BRAIN RESEARCH

ELSEVIER Brain Research 705 (1995) 315-324

Research report

Microglia in normal and regenerating visual pathways of the tench ( Tinca tinca L., 1758; Teleost)" a study with tomato lectin

Almudena Velasco, Elena Caminos, Elena Vecino, Juan M. Lara, Jose Aij6n * Departamento de Biologia Cellular y Patologia, Facultad de Medicina, Universidad de Salamanca, Avda. del Campo, Charro, s / n, 37007 Salamanca,

Spain

Accepted 12 September 1995

Abstract

We have studied the microglial cells in the normal and regenerating visual pathways of Tinca tinca (Cyprinid, Teleost) by using the lectin from Lycopersicum esculentum (tomato), which, in our case, has been demonstrated as a specific marker for teleost microglia. In the normal fish, there are tomato lectin positive microglial cells in the retina, optic nerve, and optic rectum. Following optic nerve crush, we observed a more extensive labeling of the microglia in the crushed optic nerve and in the contralateral optic tectum affecting the stratum opticum and stratum fibrosum et griseum superficiale. In both cases, there was an increase of rounded and i,~ss ramified microglial cells, and granular cells. This response of a more extensive labeling of microglial cells increases to a maximum at 2-3 weeks after the crush; the density of labeled microglial cells decreases after 3 months after crushing. However, in the retina no changes were observed after optic nerve crush. These results suggest that the microglial cells could play an important role in regeneration of fish optic pathway, as other neuroglial cells do.

Keywords: Fish; Microglia; Regeneration; Tomato lectin; Visual pathway

I. Introduct ion

Fish and amphibian retinal ganglion cells, in contrast to those of mammals, have the capacity for regenerating their axons after optic nerve lesion [2,12]. The axons of the visual system re-establish synaptic contacts with their pre- vious targets in the tectum, restoring vision [38,39].

The principal obstacles to axonal regeneration in the CNS of mammals appear to be the reactive astrocytes, oligodendrocytes and microglial cells which possess differ- ent factors that inhibit neurite growth. Reactive astrocytes form a glial scar in rat optic nerve that impedes axonal elongation [28], and oligodendrocytes possess myelin asso- ciated proteins that inhibit the axonal growth [6,10]. More- over, there are great differences between mammalian and lower vertebrates in the time-scale of removal of the degeneration products by means of microgliai cells and macrophages. In mammals, axonal debris disappeared in 5 months [3], whereas in lower vertebrates the removal of degenerating axons occurred rapidly within one week after optic nerve crush [9,46].

* Corresponding author. Fax: (34) (23) 294549.

0006-8993/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0006-8993(95)01204-4

Macrophages and microglial cells appear to play an important role in gliogenesis and neuronal repair. In mam- mals, they interact with the proliferation of astroglial cells by releasing several factors such as cytokines [17,18], and there are factors secreted by microglia involved in the response to lesion in the fish visual system [13]. The descriptions of microglial cells in the brain of the lower vertebrates have been scarce and these cells show a slightly different morphology to that of mammals [9,40]. Different techniques have been used to study microglial cells in vertebrates including silver impregnation techniques [35], enzyme histochemistry [7,20,29] and immunohistochem- istry [32,33]. In the last years, lectins have been utilized as histochemical markers for the study of microglial cells since they bind to specific sugar groups on most mem- branes [26,43]. Some examples are wheat germ agglutinin (WGA), Ricinus communis agglutinin (RCA120) , and Grif-

fonia simplicifolia B 4 isolectin (GSA-I B4) [8,26,44]. We have utilized a lectin obtained from Lycopersicum

esculentum (tomato), with affinity for poly-N-acetyl lac- tosamine sugar residues [47], since this has permitted the identification of ameboid and ramified microglial cells in postnatal and adult rat brain [1]. In this paper, we describe for the first time the morphology and distribution of

316 A. Velasco et al. /Brain Research 705 (1995) 315-324

microglial cells in the Tinca tinca visual system in normal conditions and after optic nerve crushing.

2. Materials and methods

In this study we used 15 adult tench (Tinca tinca L., 1758) with body weights between 100-150 g obtained from a commercial hatchery. The animals were deeply anaesthetized with 0.03% tricaine methanosulfonate (MS- 222, Sandoz). The left optic nerve was crushed at 1 mm from the eyeball with a fine watchmaker's forceps for about 3 s, with the right nerve serving as control. The fish were kept in glass tanks containing aerated tap water at 18-20°C at survival from 1 to 300 days following the optic nerve crush.

Before sacrifice, animals were deeply reanesthetized with MS222 and were then perfused transcardially with 0.68% NaC1 solution followed by 4% paraformaldehyde and 15% saturated picric acid in phosphate buffer 0.1 M, pH 7.4 (PB). Retina, optic nerve and brain were postfixed for 4 h at 4°C in the same fixative solution.

After fixation, tissue was washed in PB and immersed in 30% sucrose in PB overnight, frozen in liquid nitrogen and cut on a cryostat. Sections (20 /_~m) were mounted on gelatin-coated slides, washed with 0.1 M Tris-buffered saline (pH 7.4) (TBS) and processed for tomato lectin histochemistry as follows. Endogenous peroxidase activity was blocked with 2% H202 in 100% methanol for 10 rain. After rinsing in TBS twice for 10 min, sections were immersed in TBS with 1% Triton X-100 (TBS-Tx), the sections were incubated with biotinylated lectin (L-9389; Sigma, St Louis, MO) diluted to 6 p ,g /ml in TBS-Tx, either overnight at 4°C or for 2 h at room temperature. Sections were rinsed once in TBS-Tx and twice in TBS and incubated for 1 h at room temperature with avidin- labeled peroxidase (Vector) in a 1:250 dilution in TBS. After rinsing in TBS (three times for 10 rain), the peroxi- dase reaction product was visualized with 0.05% 3.3'-di- aminobenzidine tetrahydrochloride (DAB) and 0.025% hy- drogen peroxide in TBS for 10 min at room temperature. TBS was used for stopping the reaction, and subsequently the sections were dehydrated, cleared in xylene and mounted with Entellan.

The right optic nerve of lesioned tench and visual pathway (retina, optic nerve, optic tracts and optic rectum) of the normal tench were used as controls. Histochemical controls were made omitting (1) the tomato lectin, (2) the avidin-peroxidase complex and (3) using only the DAB solution.

3. Results

Our results showed that poly-N-acetyl lactosamine residues can be labeled in visual pathway of the normal

and lesioned fish by using tomato lectin histochemistry. Moreover, in certain regions of the fish visual system, tomato lectin labeling changes greatly in the microglial cells following injury to the optic nerve.

3.1. Normal tench

Tomato lectin labeled cells with a very irregular shape exhibiting ramified and ameboid morphologies are present. These cellular types are scattered but widely distributed throughout the retina, optic nerve, optic tract and optic tectum.

In the retina, only slight labeling occurs which appears as round shapes with few processes. These were found in the outer nuclear layer, outer plexiform layer, and the inner nuclear layer. This microglial labeling is distinct from that of other cells observed in the rest of the visual pathway.

Lectin-labeled microglial cells were distributed through- out the intact optic nerve although they present several morphologies. The major microglial population was consti- tuted by elongated cells with few processes. They were aligned with the fibers of the optic nerve and tract (Fig. la). Moreover, a few rounded cells were found which were labeled in the superficial part (Fig. 4b). Occasionally, cells with short processes were also labeled. Frequently, tomato lectin labeled cells were associated with the blood vessels.

Microglial cells were also observed in the diencephalon and every optic tectum levels (Fig. 2a, Fig. 3a). They had a ramified shape with short processes, and in several cases they had a perivascular position (Fig. 4a).

In the diencephalic visual pathway, tomato lectin-labeled cells were found in the nucleus pretectalis superficialis pars parvocellularis (NPSp), area pretectalis (AP), nucleus pretectalis superficialis pars magnocellularis (NPSm), nu- cleus opticus dorsolateralis (NODL) and nucleus pretec- talis (NP). Furthermore there were also many labeled cells throughout the optic tracts: tractus opticus dorsomedialis (trodm), tractus opticus ventrolateralis (trovl) and tractus opticus accesorius (troacc) (Fig. 2a).

In the optic tectum, tomato lectin binding was found on the borders between stratum periventriculare (SPV) and stratum album centrale (SAC), SAC and stratum griseum centrale (SGC) but ramified cells were also shown in the SGC, in the deeper half of the stratum fibrosum et griseum superficiale (SFGS), stratum opticum (SO) and limit with the stratum marginale (SM). In the SPV there were small rounded cells with little cytoplasm (Fig. 3a).

3.2. Lesioned tench

We observed a more extensive labeling of the microglia in the crushed optic nerve and both in the tract and contralateral optic tectum. There was also a variation in the morphology and localization of these cells at the cited places. However, the distribution of lectin-labeled cells in the retina was similar to that of the normal animals.

"mr/ 00[ = Jeff 'qou01 letUaOU oql U! pom0sqo leql Ol ~eI!tU!S SeA~ qo!qg~ qsruo oql aolje s,(ep 0EI le 0Aa0U o!ldo oql u! ~u!Ioqe I u!loo I oletuol p "uo.tso I oql Jo ol!s oql ul. ~u!Ioqe I oql JO ,(l!suolu! tuntu!xelAI "uo!so I aolje s,(ep 0E le ~u!qsnao 0,'d0u o!:ldo :o "~u!qsruo JOlJe s~(gp L]e lleqo~(o oql ol (O) iels!p pue (d) letU! xoad '~u!qsnao oql jo sop!s qloq uo siioo ig!I~O~O!tu oql jo aoqtunu oq:l u! oseoaou! oql pug (D) qsnao jo ol!s oql u! ~u!u!els oq~ u! oseo~oop ~ql :q "lleqo~(o oql ol lels!p (O) pug IetU!xoJd s! (d) 'uo!suolxo Iepneooalso~ oql lnoq~noaql SlIOO tuaoj!snj oql jo ~u!u!elS "oMou o!ldo IetUaou :e "uo!so I JolJe ~otu!l leA!AanS luo.loJJ!p le pue q~uol IetUaou u! u!loo I olgtuol oql JO ~u.qoqe I oql u! uo!~e!aeA oql ~U!A~OqS omOU 3.lldo oql Jo suo!]3os Ieu!pnl!~uo- 1 "I "~!rl

A

- 'o ~-7~,"

,

't J

a ;

,; 2,, ~

&

m

.II-

°:

$

-2

%

;¢ ,I¢

I li

l i ' k

'i -,

,, ~ ~

'.1~ P~'£-~[£ (~66I) ~OZ 1to-w~t UlVaff /"l v la oosVlaA "V

318 A. Velasco et al. /Brain Research 705 (1995) 315-324

a

!

°' ~

• o , , .

trodrn -

N O D L

k .

' N P S m

OT

-i

b

i

_ _

d

Fig. 2. Variation of the tomato lectin labeling in the diencephalic visual pathways and the optic tectum at different times after crush in transverse sections at the same levels, a: general view of the scattered and widely distributed microglial cells in the diencephalon and optic tectum (OT) of the normal tench. NODL, nucleus opticus dorsolateralis; NPSm, nucleus pretectalis superficialis pars magnocellularis; NPSp, nucleus pretectalis superficialis pars parvocellularis; trodm, tractus opticus dorsomedialis; trovl, tractus opticus ventrolateralis, b: the maximum intensity of the staining was reached at 15 days after crushing in the dorsomedial and ventrolateral tracts, in the NPSp and surrounding the NPSm. e: 45 days after crushing a similar pattern of labeling was found, d: at 150 days a decreasing of staining within the preoptic nuclei and optic tracts was found, reaching a similar pattern of labeling as in normal tench. Bar = 100 /xm.

A. Velasco et al. /Bra in Research 705 (1995) 315-324 319

SM S O ,"

SGC " k

S ~

I ~.

• /

It

rt

"s

.! ~.

d

6

4t •

,O

, V , ; ,

Fig. 3. Transverse sections of the optic tectum showing the variation of the tomato lectin in normal tectum (a) and contralateral tectum to the crushed optic nerve at different times (b, c, d). a: tomato lectin labeled ramified (small arrows) microglial cells in different strata in normal tench. SM, stratum marginale; SO, stratum opticum; SFGS, stratum fibrosum et griseum superficiale; SGC, stratum griseum centrale; SAC, stratum album centrale; SPV, stratum periventriculare, b: optic tectum at 15 days after crushing. The number of star shaped/ramified cells increased in the SO and SFGS. Small rounded cells were observed in the SPV. c: 45 days after lesion. Tomato lectin labeling remained in the same strata as those at 15 days but the cellular morphology was modified, d: decrease of the tomato lectin labeling at 150 days after crushing, similar to the control optic tectum. Bar = 100 /xm.

320 A. Velasco et al./Brain Research 705 (1995) 315-324

a

C

b f

t ~

p,.,o

W ~

1

e

f

~J, tO

g~

t

o ~

. . . . s t I ~

mE

4,

Fig. 4. Different cellular types labeled with the tomato lectin, a: cells of star (big arrow) and bipolar (small arrow) morphology in thc optic tectum of the normal tench, b: round microglial cells (small arrows) intensely stained in the normal optic nerve, c: granular cells in the lesioned optic nerve 4 days after crushing, d: vacuolar cells (big arrow) in the lesion area 7 days after crushing, e, f: ramified cells (small arrows) (e) in the optic nerve at 30 days after lesion and (f) in the SO and SFGS of optic tectum at 15 days. Bar - 50 /xm.

A. Velasco et al. /Bra in Research 705 (1995) 315-324 321

3.2.1. Optic nerue One day after optic nerve crush, the distribution of

tomato lectin-labeled cells resembled that of the normal tench. Four days after optic nerve crush the distribution of lectin-labeled cells changed. Tomato lectin stained granu- lar cells began to populate the crushed zone. They were concentrated mainly in the vicinity of folds of the optic nerve. Granular cells were of large size (14-22 /zm) and they contained intensely stained granules (Fig. 4c).

Seven days after the optic nerve was crushed, numerous labeled vacuolar cells (Fig. 4d) were located on both sides of the place of injury but in the damaged area these cells were weakly stained. Furthermore, the granular cells re- mained strongly labeled with the lectin 7 days after crush- ing. They were observed mainly between the injured area and the retina (Fig. lb). No obvious changes were ob- served throughout the optic nerve, except for the damaged area.

Thirty days after optic nerve crush, the pattern of staining with tomato lectin showed a change consistent with an accumulation of tomato lectin labeled cells. We could observe strongly labeled cells in the crushed area and in the areas directly adjacent to the site of injury (Fig. lc). These cells had a long shape with short processes similar to those found in normal optic nerve (Fig. 4e). In the rest of the optic nerve the number of the granular cells and vacuolar cells decreased and labeling was observed similar to that in the normal fish.

After 45 days, the staining began to decrease in the crushed area and on both sides of it. We found small, round cells and granular cells in the injured areas and we observed a disruption in the distribution of the axons at this time. Throughout the optic nerve, from this period onwards round and elongated cells similar to those in normal animals reappeared. The amount of granular cells and vacuolar m~croglial cells decreased and we could see labeling almost like that in normal animals around 120 days after crushing the optic nerve (Fig. ld).

3.2.2. Diencephalic uisual pathways and optic tectum Seven days after optic nerve damage, there was an

increase in the population of tomato lectin labeled cells within some diencephalic nuclei, optic tracts and in the optic tectum contralateral to the nerve crushed.

After 15 days, the increase in the number of labeled cells was at a maximum. In the diencephalon, the stained cells were concentrated mainly in optic tracts (trodm, trovl, troacc) and in the interior nuclei NPSp, AP, NP and NODL (Fig. 2b). In the contralateral hemitectum to the optic nerve crush there was also a great increase in the number of ramified cells throughout the rostrocaudal ex- tension of the stratum opticum and stratum fibrosum et griseum superficiale (Fig. 3b). Microglial cells with rami- fied morphologies (Fig. 4f) were localized in four levels within the stratum opticum and stratum fibrosum et gri- scum superfciale, and small rounded cells were found in

the SPV of the optic tectum. In the other strata normal labeling was observed to remain.

After 4 weeks post-injury, the number and localization of microglial cells remained the same while their processes were reduced, attaining rounded morphology (Fig. 2c, Fig. 3c). The density of labeled microglial cells began to decrease 60 days after crushing. Tomato lectin labeled cells in NODL disappeared completely, while the number of microglial cells in other nuclei was reduced. At this time there were more microglial cells in the optic tract and in the rostral and caudal parts of the optic tectum, but the number decreased considerably in the lateral part of the tectum.

At 150 days, no further differences occurred between the optic pathways of normal and injured fish (Fig. 2d, Fig. 3d). Few round and granular cells were present in the optic tracts and tectum and the ramified cells reappeared.

4. Discussion

Our results with tomato lectin histochemistry revealed that poly-N-acetyl lactosamine residues in the fish visual pathway are associated with a specifc population of glial cells identified as ameboid and ramified micrt, glial cells. Labeling in neurons and other glial cells was not observed. The use of tomato lectin provides a new fish microglial marker and we could assume that poly-N-acetyl lac- tosamine residues are present in fish microglial mem- branes. The histochemical study of microglial cells with lectins has revealed specific glycoconjugates in their mem- branes, including galactose [26,45], N-acetyl neuroaminic acid [36], and mannose [4]. The presence of poly-N-acetyl lactosamine in cell membranes is not uniform and several studies have demonstrated drastic changes under different conditions [27,34]. Polylactosaminoglycans in cell mem- branes play important recognition functions, carrying vari- ous antigenic structures such as blood group antigens [11]. The exact role of poly-N-acetyl lactosamine structures in microglial cells is not yet known.

4.1. Microglial cells in normal tench

Lectin binding to poly-N-acetyl lactosamine in the vi- sual pathway of the normal tench revealed an elevated number of microglial cells related to the whole cell popula- tion of the optic nerve which was similar to that described by Dowding et al. [9], which indicates that they constitute approximately at least 30% of all cells in the optic nerve.

In normal tench, we have found differences in the staining degree and in the morphologies of the microglial cells: scarce labeling in retina, fusiform cells with short processes together with round cells in the optic nerve and tract, and ramified cells in the encephalic visual pathway. Comparing our results with studies in mammalian [21,35,43] and lower vertebrates [9,30,40], we classified

322 A. Velasco et al./Brain Research 705 (1995) 315-324

the fusiform and ramified cells as resident/ramified mi- croglia, while round cells, granular cells and vacuolar structures observed in lesioned animals correspond with ameboid microglia or macrophages.

The distribution and morphology of microglia in the visual pathway of the Tinca tinca show small differences from those described in Oreochromis using FL-1 [9]. In the retina, tomato lectin-labeling was scarce in the outer nuclear layer, outer plexiform layer and inner nuclear layer but there was no staining in the inner plexiform layer and ganglionar cell layers. These cells were much smaller, poorly ramified and much less numerous than those found in Oreochromis [9], however, these results were similar to IB4-1abeled microglial cells in Pleurodeles waltl [30]. The scarcity of labeled cells within the retina could be due to the fact that tomato lectin does not stain all the microglial cells. Another possibility is that these tomato lectin labeled cells are a specific microglial subpopulation.

4.2. Microglial cells in lesioned tench

After optic nerve crush in Tinca tinca a dramatic in- crease was observed in the number of microglia/macro- phages, and changes were observed in the morphologies of the cells throughout the visual pathway with the exception of the retina where there was no modification. However, in Oreochromis retina [9] the microglial cells, which were concentrated mainly in the vicinity of the retinal ganglion cell layer, increased in number after optic nerve crush. The absence of modification in the retina labeling may be due to the fact that few ganglionar cells die after the lesion or as has been previously stated, because the labeling with tomato lectin does not affect all the retinal microglial cells.

In the injured area of the optic nerve, as well as lectin binding cell types found in normal tench, we observed granular cells and vacuolar structures which seem to be activated or reactive microglia in comparison with the morphology of these cells in mammals [16,35]. These activated cells were similar to large 5F4 + macrophages and labeled cells which contain pigment granules within the Xenopus optic nerve at the site of the lesion at 1 h after the crush [46]. In the tench at 30 days after crushing we found an elevated number of ramified cells in the site of lesion and on both sides of it.

In the diencephalic nuclei (NPSp, AP and NODL), diencephalic optic tracts and in the SO and SFGS of the optic tectum we also found ramified microglial cells con- centrated in the vicinity of degenerating retinal axons and their terminals, reaching a maximum concentration around 15 days after the lesion. Forty days after optic nerve crush, the processes of microglial cells were reduced but the accumulation of lectin-labeled cells remained in the same areas. This increase of the microglial cells in the optic nerve and in the SO and SFGS, was consistent with the observations of labeled microglia after crushing the optic nerve in Oreochromis using FL.1 [9], in goldfish with

toxic cobalt application [40] and in Pleurodeles waltl using the lectin IB-4 [30]. The accumulation of the mi- croglial cells after crushing was later in tench than in Oreochromis [9]. In tench some days after lesion the microglial cells remained with the same ramified morphol- ogy as in the normal state, however, in Oreochromis [9] after lesion the microglial cells had a round morphology. It is interesting that in Pleurodeles waltl both the chronol- ogy of the modification of the microglial cells and the affected areas coincide with our results [30].

With regard to the origin of the activated ameboid microglial cells, many researchers have proposed that blood-borne macrophages are those which respond, enter- ing the brain after CNS injury [23] but other investigators postulate that the microglial cells transform into ameboid microglia in response to the lesion [22,35,37,42]. Besides these cells extrinsic mononuclear phagocytes could be observed [15], which indicates that round cells found after damage within the lesion site were invading blood-borne monocytes [23]. There is the possibility of migration of the ameboid microglia from other sites to the lesion site [5,25,31,35]. It was not possible in our study to determine what is the origin of the different microglial types. It is probable that one part of the ameboid microglial cells observed after the crush might originate from monocytes which have penetrated from blood vessels into the optic nerve and optic tectum. In the case of the great number of ramified microglial cells that appeared at 15 days after crushing, these could originate from the mitotic prolifera- tion of the already existing microglia which later, around 45 days after crushing, lose their processes transforming into ameboid microglia.

The cells observed in the nuclei NPSp, AP and NODL, diencephalic optic tracts and SO and SFGS at 15 days and 30 days after crushing in the optic nerve, were probably activated microglial cells though they exhibited a ramified morphology. In larval and adult frogs [19,24] and goldfish [40], activated ramified microglial cells appeared after visual deafferentation in the degenerating neuropil. In con- trast, mammalian activated and reactive microglia reduce their processes, and only resident microglia possess rami- fied processes [41]. It is not clear whether or not ramified microglia participate in brain inflammatory responses after trauma [14,35]. In goldfish at 5 days after the lesion, ramified cells moved towards the ventricles, which sug- gests that these processes aid the microglial cells to move the debris of the degenerated axons towards the ventricles and the vascular elements. In our study this movement of microglial cells towards the ventricles was not observed though there are small rounded cells in the SPV which could correspond with these cells. However, it is possible that the encephalic vascularization which exists in the tench is sufficient for the elimination of the debris. As we have previously indicated, the metabolic processes in the tench in the experimental conditions employed in this work appear notably slow in relation to other species,

A. Velasco et al. /Brain Research 705 (1995) 315-324 323

which wou ld also inf luence the s low different ia t ion o f

ramif ied microg l ia to amebo id microgl ia . It is possible that

the r a m i f i e d / r e s i d e n t mic rog l i a prol i ferate and later part

o f these t ransform into amebo id cel ls on the arrival o f

amebo id mic rog l i a or ig inat ing in the rostral zones. This

wou ld suppose an anterograde m o v e m e n t a long the opt ic

pa thway f rom the zone o f lesion. The ramif ied microg l ia

could be a pr imary react ion to the lesion, whi le the ame-

boid mic rog l i a could be related to the def ini t ive prepara-

tion o f these zones for an imminen t regenerat ion.

The recovery o f the label ing to the normal pattern at

150 days after c rushing could indicate a return to the

normal state o f the visual sys tem and suggest a possible

role o f the microgl ia l cel ls in the process o f regenerat ion.

Acknowledgements

W e wou ld like to thank Mr. G.H. Jenkins for revis ing

the Engl ish, and Dr. B. Caste l lanos and Dr. B. Gonz~lez

for their sugges t ions wi th regard to the use o f tomato

lectin. This w o r k was supported by Junta de Cast i l la y

Le6n ( S A 2 9 / 9 4 ) , Spanish D G I C Y T (PB91-0424) and

( P B 9 4 / 1 3 8 8 ) projects.

References

[1] Acarln, L., Vela, J.M., Gonzfilez, B. and Castellano, B., Demonstra- tion of Poly-N-acetyl lactosamine residues in ameboid and ramified microglial cells in rat brain by tomato lectin binding, J. Histochem. Cytochem., 42 (1994) 1033-1041.

[2] Attardi, D.G. and Sperry, R.W., Preferential selection of central pathways by regenerating optic fibers, Exp. Neurol., 7 (1963) 46-64.

[3] Bignami, A., Dahl, D., Nguyen, B.T. and Crosby, C.J., The fate of axonal d6bris in Wallerian degeneration of rat optic and sciatic nerves, J. Neuropathol. Exp. Neurol., 40 (1981) 537-550.

[4] Brauer, K., Bruckner, G., Leibnitz, L. and Werner, L., Structural and cytochemical features of perineuronal glial nets in the rat brain, Acta Histochem., 74 (1984) 53-60.

[5] Brierley, J.B. and Brown, A.W., The origin of lipid phagocytes in the central nervous system; I. The intrinsic microglia, J. Comp. Neurol., 211 (1982) 397-406.

[6] Caroni, P. and Schwab, M.E., Two membrane protein fractions from rat central myelin with inhibitory properties for neurite growth and fibroblast spreading, J. Cell Biol., 106 (1988) 1281-1288.

[7] Castellano, B., Gonzfilez, B., Dalman, I. and Vela, J.M., Identifica- tion and distribution of microglial cells in the cerebral cortex of the lizard: A histochemical study, J. Comp. Neurol., 311 (1991) 434- 444.

[8] Colton, C.A., Abel, C., Patchett, J., Keri, J. and Yao, J., Lectin staining of cultured CNS microglia, J. Histochem. Cytochem., 40 (1992) 505-512.

[9] Dowding, A.J., Maggs, A. and Scholes, J., Diversity amongst the microglia in growing and regenerating fish CNS: immunohistochem- ical characterization using FL.1, an anti-macrophage monoclonal antibody, Glia, 4 (1991) 345-364.

[10] Fawcett, J.W., Rokos, J. and Bakst, I., Oligodendrocytes repel axons and cause axonal growth cone collapse, J. Cell Sci., 92 (1989) 93-100.

[11] Feizi, T., Demonstration by monoclonal antibodies that carbohydrate structures of glycoproteins and glycolipids'are onco-developmental antigens, Nature, 314 (1985) 53-57.

[12] Gaze, R.M., Regeneration of the optic nerve in Xenopus laevis, Q. J. Exp. Physiol., 44 (1959) 290-308.

[13] Giulian, D., Peptides from the regenerating central nervous system of goldfish stimulate glia, Proc. Natl. Acad. Sci. USA, 81 (1984) 3567-3571.

[14] Giulian, D., Ameboid microglia as effectors of inflammation in the central nervous system, J. Neurosci. Res., 18 (1987) 155-171.

[15] Giulian, D., Microglia and tissue damage in the central nervous system. In G. Levi (Ed.), Differentiation and Function of Glial Cells, Wiley-Liss, New York, 1990, pp. 379-389.

[16] Giulian, D. and Baker. T.J., Characterization of ameboid microglia isolated from the developing mammalian brain, J. Neurosci., 6 (1986) 2163-2178.

[17] Giulian, D., Baker, T.J., Shih, L.N. and Lachman, L.B., Interleukin-1 of the central nervous system is produced by ameboid microglial, J. Exp. Med., 164 (1986) 594-604.

[18] Giulian, D., Young, D.G., Woodward, J., Brown, D.C. and Lach- man, L.B., Interleukin-1 is an astroglial growth factor in the devel- oping brain, J. Neurosci., 8 (1988) 709-714.

[19] Goodbrand, I.A. and Gaze, R.M., Microglia in tadpoles of Xenopus laevis: Normal distribution and the response to optic nerve injury, Anat. EmbryoL, 184 (1991) 71-82.

[20] Ibrahim, M.Z.M., Khreis, Y. and Kosayan, D.S., The histochemical identification of microglia, J. Neurol. Sci., 22 (1974) 211-233.

[21] Imamoto, K. and Leblond C.P., Radioautographic investigation of gliogenesis in the corpus callosum of young rat. II. Origin of microglial cells, J. Comp. NeuroL, 180 (1978) 139-164.

[22] Kitamura, T., Tsuchihashi, Y. and Fujita, S., Initial response of silver-impregnates 'resting microglia' to stab wounding in rabbit hippocampus, Acta NeuropathoL, 44 (1978) 31-39.

[23] Konigsmark, B.W. and Sidman, R.L., Origin of brain macrophage in the mouse, J. Neuropathol. Exp. Neurol., 22 (1963) 643-674.

[24] Lfizfir, G., Elimination of cobalt from the frog brain introduced into the optic centres through the optic nerve, Acta Biol. Acad. Sci. Hung., 30 (1979) 245-255.

[25] Ling, E.A., The origin and nature of microglia. In S. Fedoroff and L. Hertz (Eds.), Advances in Cellular Neurobiology, Vol. 2, Academic Press, New York, 1981, pp. 33-82.

[26] Mannoji, H., Yeger, H. and Becker, L., A specific histochemical marker (lectin Ricinus communis agglutinin-1) for normal human microglia and its application to routine histopathology, Acta Neu- ropathol. (Berl)., 71 (1986) 341-343.

[27] McCoy, J.P., Hanley-Yanez, K., Brumfield, A., Herberman, R.B. and Chambers, W.H., Alterations in cell surface carbohydrates of rat large granular lymphocytes associated with interleukin-2 activation, Cell Immunol., 127 (1990) 275-283.

[28] McKeon, R.J., Schreiber, R.C., Rudge, J.S. and Silver, J., Reduction of neurite outgrowth in a model of glial scarring following CNS injury is correlated with the expression of inhibitory molecules on reactive astrocytes, J. NeuroscL, 11 (1991) 3398-3411.

[29] Murabe, Y. and Sano, Y., Thiamine pypophosphatase activity in the plasma membrane of microglia, Histochemistry, 71 (1981) 45-52.

[30] Naujoks-Manteuffel, C. and Niemann, U., Microglial cells in the brain of Pleurodeles waltl (Urodela, Salamandridae) after wallerian degeneration in the primary visual system using Bandeiraea simpli- cifolia isolectin B4-cytochemistry, Glia, 10 (1994) 101-113.

[31] Oehmichen, M., Inflammatory cells in the central nervous system, Prog. NeuropathoL, 5 (1983) 277-335.

[32] Perry, V.H. and Gordon, S., Modulation of CD4 antigen on macrophages and microglia in rat brain, J. Exp. Med., 166 (1987) 1138-1143.

[33] Perry, V.H., Hume, D.A. and Gordon, S., Immunohistochemical localization of macrophages and microglia in the adult and develop- ing mouse brain, Neuroscience, 15 (1985) 313-326.

324 A. Velasco et aL /Brain Research 705 (1995) 315-324

[34] Rabinowitz, S.S. and Gordon, S., Macrosialin, a macrophage-re- stricted membrane sialoprotein differentially glycosylated in re- sponse to inflammatory stimuli, J. Exp. Med., 174 (1991) 827-836.

[35] Rio Hortega, P., Microglia. In W. Penfield (Ed.), Cytology and Cellular Pathology of the Nervous System, Paul B. Hoeber, New York, 1932, pp. 481-584.

[36] Rozemuller, J.M., Eikelenboom, P. and Stam, F.C., Role of mi- croglia in plaque formation in senile dementia of the Alzheimer type. An immunohistochemical study, Virchows Arch. (Cell Pathol.), 51 (1986) 247-254.

[37] Schelper, R.L. and Adrian, E.K., Nonspecific esterase activity in reactive cells in injured nervous tissue labeled with 3H-Thy midine

1 2 5 • , or Iododeoxyundlne injected before injury, J. Comp. Neurol., 194 (1980) 829-844.

[38] Sperry, R.W., Visuomotor coordination in the newt (Triturus viri- descens) after regeneration of the optic nerves, J. Comp. Neurol., 79 (1943) 33-55.

[39] Sperry, R.W., Patterning of central synapses in regeneration of the optic nerve in teleosts, Physiol. Zool., 23 (1948) 351-361.

[40] Springer, A.D. and Wilson, B.R., Light microscopic study of degen- erating cobalt-filled optic axons in goldfish: Role of microglia and radial glia in debris removal, J. Comp. Neurol., 282 (1989) 119-132.

[41] Streit, W.J., Graeber, M.B. and Kreutzberg, G.W., Functional plas- ticity of microglia: A review, Glia, 1 (1988) 301-307.

[42] Streit, W.J. and Kreutzberg, G.W., Colocalization of microglia and astrocytes during the axonal reaction, Tenth International Congress on Neuropathology, Stockholm, 1986, A79.

[43] Streit, W.J. and Kreutzberg, G.W., Lectin binding by resting and reactive microglia, J. Neurocytol., 16 (1987) 249-260.

[44] Streit, W.J., Schulte, B.A., Balentine, J.D. and Spicer, S.S., Histo- chemical localization of galactose-containing glycoconjugates in sensory neurons and their processes in the central and peripheral nervous system of the rat, J. Histochem. Cytochem., 33 (1985) 1042-1052.

[45] Suzuki, H., Franz, H., Yamamoto, T., lwasaki, Y. and Konno, H., Identification of the normal microglial population in human and rodent nervous system using lectin-histochemistry, Neuropath. Appl. NeurobioL, 14 (1988) 221-227.

[46] Wilson, M.A., Gaze, R.M., Goodbrand, I.A. and Taylor, J.S.H., Regeneration in the Xenopus tadpole optic nerve is preceded by a massive macrophage/microglial response, Anat. Embryol., 186 (1992) 75-89.

[47] Zhu, B.C-R. and Laine, R.A., Purification of acetyllactosamine- specific tomato lectin by erythroglycan-sepharose affinity chro- matography, Prep. Biochem., 19 (1989) 341-350.