Neurofilament and glial fibrillary acid protein-related immunoreactivity in rodent enteric nervous system

NEUROFILAMENT AND CiLIAL FIBRILLARY PROTEIN-RELATED IMMUNOREACTIVITY RODENT ENTERIC NERVOUS SYSTEM H. BJ~RKI_CJND*, D. DAHL?

and A.

ACID IN

SEIWR

Department of Histology. Karolinska Inbtitute. Stockholm, Sweden; +Department of Neuropathology. West Roxhury Vcteranq Harvard Medical School and Spinal Cord InJUry Research Laboratory, .4dministration Medical Center. Boston. Massachusetts. I!.S.A

Abstract--~ Using antisera raised against ncurofilaments and the @al tihrillary acidic protcln (
Intermediate

cell

types.

viwalix

entcrlc neurona and their wpportwc

filaments

Although

have been observed structurally

similar

gIlal cells.

suggested that most of these cells contain all three polypeptides.“’ In the astrocytes of the CNS. glial fibrillary acidic protein (GFAP) is the major component of the intermediate filament system and immunohistoch~mistry with antibodies to GFAP has been used extensively for studies of astrocytic morpholog) severltl and development (see refs 2. and 14). Recently. papers reporting the presence of GFAP-like immunoreactivity in peripheral glial cells, huch as cell:, in the rat enteric nervous system,“’ some Schwann-cells in rat sciatic nerve.“.J’ and cells in the iris of several species.j have been published. However, since immunochemical techniques have so Tar failed to demonstrate the presence of GFAP outside of the CNS” it is presently unclear whether immunohistochemistry detects an antigen identical to GFAP or whether the antisera cross-react with an unknown. structurally similar intermediate filament. Using cryostat sections, NF-like immu~oreactivit~ has been observed in

in most they

are

made up of chemically and immunologically distinct prote:ns in cells of different embryological origins (see tef. 28). In cells of neuronal origin the intermediate filament system, named neuro~lamcnt (NF) 1s composed of a triplet of polypeptides weighing approximately 70,000, 150,000 and 200,000 dalton each.“ Using antibodies against the NF triplet it has been shown that many neuronal types. both in the central and peripheral nervous systems, contain lmmunohistochemically detectable amounts of NF.Y ‘(l.liu K~‘-w~’ Recent studies using antisera spccitic

to each

ol‘ the triplet

polypeptides

have

also

*Plea\e address all correspondence to: Dr. t-1. BjBrklund, De:partment of Histology, Karolinska Institutet. Box 60 400. S-104 01 Stockholm. Sweden. .-lhhrt,r.il,/iirtt.t : GFAP. glial fihrill~ir~ acidic protein; N F. ncurolilamcn~. 271

the enteric nervous system.‘3~26However, sections do not allow detailed descriptions of the number, morphology and distribution of either the neurons or the enteric glial cells. Recently we have examined the distribution and development of NF3’- and GFAPlike3 immunoreactivity in stretch-prepared whole mounts of rodent iris. Similarly, whole mounts of muscle layers including the myenteric plexus can be used to examine the morphology and development of the enteric nerve plexus.6*7~20,23~2S.29 It was the aim of the present study to investigate in more detail the distribution and development of NF- and GFAPlike immunoreactivity in rodent small intestine using such whole mount preparations.

EXPERIMENTAL PROCEDURES

Aflit?& Aduft albino rats (Spragu~Dawley) of both sexes as well as pups and embryos from each of the four age groups. embryonic day 19 and 21 and postnatal day 1 and 6, adult male aibino mice (NMRI) and adult albino guinea-pigs were used. All animals were deeply ether anaesthetized and killed by bleeding out or cervical dislocation (foetal material from pregnant rats). fmmun+luorescencerechniques From all animals pieces of the distal part of ileum were cut out and whole mounts of smooth muscle sheets containing both the longitudinal and the circular muscle layer as well as the intervening myenteric plexus were prepared. Stretch-preparations were also made from adult rat and mouse mesenterium. All whole mounts were prepared on ethanol washed, gelatin-coated slides and air-dried from one to several hours. No fixation of the whole mounts were performed. Parts of distal ileum from adult rats and mice were further sectioned at 12 pm on a cryostat. The sections were post-fixed for 3min in 100yOacetone at room temp. The antiserum to GFAP was raised in rabbits against degraded antigen (mol.it 40,500) isolated from phosphate buffer extracts of autolysed human spinal cord* and used diluted l/l00 in 0.1 M phosphate-buffered saline. Absorbed antiserum was used as a control (see ref. 8). GFAP antiserum raised in mice against human degraded antigen was used in double labelling experiments with NF antiserum. This GFAP antiserum was diluted l/SO. NF antiserum was raised in rabbits against a 50,000 mol. wt degradation product of chicken NF protein as described earlier.’ The NF antiserum, which reacts with ail three peptides in the NF triplet,‘2,‘2 was used diiuted I/l000 with preimmune serum used for control. All slides were processed according to the indirect immunofluorescence technique.’ After 10 min rinse in phosphate-buffet saline the slides were incubated at 4°C overnight in a humid atmosphere with the primary antiserum. After rinsing in phosphate-buffered saline, incubation in either ~uores~in-~sothiocyanate conjugated or rhodamine conjugated sheep anti-rabbit antibodies (Dakonatts, Denmark) diluted ii100 (fluor~~ein-isothi~yanate) br l/510(rh~~ine) for 1h at room temp in dark&s was oerformed. All antisera contained 0.3Y v/v Triton X-100. kfter a second rinse, the slides wei; mounted in 90% glycerine in phosphate buffer. The double labelling experiments were performed over 3 days so that first the NF antiserum was applied followed by rhodamine conjugated second antib~ies as described above. The 2nd day, after rinsing, the slides were incubated over-night with GFAP antiserum raised in mouse followed by fluorescein-isothiocy~ate conjugated rabbit anti-mouse antibodies (Dakopatts, Denmark) diluted 1110.All slides were

examined in a dark field fluorescence mtcroscope. I G-X film (Kodak) was used for photography. RESULTS Adult rat enteric nervous svstem Neur~~l~men~-lake i~~n~rea~t~t?~~y. In the whole mounts a large number of myenteric ganglia were seen evenly distributed over the underlying circular muscle layer (Fig. la, b). All ganglia contained large cell bodies showing a strongly fluorescent cytoplasm and a negative, eccentrically placed nucleus. Most positive neurons were brightly fluorescent and intermingled with occasional more weakly fluorescent cells. The number of positive cells in each ganglion varied from a few cells up to ten and all together at least half the total number of perikarya seemed to be stained. NF-negative perikarya were often outlined by fluorescent processes and consequently easily discernible. Usually, long, relatively thick fluorescent processes could be seen emanating from the positive perikarya. No short dendritic-like processes were seen. Apart from the ganglia, NF-positive perikarya were also frequently found in the interconnecting strands running parallel to the circular muscle layer. whereas no cells could be seen in the longitudinal strands. In the circular and longitudinal interconnecting strands a large number of long. thin, smooth fluorescent fibres were observed. The number of such fibres were clearly higher in the strands running parallel to the longitudinal muscle layer than in those running parallel to the circular muscle layer. A large proportion of the fibres in the strands were seen passing through individual ganglia without terminating. As seen in Fig. 1b, the fluorescent processes from the perikarya contributed considerably to the fibres in the strands. In the IongitudinaI strands, and to a lesser extent in the circular strands, small nonfluorescent areas were frequently seen among the NF-positive fibres. This network of NF-positive fibre strands and ganglion cells was superimposed on a sparse regular plexus of thin, delicate NF-positive fibres running in the circular muscle layer. Fibres in this muscle layer could occasionally be seen emanating from the interconnecting strands. In the sectioned material large myenteric ganglia with fluorescent cell bodies were seen between the muscle layers. Long, thin, usually individual NFpositive fibres were also seen in the muscle layers. These fibres were much more numerous in the circular muscle layer than in the longitudinal. NF-like immunoreactivity was also observed in the submucosa, where occasional fluorescent neuronal cell bodies were intermingied with irregularly d~str;buted fibres. Individual long, thin NF-positive fibres were abundant in central parts of the villi, especially towards the base. These fibres were best visualized in oblique sections (Fig. 5a), where they formed an almost continuous fluorescent rim around the epithe&al cells.

NF and GFAP in roden: enteric nervous system GFAP-like immunoreactizity, In the whole mount preparations, fluorescent structures were seen in all ganglia and with a distribution clearly different from that observed with the NF antiserum (Fig. lc). Small. weakly fluorescent ovoid cell bodies with several short positive processes were intermingled with the negative neurons. Most of the fluorescence within the ganglia consisted of individual processes, whose cellular origin was difficult to trace due to the abundancy of such fibres. Short. thin GFAP-positive fibres were frequently seen over the negative areas that probably represented neuronal perikarya (Fig. Id). Due to the irregular shape of the relatively small GFAP-positive cells it was difficult to estimate the ratio between them and the NF-positive cells within the ganglia. Thin. smooth GFAP-positive fibres were also present in the interconnecting strands although the fluorescence intensity was low. Occasionally. GF.4P-positive cell body-like structures were seen in these strands. Only very few weakly fluorescent fibres were visualized in the underlying circular muscle layer. The distribution of GFAP-like immunoreactivitl in sectioned rat put wall largely resembled that seen with NF antiserum. Large myenteric ganglia containing irregularly arranged strongly fluorescent fibres but without fluorescent cell bodies were seen. Individual GI:AP-positive fibres were seen in both layers of the turlica nluscul~ris. As with NF antiserum. the density of fibres was higher in the circular muscle layer than in the longitudinal. Fluorescent ganglia were frequently seen in the submucosal layer. with occasional small ovoid fluorescent cell bodies. GFAP-positive fibres were also seen in central parts of the villi (Fig. 5b). although more sparse than those seen v.ith the NF antiserum.

N~~lrrofilamrnt-lilie inmunoreacti~~ity. The description given above for the rat is largely valid also for the mouse enteric nervous system. Myenteric ganglia and the interconnecting strands had the same appearance as in the rat (Fig. ?a). However, the cell bodie? were more brightly fluorescent, some NF-positive perikwrya were seen in the strands running parallel to the longitudinal muscle layer and at times also outside this network. Furthermore, the underlying circularly arranged fibres were much more sparse than in the rat. Double labelting experiments using NF and GFAP antisera showed that the two antisera stained two different cell populations (Fig. 3a, h). It was furthermore obvious that the GFAP-positive tell bodies were more numerous than the NF-positive ones and that the latter ones were to a large extent surrounded by GFAP-positive cells and processes. Secrioning of the gut wall again revealed Ilucrre,cent ll~yenteric plexuses and individual NFpositive fibres both in the circular and longitudinal external muscle layer. As in the rat the fibre denbit!

279

was higher in the circular layer. The submucosal plexus also stained positively for NF. Compared to the rat relatively few NF-positive fibres were visualized in the vik. GFAP-like irnmunor~~actitity. With GFAP antiserum. structures in the myenteric nerve plexus. Including both ganglia and interconnecting strands, were more easily visualized in the mouse than in the rat (Fig. 2, d), although the general distribution was comparable. All ganglia contained a high number of small, strongly fluorescent cell bodies from which numerous short GFAP-positive processes emanated. Such cells were also seen in the strands. In the ganglia the GFAP-positive cells surrounded and outlined the negative neurons which they clearly outnumbered, probably bq a factor of two or three. Apart from cells. the ganglia contained a dense system of fluorescent tibres of variable length. SimilarI\, the interconnecting strands contained a high number of parallel, long, thin. GFAP-positive fibres. Also in the circular muscle layer. a large number of fiuorescent fibres were visualized. Within these structures fluorescent elongated cell bodies with processe\ 01 different length were frequently observed. GFAP itnmunohistochemistr~ on sections oi mouse gut wall revealed practically the same distribution of GFAP-like immunoreactivity ah in the rat, although cell bodies were not as clearly seen. Furthermore, as with the NF ~intiseruIn fewer GF:~P-positive fibres were visualized in the villi. Adult

guimr-pi,q

iVezcfofi/ut??fnt-/ik~, itl?rrlutlorecrc.tir.;t~,. The general distribution of NF-positive cells and fibres in whole mounts of guinea-pig ileum was no dltferent from that seen in the rat and the mouse. However. in contrast to the NF-positive neurons in these latter species. which seemed to be rather uniform with onI> one large process each, the neurons in the guinea-pig myenteric ganglia seemed to be of two tlpea. Most 01 the cells resembled the above-lnentiot~ed unipolar type. A second. less frequent, brightly fluorescent cell type was observed, which apart from the single process also had several shorter NF-positive processes eman~~tjng from the perikarya (Fig SC).These cells were often quite small and ovoid. GFAP-like i/,zmunorc~uc,ti~.it~,. When GFAP immunohistochemistry was applied to whole mounts of guinea-pig ileum onI> very fet\ weakly fluorescent tibres Mere observed in the periphery of the stretchpreparations. The ganglia and the interconnecting strands were usually negative or Lery \ceakly stalned.

:Vr[c~o~j/ur)lerlt-like in~mu~~oreot~tir~it~.. .4 dense network of myenteric ganglia and .interconnecting strands were present both prenatalI>, pestationul da! 19 nnd ,71 and pastnatally, da> 1 and 6 (Fig. &I). In general. the same distribution of NF-like Immu-

280

Fig. I. Dist~bution of NF- (a and b) and GFAP- (c and d) like immunor~c~vity in whole mounts of the circular and lon~tudinal external muscle layers including the intervening myenteric plexus in distal ileum of the adult rat. In (a) large fluorescent cell bodies with eccentrically placed negative nuclei are seen within the ganglia and to a lesser extent in the circularly running strands. A large number of thin, smooth NF-positive fibres are seen both in the longitudinal and circular strands. Within these bundles small, negative areas surrounded by fluorescent fibres are seen. In the circular muscle layer thin NF-positive fibres are visualized. (b) Higher marination showing NF-positive neurons in a myenteric ganglion. Note that all cells in this ganglion have a relatively thick fluorescent process running parallel to the interconnecting circular strand. (c) A large number of GFAP-positive fluorescent fibres are seen in the strands connecting the ganglia. Within the ganglia a high density of GFAP-positive fibres and processes are seen surrounding large negative areas. (d) High magnification of a ganglion illustrating the abundancy of hbres and short processes around negative swellings (large arrows). Note weakly ftuorescent small cell bodies (small arrows). Fluorescence microphotog~phs (a) and (c) x 135; (b) and (d) x 390. Fig. 2. Distribution of NF- (a and b) and GFAP- (c and d) like immunoreactivity in whole mounts of external muscle layers from adult mouse ileum. (a) NF-positive perikarya present in the myenteric ganglia and circularly running connecting strands. The perikarya have negative nuclei and variable fluorescence intensity. Note the strongly ihtorescent cell present outside the network of ganglia and strands (arrow). In the interconnecting strands, many NF-positive fibres are seen, especially in the longitudinal strands (b) NF-positive neurons in a ganglion. Several fluorescent fibres pass through the ganglion. (c) A dense system of GFAP-positive gbres are present both in the strands and in the ganglia. Many GFAP-positive fibres are seen in the circular muscle layer. (d) &lose up showing GFAP-positive cells (arrow) and hbres in ganglia, strands and the circular muscle layer. Fluorescence microphotographs (a) and (c) x 135; (b) and (d) x 460. Fig. 3. A ganglion and connected strands from a whole mount of adult mouse myenteric plexus double labelled with antisera against (a) NF and (b) GFAP respectively. Note strongly fluorescent NF-positive cell bodies in (a) (arrows) which are negative in (b). With GFAP antiserum a large number of short fibres and small, weakly fluorescent eelI bodies surround the NF-positive perikarya. Fluorescence microphotographs x 330. Fig. 4. Whole mounts of the external muscle layer and intervening myenteric plexus from distal ileum of I day old rat pups. (a) NF-like immuno~cti~ty. Positive fibres are seen both in the ion~tudinaiiy and circularly running strands although more abundant in the former ones (vertical). Fluorescent cell bodies with negative, eccentrically placed nuclei are seen in clusters along the circularly running strands. (b) GFAP-iike immunoreactivity. A system of fluorescent fibre bundles forming a regular network similar to that seen in (a) is visualized. Fluorescence microphotographs x 135. Fig. 5. (a) NF-like immunor~ctivity in obliquely sectioned rat intestinal mucosa. Fluorescent fibres outline epithelial cells of the crypt walls. (b) Adjacent section reacted with GFAP antiserum. A similar system of fluorescent fibres surrounding the epithelial cells is observed. The staining of the luminal borders (circular profiles) is antifactual. (c) High magnification of an NF-positive neuron and fibre bundles in a whole mount of adult guinea-pig myenteric plexus. This cell has several fluorescent processes, one of which can be seen joining the adjacent fibre bundles whereas the others are shorter. (d), (e) and (f) Whole mounts of adult mouse (d) and rat (e and f ) m~enterium. {d) GFAP-like immunoreactivity showing a thin flbre bundle with small, almost negative swellings at regular intervals. (e) NF-like immunor~cti~ty. A number of fluorescent fibres are seen partly along blood vessels. (f) GFAP-like immunoreactivity. Fibre bundles partly along blood vessels. Fluorescence microphotographs (a), (b), (e) and (f) x 135; (c) and (d) x 330.

Fig.

I

281

Fig 2 282

J.

4.

s1

not-cac.ti\tt)

\\;I\

fluorescence numerous

po\~tt\ e neut-ens acre in the circularly Jilticult

running

to detine

;~nd connecting po\ltl\

c

both

the exact

in ganglia

the strands

ganglia

state.

hF-

all

inter-

in

and sparse thin fibres were seen in

the Cl1c11l;11.mll~clc

la&Y

I /‘-/iAt, i/lf,,lr,,ro~c,trc,~;~./~~,. A similar network ii:, \ t~ualt/cd \tlth NF antiserum was obser\,cd ,~ga~n: t ;I hi&r background using GFAP antiserum. (IF

WI~III ovoid

cell bodies

I~I-I>c~~,\cLUCI-c 5ccn 111Ihe postnatal \\JIcI-c.1’1 no \IIC~ \tructure:,
the g 1ng11a

;\I’-po\iti\c

could

wemed

\vith

short

stages (Fig. be visualized

,I\ in the adult

pt-cn;ttall>.

i\ithin

sections.

state

cell\

\ C’\\t‘l

A!‘-po5tti\‘c

~ithtt,

the

both

in rat and mouse

cnod

WLIII

hr-anches

adipose

werr

tl\huc.

ncgat~w

ititcr\.iI~

cdl

111the bundlc~

“

cspeciallq

With

GFAP

sciatic

mesen-

frequent antiserum.

5~1).

DISC‘1 SSION general organihatton wa\s

\yn;tptic

neuropil.

numeroii5

i\olation

processes no

Furthermore. ccll~

;I

contain

tilamcnts.’ ct~~d entcrtc bindtit~ IocaIi/ed

in

[email protected]

finding

basal

Iamina number ~tmilarity

brain.

cell\ cell5 MII~I

SUC~I

the> Inrerganglionic la>er c)I‘ rat and

of

In

intertnedtatc astroqtes

cell

types.” to

with

the

we lind many

tnyenteric

fluorescent

This

be tnalnly

agreement

technique

panplia.

has enabled

LI\ to

processes both in

and in the circular

The high number

posit ve cells in the m)enteric

glial

01‘ the acidic calciutn-

111 rodent

strands mouse.

around

enteric

between

constdercd

tcs.“ “

.21\L). the I\ hole mount

are

granulae.

is present

conspicuous.

of .les\en and Mirskq,‘”

(,t~;Il’-po5iti\e IX

glycogen

5 IO0 in both

a\troc!

interest

in shape. have

glia is the prrwnce

protein.

space

arc irregular contain

hl$

p~-ote II ih. 111 the

L~,U;I

vessels. ab-

extracellular

cells and CNS

and

‘I’ ” A further

other a dense

glial

cells ;ttid mo5t

Indtviilual

blood

small

entcric

Both cell t>l-a

extensive

from

than

include

glial cclls.‘~ Of particular bet\veen

plexus in

the CNS

similarities

tissue.

the stmilaritics ;t\tro
more

Thr~

01‘ connecttnp

\cncc ,ind

resemble\

pngl~a.

;tutonomic

of the myenteric

plexus,

all enteric

However. failed

so far

outside that

filaments further

muscle

of GF;\P-

outnumbering

true

present

in

Using Their

large

indicated

sire

techniques repot-t\

of the

01’ ha\c

intermt’di;ttc

abundant

glial

a totallv

” gl~al

:IU;II~\

ccII\

and

ditferent

only

Howwer.

eccentrically

placed

;I proportion

M;IS stained. cell

bodies

tindings.

OnI)

70”,,

The prcscncc i\

in

of

the

cells,“’

are

NF-negative

means

that

the!

specific neurons

or cells with

tdentit)

specilic

plexus

were

Indicating

extrinsic

UII-

br

unipolar.

species

origin.

could

ll~)t the

Lshich h:iLC

dreneryc.

”

anttsera dilferent

obvious

that

clearly

Inter-

\\lth

cell populations.

in thin herve

lihrc\ Intrlnslc

.le\\cn

that

~LIIIIG-

t -po\itl\e

of ht-posittic with

showed

m!01 Cc‘ll\

l’clr \

cxpertmcnt>

GFAP-positi1.c branching

Mll--

\I.

and

the t\+r, a~lt~%~.a

F’urthcrmorc. structure\

tibres

tibl-c\

111 and

and

partI!

The pt-esencc 01‘ UtT-lrl\c

the neurons.

noreactivit!

a populalloll

difference\

In agreement lahelling

:I

or riiot~~~

concern\

suggest that they are ol’hoth

double

GFAP

it, sen\or\

neut-on\

types. The abundunq

the strands sky,“’

;IIu) I\

~-eprc’~~lt

short proccsses were SSCI~tn the

with multiple neuron

to

tht\

proccsbc\

u hereas most cells in rat ;~nd m~)ux

estingl!. enteric

suggested

111 the gr;~nula~-

Wherhcr

neul-on\

extra-ganglionic

been

01’

;I\ to the tr;lnw~lttcl

of some NF-positive

earlier

CJI-IICI.

cell t)pt‘\

transinittcr~.

One indication

few fluorescent

\\lth

in cultu~.c\

nettt-on\

~LICII

of neurons.

be determined.

ncut-on;11

(~1‘ ?r t.-nc+ttl\c

” “’ k ill-tlicrmc~i-c.

Nl--nepali\e

the NF-posittvc

group

01‘ the

neuron\

i/l riro.

have

Whether

(see l-cl‘

cells ‘I’ and ccrebell;lr

such as Purkinje

clear,

nuclcu~

neuron\

agrcemcnt

dorsal root ganglia are NF-posilt\c the pcrtkarya of heveral nt‘uron;l1 CNS.

population

in the \\h~~lt~ mount\.

that they represent

neuronal

round

01‘

con-

determlnant~.“’

enteric

was vtsualized

filaments.

in the mand~.

classc\

NF antiserum.

clcarl~

stain

iri\

(this

the prc,cncc

antigenic

the

~n~~:r~nin&xi

fibres

nerve.

RecentI!.

of

III

the gut. such ;I,

experimentation.

cated that even distal

hl-‘-poyiti\e

from

01‘

reported

mesenterium

identity

neuronal perikarya by a factor of two or three, as well :I\ tll,.tr location bet\veen neurons in the ganglia .tnd MII~I

The prcscncc

olfactor>

CNS.”

;II Ic‘a\t

constltutc

has been

different

of cell bodies

pig,

111 r~~lent cIc;~r. but

immunochemtcal

share common the

cell\

to demonstrate

the

01‘ m~entert~~

15 not

glial cell\.

rat

have

Therefore

glial

tissues apart

munication).

filaments

gut

suggests that the!

nerve.“,”

pertkarha

\evcr;ll

the

in culture\

species? and rodent

indicated

tibrcs

are GFAP-positive

several peripheral

were SCCII at repulat

bodies (Fig.

Whether

with

36).

7ht

of

immunoreactivity

(Ftg. 5~. I‘). The tihres usually followed blood \ and branched frequrntl~. Such NF- and

tertum (il

prcwnt

and

in the 1 ill), l‘urthet

their abundancy

rat

tibre\

them to enteric glial cells. GFAP-posltt\c

intestine

several

cryostat

prcwnt

layers

GFAP-like

by thtn

Il\ing

L+;III.

most

4b).

Iibrt3

ILC‘K

cells

of the enteric

nrpaticc

to be outlined

cell\.

tibreh III

are

plexus.“” small

gl~al

enteric

and fluorescent

a large part

\‘c,iit~o/ricillr~~,,r~111~1 (I F.-II’-IiXc iirlt,lfitiot.c,trc~ii/.it~~, ~p:~rx 5~\tem 01‘ N t-‘- and GFAP-positive tibre

l>undl~\

plexus

throughout linking

are

GFAP-poxitivc

cell bodies have been QYI~ in the

GFAP-positive

GFAP I

they

and in both muscle layers.

fluorescent

submucosal Thus,

that

we have found

have also been demonstrated

Ih;it

\ lc\+ lluore\ccnt

suggest

and

between

present

strongly

Furthermore,

it wts

border-

were

less NF-

although

As in the adult

hundlc\

5trantlz

prenatally.

strands.

strand+.

tibre

connrt:ting

at all stages, although

ohjerved

cie~cl~~pcdv,ith weaker

111the \IIII

processes contaln in the intestinal

II \\:I\ \ur~mmuAndy-

~nrcrmcdtalc muc’~)~I :Irc‘

known to originate from gangha both wlthin and outside the gut (see ref. 36). The prenatal appearances of both NF- and CiFAPpositive cells and tibres are in agreement with recent studies on the rat iris.’ ” However. the temporal relationship between the first appearance of GFAPand NF-like immunoreactivity is still unclear. The almost complete absence of GFAP-like immunoreactivity in whole mounts of guinea-pig ileum could be due to the relative thickness of these preparations compared to whole mounts of rat and mouse gut. It is surprising that no similar problems were encountered with the NF antiserum.

These experiments demonstrate the appearance and distribution of GFAPand NF-like immunoreactivity in rodent enteric nervous system and

mesenterlum. With both antisera fluoresccnl lihrci arc found in all layers of the gut wall ~crp~ the epithelium. GFAP-positive cells are also I;~und li\ most layers whereas NF-posltivr cell bodies arc I;‘stricted to the myenteric plexus and to a lesser c.\telll to the submucosal plexus. Most if not all enterlc glial cells seem to be GFAP-positive whereab only a proportion of the neurons are NE.-positive. We COIIelude that immunohistochemistry with annbodIes against NF and GFAP is a useful means for visualization of enteric neurons and glial cell\. ,1cli,lo1r,lrdlgernent.,~-~Supported by the Swedish Medical Restarch Council grants 14X-06555, 14X-03185 and 25P-6326. the Swedish Natural Science Research Council grant B-Bll I522- 102, Magnus Bergvalls Stiftelse, the “Expressen” Prenatal Research Foundation. Karolinska Institutets Fonder. Dr. Dahl was supported by the Veterans Administration We thank L. Hultgren, A. Hultgardh and I. Stromberg for skillful technical assistance and K. Seiger for typing

REFERENCES

I. Barber P. C. and Lindsay R. M. (1982) Schwann cells of the olfactory nerves contain glial fibrillary acidic protein and resemble astrocytes. Neuro.s&~zcc 7, 3077-3090. 2. Bignami A., Dahl D. and Rueger C. (1980) Glial fibrillary acidic (GFA) protein in normal neural cells and m pathological conditions. In Adr~rtces in Crlluirr Neurobiology (eds Fedoroff S. and Hertz L.). Vol. I, pp. 2X5-319. Academic Press, New York. in the iris. Development, distribution 3. Bjlirklund H., Dahl D., Olson L. and Seiger A. (1984) GFA-like immunoreactivity and reactive changes following transplantation. J. Neurosci. (in press). G. (1976) The ultrastructure of Auerbach’s plexus in the guinea pig. II. Non-neuronal 4. Cook R. D. and Burnstock elements. J. NeurocTrol. 5, 195-206. antibody methods. In G‘enrrul C~~/oc~hernicu/ Melhods (ed. Danielli J. F.), pp. 399-422. 5. Coons A. H. (1958) Fluorescent Academic Press. New York. 6. Costa M.. Bufi R., Furness J. B. and Solcia E. ( l98Oj Immunohistochemical localization of polypeptides in peripheral autonomic nerves using whole mount preparations. Histochrmistry 65, 157- 165. innervation of the alimentary canal. 2. Zdlforsch. mikrosk. Ancrr. 122, 7. Costa M. and Gabella G. (1971) Adrenergic 357-377. properties of the glial fibrillary acidic protein. Bruin Res. 116, 150&157. 8. Dahl D. and Bignami A. (1976) Immunogenic of antisera to neurofilament protein from chicken brain and human sciatic 9. Dahl D. and Bignami A. (I 977) Preparation nerve. J. camp. Nrurol. 176, 645-658. localization of the ISOK neurofilament 10. Dahl D.. Bignami A., Bich N. T. and Chi N. H. (1981) Immunohistochemical protein in the rat and the rabbit. J. c~onlp. Neural. 195, 659-666. I I. Dahl D., Chi N. H., Miles L. E., Nguyen B. T. and Bignami A. (1982) Glial fibrillary acidic (GFA) protein in Schwann cells: Fact or artifact. J. Histochcvn. Cytochem. 30, 912~-918. of neurofibrillary tangles in Alzheimer’s 12. Dahl D.. Selkoe D. J., Pero R. T. and Bignami A. (1982) Immunostaining senile dementia with a neurofilament antiserum. J. Neurosri. 2, I 13-l 19. 13. Debus E., Fliigge G.. Weber K. and Osborn M. (1982) A monoclonal antibody specific for the 200K polypeptide of the neurofilament triplet. EMBO J. I, 41 45. 14. Eng L. F. and DeArmond S. J. (1981) Glial librillary acidic (GFA) protein immunocytochemistry in development and neuropathology. In Glial rrnd Nruronul Cell Bio/ogj~ (eds Vidio E. A. and Fedoroff S.), pp. 65-79. Alan R. Liss. New York. 15. Ferri G.-L., Probert L., Cocchia D.. Michetti F., Marangos P. 1. and Polak J. M. (1982) Evidence for the presence of S-100 protein in the glial component of the human enteric nervous system. Nurure 297, 409-410. 16. Gabella G. (1971) Glial cells in the myenteric plexus. Z. Naturf 266, 244245. 17. Gabella G. (1972) Fine structure of the myenteric plexus in the guinea-pig ileum. J. Anor. 111, 69-97. of the nerve plexuses of the mammalian intestine: the enteric glial cells. Neuroscwnce 18. Gabella G. (1981) Ultrastructure 6, 425436. 19. Gown A. M. and Vogel A. M. (1982) Monoclonal antibodies to intermediate filament proteins of human cells: Unique and cross-reacting antibodies. J. Cc// Bioi. 95, 414-424. and histochemical observations on the myenteric and submucous plexuses of mammals. 20. Gunn M. (1968) Histoiogicdl J. Anat. 102, 223-239. A., Hansson H.-A., Hydin H., Persson L. and Riinnbdck L. (1974) S-100 protein in synapses 21. Haglid K. G., Hamberger of the central nervous system. Nature 251, 532-534. 22. Haglid K. G.. Hamberger A., Hansson H.-A., HydCn H.. Persson L. and RtinnbPck L. (1975) Cellular and subcellular distribution of the S-100 protein in rabbit and rat CNS. J. Neurosci. Res. 2, 175-191. 23. Hill C. .I. (1927) A contribution to our knowledge of the enteric plexuses. Phil. Truns. R. Sot. B 215, 355-387. of axonal transport. Identification of major structural 24. Hoffman P. N. and Lasek R. J. (1975) The slow component polypeptides of the axon and their generality among mammalian neurons. J. Cell Biol. 66, 351-366. plexus. Am. J. Anat. 49, 141~166. 25. Irwin D. A. (1931) The anatomy of auerbachs