Journal of Neuroscience Methods, 23 (1988) 49-55 49 Elsevier

NSM 00771

Preservation of the ependyma of the mouse spinal cord by various vascular perfusion procedures

Roger Bjugn, Hans Kristian Haugland and Per R. Flood

Institute of Anatorny, University of Bergen, Bergen (Norwc(v)

(Received 23 February 1987) (Revised 22 June 1987)

(Accepted 29 June 1987)

Key words: Perfusion fixation; Spinal cord; Ependyma; Central canal; Mouse

Seventy-two mice were used to find out which of 13 vascular perfusion procedures gave best structural preservation of the spinal cord ependyma and central canal lumen. Best results were obtained by 3% glutaraldehyde in Tyrodes solution with 50% of normal NaC1 amount, 0.06 M sucrose and 2% dextran T-40 (556 mOsmol, pH 7.2, 0 - 4 ° C). This was perfused by a peristaltic pump at 40 m l / m i n for 10 rain through a cannula inserted in the ascending aorta. No advantage was seen by heparin pretreatment or adding a prefixation rinse. With good tissue preservation the central canal was found to be round to oval in cross-sectional profile and almost free of intraluminal material.

Introduction

Through the years various perfusion techniques have been developed for the ultrastructural pre- servation of various brain regions (e.g. Palay et al., 1962; Karlsson and Schultz, 1965; Schultz and Karlsson, 1965; Peters, 1970; Kalimo, 1976; Cragg, 1980; Friedrich and Mugnaini, 1981). However, the problems involved in obtaining good fixation of the spinal cord have been considered by very few authors (Palay et al., 1962; Hill et al., 1982). In fact, it has been claimed that the central canal of the mouse in most regions normally is slit-like and largely filled with an amorphous ground sub- stance separated by membranes (Sturrock, 1981). Such obliteration of the central canal lumen seems to be incompatible with data on the rapid spread of dyes along the central canal through the post- erior internal neuroporus to the sacral sub-

Correspondence: R. Bjugn, Institute of Anatomy, University of Bergen, Arstadveien 19, N-5009 Bergen, Norway.

arachnoidal space after their injection into the ventricular system (Bradbury and Lathem, 1965). An obliteration also seems incompatible with the continuity and movement of Reissner's fibre along the entire spinal central canal (Tulsi, 1982: Castenholz, 1984; Rodriguez, 1985).

Accordingly, as a basis for experimental studies on the ependyma of the central canal, we found it necessary to develop a fixation method that gave a well-preserved ependymal lining throughout the various levels of the spinal cord.

Materials and Methods

Albino mice (NMRI-strain, 9 male and 63 female, obtained through SIFF, Oslo, Norway) weighing from 20 to 50 g (mean 38 g), with free access to standard pelleted food and water, were anesthetized with an intraperitoneal injection of Equithesin (1 ml/100 g body weight). The animals were pinned down on a rubber mat on top of a hollow plastic plate containing circulating water

0165-0270/88/$03.50 ~ 1988 Elsevier Science Publishers B.V. (Biomedical Division)

50

(380-39 ° C). The heart was exposed by thoracot- omy and a blunt ended cannula (1.1 mm external diameter) inserted through the left ventricle into the ascending aorta. The cannula was connected to a fluid reservoir via a Masterflex (Cole Palmer) peristaltic pump equipped with a 7026-20 pump head and 0.65 mm (0.30 mm internal diameter) silicone tubing. An incision was made in the right auricle at the moment the perfusion started. Less

than 30 s elapsed from opening the thorax to the start of the perfusion. The animals were segre- gated in 13 groups, 4 of which (32 animals) were given 0.1 ml Heparin (10,000 IE/ml) , prewarmed to 37 ° C, injected into the left ventricle (injection lasted 10-15 s) before the perfusion cannula was inserted. With this procedure about 70 s elapsed before the perfusion started. The temperature of all perfusates was 37°C, except in procedure 10

T A B L E I

The appearances of the ependymal lining of the central canal at three levels of the mouse spinal cord after 13 different perJusion fixation procedures

Proce- N u m b e r Hepa- R i n s e l F ixa t ive Flow 2

dure of an imals rin used ( m l / m i n ) Cervical LM appearance *

I II I l l IV

1 10 yes A C 20 3 2 5 0 2 10 yes none C 20 2 4 4 0

3 7 yes A C 40 5 2 0 0

4 4 no B C 40 2 2 0 0

5 4 no none C 40 1 0 3 0

6 4 no none D 40 2 2 0 0 7 5 yes none E 40 5 0 0 f)

8 5 no none E 40 5 0 0 I)

9 3 4 no none E 20 4 0 0 0 10 4 4 no none E 20 1 1 1 1

11 5 no none G 40 5 0 0 0 12 5 no none H 40 3 l 1 0

13 5 no none F 10 5 0 0 0

Thorac ic LM apperance * L u m b o s a c r a l L M a p p e a r a n c e * Overal l

1 I I I I I IV I 11 I I I IV qua l i ty

0 6 4 0 0 4 6 0 poor

1 3 6 0 1 3 6 0 poor 2 4 1 0 1 3 3 0 poor

2 1 1 0 0 1 3 0 poor

0 3 1 0 0 0 4 0 poor 1 2 0 1 0 4 0 0 med ium

3 1 0 1 4 1 0 0 good

5 0 0 0 4 1 0 0 best 1 2 1 0 0 3 1 0 medium 0 l 1 2 0 0 4 0 poor 5 0 0 0 2 3 0 0 good 2 2 0 1 2 2 1 0 m e d i u m

1 1 0 3 1 4 0 0 med ium

* N u m b e r s refer to n u m b e r of spec imens be longing to each of the 4 1 min dura t ion .

2 10 rain durat ion,

3 37° C f ixat ive temperature . 4 4 ° C fixative temperature .

l ight microscopica l appea rances descr ibed in the text.

(see below), where the temperature was 4°C. The prefixation rinse lasted 60 s and the perfusion fixation lasted 10 min in all procedures used.

Further details on the perfusion procedures are given in Table I.

The following solutions, all adjusted to pH 7.2, were used for prefixation rinsing:

(A) Tyrodes solution (Romeis, 1948) contain- ing 75% of normal NaC1 amount, 0.06 M sucrose and 2% dextran T-40 (Pharmacia Fine Chemicals, Uppsala, Sweden). The osmolarity was about 300 mOsm as determined by an Advanced Osmome- ter, model 3L.

(B) As A, plus 0.1% sodium nitrite as vasodila- tor.

The following solutions, all adjusted to pH 7.2, were used for fixation:

(C) Tyrodes solution with 50% of normal NaC1 amount, 0.06 M sucrose, 2% dextran T-40 and 1% glutaraldehyde (351-359 mOsm).

(D) As C, plus 0.1% sodium nitrite as vasodila- tor (377 mOsm).

(E) As C, but with 3% glutaraldehyde (556 mOsm).

(F) As C, but with 6% glutaraldehyde (915 mOsm).

(G) Tyrodes solution with 12.5% of normal NaC1 amount, 0.18 M sucrose, 2% dextran T-40 and 3% glutaraldehyde (600 mOsm).

(H) As G, but without dextran (590 mOsm). The rinsing and fixing solutions were used in

different combinations (Table I). At the end of the perfusion, the vertebral column was dissected free and immersed in the same fixative for 24 h at 4°C. The spinal cord was then isolated after laminectomy. Transverse sections (approximately 0.7 mm thick) from the cervical, mid-thoracic and lumbo-sacral levels were cut, rinsed in the same vehicle as used in the fixative and postfixed in 1% OsO4 in the same vehicle for 3 h. After rinsing in distilled water and dehydration in increasing concentrations of ethanol, the sections were em- bedded in Agar epoxy resin after being passed through 1,2-propylenoxide. Semithin sections for light microscopy (0.5 or 1 /~m thick) were stained with 1% Toluidine blue in 1% sodium borate. Ultrathin sections, mounted on formvar- and carbon-coated grids, were stained with uranyl

51

acetate and lead citrate and examined in a Philips EM 300 electron microscope. Statistical compari- son of the results from different fixation proce- dures were made by the Wilcoxon two-sample test.

Results

The ependymal lining of the cross-sectioned central canal, as seen in the light microscope, was classified as presenting one of four structural ap- pearances:

(I) The central canal was roundish to ovoid, with the long axis in the median plane. The micro- villar lining showed fluctuation in height and several small indentations (Fig. IA).

(II) The central canal was roundish to ovoid and invariably contained spread amounts of a homogeneous, weakly stained substance. The ependymal lining was uneven with the microvilli tightly aggregated in clusters (Fig. 1B).

(III) The lumen of the central canal varied in shape, from ovoid to almost closed (with the long axis in the median plane). Processes from the ependymal surface divided what remained of the central canal lumen into compartments or micel- lae, giving it a honeycomb appearance. A homoge- neous, weakly stained substance, and in some cases strongly stained particles of different size. filled the micellae to a variable degree (Fig. 1C).

(IV) The lateral walls of the central canal were in close approximation and the lumen seemed almost obliterated (Fig. 1D).

The different fixation procedures used revealed great differences in occurrence of these 4 struct- ural appearances (Table I).

In order to know which of the 4 light micro- scopical categories represented the optimum pre- servation, 10 (or more) blocks from categories 1 through III and 3 blocks from category IV were examined by TEM.

(I) At the ultrastructural level this category was characterized by intact apical ependymal plasma- lemma and microvilli of fairly uniform diameter and spacing. There were few signs of swelling or disruption of mitochondria and endoplasmic re- ticulum (Fig. 2).

52

Fig. 1. Light micrographs of mouse spinal cord showing decreasing quality of preservation of the central canal ependyma (appearances I-IV). A: lumbo-sacral level. B-D: thoracic level. Arrows indicate supraependymal nerve fibers. ×750.

(II) Large discontinuities in the apical epen- dymal plasmalemma were seen in some cells. A dispersed fibrillar to granular substance protruded from the cells into the lumen through these gaps. The microvilli were swollen and clumped together. Swelling of mitochondria and endoplasmic reticu- lum was usually present in the apical cytoplasm (Fig. 3A).

(III) In all sections many ependymal cells re- vealed large ruptures in the apical plasmalemma which was lifted into the central canal lumen, giving it a honeycomb appearance. The floccular

substance described above was much more abun- dant than in appearance II specimens. Myelin-like, lamellated figures as well as other heterogeneous granular masses were sometimes seen in the central canal. Swelling of mitochondria and endoplasmic reticutum was usually present in the apical cyto- plasm (Fig. 3B).

(IV) Most ependymat cells were grossly dis- turbed by plasmalemma ruptures, apical cyto- plasma protrusions and microvillar lumping. The central canal was almost completely filled with the floccular substance described. Additional myelin-

53

Fig. 2. Transmission electron micrograph of mouse spinal cord from the lumbo-sacral level showing well preserved central canal ependyma, x 9000.

like, lamellated figures were frequently seen. Swelling of mitochondria and endoplasmic reticu- lure was usually present in the apical cytoplasma (Fig. 3C).

Discussion

The observed differences in appearance of the ependyma of the central canal indicated that one of the four appearances was more biologically relevant than others. When applying criteria for adequate ultrastructural preservation (Glauert,

1974; Hayat, 1981) the appearance I specimens fulfilled such criteria far better than the other appearances at all levels of the spinal cord.

The best preserved specimens were obtained from the cervical spinal cord by most of our perfusion procedures. At the thoracic and lumbo- sacral levels most procedures gave inferior results. The best total preservation of the spinal cord with almost exclusively appearance I sections from all three levels, was obtained by procedure 8. This procedure was characterized by a high glutaral- dehyde concentration (3%) and a high flow rate (40 ml /min) . At the light microscopical level this

54

Fig. 3. A-C: transmission electron micrographs of mouse spinal cord showing decreasing quality of preservation of the central canal ependyma (contrast Fig. 1B-D). A: lumbo-sacrat level. B and C: thoracic level. × 9000.

procedure was not significantly better than proce- dure 7 and 11 (P > 0.05). However, at the ultra- structural level the ependymal intercellular spaces were found to be larger in one animal using proce- dure 11. This procedure was characterized by an exceptionally high sucrose/sodium chloride ratio (5:1 regarding osmolarity). No advantage was evident by pretreating the animals with in- tracardiac heparin injection (procedure 7-8) , ad- ding a vasodilator (sodium nitrite) (procedure 5 6) or rinsing the vascular bed before fixation (proce- dure 1 2). These findings are largely in agreement with those of Karlson and Schultz (1965), Schultz and Karlson (1965) and Kalimo (1976) for brain tissue.

The central canal is earlier described as a slit- like and material filled lumen (Sturrock, 1981). In all our sections revealing good tissue preservation the central canal was found to be roundish to oval in cross-sectional profile and almost free of in- traluminal material. We accordingly conclude that this is the normal appearance of the mouse spinal cord central canal. Further studies on the normal structure of the ependyma in well preserved speci- mens will be presented later.

Acknowledgements

We thank Mrs. Edith Bru, Mrs. Bodil Hansen and Mrs. Anita Koldingsnes for all the microtomy work involved. This study was supported by the Norwegian Research Counsil for Science and the Humanities.

References

Bradbury, M.WB. and Lathem. W. (1965) A flow of cerebrospinal fluid along the central canal of the spinal cord of the rabbit and communicat ions between this canal and the sacral subarachnoidal space, J. Physiol. (Lond.), 181: 785-800.

Castenholz, A. (1984) Scanning electron-microscopic study of Reissner's fiber in the caudal spinal cord of the cat and rabbit. Cell Tissue Res., 237: 181-183.

55

Cragg, B. (1980) Preservation of extracellular space during fixation of the brain for electron microscopy, Tissue ('ell, 12: 63-72.

Friedrich, V.L. and Mugnaini, E. (1981) Preparation of neural tissues for electron microscop>. In k. Heimer and M.J. Robards (Eds.), Neuroanatomical Tract-Tracing methods, Plenum. New York. pp. 345 375.

Glauert, A.M. (1974) Fixation, dehydration and embedding of biological specimens. In A.M. Glauert (Ed.), Practical Methods in Electronmicroscopy, Vol. 3, North-Holland, Amsterdam, pp. 1 207.

Hayat, M.A. (1981 ) Fixation for Electronmicroscop),, Academic Press. New York, 501 pp.

Hill, P.K., de la Torre. J.C., Thompson, S.M.. Bullock, D.F. and Rosenfield-Wessels, S. (1982) A comparative study of EM fixation procedures for the adult rat spinal cord based on regional blood flov,, J. Neurosci. Methods, 5 :23 31.

Kalimo, H. (1976) The role of the blood brain barrier in perfusion fixation of the bruin for electron nricroscopy, Histochem. J., 8 :1 12.

Karlsson, U. and Schultz, R i . (1965) Fixation of the central nervous s>stem for electron microscopy by aldehyde perfu- sion. 1. Preservation with aldehyde perfusates versus direct perfusion with osmium tetroxidc "aith special reference to membranes and the extracellular space, J. Llhrastruct. Res., 12: 160-186.

Palay, S.L., McGee-Russell , S.M., Gordon, S. and Grillo, M.A. (1962) Fixation of neural tissues for electnm microscopy By perfusion with solutions of osmium tetroxide, J. ('ell Bioh, 12: 385-410.

Peters, A. (1970) The fixation of central nervous tissue and the analysis of electron micrographs of the neuropil, with spe- cial reference to the cerebral cortex. In W.J.H. Nauta and S.O.E. Ebbesson (Eds.), Contemporary Research Methods in Neuroanatomy, Springer, New York, pp. 56 75.

Rodriguez, S., Hein, S., Yulis, R., Delannov, L., Siegmund, 1. and Rodriguez, E. (1985) Reissner's fiber and the wall of the central canal in the lumbo-sacral region of the bovine spinal cord. Comparative imnmnocvtochemical and ultra- structural study, Cell Tissue Res.. 240:649 662,

Romeis, B. (1948) Mikroskopische Technik, 15th edn., R. Oldenbourg, Munich, 695 pp.

Schultz, R.L. and Karlsson, LI. (1965) Fixation of the central nervous system for electron microscopy b), aldehyde perfu- sion. 11. Effect of osmolarity, pH of perfusate, and fixative concentration, J. UItrastruct. Res., 12:187 2(/6.

Sturrock, R.R. (1981) An electron microscopic study of the development of the ependyma of the central canal of the mouse spinal canal, J. Anat., 132: 110-136.

Tulsi, R.S. (1982) Reissner's fiber in the sacral cord and filum terminale of the possum Trichosuru.~ ~,uipecufa: a light, scannmg and electron microscopic study, J. Comp. Neurol.. 221: 11-20.