- Email: [email protected]
Factors affecting the ultrastructural pattern of anterograde labeling in axon terminals with HRP
B,.~,;,, &,spf,l.ch
&/tc,tiu,
Vol. 16, pp. 259-265,
1986. ’ Ankho
International Inc. Printed in the
U.S.A
0361-9230186 $3.00 + .OO
Factors Affecting the Ultrastructural Pattern of Anterograde Labeling in Axon Terminals With HRP ROSEMARY
Dclpurtmcjnt of Anutomy,
C. BORKE
AND MARTIN
E. NAU
Uniformed Services University of thr Health Sciences, Bethesda, Received
MD 26X314-4799
19 July 1985
BORKE, R. C. AND M. E. NAU. Fctttors uSfi,cting the ~ltru.~trl~~,t~r~il pattern ofantmy+mfc labeling in uxon te~mitr~is BRAIN RES BULL 16(Z) 259-265, 1986.-Comparisons of anterograde labeling of axon terminals originating from short and long projection neurons were made in the hypoglossal nucleus. Injections of dilute and concentrated horseradish peroxidase (HRP) or wheat germ-agglutinin-horseradish peroxidase (WCA-HRP) were made via a glass micropipette into the nucleus reticularis parvocellularis (RPc=short projection neurons) and the Spinal V trigeminal complex (Sp. V=long projection neurons). Axon terminals in the hypoglossal nucleus, a common projection site of the two efferent systems, were evaluated ultrast~cturally using diaminobenzidine (DAB) as the chromogen for the cobalt-glucose oxidase (CO-GOD) method of HRP labeling. Labeled axon terminals from these two sources demonstrated different dist~bution patterns of the reaction product. For the short pathway, high concentrations of the tracers resulted in diffuse, agranular labeling in the majority of axon terminals. Dilute concentrations of the tracers were associated with membrane-bound, granular type of labeling. All anterograde labeling of terminals of long projection neurons (Sp. V) was membrane-bound and granular irrespective of the tracer concentration. The length of the pathway and the concentration of the enzyme tracers are factors that affect the pattern of anterograde label in axon terminals of hypoglossal afferents. \t,ith HRP.
HRP
WGA-HRP
Diaminobenzidine
Diffuse agranular label
Membrane-bound
granular label
The object of the current investigation was to explore and compare factors that may influence the types of ultrastructural HRP labeling in axon terminals, namely, the length of the CNS pathway and the concentration of the enzyme. DAB was selected as the chromogen because the ultracytochemical distribution of the reaction product reflects the processes involved in the somatofugal transfer of the enzyme.
THE macromolecules, HRP and WGA-HRP, have proven to be reliable and effective anterograde markers for tracing efferent connections at the light microscopic level [4, 9, 13, 201. The usefulness of these anterograde tracers has been expanded by their applicability to electron microscopic studies permitting the characterization of the synaptic organization of projections to specific cell groups in the CNS [7, 14, 15, 161. Using diaminobenzidine (DAB) as the chromogen results in two different distribution patterns of the reaction product in axon terminals [ 121; (I) electron dense granules contained within vesicular or tubular structures, and (2) electron dense material applied to the cytomembranes of organelles within the terminals. These two patterns of anterograde label in axon terminals are thought to represent differences in the nature of the somatofugal transfer of the enzyme. Physiological processes of endocytosis and fast anterograde axonal transport in intact neurons commonly account for the membrane-bound, granular appearance of the reaction product in axon terminals [13]. The diffuse, agranular reaction is most often the result of cytoplasmic filling of nerve cell bodies and processes after entry of HRP by mechanical disruption of neuronal membranes [1,6]. The most common cause of this type of label is irreversible damage to the neuron resulting in anterograde degeneration [ 17,181. Both types of terminal labeling with HRP can be used to specify the synaptic organization of the terminal arborization of neural efferents in the CNS, as long as the phenomenon of anterograde label of severed axons of passage is taken into consideration.
METHOD
The hypoglossal nucleus, located on each side of the medullary raphe is a mutual efferent site for neurons originating from RPc and the Sp. V nucleus (pars interpolaris and oralis) [3]. RPc cells are situated adjacent to the lateral border of the hypoglossal nucleus, whereas Sp. V neurons occupy the lateral border of the brain stem at medullary levels. Based on the differences in the length of their projections (compare stereotaxic coordinates listed below) to this common site, for the purpose of the current experiments, RPc neurons are designated short projection neurons and Sp. V neurons, whose axons travel at least two times greater distances to end on hypoglossal neurons, are termed long projection neurons. The effect of the concentration of the enzyme tracer on the patterns of anterograde labeling in axon terminals originating from hypoglossal afferents of two different lengths was determined by two sets of experiments: high concentration tracer experiments and low concentration
259
260
BORKE AND NAU
tracer experiments. In addition, two other sets of experiments were performed: lesion-HRP and control experiments. Table I provides a summary of the numbers of animals used for each set of experiments and the protocols of each set of experiments are detailed below. High Concentrution
Tracrr Experimrnts
In the first set of experiments, rats were anesthetized with sodium pentobarbital (30 mgikg) or 7% chioral hydrate (5 ml/kg) and WGA-HRP (18%) or HRP (IS-30%) was injected stereotaxically through a glass micropipette (2&40 pm tip diameter) into one of two sources of hypoglossal afferents: RPc (1 .kl. 1 mm lateral, 0.7-0.8 mm anterior and a depth of 0.6 mm from the reference point, the obex) or the Spinal V nucleus (2.5-2.6 mm lateral, 0.7-0.8 mm anterior and at a depth of 0.C0.5 mm from the reference point, the obex). At each site, 4 rats were injected with 1.8% WGA-HRP. Survival times were I and 2-3 days with two specimens being used at each interval per injection site (Table 1). Longer survival periods were omitted to avoid the possibility of labeling second and third order neurons by transneuronal transport of WGA-HRP. A total of 7 rats for each site received pressure injections of 15-3C% HRP. Of these specimens, 2 rats per injection site survived 2-3, 5 and 7 days after enzyme delivery and one rat per injection site was killed one day after administering the tracer (Table 1).
In a subsequent set of experiments, the concentration of the tracer injected into the two sites of hypoglossal afferents was reduced. A two step reduction in tracer concentration was carried out for RPc injections. In one series of rats, WGA-HRP or HRP in 0.5% concentrations was injected into RPc using the same method and amount of tracers as those of the high concentration tracer experiments. Likewise, the same survival intervals were repeated for these experiments. The number of animals for each survival interval appears in Table 1. The second series received injections of 0.16% WGA-HRP or 0.25% HRP into RPc. All parameters conformed to those of the previous experiments except survival times, Survival periods of 1 and 2-3 days were utilized and the number of rats for each time period is indicated in Table i. The lowest concentrations of the tracers (0.16% WGAHRP and 0.25% HRP) were also injected into the Spinal V nucleus. Survival times and numbers of specimens corresponded to those in which the enzyme was injected into RPc. Lesion-IIRP
~.~p~ri~l~~llts
In another series of experiments, electrolytic lesions [21, comparable in size and location to the tracer injections were
t
TABLE
Type
of Experiment
Type of Injection
High Concen-
1.8% WGA-
tration Tracer Low Conoen-
I
Survival Time (days) 2-3 5 7
*2
2
HRP
I
2
2(O)
2(O)
WGA-
2(O) 4
4(O) 4
HRP
3
6
f(O)
2
15-30%
HRP
0.5% WGA-
tration
HRP
Tracer
0.5% HRP 0.16%
2
2
2(O)
2(O)
l(O)
Ito)
i(O)
2
2
2
HRP 0.25% Lesion-HRP
Electrolytic Lesion
plus
I s-30% HRP Control
0.
I
M Tris
Buffer *Number in each block represents number of animals utilized for RPc injections. For Spinal V injections same numbers were used unless indicated otherwise by ( ). N=90.
produced in RPc. In 4 rats, HRP (15-30%) in small volumes (40-50 nlf was injected via a glass micropipette at the lesion site 5-45 min after lesion production. Survival periods and numbers of animals are indicated in Table 1.
Block staining with uranyl acetate was not performed during the processing of tissue for electron microscopy and ultrathin sections were not post-stained with uranyl and lead salts. While these precautions decreased the possibility of misidentification of normally occurring membrane-bound granules with electron dense contents as containing HRP reaction product, these measures did not eliminate the possibility. Therefore a series of control rats were included in the experimental design. Sixteen rats were used for the control series: two rats for each of the four survival times (1, 2-3, 5, 7 days) per injection site (Table 1). All procedures were identical to those followed for the experimental groups except that a small volume (40 nl) of the tracer vehicle, 0.1 M Tris buffer, was injected into RPc or the Sp. V nucleus. At predetermined survival times, experimental and control animals were reanesthetized and perfused transcardially
FACING PAGE FIG.
I. A. A dark-field micrograph of an injection site in RPc, lateral to the hypoglossai nucleus (XII). high concentration tracer experiment 2 days survival, x28. B. A dark-field micrograph of an injection site in the dorsal portion of the Spinal V nucleus. Distance to the target site. the hypoglossal nucleus (XII) is 2 times that of RPc, high concentration tracer experiment (IS% HRP). 2 days survival. x23. C. Bright-field micrograph of solid (arrowheads) and granular (arrows) types of HRP labeling of neurons at injection site in RPc, high concentration tracer experiment (30% HRP), 2 days survival, x213. D. Unlabeled axon terminals (AT) surround a labeled endstructure in which diffuse. agranular reaction product is applied to the membranes of organelles in the terminal of a short projection neuron to the Xltth nucleus. high concentration tracer (1.8% WGA-HRP) injected into RPc, I day survival, unstained, x38000. E. Another type of anterograde label consisting of membrane-bound granules of HRP reaction product (arrows) was seen less frequently in axon terminals of short projection neurons (RPc) to the Xifth nucleus, high concentration tracer (30% HRP) injected into RPc, 2 days survival, unstained, x47120. F. Axon terminals (AT) of long projection neurons (Sp. V) contained almost exclusively membrane-bound, granular labeled profiles kwrows). high concent~tion tracei (3G% HRP) injected into Sp. V, 2 days survival, ~39000.
( 15%HRP).
NATURE OF ANTEROGRADE
LABELING WITH HRP
261
762
BORKE AND NAU
with Ringer’s solution, dilute and concentrated aldehydes [ 191 and a final solution of 0.1 M phosphate buffer. The brains were removed, the medulla was isolated and mounted in agar. Coronal sections, 50 pm in thickness, were cut on a vibratome. Sample sections were stained with Fink-Heimer and Nauta silver stains [5]. Remaining sections were reacted histochemically for HRP according to the CO-GOD method [8]. Alternate reacted sections were mounted on gelatinized slides and stained with neutral red for light microscopic evaluation of the injection site and localization of areas of anterograde label. Remaining reacted sections were postfixed in osmium tetroxide, dehydrated and embedded in Epon. Processing for electron microscopy did not include block staining with uranyl acetate. Epon embedded blocks were trimmed to include only the XI&h nucleus and ultrathin sections of silver interference were cut and left unstained for electron microscopic evaluation. For the high (1.8% WGA-HRP or 1530% HRP) and iowest (0.16% WGA-HRP or 0.25% HRP) concentration experiments, the proportion of diffuse agranular, membrane-bound granular and axon terminals containing both types of label was determined for 500 randomly selected terminals originating from neurons in each of the two sources of hypoglossal afferents. From the experiments in which 0.5% HRP or WGA-HRP was injected into RPc, 200 axon terminals were classified according to the same labeling patterns. RESULTS
Light microscopic evaluation. Micropipette delivery of the enzyme HRP into RPc (Fig. IA) and Sp. V (Fig. 1B) nucleus resulted in injection sites measuring 1000 microns or less in diameter. Uptake in neuronal somata appeared confined within the limits of the two sources of hypoglossal afferents. Numerous labeled ceil bodies and processes of RPc and Sp. V. neurons were identified at the respective injection sites. Solid and granular labeling occurred in neuronal somata (Fig. IC). Nerve fibers traced medially from the injection sites appeared to be filled with HRP reaction product and terminal label was seen in the neuropii of the hypogiossal nucleus. Anterograde degeneration of axons and axon terminals, originating from RPc or Sp. V neurons was not detected in adjacent vibratome sections unreacted for HRP and silver-stained with Fink-Heimer and Nauta methods. Elwtron microscopic evaluation. Ultrastructural examination disclosed that the HRP reaction product in the XIIth nucleus was found almost exclusively in myelinated axons FACING
and their terminal endings. After RPc injections, a diffuse, agranular reaction product could be seen on the external surface of organelles contained within the majority of labeled myeiinated axons. After Sp. V injections, vesicular and tubular profiles of membrane-bound label were found primarily in labeled myeiinated axons. Axon terminals originating from RPc neurons exhibited two distinct patterns of distribution of HRP reaction product. In the majority (8%) of terminals, an electron dense label was applied to the cytomembranes of the organelles (Fig. ID). Membranebound, granular label was detected in 10% of the labeled terminals (Fig. 1E) and 1% demonstrated both types of label. These labeling patterns were identified at all survival times. In contrast to these results, almost all (98%) of the axon terminals labeled from Sp. V injections contained the membrane-bound, granular type of Iabel (Fig. IF).
Light microscopic~ r,ulurrtiou. Injection sites were comparable in diameter to those obtained from the high concentration tracer experiments. Uptake appeared limited to neurons of RPc and the Sp. V nucleus. NeuronaI somata. at both sites contained chiefly granular labeling (Fig. 2A), No anterograde axonal or terminal degeneration was seen in adjacent silver-stained sections. Ekwron microscopic c~,wlrrrrtiot~. When O.S% WGA-HRP or HRP was injected into RPc, 40% of the labeled terminals were of the membrane-bound, granuiar type and both types of label were found within 22% of the labeled terminals (Fig. 2B). Myelinated axons situated in the neuropil of the XIIth nucleus demonstrated vesicular and tubular forms of granular labeling after injections of very dilute concentrations of WGA-HRP (0.16%) or HRP (0.25%) into RPc or Sp. V nucleus. Using these concentrations, 84% of labeled terminals from RPc neurons contained membrane-bound, granular reaction product (Figs. 2C,D). Terminals labeled with the diffuse, agranular reaction product constituted 7% of the labeled population and those with both types of label made up the remaining 9%. Diffuse, agranular and double-labeled terminals were rare (1%) after the same injections into Sp. V nucleus.
Light
microscr~pic~
comparable
c~drrrrtiorl.
Lesions
produced
were
in size (- I mm) and location to those injections
HRP into RPc in the previous experiments. Neuronal somata, remaining at the lesion site. were rarely seen. Injection of HRP 5-45 min after lesion production resufted in labeling primarily in nerve cell processes of RPc. of
PAGE
FIG. 2. A. Granular labeling of neurons at the injection site (RPc) occurred when the concentration of the tracer wah reduced to O.Z’/r HRP. 2 days survival. x300. B. Unlabeled axon terminals (AT) flank a terminal which contains both types of HRP labql. diffuse, agranular,and membrane-bound, granular (arrows) when the concentration of the tracer was reduced to 0.5% HRP and injected mto RF?, I day survival. unstained. x42900. C. When the concentration of the enzyme injected into RPc was reduced more than 100 fold to 0.25% HRP. anterogradc label was stilt visible and the majority of labeled terminals in the XIfth nucleus contained membrane-botlnd. granular label (arrows). Z days survival, unstained, x30400. D, Membrane-bound. granular type of HRP labeling in tubular (arrows) and vesicular iarrowheadsf organelles in an axon terminal of a short projection neuron to the Xllth nucleus when 0.16% WGA-HRP was injected into RPc. I day survival. unstained. x 31920. E. An axon terminal containing degenerative debris (“I is labeled with diffuse. agranular type of HRP reaction product. Radio frequency lesion in RPc + 30% HRP injected into RPc 30 min post lesion. I day survival, unstained. x26910. F. An axon terminal exhibiting early degenerative changes including an apparent decrease in the number of synaptic vesicles and autophagic-like vacuoles contain\ membrane-bound, granular labeling. The difference in electron density of unlabeled dense cored vesicles (arrowheads) and HRP labeled vesicular profiles (arrows) can be easily distinguished. Radio frequency lesion in RPc + 3@% HRP injected into lesion site 20 min post lesion. 2 day survival, x33440.
NATURE OF ANTEROGRADE
LABELING WITH HRP
263
264
BORKE AND NAU
Anterograde axonal and terminal degeneration in the XIIth nucleus were identified consistently in adjacent silverstained sections of all four cases. Electron miuoscwpic~ ~waluation. For each survival period of the rats injected with HRP after lesioning, ultrastructural features characteristic of terminal degeneration were apparent in the majority of the labeled terminals. Degenerative features included marked decrease in the number of synaptic vesicles, hypertrophy of the axoplasmic reticulum, mitochondrial degeneration, autophagic-type vacuoles and loss of synaptic connections with engulfment by glial processes. Both types of label, diffuse, agranular (Fig. 2E) and membrane-bound, granular (Fig. 2F) were detected in terminals that demonstrated at least two of the forementioned degenerative criteria. In a single case, small amounts of HRP reaction product could be identified in the extracellular space of the neuropil at the lateral aspect of the hypoglossal nucleus.
Injection of 0.1 M Tris buffer into RPc and the Sp. V nucleus and histochemical treatment of sections with the CO-GOD reaction for HRP produced peroxidatic activity only within erythrocytes at the injection site and associated with sinusoids of area postrema. Peroxidatic activity was not noted in neuronal perikarya and axons surrounding the injection site. Membrane-bound granules with electron dense contents (dense core vesicles) were seen in some axon terminals of the hypoglossal nucleus. These membrane-bound, granules resembled in shape and size some of the granular HRP label in axon terminals of the tracer experiments. However, the electron density of membrane-bound granules in axon terminals of HRP injected material (Fig. 2F) was markedly intensified and served as a reliable criterion for distinguishing granules containing HRP reaction product from unlabeled dense core vesicles (Fig. 2F). DISCUSSION
The current results represent the initial ultrastructural demonstration that the length of the CNS pathway and the concentration of the tracer may be interrelated factors influencing the type of HRP distribution in anterogradely labeled axon terminals of certain projection neurons. Diffuse, agranular labeling in axon terminals is more likely to occur if high concentrations of tracer (1.8% WGA-HRP or 15-30% HRP) are used to label axon terminals in the hypoglossal nucleus originating from short projection neurons in RPc. This appearance of HRP reaction product, applied to the cytomembranes of organelles within the axon terminals commonly has been interpreted to result from passive intracytoplasmic filling of irreversibly damaged somata and axons with HRP [1,6], although this type of label has been found infrequently after somatic uptake of intact CNS neurons [7]. High concentrations of the tracer may cause mechanical disruption of the membranes facilitating the somatofugal transfer of the enzyme from RPc to axonal endings in the hypoglossal nucleus. The degree of intactness may be questioned but the negative light microscopic fmdings obtained from adjacent sections silver-stained with Nauta and Fink-Heimer suggest that the insult did not result in anterograde degeneration in the hypoglossal nucleus. Moreover, at the ultrastructural level, signs of anterograde degeneration were not apparent in labeled axon terminals even after post injection times of 5 and 7 days.
Reducing the concentration of the tracer increased the proportion of axon terminals in the XIIth nucleus, demonstrating membrane-bound, granular labeling after HRP injections into RPc. The lowest concentration of tracer (0.16% WGA-HRP or 0.25% HRP) resulted in membrane-bound, granular labeling in the majority of axon terminals, This kind of label commonly represents the incorporation and anterograde transport of the tracer by intact neurons [ 12, 13, 141. Thus, data gathered from injections of the enzyme HRP into RPc in the first two sets of experiments provided evidence that if the distance between the point of application of the enzyme and the site of axonal termination is short, the concentration of the tracer has a distinct influence on the pattern of anterograde labeling. Since the prerequisite for diffuse, agranular label is presumed to be irreversible somatal or axonal damage, electrolytic lesions were placed in RPc and HRP (15-30%) was injected at the lesion site. In these cases, the enzyme would enter the damaged ends of neuronal processes and anterogradely label those axon terminals destined to undergo degeneration. A single case, in which HRP was seen in the extracellular space of the hypoglossal neuropil was subsequently eliminated as HRP could have labeled retrogradely axon terminals belonging to both intact and irreversibly damaged neurons. Both types of label, diffuse, agranular and membrane-bound, granular were apparent in axon terminals that demonstrated several degenerative features. These findings indicated that axon terminals of short projection neurons, irreversibly damaged, are not associated exclusively with a particular type of label, namely the diffuse, agranular kind and are in agreement with occasional reports of membrane-bound, granular labeling occurring in permanently damaged neurons [ 10,l I]. The fact that characteristic signs of axonal and terminal degeneration were identified at the light microscopic level from silver-stained sections and detected ultrastructurally in labeled HRP terminals but were not seen after HRP injections into RPc without prior lesioning lends credibility to the idea that somatofugal transfer of the enzyme in those experiments without lesion production was carried out by intact or reversibly damaged neurons. However, these results cannot offer conclusive evidence for this thesis. Anterograde degeneration produced incidentally by the pipette tip and/or the pressure of the injection may have occurred at different rates and survival times and escaped the sampling procedure used here. The labeling patterns obtained from the injections of HRP into RPc, however, did not constitute a common phenomenon of hypoglossal afferent labeling. When HRP was injected into the Sp. V nucleus, membrane-bound, granular labeling was found almost exclusively in axon terminals anterogradely labeled regardless of the concentration of the injected tracer. These results pointed out that for pathways of greater distances to a mutual efferent site, the concentration of the tracer was not a determinative factor of the type of HRP label in axon terminals. These experiments also dispelled the notion that HRP uptake by damaged axons of passage accounted for some of the labeling in the RPc experiments. Sp. V axons course through RPc to reach the hypoglossal nucleus [3] but anterograde labeling after Sp. V injections occurred in axon terminals with morphological features of synaptic organization that could be distinguished from those terminals labeled by RPc injections [2]. The processes involved in the exclusive visualization of the membranebound, granular label in axon terminals of long projection neurons (Sp. V) could not be ascertained from present data.
WITH HRP
265
Uptake by endocytosis and subsequent anterograde transport of the tracer in membrane-limited structures and/or access of the marker to the cytoplasm by temporary membrane interruption and eventual sequestration of the enzyme in tubular or vesicular organelles are possible means for the
somatofugal transfer of the enzyme [ 121. The pattern of HRP labeling in axon terminals of long and short projection neurons to the hypoglossal nucleus therefore may not reflect the nature of the somatofugal transfer of the enzyme.
NATURE
OF ANTEROGRADE
LABELING
ACKNOWLEDGEMENTS
The author is grateful to Dr. Malcolm B. Carpenter for his valuable suggestions and comments. The skilled typing of Mary Sias is gratefully acknowledged. This work was supported by the Department of Defense, Uniformed Services University of the Health Sciences, Department of Defense Grant CO 7019. The opinions or assertions contained herein are the private ones of the author and are not to be construed as official or reflecting the views of the DOD or the USUHS. The experiments reported herein were conducted according to the principles set forth in the “Guide for Care and Use of Laboratory Animals,” DHEW Pub. No. (NIH) 78-23.
REFERENCES I. Beattie, M. S., J. C. Bresnahan and J. S. King. Ultrastructural identification of dorsal root primary afferent terminals after anterograde filling with horseradish peroxidase. Bruin Res 153: 127-134, 1978. 2. Borke, R. C. and M. E. Nau. The ultrastructural identification of reticula-hypoglossal axon terminals anterogradely labeled with horseradish peroxidase. Bruin Res 337: 127-132, 1985. 3. Borke, R, C., M. E. Nau and R. L. Ringler, Jr. Brain stem afferents of hypoglossal neurons in the rat. Brain Ras 269: 47-55, 1983. 4. Carson, K. A. and M. M. Mesulam. Ultr~tructural
evidence in transganglionically transported horseradish mice that peroxidase-wheat germ agglutinin conjugate reaches intraspinal terminations of sensory neurons. Neurosci Lett 29: 201-206,
1982. 5. Ebbesson,
S. 0. E. The selective silver-impregnation of degenerating axons and their synaptic endings in non-mammalian species. In: ~tttt/~~tnp[~~ur~I ~~thf~~ls in N~~~~f~~~nut~~rn~~ edited by W. J. H. Nauta and S. 0. E. Ebbesson. New York: Springer-Verlag, 1970, pp. 132-161. 6. Gwyn, D. G., P. H. Wilkerson and R. A. Leslie. The ultrastructural identification of vagal terminals in the solitary nucleus of the cat after anterograde labeling with horseradish peroxidase. Netrrosti Lett 28: 139143, 1982. 7. Holstege, J. G. and J. J. Dekker. Electron microscopic identification of mammilla~ body terminals in the rat’s AV thalamic nucleus by means of anterograde transport of HRP. A quantitative comparison with EM degeneration and EM autoradiographic techniques. Npur-osci Lrtt II: 129-135, 1979, 8. Itoh. K.. A. Konishi. S. Nomura. N. Mizuno. Y. Nakamuraand T. Sugimoto. Application of coupled oxidation reaction to electron microscopic demonstration of horseradish peroxidase: cobalt-glucose oxidase method. Bruin RPS 175: 341-346, 1979. 9. Lynch, G., C. Gall, P. Mensah and C. Cotman. Horseradish peroxidase histochemistry: A new method for tracing efferent projections in the central nervous system. Bruin Rrs 65: 373380, 1974. IO. Malmgren, L. T. and Y. Olsson. Early influx of horseradish peroxidase into axons of the hypoglossal nerve during Wallerian degeneration. Naurosci Left 13: 13- 18. 1979.
II. Mawe, G. M., J. C. Bresnahan and M. S. Beattie. Ultrastructure of HRP-labeled neurons: A comparison of two sensitive techniques. Brrtin Res Buff 10: 551-558, 1983. 12. Mesulam, M. M. Principles of horseradish peroxidase neurohistochemistry and their applications for tracing neural pathways-Axonal transport, enzyme histochemistry and light microscopic analysis. In: Tracing Neural Connections, edited by M. M. Mesulam. New York: Wiley and Sons, 1982, pp. I-1.51. 13. Mesulam, M. M. and E. J. Mufson. The rapid anterograde transport of horseradish peroxidase. N~~ur[~.~~~i~nL,~5: 12771286, 1980. 14. Mizuno, N., A. Konishi, K. Itoh, N. Iwahori and Y. Nakamura. Identification of axon terminals of cerebella-olivary fibers in the cat: An electron microscopic study using anterograde horseradish peroxidase method. Neurosc,i Lm ZO: 1I-14, 1980. 15. Mizuno. N.. S. Nomura. K. Itoh. Y. Nakamura and A. Konishi. Commi~sur~ inte~e~ro~s for m~sticatin8 motoneurons: A light and electron microscopic study using the horseradish peroxidase tracer technique. Exp Nrurol 59: 2.54-262, 1978. 16. Mizuno. N.. Y. Yasui. S. Nomura. K. Itoh. A. Konishi. M. Takada and M. Kudo. A light and electron microscopic study of premotor neurons for the trigeminal motor nucleus. J Camp Nemo1 215: 290-298, 1983. 17. Oldfield, B. J. and E. M. McLachlan. Uptake and retrograde transport of HRP by axons of intact and damaged peripheral nerve trunks. NeUN>sci tetf 6: 135-141, 1977. 18. Phillipson, 0. T. Afferent projections to the ventral tegmental area of Tsai and interfascicular nucleus. A horseradish peroxidase study in the rat. J Camp Ncurol 187: 117-144, 1979. 19. Reese, T. S. and M. J. Karnovksv. Fine structural localization of a blood brain barrier to exogenous peroxidase. J Cell Bid 34: 207-217,
1967.
20. Staines, W. A., H. Kimura, H. C. Fibiger and E. G. McGeer. Peroxidase-labeled lectin as a neuroanatomical tracer: Evaluation in a CNS pathway. Brain Rrs 197: 485-490, 1980.
&/tc,tiu,
Vol. 16, pp. 259-265,
1986. ’ Ankho
International Inc. Printed in the
U.S.A
0361-9230186 $3.00 + .OO
Factors Affecting the Ultrastructural Pattern of Anterograde Labeling in Axon Terminals With HRP ROSEMARY
Dclpurtmcjnt of Anutomy,
C. BORKE
AND MARTIN
E. NAU
Uniformed Services University of thr Health Sciences, Bethesda, Received
MD 26X314-4799
19 July 1985
BORKE, R. C. AND M. E. NAU. Fctttors uSfi,cting the ~ltru.~trl~~,t~r~il pattern ofantmy+mfc labeling in uxon te~mitr~is BRAIN RES BULL 16(Z) 259-265, 1986.-Comparisons of anterograde labeling of axon terminals originating from short and long projection neurons were made in the hypoglossal nucleus. Injections of dilute and concentrated horseradish peroxidase (HRP) or wheat germ-agglutinin-horseradish peroxidase (WCA-HRP) were made via a glass micropipette into the nucleus reticularis parvocellularis (RPc=short projection neurons) and the Spinal V trigeminal complex (Sp. V=long projection neurons). Axon terminals in the hypoglossal nucleus, a common projection site of the two efferent systems, were evaluated ultrast~cturally using diaminobenzidine (DAB) as the chromogen for the cobalt-glucose oxidase (CO-GOD) method of HRP labeling. Labeled axon terminals from these two sources demonstrated different dist~bution patterns of the reaction product. For the short pathway, high concentrations of the tracers resulted in diffuse, agranular labeling in the majority of axon terminals. Dilute concentrations of the tracers were associated with membrane-bound, granular type of labeling. All anterograde labeling of terminals of long projection neurons (Sp. V) was membrane-bound and granular irrespective of the tracer concentration. The length of the pathway and the concentration of the enzyme tracers are factors that affect the pattern of anterograde label in axon terminals of hypoglossal afferents. \t,ith HRP.
HRP
WGA-HRP
Diaminobenzidine
Diffuse agranular label
Membrane-bound
granular label
The object of the current investigation was to explore and compare factors that may influence the types of ultrastructural HRP labeling in axon terminals, namely, the length of the CNS pathway and the concentration of the enzyme. DAB was selected as the chromogen because the ultracytochemical distribution of the reaction product reflects the processes involved in the somatofugal transfer of the enzyme.
THE macromolecules, HRP and WGA-HRP, have proven to be reliable and effective anterograde markers for tracing efferent connections at the light microscopic level [4, 9, 13, 201. The usefulness of these anterograde tracers has been expanded by their applicability to electron microscopic studies permitting the characterization of the synaptic organization of projections to specific cell groups in the CNS [7, 14, 15, 161. Using diaminobenzidine (DAB) as the chromogen results in two different distribution patterns of the reaction product in axon terminals [ 121; (I) electron dense granules contained within vesicular or tubular structures, and (2) electron dense material applied to the cytomembranes of organelles within the terminals. These two patterns of anterograde label in axon terminals are thought to represent differences in the nature of the somatofugal transfer of the enzyme. Physiological processes of endocytosis and fast anterograde axonal transport in intact neurons commonly account for the membrane-bound, granular appearance of the reaction product in axon terminals [13]. The diffuse, agranular reaction is most often the result of cytoplasmic filling of nerve cell bodies and processes after entry of HRP by mechanical disruption of neuronal membranes [1,6]. The most common cause of this type of label is irreversible damage to the neuron resulting in anterograde degeneration [ 17,181. Both types of terminal labeling with HRP can be used to specify the synaptic organization of the terminal arborization of neural efferents in the CNS, as long as the phenomenon of anterograde label of severed axons of passage is taken into consideration.
METHOD
The hypoglossal nucleus, located on each side of the medullary raphe is a mutual efferent site for neurons originating from RPc and the Sp. V nucleus (pars interpolaris and oralis) [3]. RPc cells are situated adjacent to the lateral border of the hypoglossal nucleus, whereas Sp. V neurons occupy the lateral border of the brain stem at medullary levels. Based on the differences in the length of their projections (compare stereotaxic coordinates listed below) to this common site, for the purpose of the current experiments, RPc neurons are designated short projection neurons and Sp. V neurons, whose axons travel at least two times greater distances to end on hypoglossal neurons, are termed long projection neurons. The effect of the concentration of the enzyme tracer on the patterns of anterograde labeling in axon terminals originating from hypoglossal afferents of two different lengths was determined by two sets of experiments: high concentration tracer experiments and low concentration
259
260
BORKE AND NAU
tracer experiments. In addition, two other sets of experiments were performed: lesion-HRP and control experiments. Table I provides a summary of the numbers of animals used for each set of experiments and the protocols of each set of experiments are detailed below. High Concentrution
Tracrr Experimrnts
In the first set of experiments, rats were anesthetized with sodium pentobarbital (30 mgikg) or 7% chioral hydrate (5 ml/kg) and WGA-HRP (18%) or HRP (IS-30%) was injected stereotaxically through a glass micropipette (2&40 pm tip diameter) into one of two sources of hypoglossal afferents: RPc (1 .kl. 1 mm lateral, 0.7-0.8 mm anterior and a depth of 0.6 mm from the reference point, the obex) or the Spinal V nucleus (2.5-2.6 mm lateral, 0.7-0.8 mm anterior and at a depth of 0.C0.5 mm from the reference point, the obex). At each site, 4 rats were injected with 1.8% WGA-HRP. Survival times were I and 2-3 days with two specimens being used at each interval per injection site (Table 1). Longer survival periods were omitted to avoid the possibility of labeling second and third order neurons by transneuronal transport of WGA-HRP. A total of 7 rats for each site received pressure injections of 15-3C% HRP. Of these specimens, 2 rats per injection site survived 2-3, 5 and 7 days after enzyme delivery and one rat per injection site was killed one day after administering the tracer (Table 1).
In a subsequent set of experiments, the concentration of the tracer injected into the two sites of hypoglossal afferents was reduced. A two step reduction in tracer concentration was carried out for RPc injections. In one series of rats, WGA-HRP or HRP in 0.5% concentrations was injected into RPc using the same method and amount of tracers as those of the high concentration tracer experiments. Likewise, the same survival intervals were repeated for these experiments. The number of animals for each survival interval appears in Table 1. The second series received injections of 0.16% WGA-HRP or 0.25% HRP into RPc. All parameters conformed to those of the previous experiments except survival times, Survival periods of 1 and 2-3 days were utilized and the number of rats for each time period is indicated in Table i. The lowest concentrations of the tracers (0.16% WGAHRP and 0.25% HRP) were also injected into the Spinal V nucleus. Survival times and numbers of specimens corresponded to those in which the enzyme was injected into RPc. Lesion-IIRP
~.~p~ri~l~~llts
In another series of experiments, electrolytic lesions [21, comparable in size and location to the tracer injections were
t
TABLE
Type
of Experiment
Type of Injection
High Concen-
1.8% WGA-
tration Tracer Low Conoen-
I
Survival Time (days) 2-3 5 7
*2
2
HRP
I
2
2(O)
2(O)
WGA-
2(O) 4
4(O) 4
HRP
3
6
f(O)
2
15-30%
HRP
0.5% WGA-
tration
HRP
Tracer
0.5% HRP 0.16%
2
2
2(O)
2(O)
l(O)
Ito)
i(O)
2
2
2
HRP 0.25% Lesion-HRP
Electrolytic Lesion
plus
I s-30% HRP Control
0.
I
M Tris
Buffer *Number in each block represents number of animals utilized for RPc injections. For Spinal V injections same numbers were used unless indicated otherwise by ( ). N=90.
produced in RPc. In 4 rats, HRP (15-30%) in small volumes (40-50 nlf was injected via a glass micropipette at the lesion site 5-45 min after lesion production. Survival periods and numbers of animals are indicated in Table 1.
Block staining with uranyl acetate was not performed during the processing of tissue for electron microscopy and ultrathin sections were not post-stained with uranyl and lead salts. While these precautions decreased the possibility of misidentification of normally occurring membrane-bound granules with electron dense contents as containing HRP reaction product, these measures did not eliminate the possibility. Therefore a series of control rats were included in the experimental design. Sixteen rats were used for the control series: two rats for each of the four survival times (1, 2-3, 5, 7 days) per injection site (Table 1). All procedures were identical to those followed for the experimental groups except that a small volume (40 nl) of the tracer vehicle, 0.1 M Tris buffer, was injected into RPc or the Sp. V nucleus. At predetermined survival times, experimental and control animals were reanesthetized and perfused transcardially
FACING PAGE FIG.
I. A. A dark-field micrograph of an injection site in RPc, lateral to the hypoglossai nucleus (XII). high concentration tracer experiment 2 days survival, x28. B. A dark-field micrograph of an injection site in the dorsal portion of the Spinal V nucleus. Distance to the target site. the hypoglossal nucleus (XII) is 2 times that of RPc, high concentration tracer experiment (IS% HRP). 2 days survival. x23. C. Bright-field micrograph of solid (arrowheads) and granular (arrows) types of HRP labeling of neurons at injection site in RPc, high concentration tracer experiment (30% HRP), 2 days survival, x213. D. Unlabeled axon terminals (AT) surround a labeled endstructure in which diffuse. agranular reaction product is applied to the membranes of organelles in the terminal of a short projection neuron to the Xltth nucleus. high concentration tracer (1.8% WGA-HRP) injected into RPc, I day survival, unstained, x38000. E. Another type of anterograde label consisting of membrane-bound granules of HRP reaction product (arrows) was seen less frequently in axon terminals of short projection neurons (RPc) to the Xifth nucleus, high concentration tracer (30% HRP) injected into RPc, 2 days survival, unstained, x47120. F. Axon terminals (AT) of long projection neurons (Sp. V) contained almost exclusively membrane-bound, granular labeled profiles kwrows). high concent~tion tracei (3G% HRP) injected into Sp. V, 2 days survival, ~39000.
( 15%HRP).
NATURE OF ANTEROGRADE
LABELING WITH HRP
261
762
BORKE AND NAU
with Ringer’s solution, dilute and concentrated aldehydes [ 191 and a final solution of 0.1 M phosphate buffer. The brains were removed, the medulla was isolated and mounted in agar. Coronal sections, 50 pm in thickness, were cut on a vibratome. Sample sections were stained with Fink-Heimer and Nauta silver stains [5]. Remaining sections were reacted histochemically for HRP according to the CO-GOD method [8]. Alternate reacted sections were mounted on gelatinized slides and stained with neutral red for light microscopic evaluation of the injection site and localization of areas of anterograde label. Remaining reacted sections were postfixed in osmium tetroxide, dehydrated and embedded in Epon. Processing for electron microscopy did not include block staining with uranyl acetate. Epon embedded blocks were trimmed to include only the XI&h nucleus and ultrathin sections of silver interference were cut and left unstained for electron microscopic evaluation. For the high (1.8% WGA-HRP or 1530% HRP) and iowest (0.16% WGA-HRP or 0.25% HRP) concentration experiments, the proportion of diffuse agranular, membrane-bound granular and axon terminals containing both types of label was determined for 500 randomly selected terminals originating from neurons in each of the two sources of hypoglossal afferents. From the experiments in which 0.5% HRP or WGA-HRP was injected into RPc, 200 axon terminals were classified according to the same labeling patterns. RESULTS
Light microscopic evaluation. Micropipette delivery of the enzyme HRP into RPc (Fig. IA) and Sp. V (Fig. 1B) nucleus resulted in injection sites measuring 1000 microns or less in diameter. Uptake in neuronal somata appeared confined within the limits of the two sources of hypoglossal afferents. Numerous labeled ceil bodies and processes of RPc and Sp. V. neurons were identified at the respective injection sites. Solid and granular labeling occurred in neuronal somata (Fig. IC). Nerve fibers traced medially from the injection sites appeared to be filled with HRP reaction product and terminal label was seen in the neuropii of the hypogiossal nucleus. Anterograde degeneration of axons and axon terminals, originating from RPc or Sp. V neurons was not detected in adjacent vibratome sections unreacted for HRP and silver-stained with Fink-Heimer and Nauta methods. Elwtron microscopic evaluation. Ultrastructural examination disclosed that the HRP reaction product in the XIIth nucleus was found almost exclusively in myelinated axons FACING
and their terminal endings. After RPc injections, a diffuse, agranular reaction product could be seen on the external surface of organelles contained within the majority of labeled myeiinated axons. After Sp. V injections, vesicular and tubular profiles of membrane-bound label were found primarily in labeled myeiinated axons. Axon terminals originating from RPc neurons exhibited two distinct patterns of distribution of HRP reaction product. In the majority (8%) of terminals, an electron dense label was applied to the cytomembranes of the organelles (Fig. ID). Membranebound, granular label was detected in 10% of the labeled terminals (Fig. 1E) and 1% demonstrated both types of label. These labeling patterns were identified at all survival times. In contrast to these results, almost all (98%) of the axon terminals labeled from Sp. V injections contained the membrane-bound, granular type of Iabel (Fig. IF).
Light microscopic~ r,ulurrtiou. Injection sites were comparable in diameter to those obtained from the high concentration tracer experiments. Uptake appeared limited to neurons of RPc and the Sp. V nucleus. NeuronaI somata. at both sites contained chiefly granular labeling (Fig. 2A), No anterograde axonal or terminal degeneration was seen in adjacent silver-stained sections. Ekwron microscopic c~,wlrrrrtiot~. When O.S% WGA-HRP or HRP was injected into RPc, 40% of the labeled terminals were of the membrane-bound, granuiar type and both types of label were found within 22% of the labeled terminals (Fig. 2B). Myelinated axons situated in the neuropil of the XIIth nucleus demonstrated vesicular and tubular forms of granular labeling after injections of very dilute concentrations of WGA-HRP (0.16%) or HRP (0.25%) into RPc or Sp. V nucleus. Using these concentrations, 84% of labeled terminals from RPc neurons contained membrane-bound, granular reaction product (Figs. 2C,D). Terminals labeled with the diffuse, agranular reaction product constituted 7% of the labeled population and those with both types of label made up the remaining 9%. Diffuse, agranular and double-labeled terminals were rare (1%) after the same injections into Sp. V nucleus.
Light
microscr~pic~
comparable
c~drrrrtiorl.
Lesions
produced
were
in size (- I mm) and location to those injections
HRP into RPc in the previous experiments. Neuronal somata, remaining at the lesion site. were rarely seen. Injection of HRP 5-45 min after lesion production resufted in labeling primarily in nerve cell processes of RPc. of
PAGE
FIG. 2. A. Granular labeling of neurons at the injection site (RPc) occurred when the concentration of the tracer wah reduced to O.Z’/r HRP. 2 days survival. x300. B. Unlabeled axon terminals (AT) flank a terminal which contains both types of HRP labql. diffuse, agranular,and membrane-bound, granular (arrows) when the concentration of the tracer was reduced to 0.5% HRP and injected mto RF?, I day survival. unstained. x42900. C. When the concentration of the enzyme injected into RPc was reduced more than 100 fold to 0.25% HRP. anterogradc label was stilt visible and the majority of labeled terminals in the XIfth nucleus contained membrane-botlnd. granular label (arrows). Z days survival, unstained, x30400. D, Membrane-bound. granular type of HRP labeling in tubular (arrows) and vesicular iarrowheadsf organelles in an axon terminal of a short projection neuron to the Xllth nucleus when 0.16% WGA-HRP was injected into RPc. I day survival. unstained. x 31920. E. An axon terminal containing degenerative debris (“I is labeled with diffuse. agranular type of HRP reaction product. Radio frequency lesion in RPc + 30% HRP injected into RPc 30 min post lesion. I day survival, unstained. x26910. F. An axon terminal exhibiting early degenerative changes including an apparent decrease in the number of synaptic vesicles and autophagic-like vacuoles contain\ membrane-bound, granular labeling. The difference in electron density of unlabeled dense cored vesicles (arrowheads) and HRP labeled vesicular profiles (arrows) can be easily distinguished. Radio frequency lesion in RPc + 3@% HRP injected into lesion site 20 min post lesion. 2 day survival, x33440.
NATURE OF ANTEROGRADE
LABELING WITH HRP
263
264
BORKE AND NAU
Anterograde axonal and terminal degeneration in the XIIth nucleus were identified consistently in adjacent silverstained sections of all four cases. Electron miuoscwpic~ ~waluation. For each survival period of the rats injected with HRP after lesioning, ultrastructural features characteristic of terminal degeneration were apparent in the majority of the labeled terminals. Degenerative features included marked decrease in the number of synaptic vesicles, hypertrophy of the axoplasmic reticulum, mitochondrial degeneration, autophagic-type vacuoles and loss of synaptic connections with engulfment by glial processes. Both types of label, diffuse, agranular (Fig. 2E) and membrane-bound, granular (Fig. 2F) were detected in terminals that demonstrated at least two of the forementioned degenerative criteria. In a single case, small amounts of HRP reaction product could be identified in the extracellular space of the neuropil at the lateral aspect of the hypoglossal nucleus.
Injection of 0.1 M Tris buffer into RPc and the Sp. V nucleus and histochemical treatment of sections with the CO-GOD reaction for HRP produced peroxidatic activity only within erythrocytes at the injection site and associated with sinusoids of area postrema. Peroxidatic activity was not noted in neuronal perikarya and axons surrounding the injection site. Membrane-bound granules with electron dense contents (dense core vesicles) were seen in some axon terminals of the hypoglossal nucleus. These membrane-bound, granules resembled in shape and size some of the granular HRP label in axon terminals of the tracer experiments. However, the electron density of membrane-bound granules in axon terminals of HRP injected material (Fig. 2F) was markedly intensified and served as a reliable criterion for distinguishing granules containing HRP reaction product from unlabeled dense core vesicles (Fig. 2F). DISCUSSION
The current results represent the initial ultrastructural demonstration that the length of the CNS pathway and the concentration of the tracer may be interrelated factors influencing the type of HRP distribution in anterogradely labeled axon terminals of certain projection neurons. Diffuse, agranular labeling in axon terminals is more likely to occur if high concentrations of tracer (1.8% WGA-HRP or 15-30% HRP) are used to label axon terminals in the hypoglossal nucleus originating from short projection neurons in RPc. This appearance of HRP reaction product, applied to the cytomembranes of organelles within the axon terminals commonly has been interpreted to result from passive intracytoplasmic filling of irreversibly damaged somata and axons with HRP [1,6], although this type of label has been found infrequently after somatic uptake of intact CNS neurons [7]. High concentrations of the tracer may cause mechanical disruption of the membranes facilitating the somatofugal transfer of the enzyme from RPc to axonal endings in the hypoglossal nucleus. The degree of intactness may be questioned but the negative light microscopic fmdings obtained from adjacent sections silver-stained with Nauta and Fink-Heimer suggest that the insult did not result in anterograde degeneration in the hypoglossal nucleus. Moreover, at the ultrastructural level, signs of anterograde degeneration were not apparent in labeled axon terminals even after post injection times of 5 and 7 days.
Reducing the concentration of the tracer increased the proportion of axon terminals in the XIIth nucleus, demonstrating membrane-bound, granular labeling after HRP injections into RPc. The lowest concentration of tracer (0.16% WGA-HRP or 0.25% HRP) resulted in membrane-bound, granular labeling in the majority of axon terminals, This kind of label commonly represents the incorporation and anterograde transport of the tracer by intact neurons [ 12, 13, 141. Thus, data gathered from injections of the enzyme HRP into RPc in the first two sets of experiments provided evidence that if the distance between the point of application of the enzyme and the site of axonal termination is short, the concentration of the tracer has a distinct influence on the pattern of anterograde labeling. Since the prerequisite for diffuse, agranular label is presumed to be irreversible somatal or axonal damage, electrolytic lesions were placed in RPc and HRP (15-30%) was injected at the lesion site. In these cases, the enzyme would enter the damaged ends of neuronal processes and anterogradely label those axon terminals destined to undergo degeneration. A single case, in which HRP was seen in the extracellular space of the hypoglossal neuropil was subsequently eliminated as HRP could have labeled retrogradely axon terminals belonging to both intact and irreversibly damaged neurons. Both types of label, diffuse, agranular and membrane-bound, granular were apparent in axon terminals that demonstrated several degenerative features. These findings indicated that axon terminals of short projection neurons, irreversibly damaged, are not associated exclusively with a particular type of label, namely the diffuse, agranular kind and are in agreement with occasional reports of membrane-bound, granular labeling occurring in permanently damaged neurons [ 10,l I]. The fact that characteristic signs of axonal and terminal degeneration were identified at the light microscopic level from silver-stained sections and detected ultrastructurally in labeled HRP terminals but were not seen after HRP injections into RPc without prior lesioning lends credibility to the idea that somatofugal transfer of the enzyme in those experiments without lesion production was carried out by intact or reversibly damaged neurons. However, these results cannot offer conclusive evidence for this thesis. Anterograde degeneration produced incidentally by the pipette tip and/or the pressure of the injection may have occurred at different rates and survival times and escaped the sampling procedure used here. The labeling patterns obtained from the injections of HRP into RPc, however, did not constitute a common phenomenon of hypoglossal afferent labeling. When HRP was injected into the Sp. V nucleus, membrane-bound, granular labeling was found almost exclusively in axon terminals anterogradely labeled regardless of the concentration of the injected tracer. These results pointed out that for pathways of greater distances to a mutual efferent site, the concentration of the tracer was not a determinative factor of the type of HRP label in axon terminals. These experiments also dispelled the notion that HRP uptake by damaged axons of passage accounted for some of the labeling in the RPc experiments. Sp. V axons course through RPc to reach the hypoglossal nucleus [3] but anterograde labeling after Sp. V injections occurred in axon terminals with morphological features of synaptic organization that could be distinguished from those terminals labeled by RPc injections [2]. The processes involved in the exclusive visualization of the membranebound, granular label in axon terminals of long projection neurons (Sp. V) could not be ascertained from present data.
WITH HRP
265
Uptake by endocytosis and subsequent anterograde transport of the tracer in membrane-limited structures and/or access of the marker to the cytoplasm by temporary membrane interruption and eventual sequestration of the enzyme in tubular or vesicular organelles are possible means for the
somatofugal transfer of the enzyme [ 121. The pattern of HRP labeling in axon terminals of long and short projection neurons to the hypoglossal nucleus therefore may not reflect the nature of the somatofugal transfer of the enzyme.
NATURE
OF ANTEROGRADE
LABELING
ACKNOWLEDGEMENTS
The author is grateful to Dr. Malcolm B. Carpenter for his valuable suggestions and comments. The skilled typing of Mary Sias is gratefully acknowledged. This work was supported by the Department of Defense, Uniformed Services University of the Health Sciences, Department of Defense Grant CO 7019. The opinions or assertions contained herein are the private ones of the author and are not to be construed as official or reflecting the views of the DOD or the USUHS. The experiments reported herein were conducted according to the principles set forth in the “Guide for Care and Use of Laboratory Animals,” DHEW Pub. No. (NIH) 78-23.
REFERENCES I. Beattie, M. S., J. C. Bresnahan and J. S. King. Ultrastructural identification of dorsal root primary afferent terminals after anterograde filling with horseradish peroxidase. Bruin Res 153: 127-134, 1978. 2. Borke, R. C. and M. E. Nau. The ultrastructural identification of reticula-hypoglossal axon terminals anterogradely labeled with horseradish peroxidase. Bruin Res 337: 127-132, 1985. 3. Borke, R, C., M. E. Nau and R. L. Ringler, Jr. Brain stem afferents of hypoglossal neurons in the rat. Brain Ras 269: 47-55, 1983. 4. Carson, K. A. and M. M. Mesulam. Ultr~tructural
evidence in transganglionically transported horseradish mice that peroxidase-wheat germ agglutinin conjugate reaches intraspinal terminations of sensory neurons. Neurosci Lett 29: 201-206,
1982. 5. Ebbesson,
S. 0. E. The selective silver-impregnation of degenerating axons and their synaptic endings in non-mammalian species. In: ~tttt/~~tnp[~~ur~I ~~thf~~ls in N~~~~f~~~nut~~rn~~ edited by W. J. H. Nauta and S. 0. E. Ebbesson. New York: Springer-Verlag, 1970, pp. 132-161. 6. Gwyn, D. G., P. H. Wilkerson and R. A. Leslie. The ultrastructural identification of vagal terminals in the solitary nucleus of the cat after anterograde labeling with horseradish peroxidase. Netrrosti Lett 28: 139143, 1982. 7. Holstege, J. G. and J. J. Dekker. Electron microscopic identification of mammilla~ body terminals in the rat’s AV thalamic nucleus by means of anterograde transport of HRP. A quantitative comparison with EM degeneration and EM autoradiographic techniques. Npur-osci Lrtt II: 129-135, 1979, 8. Itoh. K.. A. Konishi. S. Nomura. N. Mizuno. Y. Nakamuraand T. Sugimoto. Application of coupled oxidation reaction to electron microscopic demonstration of horseradish peroxidase: cobalt-glucose oxidase method. Bruin RPS 175: 341-346, 1979. 9. Lynch, G., C. Gall, P. Mensah and C. Cotman. Horseradish peroxidase histochemistry: A new method for tracing efferent projections in the central nervous system. Bruin Rrs 65: 373380, 1974. IO. Malmgren, L. T. and Y. Olsson. Early influx of horseradish peroxidase into axons of the hypoglossal nerve during Wallerian degeneration. Naurosci Left 13: 13- 18. 1979.
II. Mawe, G. M., J. C. Bresnahan and M. S. Beattie. Ultrastructure of HRP-labeled neurons: A comparison of two sensitive techniques. Brrtin Res Buff 10: 551-558, 1983. 12. Mesulam, M. M. Principles of horseradish peroxidase neurohistochemistry and their applications for tracing neural pathways-Axonal transport, enzyme histochemistry and light microscopic analysis. In: Tracing Neural Connections, edited by M. M. Mesulam. New York: Wiley and Sons, 1982, pp. I-1.51. 13. Mesulam, M. M. and E. J. Mufson. The rapid anterograde transport of horseradish peroxidase. N~~ur[~.~~~i~nL,~5: 12771286, 1980. 14. Mizuno, N., A. Konishi, K. Itoh, N. Iwahori and Y. Nakamura. Identification of axon terminals of cerebella-olivary fibers in the cat: An electron microscopic study using anterograde horseradish peroxidase method. Neurosc,i Lm ZO: 1I-14, 1980. 15. Mizuno. N.. S. Nomura. K. Itoh. Y. Nakamura and A. Konishi. Commi~sur~ inte~e~ro~s for m~sticatin8 motoneurons: A light and electron microscopic study using the horseradish peroxidase tracer technique. Exp Nrurol 59: 2.54-262, 1978. 16. Mizuno. N.. Y. Yasui. S. Nomura. K. Itoh. A. Konishi. M. Takada and M. Kudo. A light and electron microscopic study of premotor neurons for the trigeminal motor nucleus. J Camp Nemo1 215: 290-298, 1983. 17. Oldfield, B. J. and E. M. McLachlan. Uptake and retrograde transport of HRP by axons of intact and damaged peripheral nerve trunks. NeUN>sci tetf 6: 135-141, 1977. 18. Phillipson, 0. T. Afferent projections to the ventral tegmental area of Tsai and interfascicular nucleus. A horseradish peroxidase study in the rat. J Camp Ncurol 187: 117-144, 1979. 19. Reese, T. S. and M. J. Karnovksv. Fine structural localization of a blood brain barrier to exogenous peroxidase. J Cell Bid 34: 207-217,
1967.
20. Staines, W. A., H. Kimura, H. C. Fibiger and E. G. McGeer. Peroxidase-labeled lectin as a neuroanatomical tracer: Evaluation in a CNS pathway. Brain Rrs 197: 485-490, 1980.