regenerating axons (Griffin et al., 1976), it seems appropriate to compare the available results obtained by this technique with the data derived from fiat, whole-mount nerve preparations labeled with HRP. The maximum axon outgrowth distance, as estimated in flat nerve preparations, is 8.2 and 16.2 m m at 4 and 6 days after SN crush, respectively, yielding a maximum axon elongation rate of 4 m m / d a y . These values correspond well to those obtained with the fast axonal transport technique: a maximum outgrowth distance of 16 m m by 6 days after a crush (Forman and Berenberg, 1978) and a maximum elongation rate of 3.6-4.5 m m / d a y (Grafstein and McQuarrie, 1978; Lasek et al., 1981) have been reported for regenerating motor axons in the SN of young adult rats. Such a n agreement of the results strongly suggests that H R P transport under the experimental conditions used in this work extends along regenerating axons to the level indicated as the front of outgrowth by rapidly transported radioactive proteins. Another valuable application of the method reported here is that it permits the direct evaluation of the so-called initial delay of regenerating axon outgrowth. This delay includes the time required for the neuron to recover from axotomy, for axon growth to commence, and for the outgrowing axon to reach and traverse the zone of
215 axotomy (Sunderland, 1978). By using the available techniques, it is only possible to infer the duration of initial delay by backward extrapolation of the axon outgrowth values to zero distance (Grafstein and McQuarrie, 1978). With the help of fiat, whole-mount nerve preparations, the duration of initial delay can be directly ascertained as the time necessary for HRP labeled regenerating axons to reach the nerve segment just distal to the lesion site. The method described here has an additional promising potential related to the fact that both histochemical and immunocytochemical techniques can be applied to flat, whole-mount nerve preparations. Our experiments with the histochemical technique for AChE and immunocytochemical technique for tubulin detection have demonstrated that individual regenerating axons can be clearly visualized in such nerve preparations using both techniques. This makes it feasible to obtain information concerning the nature of axonal subpopulations that elongate at different rates. Such a goal seems to be attainable, since any staining technique applicable to floating or mounted histological sections should work well in these nerve preparations, and selected segments of flattened nerves can be examined by means of electron microscopy. However, the method reported here also may have certain limitations. A major potential disadvantage of this method, which involves light microscopy, is that it may not provide a resolution sufficient enough to visualize the finest regenerating axons. Due to this, errors may be introduced in the detection of the leading edge of outgrowing axons. Despite the fact that the front of outgrowth of HRP labeled axons can be located rather precisely in fiat, whole-amount preparations of the rat SN (as indicated by the above comparison of our results with those obtained with the fast axonal transport technique), this may not be the case for systems other than the rat SN. In other systems, the position of the leading edge of regenerating axons as determined in fiat, whole-mount preparations should be verified by means of the fast axonal transport technique a n d / o r electron microscopy. Another potential disadvantage of this method, using the HRP tracing technique, is peculiar for the SN system, in which not only motor, but also sensory axons may be labeled after HRP injections into the spinal cord. Further experiments with sectioning of the appropriate ventral roots should elucidate whether or not HRP labeling of sensory axons takes place. In any case, a differential analysis of SN motor and sensory axons is achievable. HRP injections into the spinal cord can be made after dorsal root sectioning for selective labeling of SN motor axons, and HRP can be injected into the dorsal root ganglia for selective labeling of SN sensory axons. It appears, therefore, that the method of fiat, whole-mount nerve preparations has the potential to become a major tool for studying the process of regenerating axon outgrowth.
Acknowledgements The authors would like to express their gratitude to Dr. R.A. Berenberg for valuable advice concerning the techniques of laminectomy and spinal cord injections, Dr. R.E. Heller for the skilled assistance in operative procedures, Dr. R.D.
216 G o l d m a n for the g e n e r o u s supply of the antisera used a n d Dr. R.W. Berry for helpful criticism of the manuscript. This work was supported by grant N1H AG-03410, NIH 19730 a n d U S A M R D S 1782C2272.
NS-18490, NIH
NS-
References Berenberg, R.A., Forman, D.S., Wood, D.K., DeSilva, A. and Demaree, J. (1977) Recovery of peripheral nerve function after axotomy: effect of triiodothyronine, Exp. Neurol., 57: 349-363. Bisby, M.A. (1978) Fast axonal transport of labeled protein in sensory axons during regeneration, Exp. Neurol., 61: 281-300. Black, M.M. and Lasek, R.J. (1976) The use of axonal transport to measure axonal regeneration in rat ventral motor neurons, Anat. Rec., 184: 360-361. Forman, D.S. and Berenberg, R.A. (1978) Regeneration of motor axons in the rat sciatic nerve studied by labelling with axonally transported radioactive proteins, Brain Res., 156: 213-225. Geneser-Jensen, F.A. and Blackstad, T.W. (1971) Distribution of acetyl cholinesterase in the hippocampal region of the guinea pig. I. Entorhinal area, parasubiculum and presubiculum. Z. Zellforsch., 114: 460-481. Grafstein, B. and McQuarrie, I.G. (1978) Role of the nerve cell body in axonal regeneration. In C.W. Cotman (Ed.), Neuronal Plasticity, Raven Press, New York, pp. 155 195. Griffin, J.W., Drachman, D.B. and Price, D.L. (1976) Fast axonal transport in motor nerve regeneration. J. Neurobiol., 4: 355-370. Lasek, R.J. (1968) Axoplasmic transport of labelled proteins in rat ventral motoneurons, Exp. Neurol., 21 : 41-51. Lasek, R.J., McQuarrie, I.G. and Wujek, J.R. (1981) The central nervous system regeneration problem: neuron and environment. In A. Gorio, H. Millesi and S. Mingrino (Eds.), Posttraumatic Peripheral Nerve Regeneration: Experimental Basis and Clinical Implications, Raven Press, New York, pp. 59-70. McQuarrie, I.G. (1981) Acceleration of axonal regeneration in rat somatic motoneurons by using a conditioning lesion. In A. Gorio, H. Millesi and S. Mingrino (Eds.), Posttraumatic Peripheral Nerve Regeneration: Experimental Basis and Clinical Implications, Raven Press, New York, pp. 49-58. McQuarrie, I.G., Grafstein, B. and Gershon, M.D. (1977) Axonal regeneration in the rat sciatic nerve: effect of a conditioning lesion and of dbcAMP, Brain Res., 132: 443-453. Orgel, M.G. (1980) A critical review of histological methods used in the study of nerve regeneration. In D.L. Jewett and H.R. McCarroll, Jr. (Eds.), Nerve Repair and Regeneration. Its Clinical and Experimental Basis, The C.V. Mosby Company, St. Louis, pp. 141-148. Pestronk, A., Drachman, D.B. and Griffin, J.W. (1980) Effects of aging on nerve sprouting and regeneration, Exp., Neurol., 70: 65-82. Pickel, V.M. (1981) Immunocytochemical methods. In L. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum Press, New York, pp. 483-509. Sunderland, S. (1978) Nerves and Nerve Injuries. Churchill Livingstone, Edingburgh, 1046 pp. Warr, W.B., de Olmos, J.S. and Heimer, L. (1981) Horseradish peroxidase: the basic procedure. In L. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum Press, New York. pp. 207-262. Young, J.Z. and Medawar, P.B. (1940) Fibrin suture of peripheral nerves. Measurement of the rate of regeneration, Lancet, ii: 126-128.
205
Elsevier
Research Papers
Flat, whole-mount nerve preparations" a useful tool for studying the process of regenerating axon outgrowth Yuri G e i n i s m a n a n d Michael T. Shipley Department of Cell Biology and Anatomy, Northwestern University Medical School, Chicago, IL 60611 and Department of Anatomy and Cell Biology, University of Cincinnati, College of Medicine, Cincinnati, OH 45267 (U. S.A.) (Received April l l t h , 1983) (Revised July 18th, 1983) (Accepted August 2nd, 1983)
Key words: nerve regeneration--flat, whole-mount nerve preparations A method, which is based on the use of fiat, whole-mount nerve preparations, has been developed for studying the process of regenerating axon outgrowth, employing the rat sciatic nerve as a model. At various intervals after a nerve crush, animals are perfused with aldehyde fixatives, the nerve dissected out, and its epineurium removed. Next the nerve is flattened between two glass slides, removed and reacted (floating), then whole-mounted on a micro slide and cover-slipped. Regenerating axons have been labeled by means of the horseradish peroxidase tracing technique, a histochemical technique for acetylcholinesterase, or an indirect immunocytochemical technique utilizing antibodies against tubulin. With all these techniques, individual outgrowing axons and their bundles can be clearly visualized. Regenerating axons labeled by horseradish peroxidase are readily traced along their entire undulating courses from the distal margin of the crush zone to axonal tips, which mark the leading edge of several waves of outgrowing axons. It appears that such flat, whole-amount nerve preparations can be useful for obtaining: (1) accurate estimates of the rate of regenerating axon elongation, (2) values characterizing the duration of the initial delay of axonal outgrowth, and (3) information concerning the nature of axonal subpopulations that elongate at different rates.
Introduction
The process of axonal outgrowth distal to the site of an axotomizing injury is an essential part of neuronal regeneration, and the rate of elongation of regenerating axons appears to reflect the varying potential of neurons to regenerate after axonal lesion (Grafstein and McQuarrie, 1978; Lasek et al., 1981). Obtaining accurate estimates of the rate of regenerating axon elongation under various experimental conditions is, therefore, an essential step in the search for factors that limit or promote neuronal regeneration. A number of techniques are available for measuring the elongation rate of regenerating axons: light microscopy of silver impregnated preparations and electron 0165-0270/83/$03.00 © 1983 Elsevier Science Publishers B.V.
206
microscopy for locating the advancing tips of outgrowing axons (Orgel, 1980): fast axonal transport of radioactive proteins for labeling the advancing growing cones (Black and Lasek, 1976; Griffin et al., 1976; Bisby, 1978; Forman and Berenberg, 1978); and the pinch test for locating the advancing tips of the fastest growing sensory axons (Young and Medwar, 1940; Berenberg et al., 1977; McQuarrie et al., 1977). However, being mainly useful for the detection of the 'front' of regenerating axons, these techniques have serious limitations. None of them allow the visualization and measurement of the length of outgrowing axons along their entire courses. In light microscopic and, even more so, in electron microscopic preparations, different segments of a particular regenerating axon are usually found in different sections, and the entire axon course should be reconstructed from a substantial number of sections. This task is extremely difficult, if not unfeasible. The axonal transport and pinch test techniques are also clearly unsuitable for such a purpose. Related to this disadvantage is another limitation of the available techniques: they do not permit the differential direct measurement of outgrowth distances for subpopulations of regenerating axons that elongate at fast. intermediate or slow rates. Nevertheless, the axonal transport technique offers the opportunity to indirectly infer the elongation rate of various axonal subpopulations from the distribution of radioactivity in regenerating nerves, and it has been strongly suggested by such analyses that a selective change in the outgrowth rate can involve only one of several axonal subpopulations. For instance, the 'conditioning lesion' effect (an acceleration of axonal outgrowth following a previous axotomizing lesion) appears to be characteristic of the subpopulation of regenerating motor axons in the rat sciatic nerve that elongate at intermediate rates, whereas the outgrowth rate of faster and slower regenerating axons remains unchanged (McQuarrie, 1981). A slowing of the elongating rate of regenerating motor axons in the rat sciatic nerve during aging appears to be typical of the axonal subpopulations that grow at intermediate and slow rates, while the fastest growing axons do not change their rate of elongation in aged animals (Pestronk et al., 1980). It is obvious, therefore, that an adequate characterization of the process of regenerating axon outgrowth cannot be achieved without differential measurements of the elongation rate for various axonal subpopulations and an analysis of the nature of axons comprising these subpopulations. To circumvent the limitations of the techniques available for studying the process of regenerating axon outgrowth, we have developed a method which is based on the use of flat, whole-mount nerve preparations. In such preparations, axons belonging to various subpopulations of regenerating fibers can be readily visualized along their entire undulating course when labeled by horseradish peroxidase (HRP), and the actual distance of regenerating axon outgrowth can be accurately measured over time after axotomy. Moreover, histochemical and immunocytochemical techniques can also be applied to such preparations, providing resolution at the level of individual regenerating axons. This may be valuable for obtaining information concerning the nature of regenerating axon subpopulations that elongate at different rates.
207 Materials and Methods
Male Holtzman rats (70-80 days of age, 300-350 g body weight) were used. All surgical procedures were done under chloral hydrate anesthesia (300 m g / k g i.p.). The right sciatic nerve (SN) was exposed in the gluteal region and crushed at the point where it crosses the fused tendon of the obturator internus and gemelli muscles. Two 45 s crushes were made at a 60 s interval with No. 5 jeweler's forceps sheathed by polyethylene tubing. At various time intervals after the crush, animals were sacrificed by intracardiac perfusion, the right SN dissected out, and the epineurium covering the SN trunk and its tibial and peroneal divisions removed. The SN divisions were flattened between two glass slides with the help of a rubber roller, after which the slides were pressed together with a flat tubing clamp. The slide 'sandwich' containinj~, the nerve was left in a cold postfixation buffer perfusate for 2 - 4 h. The flat whole-nerve preparations obtained in this manner were then removed from the 'sandwich' and reacted, while floating, in appropriate media. For the HRP tracing technique, the enzyme was injected into the region of the cell bodies of spinal motoneurons that innervate the musculature of the right hind limb, following laminectomy of T13, L1 and L2 vertebrae. To determine the rostro-caudal extent of the motoneuronal pool innervating the hind limb musculature, the procedure developed by Forman and Berenberg (1978) was used. Injections of 0.1-0.2 ~liter saline containing 10 mM sodium glutamate were placed into the ventral cell column on the left side of the spinal cord, and contractions of muscles of the left hind limb indicated the rostro-caudal level of the spinal cord at which HRP injections had to be performed. At this identified spinal cord level, 6 injections of H R P solution (40% Sigma type VI HRP in saline containing 2% dimethyl sulfoxide) were made at 1 mm intervals. HRP injections were performed with a glass micropipette attached to a micromanipulator. To place HRP into the ventral horn region, the micropipette tip (60/am diameter) was positioned 1 mm lateral to the midline and 1.6 mm below the dura (Lasek, 1968; Griffin et al., 1976; Forman and Berenberg, 1978). The HRP solution was injected at a rate of 0.1 /~liter/45 s by means of a 2 #liter Hamilton syringe whose needle was coupled by polyethylene tubing to the micropipette (the system being filled with mineral oil) and whose plunger was driven by a Sage 341 syringe pump. Although injections of 0.05-0.5 /xliter of H R P solution appeared to provide quite similar patterns of labeling of SN axons, the relatively large amount of 0.3/~liter per injection was chosen to ensure the labeling of the majority of motoneurons. Animals were perfused 16-24 h after HRP injections. Perfusion and HRP reaction procedures were essentially those of Mesulam's tetramethylbenzidine method as described by Warr et al. (1981), with the following exceptions: postfixation perfusate was 0.1 M phosphate buffer, pH 7.4, and 0.75% gelatin was added to the preincubation and incubation solutions. Following the reaction, flat nerves were whole-mounted on glass slides subbed with gelatin-chrome alum, left to air dry for 24 h, counterstained with neutral red (2.5 min), rinsed in distilled water (15 s), dehydrated in 70%, 95% and two changes of 100% ethanol (15 s each), cleared in methyl salicylate (two changes, 1 and 5 min), washed in xylene and coverslipped with Permount. In such preparations, the
208 location of the distal margin of the crush zone and tips of the most advanced axons was marked on the coverslip under a microscope and measured with a caliper. To demonstrate acetylcholinesterase (ACHE) activity, animals were perfused with 50 ml of saline at room temperature, followed by ! liter of fixative (1% paraformaldehyde-l.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4) for 30 min at room temperature and 1 liter of the same buffer for 30 min at 4°C. Flat nerves were reacted for ACHE, using the Koelle copper thiocholine method as modified by Geneser-Jensen and Blackstad (1971). The nerves were then whole mounted on subbed glass slides, left to air dry for 24 h, dehydrated in 50, 70, 96 and 100% ethanol (2 min each), cleared in xylene (two changes, 5 min each) and coverslipped with Permount. For localizing tubulin, rats were perfused with 50 ml phosphate buffer -saline (PBS) solution (0.9% NaCI in 0.1 M phosphate buffer, pH 7.4) at room temperature, followed by 1 liter of fixative (4% paraformaldehyde -0.5% glutaraldehyde in PBS) for 30 min at room temperature and 1 liter of PBS for 30 min at 4°C. Flat nerves were reacted by means of a two-step ('sandwich'), indirect immunocytochemical technique involving covalent bonding (Pickel, 1981). A primary antiserum produced in a rabbit against tubulin was used at 1:100 dilution. Preparations were stained with 4-chlor-l-naphthol, whole-mounted and coverslipped with 90% glycerol/PBS.
Results Using the HRP tracing technique, it was possible to demonstrate the dynamics of regenerating axon outgrowth. At 16 h after the crush, virtually all HRP labeled axons were seen proximal to the crush site (Fig. 1). By 2 days after the crush, many labeled axons were observed crossing the crush zone and passing peripheral to it (Fig. 2A), the distance of regenerating axon elongation being progressively increased at 4 (Fig. 2B) and 6 (Fig. 2C) days post-lesion. The crush zone was well-defined in these preparations, especially under lower magnifications, due to an increased background staining by neutral rad. Being relatively moderate in the portion of the SN central to the crush, the intensity of regenerating axon labeling by HRP was markedly increased, starting just proximal to the crush zone and continuing all the distance down to the tips of outgrowing axons. Individual regenerating axons, as well as axonal bundles could be easily traced (while changing focal planes) along their entire undulating course from the distal border of the crush zone to the axonal tips. Axonal tips marked the leading edge of several waves of outgrowing axons most noticeable in Fig. 2B. The leading edge could be readily distinguished, since many axonal tips had larger dimensions than their parent axons and formed elongated bulbs filled with the H R P reaction product (Fig. 3A, B). Examination of flat nerve preparations at higher magnifications, such as used in Fig. 3, showed that a limited number of HRP labeled axons of apparently smaller size extended distally to the prominent waves of outgrowing axons which were detected at lower magnifications and demonstrated in Fig. 2. These fine, most advanced axons (fig. 3A) were
209
Fig. 1. Flat, whole-mount preparation of the tibial division of the rat sciatic nerve, demonstrating the pattern of axonal labeling at 16 h after the crush by the HRP tracing technique. The crush zone is indicated by the arrowhead, the nerve segment proximal to the crush is on the left of the arrowhead, x 50.
designated as the fastest growing fibers. Peripheral to the fastest growing fibers, labeled axons were not seen in any preparation examined, indicating that the crush paradigm used completely axotomized the SN. With the defined subpopulation of the fastest growing fibers, it became possible to estimate the m a x i m u m outgrowth distance and m a x i m u m elongation rate of regenerating axons in the rat SN. For this purpose, rats were sacrificed 4 and 6 days after the SN crush, 7 animals at each post-lesion time. The distance between the peripheral margin of the crush zone and the most advanced tips of the fastest growing axons labeled with H R P was measured along the tibial division of the SN. This m a x i m u m axon outgrowth distance (mean _ S.E.M.) was f o u n d to be 8.2 _+ 0.2 m m (range of 7.5-9.2 mm) at 4 days and 16.2 _+ 0.4 m m (range of 14.4-17.7 mm) at
210
~iiI¸¸Ii
Fig. 2. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating the dynamics of outgrowth of regenerating H R P labeled axons at 2 days (A), 4 days (B) and 6 days (C) after the crush. The distal margin of the crush zone is indicated by arrowheads. Calibration bar = 1 mm.
6 days after the crush. Thus, the most rapidly advancing axons grew 8 mm in 2 days with the maximum outgrowth rate of 4 m m / d a y . Additional experiments were performed to verify whether or not the length of the SN was changed due to the preparative procedures used. The SN was exposed along its course, and two 11-0 surgical sutures were passed through the SN perpendicular to its long axis, separated by an interval of 12-18 mm. The distance between the sutures turned out to be virtually the same when measured in living animals, immediately following the perfusion in situ, after the SN dissection and epineurium removal, following the SN. flattening, or after the H R P reaction, dehydration, mounting and coverslipping. The results of SN length measurements in living animals as compared to all other conditions did not vary more than + 0.5 mm. Using the histochemical technique for ACHE, the AChE activity at 3 h after the crush was found in axonal segments located in the SN portions proximal and distal to the crush, with an increased reaction intensity near the crush site (Fig. 4A). Blood vessels crossed the crush zone at this post-lesion time, but single structures resem-
211
Fig. 3. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating HRP labeled axonal tips as they advanced peripherally (to the right) at various distances by 4 days after the crush (A), many of the axonal tips being enlarged compared to parent axons and filled with the HRP reaction product (B). Dark-field illumination in polarized light, x 190 (A); × 1800 (B).
212
Fig. 4. Flat, whole-mount preparations of the tibial division of the rat sciatic nerve, demonstrating the patterns of axonal labeling at 3 h (A) and 2 days (B) after the crush by means of histochemical technique for acetylchohnesterase. The distal margin of the crush zone is indicated by arrowheads. × 36.
213
Fig. 5, Flat, whole-mount preparation of the tibial division of the rat sciatic nerve, demonstrating the pattern of axonal labeling at 7 days after the crush by means of indirect immunocytochemicaltechnique for localizing tubulin. The distal margin of the crush zone is indicated by the arrowhead. × 16.5.
bling AChE-positive axons were only occasionally observed within this zone in some preparations. By 2 days after the crush, moderate AChE activity was present in a large number of axons along the SN portion proximal to the crush; approaching the crush site, regenerating axons gave a strong positive reaction for ACHE, and they could be traced more peripherally, crossing the crush zone, passing distal to it and ending as enlarged tips (Fig. 4B). Distal to the most advanced axonal tips, no AChE activity was present in axons. Although individual axons were well resolved by this technique, their entire course was difficult to follow due to a fading of staining intensity in the proximity of axonal tips. When SN preparations were incubated without acetylthiocholine, AChE activity was entirely absent. Using the immunocytochemical indirect 'sandwich' technique, distinct and reproducible patterns of immunoperoxidase staining with antibodies against tubulin were demonstrated. Individual regenerating axons and their bundles were intensely stained against the background in the SN portions proximal and distal to the crush (Fig. 5). In the crush zone, however, the axons were not stained, even by 7 days after the crush (Fig. 5). Enlarged axonal tips were never observed, and it was difficult to trace the entire course of individual axons in immunostained nerve preparations. The immunoperoxidase staining of regenerating axons with the antiserum against tubulin appeared to be specific, since no such staining was detected in control SN preparation treated with the preimmune serum.
214
Discussion The method of fiat, whole-mount nerve preparations can be useful for obtaining accurate estimates of the elongation rate of regenerating axons, since it allows the axon outgrowth distance to be precisely assessed by means of several alternative approaches: First, the method developed by us offers a unique opportunity to measure the absolute length of outgrowing individual axons labeled by HRP. This can provide an actual measure of the outgrowth distance for regenerating axons. With the available experimental techniques, the extent of outgrowing axon elongation is estimated as the distance between two points along a nerve: the site of an axotomizing lesion and the region of advancing axonal tips. An undulating course of axons within a nerve is not taken into account by such two-point measurements, which may result in a systematic underestimation of outgrowth distances covered by regenerating axons. Second, traditional two-point measurements of axon outgrowth distances can also be reliably performed in flat, whole-amount nerve preparations, using the H R P tracing technique. Compared to other current techniques (fast axonal transport and pinch test techniques), the described method has the advantage of an unequivocal precision in locating both the distal margin of the zone of axotomy and advancing axonal tips. The accuracy of two-point measurements in flat, whole-mount nerve preparations appears to be unaffected by the preparative procedures, which cause no appreciable distortion of linear nerve dimensions. Finally, a direct differential analysis of outgrowth distances for subpopulations of regenerating neurons that elongate at fast, intermediate or slow rates can now be carried out in flat, wholemount nerve preparations, using either length or two-point measurements. It could be argued, however, that H R P might be transported to the terminal, newly formed parts of regenerating axons in an amount insufficient to render them detectable at the light microscopic level. Since the technique of fast axonal transport of radioactive proteins labels the advancing cones of regenerating axons (Griffin et al., 1976), it seems appropriate to compare the available results obtained by this technique with the data derived from fiat, whole-mount nerve preparations labeled with HRP. The maximum axon outgrowth distance, as estimated in flat nerve preparations, is 8.2 and 16.2 m m at 4 and 6 days after SN crush, respectively, yielding a maximum axon elongation rate of 4 m m / d a y . These values correspond well to those obtained with the fast axonal transport technique: a maximum outgrowth distance of 16 m m by 6 days after a crush (Forman and Berenberg, 1978) and a maximum elongation rate of 3.6-4.5 m m / d a y (Grafstein and McQuarrie, 1978; Lasek et al., 1981) have been reported for regenerating motor axons in the SN of young adult rats. Such a n agreement of the results strongly suggests that H R P transport under the experimental conditions used in this work extends along regenerating axons to the level indicated as the front of outgrowth by rapidly transported radioactive proteins. Another valuable application of the method reported here is that it permits the direct evaluation of the so-called initial delay of regenerating axon outgrowth. This delay includes the time required for the neuron to recover from axotomy, for axon growth to commence, and for the outgrowing axon to reach and traverse the zone of
215 axotomy (Sunderland, 1978). By using the available techniques, it is only possible to infer the duration of initial delay by backward extrapolation of the axon outgrowth values to zero distance (Grafstein and McQuarrie, 1978). With the help of fiat, whole-mount nerve preparations, the duration of initial delay can be directly ascertained as the time necessary for HRP labeled regenerating axons to reach the nerve segment just distal to the lesion site. The method described here has an additional promising potential related to the fact that both histochemical and immunocytochemical techniques can be applied to flat, whole-mount nerve preparations. Our experiments with the histochemical technique for AChE and immunocytochemical technique for tubulin detection have demonstrated that individual regenerating axons can be clearly visualized in such nerve preparations using both techniques. This makes it feasible to obtain information concerning the nature of axonal subpopulations that elongate at different rates. Such a goal seems to be attainable, since any staining technique applicable to floating or mounted histological sections should work well in these nerve preparations, and selected segments of flattened nerves can be examined by means of electron microscopy. However, the method reported here also may have certain limitations. A major potential disadvantage of this method, which involves light microscopy, is that it may not provide a resolution sufficient enough to visualize the finest regenerating axons. Due to this, errors may be introduced in the detection of the leading edge of outgrowing axons. Despite the fact that the front of outgrowth of HRP labeled axons can be located rather precisely in fiat, whole-amount preparations of the rat SN (as indicated by the above comparison of our results with those obtained with the fast axonal transport technique), this may not be the case for systems other than the rat SN. In other systems, the position of the leading edge of regenerating axons as determined in fiat, whole-mount preparations should be verified by means of the fast axonal transport technique a n d / o r electron microscopy. Another potential disadvantage of this method, using the HRP tracing technique, is peculiar for the SN system, in which not only motor, but also sensory axons may be labeled after HRP injections into the spinal cord. Further experiments with sectioning of the appropriate ventral roots should elucidate whether or not HRP labeling of sensory axons takes place. In any case, a differential analysis of SN motor and sensory axons is achievable. HRP injections into the spinal cord can be made after dorsal root sectioning for selective labeling of SN motor axons, and HRP can be injected into the dorsal root ganglia for selective labeling of SN sensory axons. It appears, therefore, that the method of fiat, whole-mount nerve preparations has the potential to become a major tool for studying the process of regenerating axon outgrowth.
Acknowledgements The authors would like to express their gratitude to Dr. R.A. Berenberg for valuable advice concerning the techniques of laminectomy and spinal cord injections, Dr. R.E. Heller for the skilled assistance in operative procedures, Dr. R.D.
216 G o l d m a n for the g e n e r o u s supply of the antisera used a n d Dr. R.W. Berry for helpful criticism of the manuscript. This work was supported by grant N1H AG-03410, NIH 19730 a n d U S A M R D S 1782C2272.
NS-18490, NIH
NS-
References Berenberg, R.A., Forman, D.S., Wood, D.K., DeSilva, A. and Demaree, J. (1977) Recovery of peripheral nerve function after axotomy: effect of triiodothyronine, Exp. Neurol., 57: 349-363. Bisby, M.A. (1978) Fast axonal transport of labeled protein in sensory axons during regeneration, Exp. Neurol., 61: 281-300. Black, M.M. and Lasek, R.J. (1976) The use of axonal transport to measure axonal regeneration in rat ventral motor neurons, Anat. Rec., 184: 360-361. Forman, D.S. and Berenberg, R.A. (1978) Regeneration of motor axons in the rat sciatic nerve studied by labelling with axonally transported radioactive proteins, Brain Res., 156: 213-225. Geneser-Jensen, F.A. and Blackstad, T.W. (1971) Distribution of acetyl cholinesterase in the hippocampal region of the guinea pig. I. Entorhinal area, parasubiculum and presubiculum. Z. Zellforsch., 114: 460-481. Grafstein, B. and McQuarrie, I.G. (1978) Role of the nerve cell body in axonal regeneration. In C.W. Cotman (Ed.), Neuronal Plasticity, Raven Press, New York, pp. 155 195. Griffin, J.W., Drachman, D.B. and Price, D.L. (1976) Fast axonal transport in motor nerve regeneration. J. Neurobiol., 4: 355-370. Lasek, R.J. (1968) Axoplasmic transport of labelled proteins in rat ventral motoneurons, Exp. Neurol., 21 : 41-51. Lasek, R.J., McQuarrie, I.G. and Wujek, J.R. (1981) The central nervous system regeneration problem: neuron and environment. In A. Gorio, H. Millesi and S. Mingrino (Eds.), Posttraumatic Peripheral Nerve Regeneration: Experimental Basis and Clinical Implications, Raven Press, New York, pp. 59-70. McQuarrie, I.G. (1981) Acceleration of axonal regeneration in rat somatic motoneurons by using a conditioning lesion. In A. Gorio, H. Millesi and S. Mingrino (Eds.), Posttraumatic Peripheral Nerve Regeneration: Experimental Basis and Clinical Implications, Raven Press, New York, pp. 49-58. McQuarrie, I.G., Grafstein, B. and Gershon, M.D. (1977) Axonal regeneration in the rat sciatic nerve: effect of a conditioning lesion and of dbcAMP, Brain Res., 132: 443-453. Orgel, M.G. (1980) A critical review of histological methods used in the study of nerve regeneration. In D.L. Jewett and H.R. McCarroll, Jr. (Eds.), Nerve Repair and Regeneration. Its Clinical and Experimental Basis, The C.V. Mosby Company, St. Louis, pp. 141-148. Pestronk, A., Drachman, D.B. and Griffin, J.W. (1980) Effects of aging on nerve sprouting and regeneration, Exp., Neurol., 70: 65-82. Pickel, V.M. (1981) Immunocytochemical methods. In L. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum Press, New York, pp. 483-509. Sunderland, S. (1978) Nerves and Nerve Injuries. Churchill Livingstone, Edingburgh, 1046 pp. Warr, W.B., de Olmos, J.S. and Heimer, L. (1981) Horseradish peroxidase: the basic procedure. In L. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing Methods, Plenum Press, New York. pp. 207-262. Young, J.Z. and Medawar, P.B. (1940) Fibrin suture of peripheral nerves. Measurement of the rate of regeneration, Lancet, ii: 126-128.