Cell and Tissue Research

Cell Tiss. Res. 183, 395-402 (1977)

9 by Springer-Verlag 1977

Somatopetal Transport of Horseradish Peroxidase (HRP) in the Peripheral and Central Branches of Dorsal Root Ganglion Cells* ** *** W. Neuhuber, B. Niederle, and W. Zenker II. Anatomisches Institut, Universit~it Wien, Wien, Austria (Prof. Dr. W. Zenker)

Summary. The axonal transport of H R P in both the peripheral and central branches of dorsal root ganglion cells was studied in rats. F o r studying axonal transport in the peripheral branch H R P as a dry substance was applied to the peroneal nerve injured either by teasing, by cutting or crushing. After a short survival time (22 h) mainly small spinal ganglion cells of the corresponding segments were labelled, while after a prolonged survival time (70 h) mainly large cells were labelled. These labelling differences are referred to different transport rates or to differences in the process of accumulation of H R P in neurons of various sizes. No evidence could be found for H R P transport from the peripheral into the central branch. Injection of H R P into the spinal cord (survival time 22 h) or into the dorsal column nuclei (survival time 46 h) was followed by labelling of numerous spinal ganglion cell perikarya of all sizes. Reaction product was found also within the prebifurcation segment of spinal ganglion cell processes. On the basis of light microscopic exploration only somatopetal transport could be detected. Key words: Axonal transport Somatopetal transport.

-

HRP

-

Primary sensory neurons

-

Introduction In the last few years many studies have been performed to trace different neuronal pathways using the horseradish peroxidase (HRP) technique of retrograde axonal Send offprint requests to: Prof. Dr. W. Zenker, II. Anatomisches Institut, Universit~it Wien,

WghringerstraBe 13, A-1090 Wien, Austria * This investigation was supported by the"Fonds zur F6rderung der wissenschaftlichenForschungin (~sterreich" ** The authors wish to thank Prof. Dr. H. Holl/inder and his coworkers(Neuroanatom. Abteilung, Max Planck Institut ftir Psychiatrie, M/inchen) for many helpful suggestionsto improvethe technique. Thanks are also due to Dr. E. Krammer and Dr. H. Gruber for stimulating criticism and to Miss F. Schramm for technical assistance *** Dedicated to Prof. H. Ferner with best wishes on his 65th birthday

396

W. Neuhuber et al.

t r a n s p o r t (see C o w a n a n d Cu6nod, 1975, for ref.). It is well established that p e r i k a r y a of m u l t i p o l a r nerve cells can be distinctly labelled for light microscopical detection provided the a m o u n t of enzyme applied to their axons at the periphery is great e n o u g h ( K r i s t e n s s o n a n d Olsson, 1971). The i n t r o d u c t i o n o f the H R P technique raises a n u m b e r of questions: can one take a d v a n t a g e o f the H R P m e t h o d to trace sensory pathways from the periphery t o w a r d s the spinal cord; is it possible to trace sensory pathways from the central n e r v o u s system towards the periphery; could the H R P technique replace or s u p p l e m e n t the classical d e g e n e r a t i o n m e t h o d s ? T o answer these questions it is necessary to clarify the following points: 1. W h a t is the preferential t r a n s p o r t direction of H R P in the processes i o f the p s e u d o u n i p o l a r sensory g a n g l i o n cells ? As yet only the t r a n s p o r t of H R P from the peripheral b r a n c h to the p e r i k a r y o n has been established by F u r s t m a n et al. (1975), A l v a r a d o - M a l l a r t et al. (1975), A r v i d s s o n (1975), Ellison a n d Clark (1975), H i n r i c h s e n (1975) a n d K r i s t e n s s o n a n d Olsson (1975). 2. C a n H R P be t r a n s p o r t e d from the peripheral b r a n c h into the central one a n d vice versa (and if so, transsomatically or directly from the peripheral to the central process) ? N o c o n v i n c i n g results relevant to this question are yet available. 3. Are all types of spinal ganglion cells able to a c c u m u l a t e the t r a n s p o r t e d HRP? 4. Are there differences between different types of spinal ganglion cells in t r a n s p o r t kinetics o f H R P ?

Materials and Methods Male Sprague Dawley rats (150-250g, Versuchstierfarm Himberg-Wien) were anesthetized with Nembutal (4 mg/100 g body weight) intraperitoneally. For marking dorsal root ganglion cells via the peripheral branches horseradish peroxidase (HRP, Reinheitsgrad 1, Boehringer-Mannheim; 1-2 mg) as a dry substancewas applied to the peroneal nerve of one side about 45-60 mm distal to the spinal ganglion.The application was performed immediatelyafter teasing away the connectivetissue and in most of the experimentsafter additional crushing or cutting the axons (Kristensson and Olsson, 1974; De Vito et al., 1974). Gelfoam was packed around the site of application to reduce diffusion. In one case the sciatic nerve was crushed and the spinal roots of the corresponding segments were ligated. In two animals the peroneal nerveswere exposed and HRP was applied to the uninjured nervesin the same way as described above to test the possibility of HRP-uptake and transport from the adjacent tissue. For marking dorsal root ganglion cells via the central branched 0.5-1.0 lal of a 30 ~ aqueous HRP solution was injected into the spinal cord at the levelof L4/L5, and in another experimentinto the dorsal column nuclei. A Hamilton microlitre syringe fitted with a 27 gauge needle was used. The rats were sacrificedafter various survival times and fixed by perfusion with a mixture of 2.5 glutaraldehyde and 1 ~ paraformaldehyde in 0.1 m Na-cacodylate at pH 7.2. Samples of the peripheral nerves, the corresponding segments of the spinal cord, medulla oblougata as well as spinal ganglia and corresponding dorsal roots were postfixed in the same fixative for several hours, transferred into 0.1 m Na-cacodylate containing 30 ~ sucrose and were stored at 4~C overnight. 40 ~tm frozen sections were rinsed in 0.1 m Na-cacodylate (pH 7.2) and afterwards in 0.05 m Tris-buffer(pH 7.6) for severalminutes 1 In this study the part of the spinal ganglion cell process between the perikaryon and the site of bifurcation is called the "prebifurcation segment". The medial process entering the spinal cord through the dorsal root is called the "central branch" whereas the lateral one which passes peripherally is called "peripheral branch"

HRP in Dorsal Root Ganglion Cells

397

and incubatedfor 30-60 min in the usual manner for demonstrationof HRP (Graham and Karnovsky, 1966). Diaminobenzidinewas replacedby o-dianisidine(3,3'-dimethoxybenzidine,Serva)(Graham and Karnovsky, 1966; Colman et al., 1976). Contralateral spinal ganglia were used as controls. Thereafter sections were mounted in glycerolwithout dehydration or counterstaining. Cell sizesweremeasuredby determiningthe diameterof those perikaryain whichthe nucleoluswas visible using a drawing microscopeand brightfieldillumination. Results

A. Application of HRP to the Peripheral Branch 1. The Morphology of the Site of Application to the Peroneal Nerve. 22 h after application of dry H R P to nerves which had been previously teased, crushed or cut, a diffuse staining of many nerve fibres could be seen. The reaction product was confined to the axoplasm. 70 h after application this diffuse staining of axons apparently disappeared. In general, no axonal staining could be detected 1-2 cm proximally to the site of application. Occasionally single axons showed a few granules at nodes of Ranvier. 2. HRP in the Dorsal Root Ganglion Cells. After 22 h a great number of labelled cells was found in the L4 and L5 ganglion. As shown in Fig. 1, labelling appeared mainly in cells with a diameter from 20-35 ~tm (peak at 30 lam). After a survival time of 70 h the peak shifted up to 50 Ixm. There were great differences in the labelling pattern of different perikarya. Some cells contained dense granules of reaction product whereas others displayed only slight staining. The overall labelling intensity apparently depended on cell size and survival time: after short survival times reaction product was seen mainly in smaller cells whereas after longer periods the number of labelled large cells increased, while the number of labelled small cells decreased. There were also apparent differences in the distribution of reaction product: in a great number of small cells the granules were concentrated in the perinuclear zone (Fig. 2 a). In contrast, in the large cells (Fig. 2 b), granules were evenly distributed throughout the cell body. The differences in the distribution pattern apparently depended on the cell type and not on the amount of reaction product. Remarkable differences in the size of granules were observed, but no obvious correlation could be established between the size o f granules and cell size. N o exogenous H R P activity was found in the prebifurcation segment of the spinal ganglion cells. In general, no H R P was found in any axons within the spinal ganglia. Occasionally small amounts of reaction product were revealed in the endoneurial spaces. No intraaxonal labelling was detected by light microscopical examination in the dorsal roots, not even after ligation near their entrance into the spinal cord. Furthermore, no accumulation of H R P reaction product could be seen within sensory terminals in the spinal cord after application of this enzyme to the peripheral branch. B. Application of HRP to the Central Nervous System In general, the site of application displayed the typical picture of an injected area o f the central nervous system as described by several authors (e.g. Nauta et al., 1974).

398

W. Neuhuber et al.

% 40

30

20' I

.d-J

I

I- 1

10' |

~

L,,,~ .

1'0

I

9

30

50

,

70 ~um

Fig. 1. Distribution of cell sizes of labelled cells 22h (n = 229, ) and 70h (n = 202, - - - - - ) H R P application to the peroneal nerve, n indicates the total number of labelled cells in L4 + L5

after

Fig. 2. a Small dorsal root ganglion cell 22h after H R P application to the peroneal nerve. Note the perinuclear accumulation of reaction product (• 960). b Large dorsal root ganglion cell, same experiment. No preferential location of granules ( • 960)

The unilateral injection area in the spinal cord (L4, L 5) was confined mainly to the gray matter of one side. Cross sections revealed a homogeneous brown staining in axons running throughout the gray matter, of axons in the dorsal funiculus and of dorsal roots close to the site of application. H R P reaction product spread about 6.5mm in the craniocaudal direction, corresponding to the length of the two segments L4 and L5.

HRP in Dorsal Root Ganglion Cells

399

Fig. 3. Dorsal root ganglioncell46 h after injectioninto the dorsal columnnuclei,showingHRP activity in the soma and in the prebifurcation segment (arrow); ( • 960) In the L4, L5 ganglia, many nerve cell bodies showed an accumulation of reaction product (22 h after application to the spinal cord or 46 h after application to dorsal column nuclei). However ganglia belonging to the neighbouring segments were apparently also labelled. In general, the number of labelled cells was much higher after injection into the spinal cord than after injection into dorsal column nuclei or following application to the peripheral nerve. Reaction product was found in cells of all sizes. The granules resembled those seen after peripheral application both in appearance and distribution. In a number of large neurons the prebifurcation segments were also labelled (Fig. 3). In one case it was possible to follow the granules from the perikaryon to the site of bifurcation, a distance of 800 lam.

C. Control Experiments H R P applied to the uninjured peroneal nerve was completely stopped at the perineurial barrier. There was only an accumulation in macrophages outside the connective tissue sheath. Spinal ganglia related to the site of application showed no evidence of H R P labelling. In no experiments were contralateral spinal ganglion cells labelled.

Discussion Our observations confirm the previous results of several authors (Furstman et al., 1975; Ellison and Clark, 1975; Alvarado-Mallart et al., 1975; Arvidsson, 1975;

400

W. Neuhuber et al.

Kristensson and Olsson, 1975; Hinrichsen, 1975), showing that application of HR P to the peripheral branches of sensory neurons is followed by labelling of their perikarya. After a short period mainly small perikarya displayed brown granules whereas intense labelling of most of the large neurons was achieved only by prolongation of the survival time. This confirms the previous observations of Kristensson and Olsson (1975). This observation deserves some attention in the light of the findings of several authors (La Vail, 1975; Holl/inder, 1975), who showed that, within a particular neuronal system, the survival period influences only the total number and the labelling intensity of marked cells with no relation to cell size. At the site of application, axons of all diameters are filled with reaction product within a few minutes (Kristensson and Olsson, 1975, 1976). Therefore the difference in the speed of labelling cannot be explained simply by differences in HRP uptake. More likely there are differences in (1) transport rates. It cannot be decided whether these different rates are fundamental phenomena inherently different for different types of neuron, or whether they are to be attributed to individual reactions of different neurons to the injury. Other studies show that fast retrograde transport of H R P occurs at several different rates (see La Vail, 1975). This could be due to a dependence of transport rates on the size of neurons. Also, thepossibility cannot be excluded, that (2) the accumulation of HRP to a detectable level takes place earlier in small cells than in large ones. On the other hand, (3) a higher speed of degradation of the stored HRP in small-cells may be of importance. These three components must be considered in explaining the differences shown in Figure 1. 2 In this study it was shown that there is massive transport of HRP along the central branches to the dorsal root ganglion cells. Injection into the spinal cord leads to intense labelling of a greater number of cells than is found following HRP application to peripheral nerves after the same survival time(22 h). Of course the number of sensory fibres of the peroneal nerve is obviously smaller than the number of dorsal root axons connecting the injection site in the spinal cord with the sensory ganglia L4, L5. This could explain the greater number of labelled cells following central application. Also the rich arborization of the primary sensory neurons within the application zone in the spinal cord may be of importance in this regard. This assumption is in good agreement with observations in other parts of the CNS suggesting that the density of the terminal axon field is the main determinat of HRP uptake (Nauta et al., 1974; Jones, 1975). Last but not least one should take into account that the distance between the application site in the spinal cord and the cell bodies in the spinal ganglia (25-30 ram) is rather short compared to the transport distances from the periphery (45-60 mm). Therefore a greater number of neurons transporting at a slower rate possibly become labelled within the same period, in addition to faster transporting neurons. The different pattern of distribution in large and small cel'ls could be explained by differences in their cytoplasmic organization. Those cells, in which the Nissl bodies are concentrated mainly in the peripheral parts of the cell bodies (B-cells), might accumulate HRP-containing organelles in the peri~auclear zone. (These organelles are tubules of endoplasmic reticulum and lysosomal structures as shown by Sotelo and Riche, 1974; Kristensson and Olsson, 1976). In contrast, in most of 2 A report about the influence of survival times on HR P accumulation in sensory and motor neurons will be published later

HRP in Dorsal Root Ganglion Cells

401

the large cells (A-cells) Nissl bodies are spread evenly over the entire cell body separated by wide "Plasmastral3en" (Andres, 1961a), which might be the localization of HRP-containing organelles. Perinuclear concentration ofHR P was also describedin neurons of the substantia nigra (Sotelo and Riche, 1974) and correlated with ultrastructural pecularities of the perikarya of these cells, i.e. a perinuclear concentration of the Golgi apparatus and a peripherallylocated Nissl substance, as in the Bcells of spinal ganglia (Andres, 1961 a). Holl~inder(1975) reported a perinuclear location of granules in area 17 neurons, especiallyin weakly labelled cells after injection of HRP into the superior colliculus. It has to be stressed that in our experiments perinuclear accumulation occurred only in a class of neurons having a diameter of about 30 lain. The perinuclear aggregation zone of H R P in these cells corresponds closely to a zone o f high succinic dehydrogenase (SHD) activity (personal observations). Labelled large spinal ganglion cells with a diameter of about 50 I.tm never showed perinuclear accumulation of H R P reaction product. Moreover, these large cells also lack perinuclear concentrations of SHD-positive structures. At this time it is impossible to offer an explanation for the coincidence of these two results. Taking into account that in our experiments the axons have been damaged, perinuclear aggregation of transported H R P caused by early chromatolytic changes (Andres, 1961 b; Kristensson and Olsson, 1976) cannot be excluded. The light microscopic demonstration of reaction product in the prebifurcation segment of several spinal ganglion cells was not surprising, in view of similarities between this part of the neuron and the perikaryon. The prebifurcation segment contains a highly developed axoplasmic reticulum and a striking number o f dense bodies, in this regard contrasting to other axons (Zenker and H6gl, 1976). However, no explanation can be given for the fact that such deposition was found after central application but not after peripheral application. Our results based on light microscopic studies show, that transport of H R P in sensory nerve cells always occurs in the direction towards the cell body. We recommend that the term "retrograde" transport in the case of sensory neurons should be replaced by " s o m a t o p e t a r ' transport. No evidence was found for transport of H R P into the dorsal roots following peripheral application, even after ligation of the dorsal roots. Thus no confirmation was found for the assumption of Furstman et al. (1975) that H R P can be transported beyond the perikaryon into the dorsal root. In our opinion it would be rather surprising to detect any reaction product in dorsal root axons, because only sites of dense accumulation of H R P can be revealed by light microscopy, e.g. the perikaryon and the prebifurcation segment of the cell process. Even in an informativeelectron microscopicstudy (unpublished observations)no reaction product could be found in the dorsal root fibres after peripheral HRP application. The question whether HRP application into a dorsal root ganglion would be followedby somatofugal axonal transport can also be solved only using the electron microscope. Until now no evidence has been found for transport of H R P from the peripheral branch into the CNS and vice versa. In conclusion: it seems at present impossible to use the H R P transport technique at light microscope level as a tool for tracing primary afferents from the periphery into the CNS.

402

W. Neuhuber et al.

References Alvarado-Mallart, M.R., Batini, C., Buisseret-Delmas, C., Corvisier, J.: Trigeminal representation of the masticatory and extraocular proprioceptors as revealed by horseradish peroxidase retrograde transport. Exp. Brain Res. 23, 167-179 (1975) Andres, K.H.: Untersuchung fiber den Feinbau yon Spinalganglien. Z. Zellforsch. 55, 1-48 (1961 a) Andres, K.H.: Untersuchung fiber morphologische Veranderungen in Spinalganglien wahrend der retrograden Degeneration. Z. Zellforsch. 55, 49-79 (1961 b) Arvidsson, J.: Location of cat trigeminal ganglion cells innervating dental pulp of upper and lower canines studied by retrograde transport of horseradish peroxidase. Brain Res. 99, 135 139 (1975) Colman, D.R., Scalia, F., Cabrales, E.: Light and electron microscopic observations on the anterograde transport of horseradish peroxidase in the optic pathway in the mouse and rat. Brain Res. 102, 156163 (1976) Cowan, W.N., Cu6nod, M.: The use of axonal transport for studies of neuronal connectivity. Amsterdam: Elsevier Scientific Publishing Company 1975 De Vito, J.L., Clausing, K.W., Smith, O.A.: Uptake and transport of horseradish peroxidase by cut end of the vagus nerve. Brain Res. 82, 269-271 (1974) Ellison, J.P., Clark, G.M.: Retrograde axonal transport of horseradish peroxidase in peripheral autonomic nerves. J. comp. Neurol. 161, 103-114 (1975) Furstman, L., Saporta, S., Kruger, L.: Retrograde axonal transport of horseradish peroxidase in sensory nerves and ganglion cells of the rat. Brain Res. 84, 320-324 (1975) Graham, R.C., Karnovsky, M.J.: The early stages of absorption ofinjec ed horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem. 14, 291-302 (1966) Hinrichsen, C.: Retrograde transport of HRP in afferent and efferent neurons from the masseter muscle in the rat. Naturwissenschaften 62 (10), 492 (1975) Hollander, H. : Observations on cortical neurons retrogradely labeled with horseradish peroxidase. In: Advances in neurology 12-Physiol &Pathol of dendrites (G.W. Kreutzberg, ed.), pp. 315 318. Amsterdam: North Holland Publishing Company 1975 Jones, E.G.: Possible determinants of the degree of retrograde neuronal labelling with horseradish peroxidase. Brain Res. 85, 249-253 (1975) Kristensson, K., Olsson, Y.: Axonal uptake and retrograde transport of exogenous proteins in the hypoglossal nerve. Brain Res. 32, 399-406 (1971) Kristensson, K., Olsson, Y.: Retrograde transport of horseradish peroxidase in transected axons. 1. Time relationships between transport and induction of chromatolysis. Brain Res. 79, 101-109 (1974) Kristensson, K., Olsson, Y.: Retrograde transport of horseradish peroxidase in transected axons. 2. Relations between rate of transfer from the site of injury to the perikaryon and onset of chromatolysis. J. Neurocytol. 4, 653-661 (1975) Kristensson, K., Olsson, Y.: Retrograde transport of horseradish peroxidase in transected axons. 3. Entry into injured axons and subsequent localization in perikaryon. Brain Res. 115, 201-213 (1976) La Vail, J.H.: Retrograde cell degeneration and retrograde transport techniques. In: The use of axonal transport for studies of neuronal connectivity (W.N. Cowan and M. Cu6nod, eds.), pp. 218-248. Amsterdam: Elsevier Scientific Publishing Company 1975 Nauta, H.J.W., Pritz, M.B., Lasek, R.J.: Afferents to the rat caudo-putamen studied with horseradish peroxidase. An evaluation of retrograde neuroanatomical research method. Brain Res. 67, 219-238 (1974) Sotelo, C., Riche, D.: The smooth endoplasmic reticulum and the retrograde and fast orthograde transport of horseradish peroxidase in the nigro-striato-nigral loop. Anat. Embryol. 146, 209-218 (1974) Zenker, W., H6gl, E.: The prebifurcation section of the axon of the rat spinal ganglion cell. Cell Tiss. Res. 165, 345--363 (1976)

Accepted February 11, 1977