Fluoro-gold: a new fluorescent retrograde axonal tracer with numerous unique properties

Brain Research, 377 (1986) 147-154 Elsevier

147

BRE 21574

Short Communications Fluoro-Gold: a new fluorescent retrograde axonal tracer with numerous unique properties LAURENCE C. SCHMUED and JAMES H. FALLON Departmentof Anatomy, School of Medicine, Universityof California, Irvine, CA 92717 (U.S.A.) (Accepted February llth, 1986) Key words: Fluoro-Gold - - retrograde tracer - - fluorescent dye - - multiple label A new fluorescent dye, Fluoro-Gold, has been demonstrated to undergo retrograde axonal transport. Its properties include (1) intense fluorescence, (2) extensive filling of dendrites, (3) high resistance to fading, (4) no uptake by intact undamaged fibers of passage, (5) no diffusion from labeled cells, (6) consistent and pure commercial source, (7) wide latitude of survival times and (8) compatibility with all other tested neuro-histochemical techniques.

Since the introduction of the H R P retrograde axonal tract tracing technique by Kristensson and Olsson 6, a number of alternative axonal tracers have been developed to aid the neuroscientist interested in studying neuronal connectivity. Many of these new tracers are fluorescent organic dyes, many of which were identified by Kuypers and his associates 4,5,14. Some of these fluorochromes can be combined with other retrograde fluorescent dyes to study the pattern of axonal collateralization 5:4, or with immunocytochemistry 9'15 to determine simultaneously the projection and chemistry of a neuron. Despite their superior sensitivity, limitations associated with these dyes include rapid fading, uptake by fibers of passage, limited cellular definition and diffusion of the dye from labeled neurons. It was hoped that some of these drawbacks would be eliminated when Schmued and Swanson 12 reported that the reactive fluorochrome SITS could be used to demonstrate retrograde axonal transport, that it is not taken up by fibers of passage, is resistant to fading, will not diffuse from the cell despite wide survival latitudes and is generally compatible with other neurohistochemical techniques. The major drawback with SITS is the variability and heterogeneity of commercial prepara-

tions. Our thin layer chromatographic data, and that of others s, indicate that th6 best SITS preparation is actually at least 5 separate compounds and it is probably a relatively minor impurity which is actually transported retrogradely. Chemical characterization of this active component remains elusive, mainly due to the minute quantities available for analytical study. Substituted stilbenes other than SITS were therefore screened for biological activity. Two such compounds, Stilbene Gold and Fluoro-Gold were found to undergo retrograde transport. Fluoro-Gold exhibited a much more intense emission than StilbeneGold, whose properties were described elsewhere 11. In an attempt to avoid some of the problems associated with impure commercial SITS preparation we submitted 4 batches of Fluoro-Gold for independent qualitative analysis. All batches exceeded 99.2% purity. Additionally, our thin layer chromatographic data showed only one spot associated with all preparations. In this report we will outline a series of experiments which indicates that Fluoro-Gold exhibits many advantages not seen in the majority of presently existing retrograde tracers. All experiments were performed on fully anesthe-

Correspondence: L.C. Schmued, Department of Anatomy, School of Medicine, University of California, Irvine, CA 92717, U.S.A. 0006-8993/86/$03.50 (~) 1986 Elsevier Science Publishers B.V. (Biomedical Division)

148

phate buffer at concentrations ranging from 1 to

niques commonly employed by neuroscientists. Other compatible techniques include combining FluoroGold with other fluorescent retrograde tracers as described by Kuypers and others 5'14, autoradiographic

10%. The dye was delivered in small volumes

demonstration of anterogradely transported 3H-la-

( 0 . 0 5 - 0 . 2 5 / A ) via a microsyringe, or with positive

beled macromolecules as described by Cowan et ai.l, traditional paraffin and plastic embedding procedures, H R P histochemistry according to Mesulam 7,

tized 250-300 g A l b i n o rats. Fluoro-Gold (Fluorochrome, Englewood, CO) was dissolved in distilled water, physiological saline, or 0.2 M neutral phos-

iontophoretic current (5 mA/10 min, 2% dye in lactated Ringer's solution) to various sites in the brain. Postoperative survival times ranged from 1 to 60 days. The rats were then perfused with 4% formaldehyde in 0.02 M neutral phosphate buffered saline.

A C h E histochemistry as described by Karnovsky and Roots 3, fluorescent counterstains as used by Schmued et al. 13, or combining with immunocyto-

Other fixatives can be employed as long as they do not contain high concentrations of heavy metals. The

chemistry for demonstration of specific antigens as described by Sawchenko and Swanson 9, or Van der

tracer was also compatible with unfixed tissue. Most tissue was cut at 20-30/~M on a freezing sliding mi-

Kooy and Steinbusch 15. In the first series of experiments, Fluoro-Gold was

crotome, although some sections were cut at 10 # M on a cryostat, at 6/~M following paraffin embedding, or at 2/~M after plastic embedding. Cut sections were m o u n t e d on gelatin coated slides, air dried, xylene cleared, coverslipped with D P X and observed under

injected into a variety of terminal fields in the central nervous system and retrogradely labeled neurons were observed in appropriate sites. For example, 0.1 /A of 2% Fluoro-Gold injected into the striatum led to vibrant labeling in the ventral tegmental area and the

the fluorescence microscope using wide band ultraviolet excitation (emission max: 408 nM, excitation max: 323 nM). Fluorescent immunocytochemical preparations were coverslipped with buffered

pars compacta of the substantia nigra after a survival time of 4 days. Similarly, injections of Fluoro-Gold

glycerol. The final phase of this study involved combining Fluoro-Gold with a variety of histochemical tech-

into the sensory-motor cortex led to massive retrograde labeling in the ventral nuclei of the thalamus and in the contralateral hemisphere, injections in the suprachiasmatic nucleus resulted in retrograde labeling in the ventral lateral geniculate nucleus, injec-

Fig. 1. A: a typical pressure injection of 100 nl of 4% Fluoro-Gold into the striatum. Three distinct zones can be resolved. Centrally (1) is the needle tract and small region of necrosis. Peripheral to this (2) is the bright region in which both neuropil and cells fluoresce, while myelinated fascicles exhibit little tracer. This mid-region is presumably the zone of active terminal uptake. The most peripheral zone (3) exhibits weak cellular label. 51 x, U.V. excitation. B: motor cortex layer V pyramidal cells label retrogradely following injection of ventral horn cells at the lumbar region of the spinal cord. This demonstrates that Fluoro-Gold is suitable for demonstrating long distance connections, and is capable of extensively filling axons (solid arrow) and distal dendrites (open arrow). The superimposition of many vesicles may result in an apparently homogeneously stained cytoplasm with ethidium bromide counterstain. 328 ×, combined green U.V. excitation. C: a gigantocellular neuron of the reticular formation indicates weak retrograde label following spinal cord injection. This weaker label allows visualization of the individual cytoplasmic granules. Also typical is stain in the nucleolus but not in the nucleus. D: a horizontal section of lumbar spinal cord ventral horn motor neurons which indicate uptake and retrograde transport of Fluoro-Gold following soaking the cut sciatic nerve in the tracer solution for two hours and allowing a 4-day survival period. Soaking the intact sciatic nerve in the tracer did not result in any central label. 123 x, U.V. excitation. E: labeled axons can be observed traversing somatosensory cortex following injection of Fluoro-Gold into the contralateral cortex. Although these fibers have not been followed back to their cells of origin. 82x, U.V. excitation. F: horizontal section of the red nucleus following spinal cord injection of Fluoro-Gold. The tissue was then processed for paraffin embedding. Prolonged exposure to organic solvents and heat did not diminish the intensity or fastness of the fluorescence. The sections were mounted with low pH buffer, resulting in a blue cast. 128x, U.V. excitation. G: multiple label can be seen in the medial substantia nigra following injection of HRP into the caudate-putamen and FluoroGold into the nucleus accumbens. Single labeled cells can be seen with HRP (solid arrow) and Fluoro-Gold (open arrow) is also visible in the zona compacta region. A double labeled neuron (arrowhead) is also visible. The zone reticulata (r) indicates many anterogradely HRP labeled striatal-nigral fibers. 1~23x, combined U.V./brightfield illumination. H: these single and double labeled neurons in the pontine reticular formation are the result of Fluoro-Gold injected into spinal cord and [3H]adenosineinjected into motor cortex. After a one-week survival, the tissue was processed for autoradiography. The cell on the left contains only Fluoro-Gold, while the cell on the right contains the dye as well as transneuronally transported [3H]adenosine as indicated by overlying black silver grains. Combined brightfield/U.V, illumination may be used as in G. 328x, U.V. illumination.

i

i

150 tions in the dentate gyrus resulted in labeled cells in the contralateral hilar region and injections at thoracic and lumbar levels of the spinal cord labeled cells in a wide variety of sites from the medullary reticular formation (Fig. 1C) to the cerebral cortex (Fig. 1B). No cases were found in which Fluoro-Gold failed to label a known projection. No cases were found in which Fluoro-Gold failed to label a known projection. The appearance of the retrogradely transported label is characterized by fluorescent gold colored granules within the cytoplasm and processes (Figs. 1C,D). Incubating the tissue briefly at a low pH (acetate-HC1 buffer, pH 3.5) will shift the emission color from gold to blue (Fig. 1F). The gold color is characteristic of the tracer at neutral and basic pH. Plasma membranes and the nucleolus also sequester some of the dye. The nucleus, however, is devoid of stain. Likewise, adjacent neurons and glia cells show an absence of tracer regardless of the survival times used. Shorter survival times (1-2 days) typically show distinct vesicles within the cytoplasm and axon. Longer survival periods (4 days-4 weeks) result in an accumulation of vesicles in the soma and an extensive filling of dendritic processes, which could often be followed for hundreds of microns past numerous branches throughout their arborizations (Fig. 1B). Injections of high concentrations (5-10%) of large volumes (0.2-0.5/zl) result in massive and intense retrograde labeling, but may also produce undesirable necrosis at the injection site. Good retrograde label with negligible lesion at the injection site can be obtained with lower concentrations ( 2 - 3 % ) of the dye. Mixing the dye in neutral 0.2 M phosphate buffer results in the formation of a suspension which exhibits slower uptake and diffusion. Prolonged exposure of labeled tissue to ultraviolet light causes little fading, despite months of intermittent examination. The injection sites (Fig. 1A) typically show a small necrotic zone adjacent to the needle tract surrounded by a well-circumscribed area in which all tissue components exhibit a brilliant fluorescence. Outside this region is a limited zone in which there is some cellular staining. Myelinated fiber tracts near the injection site, such as the anterior commissure or fascicles of the caudate-putamen, do not appear to sequester tracer. This latter finding relates to the second part of this

study in which we attempted to determine some of the uptake and transport characteristics of FluoroGold. In the first experiments, the tracer was injected into either cut or intact fiber tracts. Included were the corpus callosum, fimbria and sciatic nerve. Cut or damaged axons appear to take up and transport the dye retrogradely (Fig. 1D) while intact fibers show no uptake and subsequent retrograde transport. Thus, label can be found in relatively few neurons in the rostral hippocampus only following large volume or high concentration injections, which produce a visible lesion. Fluoro-Gold was also injected directly into the lateral ventricle, which, in high enough concentrations, could produce a hazy fluorescent zone in areas immediately surrounding the ventricle. This diffuse, primarily astrocytic and ependymal label is easy to differentiate from retrograde neuronal label and we have seen no evidence of subsequent retrograde transport of the tracer from these circumventricular regions. When the dye was injected intraperitoneally (40 mg/kg), it could be visualized an hour later throughout brain endothelial cells, but unlike SITS 1°, Fluoro-Gold does not appear to pass the blood-brain barrier. We also addressed the question of possible bi-directional transport of this tracer, as has been observed for H R P and many other fluorescent dyes. Injections into the striatum result in punctate labeling in the pars reticulata of the substantia nigra, while motor cortex injections result in axonal label in the pyramidal tract and contralateral cortex. This indicates that Fluoro-Gold, like many other tracers, can undergo bi-directional transport. In a final set of experiments we determined which other methodologies, commonly used by neuroscientists, were compatible with Fluoro-Gold. All methods tested were found to be compatible. [3HIAdenosine was injected into the motor cortex, while the contralateral spinal cord gray matter was injected with Fluoro-Gold. The animals were perfused 2 weeks later and either frozen sections (at 30-20 #M) or paraffin embedded sections (at 5-10 ktM) were processed according to standard autoradiographic procedures. Cells in the red nucleus and gigantocellular neurons of the brainstem reticular formation were covered with silver granules indicating anterograde transneuronal transport of [3H]adenosine from motor cortex (Fig. 1H). These same neurons also contained Fluoro-Gold transported retrogradely

151 from the spinal cord. HRP histochemistry was also compatible. When HRP was injected into the nucleus accumbens and Fluoro-Gold into the striatum, populations of both double and single labeled neurons could be found in the substantia nigra, pars compacta, while HRP-containing terminals could be observed in the pars reticulata (Fig. 1G). Simultaneous visualization of TMB reaction product and the dye fluorescence requires a combination of brightfield and ultraviolet darkfield illumination. Fluoro-Gold is also fully compatible with immunocytochemical techniques (Figs. 2 E - H ) . Fluoroscein and Rhodamine both fluoresce in response to longer wavelength excitation than that required for Fluoro-Gold visualization. This allows for easy differentiation of the two fluorescent labels. Intense Fluoro-Gold label may, however, faintly bleed through some filters used for visualizing FITC. In such situations the Fluoro-Gold staining may be suppressed, as discussed shortly. PHA-L anterograde immunocytochemical tract tracing techniques, such as those described by Gerfen and Sawchenko 2, are compatible with Fluoro-Gold retrograde tracing, sometimes allowing visualization of anterogradely labeled axons and terminal-like boutons adjacent to retrogradely filled dendritic shafts of Fluoro-Gold labeled neurons (Fig. 2H). This combined anterograde/retrograde technique is useful for identifying potential synaptic sites where PHA-L (or WGA/HRP) filled axons and terminallike puncta closely abut on Fluoro-Gold filled somata and dendrites. The tracer will also withstand processing for paraffin or plastic embedding (Fig. 1F), allowing for thin sections with excellent preservation of morphological detail. Paraffin processing will, however, increase the background fluorescence of any tissue but the specific fluorescent signal is sufficiently intense that this is not a major problem. Fluoro-Gold may also be combined with other fluorescent retrogradely transported dyes. To demonstrate axon collateralization, we examined the collateralization of substantia nigra pars compacta neurons to both the nucleus accumbens and the caudate-putamen. Fluoro-Gold was injected into one terminal field, while either Propidium Iodide, (1-2 day survival) True Blue (2-14 days survival) or Bisbenzimide (12-24 h survival) was injected into another terminal zone. Multiple label with Propidium Iodide thus appears as red cytoplasm when excited with green light and gold

colored granules result from ultraviolet (U.V.) excitation of the Fluoro-Gold (Fig. 2C,D). When combined with True Blue, U.V. excitation reveals double labeled neurons as pale lavender or white (Fig. 2B). Bisbenzimide can be combined to yield an easily distinguishable double labeled cell following U.V. excitation, displaying a blue nucleus with gold cytoplasmic granules (Fig. 1A). Generally Bisbenzimide and Propidium Iodide were found to produce the most definitive double labeled neurons when combined with Fluoro-Gold. Photography of Fluoro-Gold required half the exposure time needed for comparable print densities of the aforementioned fluorescent tracers. Thus, Fluoro-Gold was judged to be the most intensly fluorescent of the tracers at least within the wavelengths recorded by Ektachrome or Tri-X-film. The Fluoro-Gold protocol can be modified or augmented to accommodate a wide variety of constraints imposed by experimental paradigm or by a second methodology. At times we reduce the Fluoro-Gold staining to insure against masking of a second weaker fluorochrome which is also excited by U.V. This suppression of label can be accomplished by cutting down either the survival time (2 days), dye concentration (1%), or volume injected (50 nl). Although the dye does not appear to leak out of labeled cells despite long survival times, by varying this period, one can obtain interesting differences in terms of the degree of axonal and dendritic fill. Thus, although there is no rigid protocol, we typically pressure inject 0.1 pl of a 2 - 4 % solution and allow for a 2-7 day survival period, using shorter survival times for shorter pathways and longer survival periods for large animals or greater dendritic fill. With regards to the mechanisms involving retrograde transport of Fluoro-Gold, little is known. We suspect that certain physical properties of the dye (charge, aqueous/lipid solubility, polarity, size, etc.) prevent the tracer from passively crossing the plasma membrane. It also seems probable that certain reactive groups facilitate active endocytotic vesicular uptake of the tracer at terminals. These pinocytotic vesicles seem to be retrogradely transported to the cell body where they appear to undergo little degradation. With longer survival times the tracer concentration within axons appears to be depleted. Concomitantly, the tracer accumulates in the cell body and may be actively transported distally along the den-

153 drites. In summary, F l u o r o - G o l d labeling is characterized by an intensely fluorescent golden color seen primarily within what appears to be vesicles, and secondarily associated with the plasma m e m b r a n e and the nucleolus of retrogradely labeled neurons. Extensive retrograde filling of dendrites characterizes Fluoro-

cytochemistry, plastic or paraffin embedding, and other fluorescent retrograde tracers. The tracer is also remarkably p e r m a n e n t , fading little over time or from ultraviolet irradiation. The aforementioned characteristics demonstrate the potential usefulness of Fluoro-Gold for n e u r o a n a t o m i c a l tract-tracing studies.

Gold label. No tracer appears to be taken up by intact u n d a m a g e d fibers of passage. Once the tracer is internalized and retrogradely transported, it will not leak out of labeled cells, despite prolonged sur-

This work was supported in part by N I H G r a n t NS15321 and by N I M H Predoctoral G r a n t M H O

vival periods. Survival periods can be of wide latitude, ranging from 2 days to 2 months. The perma-

9121. The authors wish to express their gratitude to Dr. Ken Shea and Mr. Greg Stoddard for spectro-fluorometric analysis, Dr. Kim Seroogy for immunocytochemical preparations and to Ms. Natalie Sepion

nence of the c o m p o u n d allows it to be c o m b i n e d with numerous other histochemical techniques including autoradiography, enzyme histochemistry, i m m u n o -

1 Cowan, W.M., Gottlieb, D.I., Hendrickson, A.E., Price, J.L. and Woolsey, T.A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 2 Gerfen, C.R. and Sawchenko, P.E., An anterograde neuroanatomical tracing method that shows 'the detailed morphology of neurons, their axons and terminals: immunocytochemical localization of an axonally transported plant lectin, phaseolus vulgaris leucoagglutinin (PHA-L), Brain Research, 290 (1984) 219-238.

for typing the manuscript.

3 Karnovsky, M.J. and Roots, L., A 'direct coloring' thiocholine method for cholinesterases, J. Histochem. Cytochem., 12 (1964) 219-221. 4 Keizer, K., Kuypers, H.G.J.M., Huisman, A.M. and Dann, O., Diamidino yellow dihydrochloride (DY.2HCI): a new fluorescent retrograde neuronal tracer which migrates only very slowly out of the cell, Exp. Brain Res., 51 (1983) 179-191. 5 Kuypers, H.G.J.M., Bentivoglio, M., Catsman-Berrevoets, C.E. and Bharos, A.T., Double retrograde neuronal

Fig. 2. A: a cell in the medial substantia nigra indicates collateralization to the striatum and nucleus accumbens where Fluoro-Gold and Bisbenzimide were injected respectively. A double labeled cell with a blue nucleus and gold cytoplasmic granules makes the two tracers easy to distinguish. Adjacent single labeled Bisbenzimide cells are most likely the result of leakage, which is not seen with Fluoro-Gold. 123x, U.V. excitation. B: similar paradigm to A except True Blue is injected into the striatum. Three neurons have sequestered retrograde tracer. Single label with Fluoro-Gold is indicated by solid arrow, single True Blue label is marked with an open arrow, while a double labeled neuron appears as off-white (arrowhead). Section counterstained with Ethidium Bromide. 128x, combined U.V./green light illumination. C: Propidium Iodide is retrogradely transported from the striatum to label these two cells in the substantia nigra. The cell on the left also sends an axon collateral to the nucleus accumbens which was injected with Fluoro-Gold. This is demonstrated in D. 328x, green light excitation. D: same view as Fig. 2C with ultraviolet illumination. It is apparent from comparing C and D that the cell on the left contains both Fluoro-Gold and Propidium Iodide. Once again the more extensive cellular filling with the Fluoro-Gold compared to the other tracers is observed. 328x, U.V. excitation. E-G: these figures illustrate retrograde labeling in the substantia nigra with Fluoro-Gold combined with immunocytochemistry. E: the Fluoro-Gold labeled cells under ultraviolet illumination. F: represents the same section which has been processed immunocytochemically. Cells containing tyrosine hydroxylase immunoreactivity are indicated by a FITC tag which is visible with blue light excitation. G: the same section which has also been treated immunocytochemically to reveal cells containing the peptide CCK as indicated by a rhodamine conjugated secondary antibody viewed with green light. This series indicates that Fluoro-Gold is useful in multiple labeling experiments involving both Rhodamine and Fluorescein as a second fluorescent tag. Some cells can be observed to contain only the retrograde tracer (solid arrow), others indicate only antibody labeling (open arrow) while the majority indicate colocalization of all 3 tracers (arrowhead). A process containing FluoroGold and tyrosine hydroxylase, but not CCK, is indicated by curved arrow. 205x, H: another combined immunocytochemical/retrograde transport methodology is demonstrated. Here the kidney bean lectin, PHA-L, is iontophoresed into the pallidum at the same time as Fluoro-Gold is injected into the spinal cord. The sections were subsequently processed for immunofluorescent identification of axons which contain the anterogradely transported lectin. This field shows such a retrogradely labeled cell in the rostral pontine reticular formation which is surrounded by presumably descending pallidal axons. An arrow indicates a potential synaptic site where PHA-L containing terminals appear to align with the dendrite of a cell retrogradely filled with Fluoro-Gold. 128x, combined U.V./blue light illumination.

154 labeling through divergent axons, using two fluorescent tracers with the same excitation wavelengths which label different features of the cell, Exp. Brain Res., 40 (1980) 383-392. 6 Kristensson, K. and Olsson, Y., Retrograde axonal transport of protein, Brain Research, 29 (1971) 363-365. 7 Mesulam, M.-M., Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction product with superior sensitivity for visualizing neural afferents and efferents, J. Histochern. Cytochem., 26 (1978) 106-117. 8 Payne, J.M., Lawes, I.N.C., Proctor, G.B. and Horobin, R.W., Variation between different samples of SITS with respect to axonal transport and toxicity, Neurosci. Lett., 42 (1983) 229-234. 9 Sawchenko, P.E. and Swanson, L.W., A method for tracing biochemically defined pathways in the central nervous system using combined fluorescence retrograde transport and immunohistochemical techniques, Brain Research, 210 (1981) 31-51. 10 Schmued, L.C. and Fallon, J.H., Selective neuronal uptake of SITS and other related substituted stilbenes in vivo: a flu-

11

12

13

14

15

orescent whole cell staining technique, Brain Research, 346 (1985) 124-129. Schmued, L.C. and Fallon, J.H., Stilbene Gold: a new fluorescent retrograde axonal tracer with several unique advantages (abstr.), Anat. Rec., 211 3 (1985) 170A. Schmued, L.C. and Swanson, L.W., SITS: a covalently bound fluorescent retrograde tracer that does not appear to be taken up by fibers of passage, Brain Research, 249 (1982) 137-141. Schmued, L.C., Swanson, L.W. and Sawchenko, P.E., Some fluorescent counterstains for neuroanatomical studies, J. Histochem. Cytochem., 3012 (1982) 123-128. Van der Kooy, D., Kuypers, H.G.J.M. and Catsman-Berrevoets, C., Single mammillary body cells with divergent axon collaterals. Demonstration by a simple fluorescent retrograde double labeling technique in the rat, Brain Research, 158 (1978) 189-196. Van der Kooy, D. and Steinbusch, H.W.M., Simultaneous fluorescent retrograde axonal tracing and immunofluorescent characterization of neurons, J. Neurosci. Res., 5 (1980) 479-484.