вход по аккаунту


The architecture of the thyroid glandA 3-dimensional investigation.

код для вставкиСкачать
The Architecture of the Thyroid Gland:
A 3-Dimensional Investigation '
Department of Anatomy, McGill University and Department of Biological
Sciences, University of Montreal, Montreal, Canada
The three-dimensional structure of the thyroid gland was investigated
by reconstructing three-dimensional models from serial sections of the rat thyroid,
stained with the periodic acid-Schiff technique to identify the basement membrane. It
was found that the follicles were not completely separated from each other by connective
tissue, reticular fibers, or basement membranes. At some points, the base of the epithelium of a follicle was in direct contact with the base of the epithelium of a n adjacent
follicle. A n area of contact involved about a dozen follicular cells in each follicle. One
follicle usually had two or more areas of epithelial contact. All follicles examined in
three dimensions (over loo), showed epithelial contacts, in one-year-old as well as in
younger rats. The area of contact between parenchyma and stroma, as marked by the
basement membrane, consisted of a branching system of anastornosed and partially
pinched sacs, within which the epithelia of the follicles were in direct contact. Thus,
each follicle can no longer be considered as being completely surrounded by connective
tissue or even wrapped by a basement membrane discrete from those of neighbour
follicles, and therefore its individuality is essentially due to the fact that it owns a
discrete lumen.
Since Jones (1836) described the thyroid follicles as closed discrete units made
of one layer of cells surrounding a mass
of colloid, thus opposing Lalouette's (1750)
concept of the follicles as branching sacs,
a controversy developed which was not
dispelled by subsequent reconstruction
studies of various orders (Streiff, 1897;
Norris, '16, '18; Wilson, '27; Rienhoff, '29;
Moritz, '31; Hammer and Loeschcke, '34;
Ewe, '36; etc.) Heidenhain's ('21) observation that the follicles could exist side by
side without being separated by connective
tissue was confirmed by Yagizawa ('56).
However, these workers did not investigate
whether or not a basement membrane
existed between follicles which showed no
apparent separation by connective tissue.
In the present work, the three-dimension structure of the thyroid follicle was
reinvestigated on models reconstructed
from thin sections of the rat thyroid,
stained with the periodic acid-Schiff technique to identify the basement membrane.
The animals were killed with ether and the
thyroid attached to the trachea was fixed
in Bouin's fluid, embedded in paraffin, and
sectioned serially at 5 p. The sections
were routinely stained with periodic acidSchiff and hematoxylin. Some were
stained with silver for reticular fibers,
according to the method of Gordon and
Sweets ('36) using the silver diaminohydroxide solution of Wilder ('35). A few
sections were stained with both the silver
and periodic acid-Schiff techniques.
The selection of a field for three-dimensional reconstruction was guided by the
following conditions : ( a ) simplicity of
follicular structures in order to avoid bizarre shapes and overtly complex morphology which might have obscured the
basic design; (b) medium size follicles
permitting reconstruction to be completed
within 10 to 15 levels (50 to 75 p ) , the
most suitable size; ( c ) prominent landmarks such as blood vessels, allowing
accurate tracing from one section to the
next; (d) absence of artefacts.
In each section the selected field was
Female albino rats of Sprague-Dawley
photographed, with a magnification of
strain, ranging in age from 12 hours to
x500, on Panatomic 3%" x 4%" films.
12 months, were obtained from the Holtz1This work was supported by grants from the Naman Co., Madison, Wisconsin, and the tional
Cancer Institute of Canada to Dr. C. P. Leblond
Quebec Breeding Farm, St-Eustache, P. Q. and to Dr. H. Isler.
ANAT. REC., 161: 325-336.
Each resulting negative was then projected
onto a sheet of graph paper with a further
magnification of x6; the basement membrane of the follicles as well as the outline
of the intrafollicular colloid were traced
on the paper with a pencil. All tracings
were checked by microscopic examination
of the slide at the time of drawing. Proper
superimposition was obtained by projecting the negative onto the tracing of the
preceding serial section. Using landmarks such as blood vessels, colloid and
basement membrane, a “best fit” was
achieved by moving around the board
supporting the graph paper. Then without
moving the paper, another sheet was
pinned over the preceding one, and the
landmarks and structures outlined on it.
To help keeping an accurate relationship
between the layers while later assembling
them into three-dimensional models, four
arbitrary points were chosen on the first
tracing, and inverted thumbtacks were
placed through them. As each new sheet
of paper was added representing the following level, they were also pierced by the
thumbtacks, thus maintaining four constant reference points throughout the
series of tracings.
Three different materials and techniques
were used to transfer and superimpose the
two-dimensional tracing into three-dimensional models : wire, transparent plexiglass
sheets and foam plastic sheets.
W i r e models. Solid copper electric wire,
insulated with plastic coating: green for
the basement membrane and red for the
colloid outline, was used. The first models
were made of 18 gauge wire with an outside diameter of the coating of 1.5 mm.
As these showed a tendency to collapse,
more rigid models were subsequently made
of 12 gauge wire with TW plastic coating
and outside diameter of 4 mm. Wire replicas were superimposed and held in proper
relation by a system of metal supports and
soldering. Through the four thumbtacks
reference points mentioned above, vertical
support wires were run, to which the wire
outlines for each level were attached by
horizontal connections at i 5 mm intervals.
Four wire models were thus constructed :
two from three-month-old and two from
12-month-old rats, each 15 level-high.
These wire models proved to be very adequate for an accurate analysis of the threedimensional structure of the follicles, and
in fact they yielded most of the information
related in this paper. However, they did
not show the structural relationship at one
glance, such as is required from a good
demonstration model. Hence the other
types of models.
Transparent plexiglass model. One
such model was built to illustrate the spatial configuration of the area of two follicles in direct contact. Plexiglass plate 35
cm x 47.5 cm x 5 mm were used. When
received from the dealer, these plates were
protected with a sheet of adhesive paper
on each side. On this paper, the basement
membrane and colloid outlines were drawn.
Three plates were used for each histological section. The difference in outline from
one section to the next was made in three
equal steps, one step per plate. When the
basement membrane was cut obliquely, a
better appearance of continuity was obtained by giving the lines a certain thickness-for each plate one third of the difference in outline between two histological
sections. Then the paper covering the colloid area was cut out with a razor blade and
the exposed part sprayed with red fluorescent paint. After this area had been covered, the surface of the basement membrane was exposed and sprayed with green
fluorescent paint. Only one side of the
plates was painted. The adhesive papers
were then removed completely on both
sides and the plates assembled in a metal
frame. The model was illuminated from
the bottom and four vertical sides by means
of fluorescent lights, while it was observed
from the top.
Foam plastic model. This model was
designed to demonstrate the spatial configuration of a group of several follicles
attached to each other by direct contacts.
The outline of the basement membrane,
drawn originally on graph paper, was
transposed to the surface of gray, opaque,
1.5 cm thick foam plastic sheets. The
sheets were sewn along this outline and
glued on top of each other.
Examination of periodic acid-SchifE and
hematoxylin stained sections of the thyroid
gland showed two lobes surrounded by a
capsule of connective tissue incompletely
divided into lobules by connective tissue
septa. The septa contained blood vessels,
lymphatic and nervous elements in addition to connective tissue cells and fibres.
These lobules were irregular in size and
shape, and contained variously shaped but
mostly oval follicles. The follicles were
coated by a purple basement membrane,
but were not always completely surrounded
by it, as epithelia of adjacent follicles were
seen in close contact without apparent intervening basement membrane (fig. 1 ). In
these cases, the membrane surrounded
two or a group of follicle sections. The
three-dimensional models showed that in
most cases these were actually distinct follicles with discrete colloid masses. These
masses were oval or pear-shaped, that is
with a blunt and a pointed end. Only unfrequently did the colloid branch into short
Solid masses of epithelial cells devoid of
colloid-filled lumen were also observed in
single sections; they were surrounded by a
basement membrane (fig. 2) or enclosed
within the same basement membrane as a
neighbour follicle (fig. 3). Most of these
masses consisted of follicular cells from
follicles cut tangentially. Each of these
cells was in contact with the colloid as
shown by the models. Some, however,
proved to be made of light cells (figs. 3 and
4), an epithelial cell type different from
the follicular cells (Stux et al., '61) and,
as the models showed, not in contact with
the colloid, but always enclosed within the
basement membrane of follicles. In adult
rats light cell buds of considerable size were
occasionally encountered (fig. 5).
The close contacts between two neighbour follicles were investigated in some detail. At these locations, the periodic acidSchiff positive basement membrane was
absent (figs. 4 and 6). The basement membranes of two follicles occasionally merged
and penetrated to some distance between
the follicles, appearing as a single line, as
if there were only one membrane, common
to the two follicles, which became gradually
finer and vanished. More often, the basement membranes joined around a capillary common to the two follicles, thus outlining an 8-shaped epithelial structure
(figs. 4 and 6). No silver reticulin-stained
structures were present in the area of contact (fig. 7). Incidentally, while both the
periodic acid-Schiff technique and the silver reticulum stain show the outside surface of follicles and capillaries, the two procedures may not stain the very same structures. Indeed, the space between a capillary and the follicular cells showed as a
thin, pale line with the periodic acid-Schiff
technique (figs. 6 and 9), and a thick,
dense line with the reticulin stain (fig. 7).
When both procedures were used on the
same section (fig. 8 ) , the space appeared
as a three-layered structure: two thin lines,
presumably the basement membranes of
epithelium and endothelium, separated by
a space which was occasionally stained
with silver. The wire and plexiglass models
revealed the area of contact as an oval
shaped surface involving about a dozen
cells in each follicle (fig. 11). Such direct
contacts could readily be found in single
thyroid sections of new born rats, as 92%
of the follicles showed one or more contacts. As the animals grew older and the
diameter of the follicles increased, this
proportion decreased (to 60% in onemonth old rats). Large groups of follicles
were frequently observed in contact with
each other (figs. 7-10). As many as nine
follicles were observed within the same
basement membrane in a single section
(fig. 10). The foam plastic model illustrates a group of eight follicles attached
to each other (fig. 12). This group was
actually only a part of a larger one, since
contacts were present at its periphery.
An attempt was made to determine the
size of such groups of follicles entirely surrounded by a basement membrane. Without actually constructing models, groups
were followed in serial sections, but these
groups seemed to have no end, as almost
at each level new contacts were appearing.
At the periphery of the thyroid lobe, particularly in the vicinity of the trachea,
single follicles were frequently observed,
N hich, in one section, were separated from
the other follicles by a thick layer of connective tissue. Over a hundred of such
follicles were followed in serial sections of
one-month old rats' thyroids. All of them,
without a single exception, proved to be
in contact with other follicles at one level
or the other. Neither was any discrete Our own observation on sections stained
follicle found in older rats up to one year successively with both procedures is in
of age. Each follicle was in epithelial con- keeping with this view. Anyway, the abtact with one, two or more neighbour fol- sence of basement membrane and reticulin
licles. It was wrapped by a basement indicate that the base of the epithelium of
membrane which was not discrete from one follicle is, in some points, in direct
those of other follicles, but which rather cellular contact with the base of the epiconsisted of a system of interconnecting thelia of neighbour follicles. A summary
and branching constricted tubes, each con- of this observation was presented as a demstriction only partially separating two fol- onstration at the VIIth International Congress of Anatomists in 1960 (Isler et al.).
licles from each other.
At this stage, the question arises whether or
not a very thin basement membrane, or
Although the first description of the scant connective tissue elements too thin
microscopic structure of the thyroid gland or scarce to be detected with the periodic
appeared over 200 years ago (Lalouette, acid-Schiff technique or the silver reticulin
1750), a number of the features of this stain, may still be present at the location of
gland are still controversial. Lalouette's these seemingly direct contacts. This point
concept of the thyroid parenchyma as be- has been clarified by a subsequent electron
ing formed of branching sacs had been microscopic investigation (Messier and
rightly refuted by Jones in 1836. Indeed, Isler, unpublished) showing that there are
all our three dimensional models show that areas indeed, where two follicles are in
the lumina and colloid masses of the fol- contact without intervening basement
licles branch only unfrequently, so that membrane.
almost all of them are discrete and entirely
Thus the concept of a discrete thyroid
surrounded by a layer of epithelial cells. follicle completely separated from neighHowever, Jones went further and described bour follicles by extracellular material, as
not only the lumen, but the whole follicle advanced by Jones, and still retained by
as discrete. The follicle was then visual- most modern text-books, should be abanized as a shell of cells surrounding a mass doned for the more accurate one of chains
of colloid, which was itself surrounded by of follicles incompletely surrounded by conconnective tissue. Later, with the advent nective tissue and basement membrane.
of the periodic-acid-Schiff method, the Although each follicle is attached to one or
basement membrane, which this technique several neighbour follicles, it still retains
shows off, was considered as part of the its individuality: i t owns a shell of cells and
follicle and was believed to completely a lumen which are not shared with other
wrap each follicle (Leblond, personal follicles. The epithelial contacts merely incommunication). Neither this, nor Jones' dicate a very close association between folconcept is accurate. Heidenhain ('21) ob- licles. The resulting groups of follicles are
served that the epithelia of some follicles certainly quite large, as their limit has
were only incompletely separated by con- never been reached in our investigation. It
nective tissue. Yagizawa ('56) confirmed is likely that the whole thyroid consists of
this observation and furthermore found one single group of associated follicles.
that even the silver reticulin positive mateWhile our observations are limited to the
rial did not completely surround all folli- rat, Yagizawa ('56) reported an incomplete
cles. The observations of these two authors wrapping of follicles by connective tissue
have been confirmed in the present work, in several other species (rabbit, guinea pig,
which, in addition, shows that not only the hamster, cat, dog, pig, horse, cattle, monconnective tissue and reticulin, but also the key). The frequency of direct epithelial
extra cellular basement .membrane may be contacts varied with the species, which may
absent between two follicles. The base- have been due to differences in follicle size,
ment membrane, as visualized by the perio- as the probability of finding direct contacts
dic acid-Schiff technique, is probably differ- in single sections decreases as the diameter
ent from the material stained by the silver of the follicle increases. Heidenhain ('21)
reticulin method, as shown by Lillie ('51). had also investigated several species and
had found direct follicular contacts in species other than rats. While his observation
concerning differences between species are
not in complete agreement with Yagizawa’s
( ’ 5 6 ) , it is probable that the few discrepancies existing between them are more of a
quantitative than of a qualitative nature,
Heidenhain finding no contact when Yagizawa found only few. The difference in
the staining techniques used may well account for these differences. Thus, the concept of incomplete splitting of the parenchyma into independent follicles may well
be valid for most, if not all, mammals. In
fact such an arrangement is not surprising,
considering the process by which the histogenesis of the thyroid gland is accomplished. At first, the gland consists of continuous epithelial trabeculae embedded in
connective tissue (Nelsen, ’53). Later, colloid appears in intercellular spaces, within
the trabeculae. These nascent follicles
then increase in size, giving the trabeculae
a pinched-like appearance. The present
work shows that they never split apart from
the trabeculae, but that they remain attached to each other, thus retaining the
general trabecular architecture. Only the
nature of the trabeculae has changed; instead of solid cords of epithelial cells, they
are made, in the post-natal life, of a succession of colloid-filled vesicles.
Another controversial topic is that of
“follicular budding” (Wilson, ’27; Bargmann, ’39), which derives from the occurrence of several layers of epithelial cells on
one side of a follicle. The present work
demonstrates that most of these so-called
“buddings” observed in single sections are
due to tangential sectioning of follicles,
some of them irregularly shaped, since the
cells, as visualized in three dimensions, actually border the lumen of the follicles. Occasionally, however, cells or groups of cells
are present which, when followed in serial
sections, prove to be separated from the
lumen by a layer of follicular cells or of
follicular cytoplasm, while still contained
within the basement membrane of a follicle. These cells are not identical to the
cells bordering the lumen, as their cytoplasm and nuclei are larger and paler, and
do not contain colloid droplets after injection of thyroid stimulating hormone. These
properties identify them as light cells (Stux
et al., ’61). Thus, while there are no true
follicular cell buds, groups of light cells are
present which give some support to the
concept of budding, especially since there
is evidence that these iight cells originate
from the follicular cells (Sarkar and Isler,
’63). However, there is no clearcut evidence indicating that these groups of light
cells participate in the process of enlargement of an existing follicle, or in the
growth of a new follicle, as believed by
Ponse (’38).
The help of Jean Garneau in making the
photographs is gratefully acknowledged.
Bargman, W. 1939 Schilddriise. In: von Mollendorf. Part 2. Handb. d. mikr. Anat. d. Menschen, 6: 1.
Ewe, H. 1936 Ueber die Pathogenese der Proliferationsknospen in der Schilddriise. Beitr. z.
Path. Anat., 97: 195-204.
Gordon, H., and H. H. Sweets, Jr. 1936 A simple
method for silver impregnation of reticulum.
Am. J. Path., 12: 545-551.
Hammer, E., and H. Loeschcke 1934 Der
feinere Bau der Schilddruse und die sich aus
ihm ergebenden Vorstellungen iiber das Wesen
der sogenannten Proliferationsknospen. Verhandl. d. Deutschen Path. Ges. Rostock., 27:
Heidenhain, H. 1921 Ueber veschiedene Typen
im Bau der Schilddriise. Verhandl. d. Anat.
Ges., 141-151.
Isler, H., B. Thompson, R. Tonkin and S. K. Sarkar
1960 Three dimensional structure of the thyroid gland, with reference to the basement
membrane of the follicles and to the location of
the “light cells”. Anat. Rec., 136: 338.
Jones, C. H. 1836 In: Todd. Cyclopaedia of
Anatomy and Physiology, 4: 1102-1118.
Lalouette, P. 1750 Recherches anatomiques sur
la glande thyro’ide. I n : The thyroid gland in
medical history. By A. H. Iason. Forben Press,
New York, 1946.
Lillie, R. D. 1951 The allochrome procedure. A
differential method segregating the connective
tissues collagen, reticulum and basement membranes into two groups. Am. J. Clin. Path., 21:
Moritz, A. R. 1931 Interacinar epithelium of
the thyroid gland. Am. J. Path., 7 : 37-46.
Nelsen, 0. E. 1953 Comparative Embryology of
the Vertebrates.
The Blakiston Division,
McGraw Hill Book Co. Inc.
Norris, E. H. 1916 The morphogenesis of the
follicles in the human thyroid gland. Am. J.
Anat., 20: 411-48.
1918 The early morphogenesis of the
human thyroid gland. Am. J. Anat., 24: 443465.
Ponse, K. 1938 Histophysiologie de l’activation
thyroidienne. Rev. Suisse de Zool., 45: 4 4 1 4 4 9 .
Rienhoff, W. F. 1929 Gross and microscopic
structure of the thyroid gland in man. Arch.
Surg., 19: 986-1036.
Sarkar, S. K., and H. Isler 1953 Origin of the
“light cells” of the thyroid gland. Endocrinol.,
73: 199-204.
Streiff, J. J. 1897 Ueber die Form der Schilddruse Follikel des Menschen. Arch. f. milmsk.
Anat., 48: 579-586.
Stux, M., B. Thompson, H. Isler and C. P. Leblond
1961 The light cells of the thyroid gland in the
rat. Endocrinol.y 68: 292-308.
Wilder, H. c. 1935 A n improved technique for
silver impregnation of reticular fibers. Am. J.
Wilson, G. E. 817-819*
1927 The thyroid follicle in man.
Its normal and pathological configuration. Anat.
Rec., 37: 31-61.
Yagizawa, T. 1956 The follicular pattern in the
thyroid gland of maturing and mature mammals. Okajima Fol. Anat. Jap., 29: 93-115.
basement membrane
direct cellular contact between follicles.
follicular cell
light cell
solid mass of cells
thin line i n three-layered structure
1, 2, 3 . . etc., numbering of follicles in contact
with each other.
Sections of rat thyroids stained with the periodic acid-Schiff technique and hematoxylin.
Several thyroid follicles are not completely surrounded by the basement membrane
( b ) but show direct epithelial contacts ( c ) with adjacent follicles. x 385.
Solid mass of epithelial cells ( s ) surrounded by a basement membrane ( b ) and
capillaries (ca). Contiguous serial sections showed that these cells were lining a
follicular lumen and were a mere tangential section of a follicle. x 725.
Solid mass of epithelial cells ( s ) enclosed within the basement membrane ( b ) of a
follicle ( F ) . Contiguous serial sections showed that these are follicular cells lining
the lumen of follicle F. Other cells however are present ( L ) , which are not in contact with the colloid, but are budding outwards, while contained within the follicular
basement membrane (b). These are the “light cells” (L). x 725.
Area of direct epithelial contact ( c ) between two follicles. Instead of penetrating
into this area, the basement membranes ( b ) of the two follicles follow the outline of
capillaries ( c a ) present at this level and join there to become a continuous membrane from one follicle to the adjacent one. Light cells ( L ) are bulging on the surface of follicles, within the basement membrane ( b ) of the follicle. These cells are
usually not seen in the contact areas. x 620.
Large light cell bud. Note the difference between the nuclei of light cells ( L ) and of
follicular cells (FC), and the absence of basement membrane ( b ) between the latter
and the bud. x 725.
Two follicles ( F ) in direct epithelial contact. The same basement membrane ( b )
which surrounds them both follows the outline of the capillaries ( c a ) limiting the
contact area ( c ) . x 725.
H. Isler, S. K. Sarkar, B. Thompson and R. Tonkin
7 Section stained with the silver reticulin stain. The layer of reticulin
( I ) does not completely separate the follicles from each other; it is
lacking at the areas of direct contact ( c ) between follicles. ca = capillary. x 860.
Section consecutively stained with the periodic acid-Schiff and the
silver reticulin techniques. Where two follicles are in direct contact
( c ) no basement membrane or reticulin is present. Where capillaries
abut to follicles a three-layered structure is present: two thin lines ( t )
separated by a space occasionally filled by the silver stain. x 860.
Periodic acid-Schiff and hematoxylin stained section showing five
(numbered) follicles arranged in a row, each in contact with the following and preceding one. Note the absence of basement membrane
in the areas of contact. x 860.
10 Section stained with the periodic acid-Schiff and silver reticulin techniques showing nine (numbered) follicles attached to one another.
Note the absence of basement membrane in the areas of contact.
x 460.
H. Isler, S. K. Sarkar, B. Thompson and R. Tonkin
11 Two thyroid follicles reconstructed into a three-dimensional model
made of superimposed transparent plexiglass sheets onto which the
luminar colloid (col) and basement membrane ( b ) from serial sections
have been painted. The cells (not shown) occupy the space between
these two structures. The clear “window” in the center indicates the
area of direct epithelial contact ( c ) of the two follicles. In this area,
which involves a dozen cells in each follicle, the basement membrane
is absent.
12 Three-dimensional model made of superimposed opaque foam plastic
sheets cut along the basement membrane outlining a group of eight
(numbered) follicles attached to one another.
H. Isle+, S . K. Sarkar, B. Thompson and R. Tonkin
Без категории
Размер файла
938 Кб
architecture, investigation, dimensions, gland, thyroid
Пожаловаться на содержимое документа