- Email: [email protected]
Brain extracellular space fixed for electron microscopy
N ~ Letters, 15 (1979) 301--306 © Elsevier/North-Holland Scientific Publishers Ltd.
301
BRAIN EXTRACELLULAR SPACE FIXED FOR ~ M I ~
B. CRAGG Department of Pkydolo~, Monash University, Clayton. V'~toria 3168 (Australia)
(Received A~qpmt21st. 1979) (Revised vevdcmreceived September 10th, 1979) (A©ceptedSeptember 13th, 1979)
SUMMARY
Adult mammalian brain contains 17--20% extraceUulat space, but fixatives cause the cellular e~ements to ingest the extracellular fluid so that the space is reduced to less than 5% with all ¢~nventional methods of fixation~ This can be prevented by washing out the ~tracellular fluid with isotonic sucrose, which does not penetrate cells. Subsequent fixation with aldehydes and osmium and conventional processing leads to the preservation of extmcellulat space in electron micrograpl~. Ext~acellular space was found to be mzevenly distributed, widely separating some cellularprocesses while leaving other groups of ~ contiguous. Three independent ltne~ of evidence indicate that about 20~ of the volume of adult brain consists of extraceilular space. This evidence was o~d by keeping a known concentration of a chemical marker in the cembrospinal fluid until the concentration of marker in the brain reached a plateau a~d then expneming this concentration as a percentage of the concentration in the cerebrospinal fluid. When sucroae or i n . i n were used as markers, values of 17--'209b for brain extracellular space were obtained in rabbit, cat, dog and monkey [1]. Independent confirmation has been obtain. ed by measuring the electrical resistance of brain tissue and of cerebrospinal fluid, the f o t m ~ being four or five times greater than the latter. Because of the high m e m b r ~ e msktance of cells, most of the electrical me~tri_ng current is carried by extm~.]ltdar fluid and an equation due to Maxwell shows that 15--20~ extn~Ilulax space is consistent w i ~ the electrical
msist~ce mea~ecl [4], Extm~ihdaz space has been obeyed by electron micro!copy in ~ ~ prepared by extremely rapid freezing,substitution of ice by e t h ~ o l o r ~ saturated with osmium tetroxide and embedcling in plast~ [5]. However. th~ region that is frozen fast enough to a,-~id
by ~ crystals is ~ y 15 pm t ~ k , and ~ has not ~ e p ~ E~ fixsfion as the ~ m r ~ i pl,ocedum for electron microscopy. S t ~ d m ~ fixation ~ c c e & n ~ lead to an e x ~ D u l a r space of less than 5% in brain tissue because fi~_a~es probably ~ a m~,ement of ~ m and c M ~ ions from the extmcellul~ fluid into the nem-onal and gZial cell p r c c e ~ [6]. The ionic d i f f u ~ n is accompsnied by water ~ h e~pmr~ls the cellular processes to occupy the extracenulm' space. quenfly an conventional e k ~ o n m i ~ h s of br~n display swollen cellular processes end obliterated e x ~ u l a z ~ , but th~ distortion is not acknowledged, even in the standard a t ~ of ultmstmctm~ [2]. It is possible to stop this distortion o c c ~ by r e ~ the sod~m chloride in the extracellular fluid by a s u i t e that is unable to enter cells during fixation. Treatment with aldehyde fixatives and osmium tetroxide and conventional processing then leads to elec~on micrographs with pzese~ed extzacelluhr space (Fi~. 1--3). In order to wash out the extzacellut~r space with an impc~neant solution it is necessary to open the blood-brain barrier before perf~ing the va~cuiaturn, or else to bypass the blood-brain barrier by ~mers~ng thin ~_ices of fresh brain in the impermeant solution. The ~ can be opened by perfusingat a pte~nm of 300 m m H g [3] and thisdoes not ~ to damage the tissue ~ound the c a p ~ (Fig. 1). A l ~ a t i v d y , slices of ~ less than 1 m m thick can be cut and imm~rsei in the i m ~ ~ ~olution. If a non-ionic substance is used for the latter,the electricalc~nduct~vity of the solutioncan be measured in o~ler to fc~ow ~he e|u~cm of the ,.~trao cellularsaline.The r e ~ c e between ch]ofidedsH~r eiectro~ i m m ~ in a ma~et~caUy stirr~ beaker of suero~ ~ o l u ~ w ~ m e . r e d with a digitalresistancemeter pas~ng 2 p A and calibratedby ~]d~g succe~ve 10 pl aliquots of 0.9% NaCL The conductivity was found to riseand approach a plateau exponentiallywith a time c o ~ t of 75 ~ec and 90~ 120 sec of immersion w ~ mfficient to p~esct~e e x ~ t [ ~ ~pacc. ~ i n slicescan be cut even from the centms of rag or mouse brainsand good fixationobtained (Fig. 2). Moreover, immemion fixationallowse~tracelluiar space to be preserved in human cerebraland ~ b e l l ~ cor'~ ( ~ . 3). The impenneant ~ution was ~ at a concentration thatis~tonic with ceteb~ospinal fluid (about 0.28 M). It can be expected to ~ptace the ~ t m cellular fluid without changing the size of the c e l l ~ ~ . In p m l ~ i nary experiments,glucose,HEPES, calciun~c.hlorideand ~crose were found to act as impermeant solutesthat preserveextracellul~space and I have u ~ isotonics u c ~ in most subsequent experiments. Addition of aldehyde fixative was found to lead to electron miczognkuhs s h o ~ g much more ~ a n 20% extraceIl~arspace and severelyshrunken ce~ularprocedures. ~ w~ p~obably bec~use aldehyde fixative increases the permeability of the cell membrane and small molecules then di~se out of the ceils.Sucrose d ~ not enter the cells,so the cytop~az~n becomes hypotonic and t~e ce~s shrink. This a h ~ e can be avoided by replacing part of the sucrose with another
303
Fig. S. An ~
~
z
~
~ 15% ~ p m ~ d
is ~ e ~ u ~
of a m o u ~ ~ r
~4
~euro~u~c~l ~peeh'~en of h u ~
~II~
~
f~u:~ by ~
m
~
~ t ~
~ c e that is able to d i f f ~ into the ce~Is in the pre~'~ce of a l ~ h y d e f~ative~° ~ r a phosphate, cac~:~y~te, ~ l p b a ~ , ~ e ~ d ~ h t o m a ~ ~,Jd u ~ a ~ d glycerol b ~ e ~ tried ~ ~ ~ ~ ~o~ ~ d ~ u ~ ph~ p h a ~ wa~ found to give ~ most ~tt~factor] ~ ~ U o a of ~ ~ c ~ j ~ The ~ize of the extraceUu]~ ~ p ~ ia the elect~n m ~ r ~ ~ d~end~ on the ~ o of p e r m e ~ t to ~ p e ~ n ~ t ~ t u t e ~ , and a one,h o r s e ~ o ~ ra~o w ~ found e m ~ y ~ g~ve an e x ~ m e ~ ~ of 1 5 - - 2 ~ c o ~ sis~ent~ t h the ~ t ~ of ~ e chemic~ m ~ ~ ~ [1]. ~ ~ of ~ e ex~zacellu~ s p ~ e was e s ~ a ~ d ~ a grid of 1 0 0 red ~ i n t ~ 2 cm a p ~ on a transparent acetate ~ wh~h w ~ ~ d ~ e r each photomic~e~ graph ~ d ~ e number of ~ m ~ lying over e ~ t ~ e ~ l u ~ r space counted for te~ placements of the grid. Myefin ~ the m ~ t die.cults~mctt~e to ~ ~thout dmtortion,~ ~tis conven~o~ m e ~ ~ d the ~dd~fionof 2 m M ~ u m chloride~ m e d help the a d h e ~ of the myelin l a m e ~ . For ~ e same m ~ n , I have prefer~d ~ dehydrate ~ e t ~ e by ~dL~g alcohol slowly to a s t ~ and ow~rflowing d~sh c o n ~ a h ~ ~ e tissue m water, so ~h~ t h e concentration of ~ c ~ o l ~ses ~ f l y from zero to 8~% over ab~zt 15 mira The addition of ~dehyde fi.xat~ve~ ~ the peffu~te or ~ m e r s i o n fluid h ~ al~o been made in a c ~ t h ~ u ~ mann~, sLnce the c ~ a o ~ t y of the commonly used a ] d e h y ~ fixatives ~s large and the~ sudden addition can d ~ r t ceHu~.ar profiles. However, extm~Uular ~p~ce is p ~ e s e r ~ whether con~uou:~ or abrupt c h ~ g e s of aldehyde and ~ o h o l concen~ation are employed°
395 A ~ e d u l e for fixation has been developed in i 9 4 experiments in which the impermean~ p e r ~ e a n t and ,fixative chemicals were systematically varied and the method d e s c r i ~ below is as early to use as conventional procedures. It is fikely to be c a p a b k of f u , ~ e r improvement because the fixation of myelin z e m ~ imperfect. ~ t h ~ e perfusion m e t h o d and the immersion method i n v ~ a b l y preserve extraceUular space, although the same solutions p e r f ~ a ~ ~ lOO m m H g presvaze without opening the blood-brain barrier do n o t and immersion o f fresh brain slices directly into the final fb ative does n o t preserve extzacellular space, in a deeply anaesthetized rat, the apex of ~ e heart is cut off and a cannula made by squaring off t~e end of a No. 14 hypodermic needle is clamped in the ascending aorta. A roller pump cmme,::~e~'!to the c a l m e r is brought up to a speed sufficient to supply per~sW ;- ~:~m a stirred beaker at a pressure of 300 mmHg. When 100 mI pref~asate has been passed and lOO ml is left in the beaker, a second solution of the same composition as the first but containing aldehyde fixative is run into the beaker to keep this level of fluid roughly constant. The presydre is ~ u c e d to 199 m m H g and 2 ~ ml of fixative solution perf u ~ . After 2 h ~ e b ~ is removed and slices of ~ssue about I mm thick Lmmersed in the ~ e fixative for 2 days. As an alternative to perfusion, nec~-~ry when dealing with human cortex o b ~ e d at neurosurgery, small s t i ~ less ~ 1 mm thick ~ c a t from fresh brain and immersed in 100 m| of the impermeant s o l u f i ~ in a stirred ~ a k e r . After 90 sec, 100 ml of the ~ o n d solution containing i~ative is run i n ~ the stirred beaker slowly over 3 rain, After 2 days, the ~ r f u s e d or ~ m e r s i o n o f i x ~ ~is~ue in a vi~ con° t~ning t ml of fixative ~ c ~ i ~ to 4°C and 0.2 ml of a 6% solution of ~ m i u m tetroxide in wate~ ~iu added i o a fin~ osmium concentration of 1% the ~ o n d fixative ~oiution. After 24 h, the ti~ue i~ washed in 0.1 M ~¢~dium a ~ t ~ buffer at pH 4.5, stained i~ 2% ~ y ! acetate for 2 h and dehydrated by slow co~tinuou~ addition of alcohol. The tissue is then embedded i~ A~aldite, Foz ~ ~ u ~ . o n ~ d imme~io~ f l e a , on, the. fi~t, i m ~ r m e a n t solution is 0.135 M s-~o~e with 0.135 ~ M a ~ i u m phospha~ buffer at pH 7.3 and 0.002 M calciura chloride. ~ e ~ c o ~ d ~olution i~ the ~ m e wifh the addition of 1% g~utaxaldehyde for pert~sion or 2% S u ~ I d e b y d e for im~ner~io~ fi):ation. F o ~ a l d e h y d e or aero~m can ~ added wi~h ~he giut~zafdehyde. The two ~ l u t i o n a ~¢e made by di~olving 18.5 g sucrose a~d 4.1 g NaH~POa * 2H~O in 400 ~ water, adding 9~5 ml of a 5% ~ l u t i o n of c~¢iura chloride and bringing the p H to 7,3 by ~dition of a sa~ur-ated~lu~;ion of sodiu~ hydroxide. The d i s t r i b ~ o s of exizace~utar ~pace h ~ ~ e v e ~ in~re~ting fea~ure~. Diffusion ~o and from b~)d ca~)i_'Harie~'~ i~ r~ot restricted to narrow c~efts between swo~en c~ul,ar p ~ o c e ~ , bu~ .~ ~iowc~l ~i~ac~.ou~eharme~ {Fig. 1 )~ Of ~ the ce~uiar elemen%, ~.~t~oc~,tic ptc~c~e~ ~ d n e u r o ~ d e ~ d r i ~ are ~'.o-~em ~ t ~wol!en by convengon~ F~ation. The *~hinfinge~ of ~et~o., c y s t : t ha~ are ~ n when exlx~ellular spaze N preserved o f ~ n do hog
~ inve~ ~naptic a p p ~ t i o ~ but leave synaptic clefts in dose c ~ tintfity with !z~--~ e z t m c e ~ d ~ ~ The di~fibut/on of space is uneven and ~ s ~ s ~ y . In a sym~tic g l o m e ~ in the ~ or ce~bdlar ¢ ~ , e x ( ~ 8) most of'~he cenul~ ~ ire contiguous, for ~ glomem/us ~ ~ space, but is widely ~ from the gnmule c e b ~ exUacellul~ space. ~ ~ o n s md snudl~ ~ ~have the same widths as in conventional p~peratiom, but both appear unusually prominent because cell t ~ ~ are w~2~y separated everywhere else. ~ study of the ~ . ~ t / o n of ext~mcellular space can be expecte~ to help our understanding of the structural bias of ~uronal interactions. REFERENCES I I,evln, V ~ . , ¥ ~ e m e h e r , J.D. and PatJ~k, C.8o,Suborn msdinulin sSpsce~ ments of eenebnd cor~z in four nutmmal/sn epeeiu, Amer. J. Pbye~l., 219 (1965) 1528---1533. 2 Peters, A., Palay, S.L. 8nd W ~ , H. de F., The Fine 8tn~ct.ure of the Nervous System, 8aunders, Philadelphia, 1976. Rapport, 8.I., Opening of the blood-bnd_nbarrier by Lute hypertension, F,~p. NeuroI., 52 (1976) 467--479. 4 Van ~ e l d , A., The ez~sceHular slptce in the ~ eentrzl nervous system. In G.H. Bourne (Ed.), The Structure and Function of Nervous Tissue, VoL 4, Academic P r ~ , New York, 1972, pp. 447-511. 5 Van Htrreveld, A. and $teiner, J., ~ t r ~ e U u l l r tp~ee h~ frozen and e ~ o l ~bttituted central nervo~ thmueoAnat. l~ec., 116 (1970) 117--130. 6 Van H ~ v e i d , A. and Khattab, F.|., Perfm~on f l z ~ o n with ~ u t u ~ _d~byd~amd p o ~ f~atton wi~ ountum tetrog~de for e ~ c t ~ n microcopy, J. C41ik L , ~ ()965} 479~494.
301
BRAIN EXTRACELLULAR SPACE FIXED FOR ~ M I ~
B. CRAGG Department of Pkydolo~, Monash University, Clayton. V'~toria 3168 (Australia)
(Received A~qpmt21st. 1979) (Revised vevdcmreceived September 10th, 1979) (A©ceptedSeptember 13th, 1979)
SUMMARY
Adult mammalian brain contains 17--20% extraceUulat space, but fixatives cause the cellular e~ements to ingest the extracellular fluid so that the space is reduced to less than 5% with all ¢~nventional methods of fixation~ This can be prevented by washing out the ~tracellular fluid with isotonic sucrose, which does not penetrate cells. Subsequent fixation with aldehydes and osmium and conventional processing leads to the preservation of extmcellulat space in electron micrograpl~. Ext~acellular space was found to be mzevenly distributed, widely separating some cellularprocesses while leaving other groups of ~ contiguous. Three independent ltne~ of evidence indicate that about 20~ of the volume of adult brain consists of extraceilular space. This evidence was o~d by keeping a known concentration of a chemical marker in the cembrospinal fluid until the concentration of marker in the brain reached a plateau a~d then expneming this concentration as a percentage of the concentration in the cerebrospinal fluid. When sucroae or i n . i n were used as markers, values of 17--'209b for brain extracellular space were obtained in rabbit, cat, dog and monkey [1]. Independent confirmation has been obtain. ed by measuring the electrical resistance of brain tissue and of cerebrospinal fluid, the f o t m ~ being four or five times greater than the latter. Because of the high m e m b r ~ e msktance of cells, most of the electrical me~tri_ng current is carried by extm~.]ltdar fluid and an equation due to Maxwell shows that 15--20~ extn~Ilulax space is consistent w i ~ the electrical
msist~ce mea~ecl [4], Extm~ihdaz space has been obeyed by electron micro!copy in ~ ~ prepared by extremely rapid freezing,substitution of ice by e t h ~ o l o r ~ saturated with osmium tetroxide and embedcling in plast~ [5]. However. th~ region that is frozen fast enough to a,-~id
by ~ crystals is ~ y 15 pm t ~ k , and ~ has not ~ e p ~ E~ fixsfion as the ~ m r ~ i pl,ocedum for electron microscopy. S t ~ d m ~ fixation ~ c c e & n ~ lead to an e x ~ D u l a r space of less than 5% in brain tissue because fi~_a~es probably ~ a m~,ement of ~ m and c M ~ ions from the extmcellul~ fluid into the nem-onal and gZial cell p r c c e ~ [6]. The ionic d i f f u ~ n is accompsnied by water ~ h e~pmr~ls the cellular processes to occupy the extracenulm' space. quenfly an conventional e k ~ o n m i ~ h s of br~n display swollen cellular processes end obliterated e x ~ u l a z ~ , but th~ distortion is not acknowledged, even in the standard a t ~ of ultmstmctm~ [2]. It is possible to stop this distortion o c c ~ by r e ~ the sod~m chloride in the extracellular fluid by a s u i t e that is unable to enter cells during fixation. Treatment with aldehyde fixatives and osmium tetroxide and conventional processing then leads to elec~on micrographs with pzese~ed extzacelluhr space (Fi~. 1--3). In order to wash out the extzacellut~r space with an impc~neant solution it is necessary to open the blood-brain barrier before perf~ing the va~cuiaturn, or else to bypass the blood-brain barrier by ~mers~ng thin ~_ices of fresh brain in the impermeant solution. The ~ can be opened by perfusingat a pte~nm of 300 m m H g [3] and thisdoes not ~ to damage the tissue ~ound the c a p ~ (Fig. 1). A l ~ a t i v d y , slices of ~ less than 1 m m thick can be cut and imm~rsei in the i m ~ ~ ~olution. If a non-ionic substance is used for the latter,the electricalc~nduct~vity of the solutioncan be measured in o~ler to fc~ow ~he e|u~cm of the ,.~trao cellularsaline.The r e ~ c e between ch]ofidedsH~r eiectro~ i m m ~ in a ma~et~caUy stirr~ beaker of suero~ ~ o l u ~ w ~ m e . r e d with a digitalresistancemeter pas~ng 2 p A and calibratedby ~]d~g succe~ve 10 pl aliquots of 0.9% NaCL The conductivity was found to riseand approach a plateau exponentiallywith a time c o ~ t of 75 ~ec and 90~ 120 sec of immersion w ~ mfficient to p~esct~e e x ~ t [ ~ ~pacc. ~ i n slicescan be cut even from the centms of rag or mouse brainsand good fixationobtained (Fig. 2). Moreover, immemion fixationallowse~tracelluiar space to be preserved in human cerebraland ~ b e l l ~ cor'~ ( ~ . 3). The impenneant ~ution was ~ at a concentration thatis~tonic with ceteb~ospinal fluid (about 0.28 M). It can be expected to ~ptace the ~ t m cellular fluid without changing the size of the c e l l ~ ~ . In p m l ~ i nary experiments,glucose,HEPES, calciun~c.hlorideand ~crose were found to act as impermeant solutesthat preserveextracellul~space and I have u ~ isotonics u c ~ in most subsequent experiments. Addition of aldehyde fixative was found to lead to electron miczognkuhs s h o ~ g much more ~ a n 20% extraceIl~arspace and severelyshrunken ce~ularprocedures. ~ w~ p~obably bec~use aldehyde fixative increases the permeability of the cell membrane and small molecules then di~se out of the ceils.Sucrose d ~ not enter the cells,so the cytop~az~n becomes hypotonic and t~e ce~s shrink. This a h ~ e can be avoided by replacing part of the sucrose with another
303
Fig. S. An ~
~
z
~
~ 15% ~ p m ~ d
is ~ e ~ u ~
of a m o u ~ ~ r
~4
~euro~u~c~l ~peeh'~en of h u ~
~II~
~
f~u:~ by ~
m
~
~ t ~
~ c e that is able to d i f f ~ into the ce~Is in the pre~'~ce of a l ~ h y d e f~ative~° ~ r a phosphate, cac~:~y~te, ~ l p b a ~ , ~ e ~ d ~ h t o m a ~ ~,Jd u ~ a ~ d glycerol b ~ e ~ tried ~ ~ ~ ~ ~o~ ~ d ~ u ~ ph~ p h a ~ wa~ found to give ~ most ~tt~factor] ~ ~ U o a of ~ ~ c ~ j ~ The ~ize of the extraceUu]~ ~ p ~ ia the elect~n m ~ r ~ ~ d~end~ on the ~ o of p e r m e ~ t to ~ p e ~ n ~ t ~ t u t e ~ , and a one,h o r s e ~ o ~ ra~o w ~ found e m ~ y ~ g~ve an e x ~ m e ~ ~ of 1 5 - - 2 ~ c o ~ sis~ent~ t h the ~ t ~ of ~ e chemic~ m ~ ~ ~ [1]. ~ ~ of ~ e ex~zacellu~ s p ~ e was e s ~ a ~ d ~ a grid of 1 0 0 red ~ i n t ~ 2 cm a p ~ on a transparent acetate ~ wh~h w ~ ~ d ~ e r each photomic~e~ graph ~ d ~ e number of ~ m ~ lying over e ~ t ~ e ~ l u ~ r space counted for te~ placements of the grid. Myefin ~ the m ~ t die.cults~mctt~e to ~ ~thout dmtortion,~ ~tis conven~o~ m e ~ ~ d the ~dd~fionof 2 m M ~ u m chloride~ m e d help the a d h e ~ of the myelin l a m e ~ . For ~ e same m ~ n , I have prefer~d ~ dehydrate ~ e t ~ e by ~dL~g alcohol slowly to a s t ~ and ow~rflowing d~sh c o n ~ a h ~ ~ e tissue m water, so ~h~ t h e concentration of ~ c ~ o l ~ses ~ f l y from zero to 8~% over ab~zt 15 mira The addition of ~dehyde fi.xat~ve~ ~ the peffu~te or ~ m e r s i o n fluid h ~ al~o been made in a c ~ t h ~ u ~ mann~, sLnce the c ~ a o ~ t y of the commonly used a ] d e h y ~ fixatives ~s large and the~ sudden addition can d ~ r t ceHu~.ar profiles. However, extm~Uular ~p~ce is p ~ e s e r ~ whether con~uou:~ or abrupt c h ~ g e s of aldehyde and ~ o h o l concen~ation are employed°
395 A ~ e d u l e for fixation has been developed in i 9 4 experiments in which the impermean~ p e r ~ e a n t and ,fixative chemicals were systematically varied and the method d e s c r i ~ below is as early to use as conventional procedures. It is fikely to be c a p a b k of f u , ~ e r improvement because the fixation of myelin z e m ~ imperfect. ~ t h ~ e perfusion m e t h o d and the immersion method i n v ~ a b l y preserve extraceUular space, although the same solutions p e r f ~ a ~ ~ lOO m m H g presvaze without opening the blood-brain barrier do n o t and immersion o f fresh brain slices directly into the final fb ative does n o t preserve extzacellular space, in a deeply anaesthetized rat, the apex of ~ e heart is cut off and a cannula made by squaring off t~e end of a No. 14 hypodermic needle is clamped in the ascending aorta. A roller pump cmme,::~e~'!to the c a l m e r is brought up to a speed sufficient to supply per~sW ;- ~:~m a stirred beaker at a pressure of 300 mmHg. When 100 mI pref~asate has been passed and lOO ml is left in the beaker, a second solution of the same composition as the first but containing aldehyde fixative is run into the beaker to keep this level of fluid roughly constant. The presydre is ~ u c e d to 199 m m H g and 2 ~ ml of fixative solution perf u ~ . After 2 h ~ e b ~ is removed and slices of ~ssue about I mm thick Lmmersed in the ~ e fixative for 2 days. As an alternative to perfusion, nec~-~ry when dealing with human cortex o b ~ e d at neurosurgery, small s t i ~ less ~ 1 mm thick ~ c a t from fresh brain and immersed in 100 m| of the impermeant s o l u f i ~ in a stirred ~ a k e r . After 90 sec, 100 ml of the ~ o n d solution containing i~ative is run i n ~ the stirred beaker slowly over 3 rain, After 2 days, the ~ r f u s e d or ~ m e r s i o n o f i x ~ ~is~ue in a vi~ con° t~ning t ml of fixative ~ c ~ i ~ to 4°C and 0.2 ml of a 6% solution of ~ m i u m tetroxide in wate~ ~iu added i o a fin~ osmium concentration of 1% the ~ o n d fixative ~oiution. After 24 h, the ti~ue i~ washed in 0.1 M ~¢~dium a ~ t ~ buffer at pH 4.5, stained i~ 2% ~ y ! acetate for 2 h and dehydrated by slow co~tinuou~ addition of alcohol. The tissue is then embedded i~ A~aldite, Foz ~ ~ u ~ . o n ~ d imme~io~ f l e a , on, the. fi~t, i m ~ r m e a n t solution is 0.135 M s-~o~e with 0.135 ~ M a ~ i u m phospha~ buffer at pH 7.3 and 0.002 M calciura chloride. ~ e ~ c o ~ d ~olution i~ the ~ m e wifh the addition of 1% g~utaxaldehyde for pert~sion or 2% S u ~ I d e b y d e for im~ner~io~ fi):ation. F o ~ a l d e h y d e or aero~m can ~ added wi~h ~he giut~zafdehyde. The two ~ l u t i o n a ~¢e made by di~olving 18.5 g sucrose a~d 4.1 g NaH~POa * 2H~O in 400 ~ water, adding 9~5 ml of a 5% ~ l u t i o n of c~¢iura chloride and bringing the p H to 7,3 by ~dition of a sa~ur-ated~lu~;ion of sodiu~ hydroxide. The d i s t r i b ~ o s of exizace~utar ~pace h ~ ~ e v e ~ in~re~ting fea~ure~. Diffusion ~o and from b~)d ca~)i_'Harie~'~ i~ r~ot restricted to narrow c~efts between swo~en c~ul,ar p ~ o c e ~ , bu~ .~ ~iowc~l ~i~ac~.ou~eharme~ {Fig. 1 )~ Of ~ the ce~uiar elemen%, ~.~t~oc~,tic ptc~c~e~ ~ d n e u r o ~ d e ~ d r i ~ are ~'.o-~em ~ t ~wol!en by convengon~ F~ation. The *~hinfinge~ of ~et~o., c y s t : t ha~ are ~ n when exlx~ellular spaze N preserved o f ~ n do hog
~ inve~ ~naptic a p p ~ t i o ~ but leave synaptic clefts in dose c ~ tintfity with !z~--~ e z t m c e ~ d ~ ~ The di~fibut/on of space is uneven and ~ s ~ s ~ y . In a sym~tic g l o m e ~ in the ~ or ce~bdlar ¢ ~ , e x ( ~ 8) most of'~he cenul~ ~ ire contiguous, for ~ glomem/us ~ ~ space, but is widely ~ from the gnmule c e b ~ exUacellul~ space. ~ ~ o n s md snudl~ ~ ~have the same widths as in conventional p~peratiom, but both appear unusually prominent because cell t ~ ~ are w~2~y separated everywhere else. ~ study of the ~ . ~ t / o n of ext~mcellular space can be expecte~ to help our understanding of the structural bias of ~uronal interactions. REFERENCES I I,evln, V ~ . , ¥ ~ e m e h e r , J.D. and PatJ~k, C.8o,Suborn msdinulin sSpsce~ ments of eenebnd cor~z in four nutmmal/sn epeeiu, Amer. J. Pbye~l., 219 (1965) 1528---1533. 2 Peters, A., Palay, S.L. 8nd W ~ , H. de F., The Fine 8tn~ct.ure of the Nervous System, 8aunders, Philadelphia, 1976. Rapport, 8.I., Opening of the blood-bnd_nbarrier by Lute hypertension, F,~p. NeuroI., 52 (1976) 467--479. 4 Van ~ e l d , A., The ez~sceHular slptce in the ~ eentrzl nervous system. In G.H. Bourne (Ed.), The Structure and Function of Nervous Tissue, VoL 4, Academic P r ~ , New York, 1972, pp. 447-511. 5 Van Htrreveld, A. and $teiner, J., ~ t r ~ e U u l l r tp~ee h~ frozen and e ~ o l ~bttituted central nervo~ thmueoAnat. l~ec., 116 (1970) 117--130. 6 Van H ~ v e i d , A. and Khattab, F.|., Perfm~on f l z ~ o n with ~ u t u ~ _d~byd~amd p o ~ f~atton wi~ ountum tetrog~de for e ~ c t ~ n microcopy, J. C41ik L , ~ ()965} 479~494.