вход по аккаунту


The fine structure of brown adipose tissue in the rat with observations on the cytological changes following starvation and adrenalectomy.

код для вставкиСкачать
Department of Anatomy, Cambridge University, England
Browii adipose tissue occurs at a number of sites in many
mammalian species, but is best known as the lobulated, brown
coloured fatty tissue situated subcutaneously between the
scapulas of hibernating and some non-hibernating (mouse, rat)
rodents. Cramer ('20) used the term "lipoid gland" in suggesting an endocrine function f o r this interscapular body; a
concept recently endorsed by Wells ( '40). However the main
body of opinion is that brown fat is merely a form of "adipose
tissue"; but there is disagreement as to whether it is histogenetically distinct (Cramer, '20 ;Rasmussen, 'as),or whether
there are intergrades and transformations between brown and
white fat (Hammar, 1895; Sheldon, '24 ; Sidman, '56a).
The work of Wertheimer and Shapiro ('48), Fawcett ('47
and '49) and Renold, Marble and Fawcett ('50) suggests the
active synthesis of glycogen and lipid within fat cells. It is
also widely believed that the metabolism of adipose tissue is
subject to hormonal control. Thus lipid depletion from adipose
tissues occurs after adrenalectomy (Wertheimer and Shapiro,
'48) and following hypophysectomy (Fawcett and Chester
Jones, '49). Further, a number of investigators have observed
the appearance of glycogen and a lipid increase in brown adipose tissue following insulin administration and Sidman
JULY 1957
128, NO. 3
( ’56b) believes this to be a direct effect since he has observed
these changes in vitro.
Rasmussen ( ’23) observed “ secretion-like ” fuchsinophile
granules ranged around the f a t vacuoles in brown adipose tissue, and added that “it is altogether probable that they represent different stages in the transformation of some lipoidcontaining product’’ Sheldon ( ’24) believed that brown fat
mitochondria gradually enlarged and were transformed into
lipid droplets.
Recent electron microscopic studies indicating that the
mitochondria are very likely the sites of appearance of lipid
in both the adrenal cortex (Lever, ’55 and ’56a) and the corpus
luteum (Lever, ’56b) have prompted the present investigation.
This aims to correlate certain light and electron microscopic
observations on brown adipose tissue.
At each biopsy several small pieces of brown fat were removed to provide f o r separate investigations :
Electron microscopy. Fixation (40-60 mins.) in Dalton’s
(’55) dichromate-osmic solution (pH 7.2) was followed by a
5 min. wash in distilled water, rapid ethyl alcohol dehydration
(three 5 min. changes of first 95% and then 100% alcohols).
After 30 mins. in equal parts of monomer and absolute alcohol,
the tissue was treated in three 30 min. changes of monomer
alone ( 7 parts of butyl methacrylate :1part of methyl methacrylate) ; and finally in a 0.2% solution of benzoyl peroxide in
monomer. Some of this catalyst-containing solution was partially polymerised to a syrupy consistency by heat (60°C.) f o r
30 mins. The tissue was then transferred to gelatin capsules
containing the pre-polymerised methacrylate and final hardening was effected in a 45°C. oven overnight. Sections in the
order of A150-200 were cut with glass knives on a PorterBlum microtome and viewed with a Siemens microscope at
80 K.V.
Light microscopy. ( a ) Fixation (60 mins.) in Zenker’s fluid
was followed by dehydration and paraffin embedding: 6 I.I sections were stained by haematoxylin and eosin. (b) Fixation in
Baker calcium formol solution was followed by gelatin embedding: 15 p sections were stained by Baker’s acid haematein
method for phospholipids and (c) by Sudan black and Scharlach R. ( d ) After fixation in Bensley’s fluid, tissue was
paraffin-embedded, sectioned at 6 p and stained by the methyl
green acid-fuchsin method to demonstrate mitochondria.
I n the following two experiments ether anaesthesia was
used and aseptic surgical techniques employed on white
Wistar rats : (1)12 animals were bilaterally adrenalectomised
and a control biopsy of interscapular brown fat was simultaneously taken. They were maintained on 0.9% NaCl in the
drinking water and received a full diet. Biopsies of brown
fat were taken from each rat at 7, 14, 21 and 28 days after
adrenalectomy. ( 2 ) I n a second series control biopsies of
interscapular brown fat were taken from 12 rats. The animals
then had free access to water but were given no food for 7
days when further f a t biopsies were taken.
Cytological observations 092 normal brown f a t by
light rnicroscopy
The lipid within brown fat cells is typically multilocular in
distribution (Hammar, 1895)’ and in any cell the droplet
diameter ranges from 15-$ p : this can be directly observed in
Sudan or osmium-stained material (fig. 1) but only inferred
in haematoxylin and eosin (H. and E.) preparations from
which the lipids are removed during dehydration and clearing. As a result of this leaching out process sites previously
occupied by lipid droplets appear as cytoplasmic vacuoles or
fenestra. I n addition to a number of normal sized drops some
cells contain single large (15 p ) drops (fig. 12). The smallest
sudanophil droplets or, in H. and E. preparations, the smallest
fenestra have a diameter range of 4-1 p and are disposed be-
tween and around the larger lipid droplets (fig.l), often in a
rosette fashion. They (the smallest lipid droplets) are comparable in size and position to the fuchsinophile bodies visualised with the acid fuchsin methyl green stain (fig. 2) and to
the phospholipid-positive bodies depicted by the Baker acid
haematein method (fig. 4).
As noted by Fawcett ('52) the ordinary f a t droplets contain
only faint traces of phospholipid, all of which, according to
this investigator, is located between the droplets within the
mitochondria. Certainly there are numbers of acid haemateinpositive granules present but larger irregular areas of staining
are also demonstrated between the unstained neutral fat
droplets: these areas may in fact be granule aggregates (fig.
4). The acid fuchsin methyl green stain provides an accredited
demonstration of mitochondria (fig. 2) and these appear either
as homogeneously stained bodies or they may have lucid
centres : in diameter they range from 8 to 1p.
Because of the higher resolutions and magnifications possible with the electron microscope much further information
can be obtained by this means.
Electron microscopic appearances in mormal brown f a t
As the emphasis in this paper is on the lipid bodies and
mitochondria features other than these are only briefly discussed in the ensuing paragraphs.
The brown fat cell plasma membrane is relatively simple
apart from occasional short invaginations into the cell. It
appears to be covered throughout its extent by a basement
membrane (fig. 13): this arrangement differs from that observed in certain endocrine tissues (Lever, '56c) in which a
basement membrane lines each side of the interval between
capillary endothelium and parenchymal plasma membrane but
does not extend between parenchymal cells.
The endoplasmic reticulum of brown fat cells is inconspicuous and consists of tubular and saccular elements. A granular
component (150 d) occurs both separately in the cytoplasm
and ranged along the outer walls of the tubular and saccular
elements (figs. 6 and 18). The arrangement thus conforms to
the general descriptions of Porter ('53) and Palade and Porter ('54). The Golgi apparatus has not been demonstrated
convincingly in this study of brown adipose tissue.
I n general there are no special nuclear features : the nucleoplasm has a granular consistency and the nuclear membrane
is bilaminar.
Mitochoizdria a d lipid droplets
From a large number of electron micrographs of known
magnification an average cross sectional diameter (the longest
dimension of each mitochondrial outline was taken) of brown
fat mitochondria was calculated to be 0.85 p. This figure is of
no absolute significance because of mitochondrial pleomorphism but it is an index of mitochondrial size. Although they
are found anywhere in the cytoplasm, mitochondria are commonly grouped in a rosette fashion around lipid droplets.
The mitochondrial enclosing membrane is bilaminar in
places while elsewhere it is resolved as a single line (figs. 11
and 17). At points of contact with lipid droplets mitochondrial
limiting membranes may be completely deficient (figs. 8 and
10). I n general the cristae rnitocholzdrales are double membranes disposed for the most part in a parallel fashion (figs. 6,
7,8 and 9). Although in many instances they appear to extend
completely across the mitochondrion yet shorter cristae are
not uncommon (fig. 6). The intercristal distance shows some
variation within individual mitochondria but more variation,
over a range of 250-850 A, from mitochondrion t o mitochondrion. The thickness of the cristae (double membrane plus
contained space) shows similar individual and general variation between 150-350 d. The broadest cristae are usually the
most widely separated and frequently have a beaded appearance (figs. 7, 8 and 11). Between the cristae is a granular
semiopaque material and in some instances, presumably if the
plane of section is intercristal, this occupies all, or most, of
the mitochondrial outline (fig. 6).
I n a number of mitochondrial outlines the internal membranes are vesicular in arrangement but as in the parathyroid
glands (Lever, ’57) it can often be shown that a mitochondrion may contain internal cristae at one end and vesicles
at the other (figs. 6 and 7). I n addition to obvious lipid droplets and mitochondria there are other bodies of an intermediate appearance (fig. 7 ) : these are frequently to be found in
proximity to the lipid droplets and they contain intensely
osmiophile vesicular components. There is evidence of an enclosing membrane (albeit incomplete) to some of these bodies
(fig. 7) which in size and shape resemble mitochondria. As
already noted many of the mitochondria are freely open t o the
lipid droplets (figs. 8 and 10) since the latter do not appear to
have a limiting membrane and mitochondrial enclosing membranes are often deficient at points of contact with lipid droplets. Some of the mitochondria communicating with lipid
droplets in this way appear to contain a material identical in
electron density with the lipid (figs. 7 and 8). Fat droplets lie
freely in the cytoplasm and permeate irregularly between the
other constituents: in figure 6 several small drops are seen t o
be interconnected. Although the lipid itself is homogeneous in
electron micrographs yet some of the droplets may contain
vesicular elements (figs. 7, 9 and 17) which resemble those
within the “intermediate” bodies described above (fig. 7 ) .
Some e f e c t s of adrenalectonay and starvation on
the cytology of brown fat
The changes in brown fat evident by light microscopy with
the techniques already detailed are essentially similar following adrenalectomy and during starvation, but they are more
rapid in onset and more pronounced in the latter circumstance.
The following description applies to tissue examined after
7 days of starvation and 21 days (and subsequently) after
As judged from the cell-field as a whole in Sudan-stained
preparations, there is a considerable reduction in total lipids,
but this is an uneven reduction and some cells are less affected
than others (fig. 3). I n general there is a diminution in drop
size and an increase in the number of droplets (fig. 3). In acid
haematein preparations (fig. 5 ) the cytoplasm of those cells
with few or no lipid drops contains large numbers of small
(+ v) phospholipid-positive granules throughout their cytoplasm :while in those cells containing obvious lipid, these same
granules are densely grouped between the drops. The fuchsinophile bodies demonstrated in the acid-fuchsin methyl green
preparations are of similar size (4-1 1-1) and identical in distribution to the phospholipid-positive granules. These fuchsinophile (and phospholipid-positive) bodies are almost certainly
mitochondria and their number would appear to be increased
during starvation and after adrenalectomy (see below).
By direct measurement of large numbers of cell “diameters” (maximum cross sectional dimensions), in control and
experimental series “average radius” measurements, and
hence the hypothetical average volumes can be estimated (assuming the cells to be spherical). I n this manner it was estimated that 28 days after adrenalectomy a 5 0 4 0 % reduction
in cell volume had occurrcd. Sheldon (’24) estimated a 25%
volume reduction after 10 days starvation.
Electron rnicroscopic appearames
Seven days after adrenalectomy brown adipose tissue shows
no obvious electronmicrographic change except for an overall
reduction in cell fat. Some of the fat droplets lack the homogeneous lipid constituent while thcir contained osmiophilic
vesicular element (described in an earlier section) are much
in evidence (fig. 17).
Brown fat 31 days (and later) after adrenalectomy has a
comparable cytology to that of rats starved for 7 days. I n both
conditions there is marked overall lipid reduction, a diminution
in lipid droplet size and, also, an increased number of mitochondria per unit area of cytoplasm (fig. 16) : these effects are,
however, more pronounced after starvation. From direct
measurements on electron micrographs of brown fat at the
same magnification it was found that the number of mitochondria per unit area of cytoplasm after 7 days starvation was
just over double the normal control figure (figs. 15 and 16).
I n some animals after 7 days starvation, but particularly in
one rat starved for 10 days, many of the brown fat mitochondria had a greatly disorganised internal structure. There
cristae were irregularly disposed often with very wide intercristal distances (fig.14).
Cell bodies other than the lipid droplets and mitochondria
showed no obvious changes throughout these experiments.
One of the aims in this paper is to illustrate the value of the
combined use of light and electron microscopy. Since errors
of sampling are very liable to occur in an average electron
microscopic study it is difficult to make quantitative estimations of a whole tissue by this means. Within individual cells
however, accurate measurements of organelles can readily be
made from electron micrographs. It is of considerable significance that the size range as measured by the light microscope
of i-1 p for the phospholipid and fuchsinophile granules and
the smallest sudanophile bodies, should be comparable to mitochondrial size (0.865 p d ) as estimated from electron micrographs.
The occasional observation that the internal membranes
may be arranged both as cristae and vesicles within the same
mitochondrion, is supported by a similar finding in parathyroid mitochondria (Lever, ' 5 7 ) . Conceivably the beading
(figs, 7, 8 and 11) of mitochondria1 cristae, frequently
observed in brown fat, may precede their fragmentation into
filaments or vesicles. In this work on brown f a t intensely osmiophile vesicular bodies can often be observed within the
lipid droplets of brown f a t (figs. 9 and 17) and also within
bodies of an appearance intermediate between mitochondria
and lipid droplets (fig. 7). It has been suggested that lipid
appears within the mitochondria of the adrenal cortex (Lever,
’55), the corpus luteum (Lever, ’56b) and possibly the parathyroid (Lever, ’57). From the evidence of the present investigation it is very likely that mitochondria1 internal membranes
in the brown fat cell become modified from a predominantly
cristal to a vesicular form, with a concurrent appearance of
lipid within the mitochondria,
As stated earlier the lipid droplets within brown adipose
tissue lie freely within the cytoplasm. Therefore if the postulate that lipid appears within the mitochondria is correct then
these bodies either discharge it freely into the cytoplasm
through breaches in their enclosing membranes or these membranes disintegrate around a contained droplet. It can be
appreciated from figure 6 how droplets in considerable numbers might coalesce into larger drops. Such a physical process
could account for the appearance of the numerous large drops
in brown fat cells when a starved animal is re-fed (Fawcett,
’52) and could readily explain the presence within some normal brown fat cells of occasional very large drops (fig. 12).
The observed increase in the number of mitochondria during starvation and following adrenalectomy must be viewed in
the light of a concurrent cell shrinkage. I n ordinarv white
adipose cells the cytoplasm and nucleus are displaced peripherally by the mass of fat and the crescentic shape of the nucleus
suggests some degree of compression. I n considering brown
adipose tissue, if it be assumed that because of the presence
of intracellular fat droplets, the other cell constituents are
displaced or “compressed,” then if most of this f a t material is
released from the cell, as occurs for example during starvation, the other cell constituents will be less “compressed.”
This explanation, does not take into account the elasticity of
the plasma membrane nor does it allow for any other change
of state within the cell which might accompany lipid release.
Arguing thus and making these assumptions it would be expected that if the cytoplasm is less “compressed” in the brown
f a t cell of starvation, then the population density of mito-
chondria would decline. I n fact there are many more mitochondria per unit area than in the normal controls and from
what has been said it is reasonable to postulate an actual
increased production of these bodies during starvation and
following adrenalectomy.
From a morphological standpoint it is difficult to classify
brown fat as an endocrine tissue. Thus it does not appear t o
contain a regzclur system of intercellular spaces communicating with each other and with the subendothelial space as is
the case in typical endocrine tissues (Lever, ,564. Intervals
between brown fat parenchymal cells are not easily visualised
but are apparently lined on each side by a basement membrane
(fig. 13) : in fact it would seem that brown f a t cells are completely invested by such a membrane, while endocrine cells are
covered only on that aspect exposed to a subendothelial space.
From the evidence presented there is no qualitative difference between the reactions of brown f a t to starvation and to
adrenalectomy. Further work is indicated to investigate hormonal influences on brown adipose tissue.
1. Brown adipose tissue in the normal, starved, and adrenalectomised rat was studied by both light and electron microscopy. Preparations variously stained to show mitochondria
neutral fat and phospholipids were examined by the light
2. As judged by both light and electron microscopy the
mitochondria lie between the fat droplets and range from 0.5
to 1p in cross sectional diameter.
3. The majority of mitochondria contain bilaminar internal
cristae with an intercristal matrix substance. Particularly in
mitochondria with wide intercristal distances the cristae themselves are thicker and tend to be beaded. Some mitochondria
contain vesicular internal membranes in addition to cristae.
4. Obvious mitochondria may contain small quantities of an
intensely osmiophile material : and bodies intermediate in ap-
pearance between lipid droplets and mitochondria are observed.
5. Mitochondria1 limiting membranes are often deficient,
particularly at points of contact between mitochondria and
lipid droplets, which lie freely within the cytoplasm.
6 . Following starvation and adrenalectomy there is a reduction in total lipid and droplet size : and the number of mitochondria is markedly increased. All the above changes are
more pronounced and of more rapid onset in starvation, and
internal disorganisation of mitochondria can occur if this is
I am greatly indebted to Professor J. D. Boyd for his interest and criticism. The electron microscopy was performed
at the Cavendish Laboratory, Cambridge, and I am grateful
to Dr. V. E. Cosslett for the afforded facilities and to Mr. R.
Home for technical advice. I also wish to thank Mr. J. F.
Crane for the light micrographs.
W. 1920 On glandular adipose tissue and its relation-to other endocrine
organs and to the vitamine problem. Brit. J. exp. Path., 1: 184-195.
DALTON,A. J. 1955 A chrome-osmium fixative for electron microscopy. Anat.
Rec., I 2 1 : 281 (abstract),
D. W. 1947 Differences in physiological activity in brown and white
fat as revealed by histochemical reactions. Science, lob: 123.
1949 Histochemical and experimental studies on brown and white
adipose tissue. Anat. Rec., 103’:450 (proc.).
1952 A comparison of the histological organisation and cytochemical
reactions of brown and white adipose tissues. J. Morph., 90.- 363406.
I. CHESTER 1949 The effects of hypophysectomy,
adrenalectomy and of thiouracil feeding on the cytology of brown
adipose tissue. Endocrinology, 45 :609-691.
J. A. 1895 Zur Renntniss des Fettgewebes. Arch. mikr. Anat., 45:
J. D. 1955 Electron microscopic observations on the adrenal cortex. Am.
J. Anat., 97: 409-430.
1956a Cytological studies on the hypophysectomised rat adrenal cortex: the alterations of its fine structure following A.C.T.H. administration and on lowering the Na/K ratio. Endocrinology, 58: 163-180.
LEVER,J. D. 1956b Remarks on the electron microscopy of the rat corpus luteurn:
and comparison with earlier observations on the adrenal cortex. Anat.
Rec., 124: 111-126.
1956c The subendothelial space in certain endocrine tissues. J.
Biophys. Biochern. Cytol., 2: 293-296.
1957 Fine structural appearances in the rat parathyroid. J. Anat.
Lond., 9 1 : 73-80.
1954 Studies on the endoplasmic reticulum.
I. I t s identification i n cells in situ. J. exper. Med., 100: 641-655.
K. R. 1953 Observations on a submicroscopic basophilic component of
cytoplasm. J. exper. Yed., 97: 727-749.
A. T. 1923 The so-called hibernatiiig gland. J. Morph., 58: 147193.
1950 Action of insulin on depoAND D. W. FAWCETT
sition of glycogen and storage of f a t in adipose tissue. Endocrinology,
46: 55-66.
E. F. 1924 The so-called hibernating gland in mammals: a form of
adipose tissue. Anat. Rec., 2s: 331-343.
R. L. 1956a Histogenesis of brown adipose tissue in vivo and i n organ
culture. Anat. Rec., 124: 581-595.
1956b The direct effect of insulin on organ cultures of brown
fat. Anat. Rec., 124: 723-734.
E., AND E. WERTHEIMER1942 Glycogen and adipose tissue. J.
Physiol., 100: 385409.
WELLS,H. G. 1940 Adipose tissue a neglected subject. J. Amer. med Ass., 114:
WERTHEIMER,E., AND B. SHAPIRO 1948 The physiology of adipose tissue.
Physiol. Rev., 28: 451-464.
All the figures in both plates 1 and 2 are taken from sections of brown adipose
tissue i n the rat.
1 Frozen section of normal brown f a t stained with Sudan black. Sudanophile
bodies range in size from 7-0.5 p d. Light micrograph, X 1,300.
2 ParaEn section of normal brown f a t stained by the acid fuchsin methyl green
method. Fuchsinophile bodies 1-0.5 p d are clustered around the large osmiophile lipid drops. Light micrograph, X 1,300.
Frozen section of brown f a t after 7 days starvation, stained with Sudan black.
Lipid depletion is marked but patchy: cell A contains a few tiny osmiophile
bodies (1-0.5 p d) while other cells still retain numbers of small droplets
(compare with fig. 1). Light micrograph, X 1,300.
4 and 5 Frozen sections stained by the Baker acid haematein method. I n figure 4
(normal brown f a t ) phospholipid-positive bodies (1-0.5 p d ) lie between unstained f a t drops. I n figure 5 (28 days after adrenalectomy) many of the
brown f a t cells contain no obvious f a t drops and the phospholipid-positive
bodies are scattered throughout the cytoplasm. Light micrographs, X 1,300.
Figs. 6-10 inclusive are all electron micrographs.
Electron micrograph of a normal brown f a t cell. Mitochondria (M) for the
most part have regular internal cristae many of which are beaded: some of
the cristae do not extend right across the mitochondrion (A). Sections between mitochondria1 cristae may account for the appearance of such bodies a s
B, the semiopaque material being the intercristal matrix. Some mitochondria
may have vesicular internal membranes a t one pole (V) and tiny vesicles can
sometimes be seen between cristae (X). Lipid droplets lie freely in the cytoplasm and often interconnect. The rosette grouping of mitochondria around
lipid droplets (L) is a common feature. The endoplasmic reticular elements
include a saccular component (S) ; the granular component ( G ) occurs independently, and in association with the saccular component, X 45,000.
The mitochondrion A contains both cristae and vesicles. The body B contains
comparable vesicles, and in the body C vesicles of similar appearance are intensely osmiophile. As well as the homogeneous lipid constituent in the f a t
drop F, there is a contained vesicular component. The mitochondrion M appears to lack an enclosing membrane along the line of contact with the f a t
drop L, since osmiophile material lies both within the mitochondrion and the
f a t drop (see fig. S), X 35,000.
8 The internum of this mitochondrion is widely open to the f a t droplet through
a breach in the mitochondria1 limiting membrane. Note beading of some of
the cristae, X 35,000.
A vesicular component can be seen within this lipid droplet, X 35,000.
10 As in figure 6 a rosette of mitochondria around a lipid droplet is depicted.
Note the absence of an obvious membrane enclosing the lipid. Mitochondria1
limiting membranes are deficient at D, X 35,000.
11 From right to left, a caoillary lumen; part of an endothelial lining cell containing numerous vesicles; a basement membrane (M) ; the subendothelial
interval (S) ; a basement membrane (B) applied t o the f a t cell plasma membrane (ill defined). Note beading of some of the mitochondrial cristae. Eleetron micrograph, X 35,000.
Light micrograph of a brown f a t ccll containing a large f a t vacuole (15 p
Paraffin section stained with hacmatoxylin and eosin, X 1,300.
Figs. 13-18 inrlusive are electron micrographs of brown fat.
13 The interval (A) between two f a t cells is lined on each side by a basement
membrane, X 18,000.
14 After 10 days starvation, the internal cristae of these brown f a t mitochondria
are irregularly disposed and the intercrivtal matrix material is reduced below
normal (compare with figs. 6-10), X 35,000.
15 Is a representative field of a normal brown f a t (control) cell while figure 16
is a comparable field from the same animal after 7 days starvation. Note the
marked reducton in the size and number of lipid drops and the increase i n the
number of mitochondria which occurs during starvation. Both figures 15 and
16, X 18,000.
17 Seven days after adrenalectomy there is some reduction in lipid.
leached-out lipid drops contain an osmiophile vesicular residuum,
18 From the brown f a t of a rat starved for 7 days. Note the deficiency in the
mitochondria1 enclosing membrane at M. See figures 15 and 16 and compare
with figure 6, X 35,000.
Без категории
Размер файла
900 Кб
structure, starvation, change, observations, following, adrenalectomy, brown, rat, adipose, tissue, fine, cytological
Пожаловаться на содержимое документа