close

Вход

Забыли?

вход по аккаунту

?

Observations on the fine structure of the carotid body.

код для вставкиСкачать
OBSERVATIONS ON THE F I N E STRUCTURE
O F T H E C14ROTID BODY1
CURTIS M. G A R K E R 2 . ~ N D DONALD DUNCAN
Department of Anatomy, The University of Texas Medical
Branch, Galveston
SEVENTEEN FIGURES
INTROI>UCTION
Present ideas on the structure and function of the carotid
body date from the work of DeCastro ( '28). Prior to his work,
the carotid body was widely regarded as a part of the chromaffin system and thus by implication as an effector of some kind.
For much earlier views and a history of the discovery of the
carotid body, see Luschka (1862). DeCastro disproved the
chromaffin idea, demonstrated the abundance and predominance of afferent nerve endings and suggested a receptor
function for the carotid body. Such function was established
shortly thereafter by Heymans et al. ('30). I n DeCastro's
opinion, as expressed in 1951, the glomus cells are all of one
type and are in direct contact with afferent nerve fibers. He
pictures the nerve fibers as entering the cytoplasm of the
glomus cells and expanding in meniscus-like fashion around
the nucleus. Other workers have described a more elaborate
architecture featuring interstitial cells, a variety of nerve
endings and at least two types of glomus cells. Such opinions
are exemplified and reviewed in a paper by de Kock ( '54).
Hollingshead ('43, '45) has done extensive work on the
cytology of the carotid body with particular emphasis on the
'Supported in part by USPHS Grant B-690(C2).
Supported by summer fellowships from the National Foundation of Infantile
Paralysis (1956) and the Lederle Company (1957).
691
692
CURTIS M. GARNER AND DONALD D U N C A N
nature of the cytoplasm. He describes the glomus cells as
containing very fine granules that are neither mitochondria
nor lipid droplets and his experiments indicated a depletion
in the number of these granules in the carotid bodies of mice
subjected to prolonged anoxia.
Tlie present study was started with intent to repeat with
the electron microscope the observations of previous investigators using the light microscope, especially those of Hollingshead on the cytoplasmic granules and their alteration by
anoxia. Recently brief abstracts on the electron microscopic
appearances of the carotid body have been published by Ross
( '57) and Duncan and Garner ( '57). These descriptions agree
in general with those presented here.
MATERIALS AND METHODS
Carotid bodies of mice and cats were exposed in living
animals and fixed in various ways. Some were excised and
immediately dropped in fixative. With others the fixative was
injected into the common carotid artery prior to excision and
further fixation. I n a few animals the common carotid arteries
were perfused with Muller's fluid or with 3% K,C,O, and after
a few minutes excised and placed in an osmium containing
mixture. Both Palade's ( '52) and Dalton's ('55) buffered
osmic acid mixtures were used with essentially equal success.
However, due to the difficulties of locating and removing
these bodies and their apparent unusual sensitiveness to any
fixation procedure, it was infrequent rather than usual to
obtain a good preparation.
Following fixation, the material was washed briefly in
distilled water and rapidly dehydrated in graded alcohols
starting with 70%. From absolute alcohol the specimens were
transferred to a mixture of 15% methyl methacrylate in
n-butyl niethacrylate without catalyst. After three changes in
this mixture the material was placed in N 0.00 gelatin capsules
containing the methacrylate mixture and about 1%Luperco
FINE S T R U C T U R E O F THE CAROTID BODY
693
CDB3 as a catalyst. The capsules were placed in a 43' oven
for 24-48 hours. Sections were cut in a Porter-Blum Servall
microtomC, ming glass knives as originally advocated by Latta
and Hartmann ( 'SO). The sections were picked up on collodion
coated grids or on uncoated grids which were subsequently
sprayed lightly with carbon, as recommended by Low ('36).
Sections were studied and photographed in an RCA EbIL 1B
electron micro~cope.~
RESULTS
The carotid body in both the cat and the mouse appears
a s an aggregate of small clusters of glomus cells scattered
throughout a mass of blood vessels, nerve bundles and loose
connective tissue. I n the cat as shown in figure 1,each cluster
is more or less surrounded by thin walled sinnsoidal blood
spaces. Reduplications of the endothelium which project into
the lumen and suggest a system of valves or baffles are conspicuous features of many of the sinusoids (fig. 2). I n the
mouse (fig. 3) the smallest blood vessels are more apt to be of
capillary size although wider spaces with no wall except
endothelium are present also. The endothelium of all vessels
appears continuous and its surface cytoplasm is rather dense.
The glomus cells are always separated from the endothelium
by a distinct space containing fine connective tissue fibers and
frequently nerves. Nerve fibers between blood spaces and the
glomus cells are much more conspicuous in the mouse than the
cat (compare figs. 1and 3).
While most of the glomus cells at first glance appear to
be ovoid o r polygonal, one does see occasional processes of
considerable length and closer inspection of any glomus cluster
shows numerous interlocking masses of cytoplasm. Many of
these are undoubtedly derived from glomus cells, others are
definitely nerve, and others cannot be identified with certainty.
This feature is illustrated in figure 4.
Contributed by Lucidol Division, Wallace and Tiernsn, Inc., 1740 Xilitary Road,
Buffalo, 5, N. Y.
Purchase made possible by a donation from the Brown-Lupton Foundation.
694
CURTIS M. GARNER AND DONALD DUNCAN
I n the early stages of this study it was believed that the
cytoplasm of the carotid body cells contained numerous minute
droplets of high electron density (see figs. 4 and 5). Since
these particles were limited to glomus cells, they were thought
t o represent the fine granules described by Hollingshead. As
further material became available it was noticed that poorly
fixed material was loaded with these bodies and what was
seemingly the best fixed material (fig.4) had few if any such
granules. Then it was found that a carotid body that had been
perfused with Xuller’s fluid prior to removal and osmic
fixation was filled with electron dense granules (figs. 5-6)
Repetition of dichromate perfusion in another animal gave the
same result; i.e., a cytoplasm loaded with granules and quite
good preservation of the mitochondria. The nodose ganglion
from the same animal was examined and although the nuclei
have the typical dichromated appearance, the cytoplasm is
better preserved than that of carotid body cells. I n nerve
cells, the cytoplasm is entirely devoid of the characteristic
granules, and this is likewise true of connective tissue cells
and endoihelial cells. Thus it appears that glonius cell cytoplasm is far more sensitive t o fixatives than that of other cells.
The most probable arrangement of the cytoplasm is in
the form of endoplasmic reticulum of a tubular variety. These
tubules are occasionally seen in parallel array as shown in
figures 1 , 2 and 8 but they are f o r the most part a tangled and
randomly arranged network. Attached to and between the
tubules are numerous minute granules identified a s RNA
protein by Palade ( ’55).
Finer parallel membranes have been seen occasionally.
These probably represent the Golgi apparatus. However, the
lack of vacuoles and the paucity of the very fine membranes
suggest that the Golgi apparatus is feebly represented in the
glomus cells.
As previously mentioned it has been difficult to identify
nerves and nerve endings in relation to the glomus cells. This
has been most true of well fixed material using Dalton’s mixture. I n this material many fine cytoplasmic processes exist
FINE STRUCTURE O F THE CAROTID BODY
695
between the bodies of glomus cells which lack either discernible
neurofilaments o r well defined synaptic vesicles. Figure 4
is an example of this.
I n Palade-fixed material, particularly if it is somewhat
spread out or swollen, nerve elements are recognizable in
abundance and are seen to bear a membrane to membrane
relationship to the cytoplasm of the glomus cell. At least
this is true in some instances, while in others there are a
series of membranes interposed between the neuroplasm and
glomus cytoplasm (figs. 8-12).
While the great majority of glomus cells are of the same
type, occasional cells are encountered with less and a more
dense cytoplasm than is ordinarily the case (figs. 13, 14, 15).
Sometimes these cells bear a special relationship to nerve
fibers as shown in figures 16 and 17. One could attach much
importance to such cells if they were more numerous and
invariably invested nerve fibers as shown in these figures.
On the other hand their scarcity and the frequent lack of
intermediate membranes between glomus cytoplasm and nerve
fibers make it possible to discount their essentiality. They
may be a phase in the liie cycle of glomus cells and the
possibility that they are Scliwann cells must be kept in mind.
DISCUSSION
Because the carotid body is exquisitely sensitive to alterations in the chemical composition of the blood, it was thought
that the electron microscope might reveal a peculiar and intimate relationship between the blood stream and the carotid
body cells. If any such special arrangements do exist they
have escaped observation in this study. Much observation and
several hundred photographs have presented a uniform picture of a continuous and relatively thick endothelium and
a definite perivascular space which varies in width and content
but seems invariably present. The only feature of the vascular
system that might be called special is the presence of numerous endothelial folds. These and the large caliber of the
smallest vessels might produce a sluggish circulation and
696
CURTIS M. GAXNER A X D DOKALD D U N C A S
thus a relatively poor oxygen supply and a slower than usual
CO, removal.
Like the vessels, the relationships between nerve and the
glomus cells are either quite simple or the more subtle features have not been seen in this study. The latter possibility
should not be discounted since there has been difficulty in
identifying nerves and nerve endings, particularly in what
appears to be the best fixed material. For identification as
nerve two criteria have been relied on, either a wispy appearance of the cytoplasm suggestive of neurofilaments, or cytoplasmic masses containing bodies of approximately the diniensions and appearance of synaptic vesicles as first described by
Dc Robertis and Bennett ( ’55). Assuming such identification
is correct, one can say that nerves and glonius cells are in
abundant and intimate contact with each other. I n many
places, especially in the mouse, small nerve fibers are packed
in almost solid array in the perivascular spaces, see figure 3.
I n addition to a perivascular position, nerve fibers may
be found between glomus cells (figs. S-9) and some are apparently surrounded by the cytoplasm of individual glomus cells
(figs. 3). Frequently, as in figure 8 and 9, a very narrow band
of cytoplasm separates a nerre from the nucleus of a glomus
cell. This finding supports the contention of DeCastro (’51)
on penetration of the carotid body cells by afferent endings in
so far as they can be seen with the light microscope. I n all
cases, however, a pair of limiting membranes separate the
cytoplasms of nerve and glomus cells. Since some of the
elements identified as nerve contain small vesicles, it should be
pointed out that such vesicles are “synaptic” in the sense of
being present in nerve endings but with reverse implications,
if any, as to direction of impulse transfer when compared t o
centrally placed boutons.
The gloinus cells are unquestionably intricately interlocked
with each other by means of processes of unknown number and
degree of elaboration. Without serial reconstruction of a
number of these cells, the best that can be said is that observation of many single sections suggest relatively few and
F I K E S T R U C T U R E O F T H E CAROTID BODY
697
relatively short processes that do not subdivide extensively.
While fine connective tissue fibers surround and to some extent permeate each cluster of cells much of the surface of any
cell appears t o be in direct plasma membrane contact with the
adjacent ones.
An interesting and unelucidated feature of the carotid
body is the question of the presence of elements in addition to
typical glomus cells, nerve fibers and connective tissue. Our
observations suggest that in addition to the typical glomus
cells, there are lesser numbers with similar nuclei but less
abundant and more dense cytoplasm. These cells are additionally characterized by long processes extending between the
typical glomus cells. I n a few instances they seem to bear a
special relationship t o the nerve endings as suggested in
figures 16 and 17. They may be a distinct type of glomus cell,
a phase in the life cycle of these cells, or conceivably Schwann
cells. Since they are not invariably interposed between the
major glomus cells and the nerve endings, they do not seem to
be constant and, therefore, probably not an essential link in
the translation of alterations in the blood to nervous energy.
The most striking observations made in this study are
those that suggest a rather specific instability of glomus cell
cytoplasm. As mentioned in the statement of results, there are
indications that the endoplasmic reticulum of these cells is
very easily converted to small spherules of high electron
density. These particles are quite uniform in size and around
0.1 micron in diameter. These droplets and mitochondria are
the only recognizable elements in the cytoplasm of obviously
poorly fixed material. They are absent or very nearly so in
preparations that show the best mitochondria1 detail, the
least clumping of nuclear content, and minimal separation of
one cell from another. These droplets are by far the most conspicuous feature of glomus cells that have been initially fixed
with 3% potassium dichromate. Such treatment also produces
a characteristic nuclear pattern, but maintains the mitocliondria to a surprising degree. The contention that there is
considerable specificity to the reaction rests on the observation
698
CURTIS M. GARNER AND DONALD DUNCAN
that it does not occur in similarily treated nerve cells, Schwann
cells o r endothelinm. This lability is the only evidence we have
of the morphology accompanying the special sensitivity of
the carotid body. The mitochondria1 preservation found here
is at variance with observations of Low and Freeman ('56),
on the use of dichromate mixtures f o r fixation. This discrepancy may be due to differences in times between contact
with dichromate and subsequent exposure to osmic acid.
Concerning possible mechanisms of carotid body function,
it can be pointed out first that in the carotid body there is far
more direct contact between unmyelinated nerve and endothelium than in most parts of the body. Because of this it may be
that the nerve fibers are directly stimulated by pH and other
changes in the blood stream. However, this seems unlikely
because there is the same, although less abundant, nerve to
vessel intimacy throughout much of the body; also, such an
explanation would assign but a passirc role to the glomus
cells. An alternative is to suppose that the glomus cells exude
some specific product in response to changes in vascular
chemistry which in turn excites the nerve endings. An attractive speculation from an anatomical point of view is t o
suspect that the interlocking glomus cells are capable of expanding o r contracting in response to minor changes in their
environment and thus stimulate the nerve endings mechanically. But it must be admitted that this initial electron microscope study furnishes little but negative evidence on the functional mechanisms of the carotid body. There seem to be no
special features beyond very close contact between the blood
stream, nerves, and cells with cytoplasm that appears to
be exceptionally sensitive t o thc action of fixatives.
SUMMARY AND CONCLUSIONS
Carotid bodies of cats and mice were prepared f o r study
in the electron microscope using buffered osmic acid fixatives
and ultrathin sections. I n both species the major features are
similar but there are sufficient differences between cat and
mouse carotid bodies to permit their distinction in the electron
microscope. I n general the cell clusters of the mouse a r e
smaller aiid large irregular endothelial-lined blood spaces a r e
less prominent fcatures than in the cat.
The distinctive cells of the carotid body a r e of two types.
Most of them are rnoderately large cells with abundant cytoplasm and a few interlocking processes. The other type is
infrequent with less and more dense cytoplasni. The occasional smaller cells liave loiig and numerous processes. Tlic
cdoplasrriic reticulum of the glomus cells is very sensitive to
fixation procedures. I n all but optimally fixed preparations it
lias disintcgratcd into small xplierulcs of high clecti*ondensity.
Nerve fibers and ncrve endings are located between the
glomus cells and blood vessels, between acljaceiit glomus cells
and sonic a r e surrounded by the cytoplasm of a single cell.
Nerve cytoplasm is separated from carotid body cell cytoplasm
by a pair of single closely opposed membranes in some
instances and a s many as four double membranes in others.
Tlic significance of thes:’ vaihtioiis is unknown.
It is concluded that specific cells of the carotid body are in
close and relatively simple contact with abundant nerve
endings, that the carotid body cells have a system of iiiterlocking cytoplasniic processes and that the endoplasmic reticulum of tliese cells is unusually sensitive to fixatives. This
organelle readily disintegrates into small granules or splierules of high density which have been described in light niicroscope studies as a characteristic feature of carotid liody cells.
LITERATURE CITEI)
DALTON,A. J. 1955 A chrome-osmium fixative f o r electron microscopy. Anat.
Rec., 131: 281.
UE CASTRO,F. 1928 Sur la structure ct l’innervatioii du sinus rarotidicii tle
l’homme e t drs mammifcrcs, etc. Trav. du Lab. de Recli. Biol. de
L ’TJniv. de Madrid, 2.5: 331-380.
1951 S u r la structure de l a synapse clans les chemorecepteurs. Acta
Pliysiol. Sc:ind., 2 2 : 14-43.
L)E I<OBERTIS, E. 11. P., A N D s. BENNETT 1955 Some features Of the submicroscopic morphology of the synapses i n f r o g and earthworm. J. Biopliys.
Cytol., 1: 47-58.
700
CURTIS M. GARNER AND DONALD DUNCAN
C. M. GARNER 7987 An clwtron niicroscopc study of the cnrotid
body. Anat. Rev., 197: 285.
1930 Sinus enrotidieii et
HEPNAXS,C., J. J. NOIUXARRTAND L, DAUTREBANDE
reflcscs rcspir:itorics. Arch. Ink Plinninincod~n. 39 :400448.
~TOLLINGHHEAD, W. If. 194.3 A cytologicnl study of tlic cnrotid body of tllc
cat. Am. J. Annt., 73: 1857211.
t ~:inoxi:t upon cttroticl body inorpliology. Anat. Rw.,
1945 P f f ~ of
X)UN(*.\N, D., AND
9.9: 255-261.
KOCK,1,. I,. I)& 1954 TIIVintr~i-Klo,i,c.riiltirtixsiics of tlw crlrotitl I)otly. Act:!
Ailtit., 32: lOl-ll(i.
TATTA, IT., AND J. F. HAHTJCANN
1950 Use of n g1n.w edge in thin sectioning for
clwtron iiiierosc!opy. Proc. 8oe. Esp. Riol. and bicvl., 7.f: 130-139.
T.OIV, F. N. 1956 Porsonnl conimunicntion.
TJOW, F. N.,A N D J . A. PHEE:MAN 1956 8oiiic expriuicwts with chorniiuni coalpounda as fiscrx for c*lcwtron inic*roxcopT. J . Riopltys. Cytol., 2 : (iQ9-(i32.
T,~T~CITKA,11. 1862 olwr die clriiaenttrtigc~Nntur dcs sogeiliinntai G:inglion
Tiit(~rciirotircitii.A r c k f. Aii:it. Phpiol., 405114.
P.\L:\PE, Ci. E. 1952
A stiidy of fistition for c k t r o n niicroxcopy. J. Exp. Mcil.,
ROSS, TI.
95: 285-298.
1955 A sinall pnrtieulnte coniponcnt of tlic cytoplnsin, J . l3iopliyx.
Cytol., 1: 59-08.
1957 An electron ntirroncopc study of cnrotid body clicInioreccptorn.
Aunt. Rcc., 127: 481.
Abbrcriationr for pldcs
TW, blood vessel
NE, nerve ending
E,i*ytoplasm iclontificvl ns n very large
M, initocliondria
MM,niultiplc nmiibraiics
uerre ending
F X , enc1ol)lnninic rcticuliini
P, process of gloniiis cell
ci, glo111us cell
N, nerve film
s c . uuc~leuu
8, sinusoid
IT, "interstitial ' ' cc4l
PLATE 1
1 A nection from 11 nonnnl cat cnrotid body sl~owingthin \r:rlld sinusoids in
reltitioii to a clump of typical gloniua cells. Xote cliaructeristic pitch of
cucloplasinic reticulum in parallel nrrny, nnd process of gloniux cell. Tliix
picture iudicnttws t.liat tlic glonius cells nre not truly oroiil or polygonal but
r:itlier bluntly stelhte, 1)alton fisntive. X 5,400.
2
The arrows indicate two of the nniiieroua mhre-like folds that project into
the luniinn of carotid body siiinsoids, 1)nlton fixative. x 3,600.
F l S E STEUCTURE O F THE C d R O T I D BODY
PLATE 1
CURTIS 31. GARNER A ND DONALD Dl J N CA N
701
PLATE 2
BSPIASATION OF FIOURES
3 Mouse carotid body to illustratc that the s~nnllestblood vessels in this forin
approach capillary size more than they do in the cat. The picture also sliowx
tho minute dense granules seen in imperfwtlr fixed cytoplasm and the
clitiracteristie til)unclaiice of nerves between the glonius cells :ind tlie I~loocl
vessels, Pnlade fixative. X 6,400.
4
Cat carotid body to drom n process o f one glomus cell and nuiiieroux interloc.kiag yrocesnes of otlier cells. One of these inclicnted by nn :irrow itlay be
Iierve. Note a niiiiiinuni of dense gr:innles ns eomp:ircd to figures 3, 5, B :incl i ,
Drilton fixative. X 24,600.
702
PLATE 2
703
5
A section of Pirl:tdc-fixed iiioiise wirotid body t o 811ow tlic iiuiiierous deiisc.
granules found i n gloiiius cell cytoplasni when fixation is good b u t not optinial.
Tlie granules are regarded iis disintegrated endoplasniic wticuluni and they
indicate a special seiisitiveiicss of glomus cell cytop1:isiii. A-crve fibers in
closc relationsliip t o glomus cell c,ytopl:tsiii :ire also s l i o \ v ~ i . X 29,600.
(i lAow power vicw of eat carotid body perfused with 3 % K,Cr,O, prior to
fixation in Dalton ’s mixture. This treatment markedly niiginc~itst l ~ ciiuliil)cr
:tiid size of cytoplasinic gr:iiii~lcs i n glomus cclls. X 3,200.
i
A b i t of glomus cell cytoplasm from n eat carotid Imly perfnsril with
3% K,Cr,07 and subsequently fixed i n Palade’s inixturc.. K o t e t h a t niitoclioiidrial cristae a r e preserved xiicl the cytoplasmic g r m u l w f r q u e n t l y consist
of :I very delis? c,ore siirroiuitled h~ n less ~lctiec~
Iinlo. x 30,000.
PL.\TE 3
705
ESI’L,\N.\‘J’ION
8
9
OF PIGIJRES
A cytoplasmic process containilig iniiiutc vcsiclcs :ind coiisrqueiitly ideiitifid
as a nerve ending. This ending is very close to the nucleus of a gloinus cell.
The glomus cell cytoplnsiii occupying most of the figure shows niitoclio~iilri;~,
eiidoplasniic r~ticiiluniencrusted with minute grmules, a number of circular
profiles tliouglit to be cross sections of eiidoplasniic tubules, clusters of niinntc
ociatrtl with cwloplasmir nienibranes, and a n oceassioiinl
dense droplet as d(wri1)cd aiid i1lustr;itrd in prwrdiiig figures, c a t carotid hotly,
Palade fisativr. X 16,000.
A relatively lnrgc, iirrre ciitliiig 1:itlcii witli rc~sicles, : i i i ( l i n ~ i i r n ~ l ) r : i ~
t oi r
c,oiitact, with glomus crll cytoplasm. X 24,000.
iii(wi1)r:inc
10
3rultiplti i i i e ~ ~ i I ) r ; ~lwtweeii
~irs
gloiiius cc,ll cytopl;isin :rnd iiervr. With grc;it(xr
~ri:~gnific;~tion
w c l i of tlie tlirer liiic~s iiit1ic:itrrl l ~ ytlir arrow is tlistiiictly
x fi:$,oon.
t~l)Il~)~(2.
1I
A portion of IICI’VC fibrr scyitxtcvl f r o m glomus crll cytop1:isni 1)y
distinct niriiibraiirs. x 25,000.
12
A l o w - p o w t ~view of iiioiise rarotitl botly sliowii~g:I wrll tlrli~ic~atetl
lil:~ss,
w l i i c ~ l i iii:iy l w :I I:irgc new(’ rii(liiig c.inl)rddrd i n glo1irus rca1I cytop1:cslii.
x
12,000.
folly
PLATE 5
EXPLAhTSTION OF FIGURXS
13
Illustrating a second cell type encountered less frequently than the typical
glomus cells, and with long narrow processes insinuated between its more
abundant neighbors. Carotid body of a cat, Palade fixative. X 6,400.
14
T o show a process of the dense cell type between two typical glonius cells.
Cat carotid body, Dalton fixativc. X 12,500.
15 Demonstrating a cytoplasmic process in cross section which appears identical
to P in figure 14, but also resembles a nerve ending. Cat carotid body,
Dalton fixative. X 12,500.
16 A low power view of a dense glomus cell with a multiple menibraned process
partially surrounding a iierve fiber. Cat carotid body, Dalton fixative. x 5,400.
17
A higher power view of p a r t of figure 16 to show four double niembraiies
derived from a dense carotid body cell between typical glomus cell cytoplasm
arid n nerve fiber. x 10,800.
‘i08
F I S E S T K T C T V R E O F TIIE CMLOTID BODY
PLATE 5
C C R T I S 11. G A l < S L R A S D D V S l L D D C N C A S
709
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 248 Кб
Теги
structure, carotid, observations, body, fine
1/--страниц
Пожаловаться на содержимое документа