Download Primary cultures of unilocular fat cells

Differentiation
Differentiation (1986) 31 :42-49
(cSpringer-Verlag 1986
Primary cultures of unilocular fat cells :
Characteristics of growth in vitro and changes
in differentiation properties
Hajime Sugihara *, Nobuhisa Yonemitsu, Shinichi Miyabara, and Kankatsu Yun
Department of Pathology, Saga Medical School, Saga 840-01, Japan
Abstract. Mature white fat tissue consists primarily of unilocular fat cells. Clearly, the study of the biology of these
cells would be most helpful for the elucidation of the mechanism of obesity. We describe a new method termed ‘ceiling
culture’ for culturing in vitro unilocular fat cells obtained
from humans and rats. These cells can be maintained in
culture for long periods of time and, under such conditions,
continue to exhibit specific functions such as lipogenesis
and lipolysis. Under the culture conditions described, unilocular fat cells change into multilocular fat cells or cells
with a fibroblast-like appearance. These cells then proliferate, form a cell monolayer attached to the substratum, and
after becoming confluent, exhibit accumulations of intracytoplasmic lipid droplets. These attached dedifferentiated
cells continue to exhibit lipogenesis and lipolysis.
Introduction
Obesity is one of the greatest problems of public health
and nutrition, and it is clearly an old one, as witnessed
by the Venus of Willendorf [3]. The establishment of a culture system of cells obtained from mature fat tissue is clearly
a prerequisite for studies of the cell biology and metabolism
of fat cells [S]. Mature white fat cells typically have a large
single lipid droplet and a peripherally located nucleus, and
are therefore called ‘unilocular fat cells’. These fat cells
do not attach to the substratum, e.g., the plastic surface
of culture flasks when grown in vitro, but tend to float
on top of the medium because of their high lipid content.
Poznanski et al. [20] and Adebonojo [l] have tried to culture unilocular fat cells, but their methods have been found
to be technically difficult to repeat by other investigators.
For this reason, only preadipocytes, i.e., precursor cells with
a relatively low lipid content, have been used as the material
for cell culture [2, 4, 8, 14, 15, 18, 19, 22, 24, 261. Biochemical research has been based on the method of Rodbell [21],
which consists of the preparation of dispersed mature white
fat cells; however, these cells are only able to survive for
a few hours.
In the present study, we describe conditions under whch
unilocular fat cells are able to survive and function for
an extended period of time in a cell-culture system. Furthermore, we described certain changes in the differentiation
*
To whom offprint rcqucsts should be sent
properties and proliferative ability of these cells. These findings are suprising, as unilocular fat cells are generally considered to be in the terminal stage of differentiation [8].
We hope that the culture system described here will be valuable in studies of the biology of fat cells and the mechanism
of obesity.
Methods
Cell culture
Epididymal fat pads from 1- to 4-week-old Wistar rats that
were not obese were used. For the culture of human fat
cells, subcutaneous fat tissue from newborn babies, children
of various ages, and 20-year-old adolescents was obtained
during surgery. The human tissue was obtained from the
following individuals : three newborns (one male and two
females: height, 50 cm; body weight, 3.5 kg), six I-year-old
children (three males: height, 75 cm; weight 9.5 kg; three
females: height 73-75 cm, weight, 9 kg), three 5-year-old
children (three males: height, 10G105 cm; weight, 17 kg),
and three 20-year-old adolescents (three females : height
154160 cm; weight, 44-48 kg). All were Japanese and not
obese.
The fat tissue was chopped into pieces and digested with
collagenase solution [21]. After filtration and centrifugation
(1 86 g for 10 min) of the digestion fluid, unilocular fat cells
were obtained in a thin, white. floating layer; the sediment
consisted of fibroblasts, erythrocytes, and preadipocytes.
The fat cells were taken up and lo5 unilocular fat cells
per culture flask (Falcon 3012; 25 cm2) were incubated at
37” C in a medium consisting of equal volumes of Ham
F12 and Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum (pH 7.2-7.4). The flasks
were completely filled with medium, thus providing the fat
cells with an air-free environment. Initially, the flasks were
turned upside down, and the cells floated up in the medium
and adhered to the top inner surface (ceiling surface) of
the flasks. When the cells had become firmly attached, the
flasks were turned so that the cells were on the bottom
inner surface and could be easily observed. These procedures are shown schematically in Fig. 1. To obtain cultures
consisting of only unilocular fat cells and to avoid contamination by preadipocytes, fibroblasts, and other stromal
cells, we digested sufficiently the fatty tissue and repeated
the pipetting, washing, and centrifugation of the dissociated
unilocular fat cells two times.
43
Centrifuge
@
Collagenase
1
r
unilocular
f a t cells
f a t t y tissue
preadipocytes.
not
fibroblasts
used.
I
Culture
I
c
t
Fig. 1. The method used for culturing unilocular fat
cells. Mature white fat tissue was digested with
collagerase according to Robdell’s method [21].
Unilocular fat cells were obtained by filtering and
centrifuging the digested material. The separatcd
unilocular fat cells were incubatcd in culture flasks
that were completely filled with medium. The
unilocular fat cells floated to the top of the
medium and adhered to the top inner surface
(ceiling culture) of the flask. When the cells had
attached firmly, the flask was placed upside down
to allow regular observation
Fig. 2. Mature white fat cells containing unilocular lipid droplets in culture (unilocular fat cells). These cells were disaggregated by
digesting the fatty tissue of young rats. The cells on the 2nd day of culture adhered to the ceiling surface of flask. Arrowheads, nuclei.
Oil-red-O/hcmatoxylin stain. Bar, 10 pm
The following chemicals were added to the medium of
some cultures: (1) 0.1-1 mIU/ml insulin (Sigma, St. Louis,
USA), (2) 10- M norepinephrine (L-arterenal bitartrate;
Sigma), (3) lo-’ to
M ACTH (Cortrosyn, OrganonDaiichi Seiyaku, Tokyo), and (4) l o p 4 to 1 O - j A4 N6-2‘dibutyryl adenosine-3’-5’-cyclic monophosphate Na-salt
(db-CAMP; Seika-Gaku Kogyo, Tokyo).
Fat cells growing in culture were examined using the
following procedures :
1. Oil-red-O/hematoxylin staining for lipid droplets in cells
was performed as described by Sugihara et al. [25].
2. Enzyme histochemistry of triglyceride lipase according
to Gomori’s method was performed as described by Pearse
3. Enzyme histochemistry of glycerophosphate dehydrogenase was performed according to Pearse’s method [17].
4. Labeling with H-thymidine and autoradiography were
performed as follows: after 6 h of labeling with 2 pCi/ml
H-thymidine (Amersham-Japan, Tokyo) added to the culture medium, cultured cells were fixed for 10 min in 2%
formaldehyde solution made from paraformaldehyde
(Sigma) in 0.05 M cacodylate buffer. Each specimen was
44
coated with Sakura NR-M2 emulsion (Konishiroku, Tokyo) and exposed for 2 weeks before being developed using
Konidol X (Konishiroku). The specimens were then stained
with oil red 0 and hematoxylin. To examine the specific
incorporation of H-thymidine into DNA, two control procedures were followed. First, after labeling with 3H-thylllldine and fixation with 2% formaldehyde solution containing 0.05% saponin, cells were treated with 100 mg/ml
DNA-ase (Sigma) for 2 h at 37" C. Second, cells were labeled with 3H-thymidine in the presence of 0.01 M hydroxyurea. In determinations of the labeling indices, 1,000
cells were examined, and cells containing more than five
silver grains over the nucleus were considered to be significantly labeled.
5. For the determination of levels of glycerol in the medium,
the enzymatic method of Eggstein et al. [5] was applied.
In brief, glycerol was reacted with glycerokinase, pyruvate
kinase, and lactate dehydrogenase. As the amount of
NADH consumed in these reactions is stoichiometric with
the amount of glycerol present, the amount of NADH was
determined by its absorption at 340 nm.
6. For assays of the level of cyclic AMP (CAMP) in cells,
106 cells from each flask were treated with 6% trichloroacetic acid (TCA) and homogenized. The precipitates were
removed by centrifugation, and the supernatant was extracted using water-saturated ether. After succinylation,
cAMP levels were determined using a commercially available cAMP radioimmunoassay kit (Yamasa Shoyu, Choshi,
Japan) according to the method of Honma et al. [7].
7. For scanning electron microscopy, cultured cells were
fixed in 1.5% glutaraldehyde (in 0.05 M cacodylate buffer,
pH 7.2), dehydrated in alcohol, critical-point dried with
C 0 2 , and sputter coated with gold for 40 s. The specimens
were observed using a Hitachi S700 scanning electron microscope.
Results
Dispersed unilocular fat cells obtained from mature fat tissue of young rats and of children and adolescents were
used for primary cultures. In culture flasks which were completely filled with medium, these cells floated to the top
and adhered well to the top inner surface (ceiling surface)
of the flask on which they had accumulated. This 'ceilingculture' method developed in our laboratory is particularly
useful for studies of unilocular fat cells in vitro.
On day 2 of culture, unilocular fat cells obtained from
rats strongly adhered (at their cytoplasmic peripheries) to
the ceiling surface of the flasks (Figs. 2, 3a). On day 3,
the cytoplasm exhibited spreading of the flattened cytoplasmic rim, although the cell body remained spherical
(Fig. 3 b). By day 3 or 4, the cytoplasmic rims, which looked
like tentacles, had spread and extended further, and finely
granular lipid droplets were visible in the cytoplasm. At
this stage, cultured fat cells still contained a large lipid droplet at the center as well as numerous smaller lipid droplets
at the periphery (Fig. 4). Subsequently, these cells hecame
multilocular or assumed a fibroblast-like appearance, containing a large number of small lipid droplets (Fig. 5). This
morphological change occurred within a few days in rat
cells, but it took 7-10 days in human cells.
The average plating efficiency of the cells used as culture
material was over 95% for epididymal fat tissue of young
rats and 90% for subcutaneous fat tissue of human adoles-
Fig. 3a, b. Scanning electron micrograph of cultured fat cells. a
Mature white fat cell containing unilocular lipid droplets (unilocular fat cell). This cell corresponds to the fat cells of Fig. 1. The
cellular surface is almost smooth, with a few short microvilli. b
Mature white fat cell on the 3rd day of culture. This cell can
be seen to adhere tightly to the ceiling surface of flask. The adhered
part of the cytoplasm has spread out, and a large part of the
cell is still spherical. Bars, 5 prn
cents. To obtain high plating rates, over 10' fat cells had
to be plated in 25-cm2 flasks containing 40 ml medium.
No contamination by fibroblasts or preadipocytes was
found on day 2 of culture. At this time, we carefully checked
that all of the cultured cells attached to the ceiling surface
of the flask were unilocular fat cells; this was necessary
because the fat cells changed their appearance during the
course of further culture (see Fig. 2).
In order to determine specific cellular functions such
as lipogenesis and lipolysis [9], rat fat cells were examined
on day 3 of culture. Using histochemical procedures, triglyceride lipase and glycerophosphate dehydrogenase activity
[16, 171 were found to be positive in the cytoplasm. The
45
Fig. 4. Phase-contrast micrograph of mature white fat cells on the 4th day of culture. Their cytoplasm has extended. The cells have
a large lipid droplet at their center and finely granular lipid droplets at their periphery. The cell nuclei cannot be seen clearly in
this figure. Bar, 10 pm
addition of insulin to the medium enhanced the lipogenetic
activity. When 0.1 mIU/ml insulin was added to the medium, cells grown for 4 days exhibited more and bigger cytoplasmic lipid droplets in the resulting multilocular and
fibroblast-like cells. Norepinephrine, ACTH, and dbCAMP, on the other hand, caused lipolysis in the fat cells.
For example, 10- M norepinephrine induced cellular retraction after 30 min, as well as an increase in the numbers
of microvilli on the cell surfaces (Fig. 6 ) . After 3 4 h, dispersion of the lipid droplets was observed (Fig. 7); at the
same time, the glycerol content an index of lipolysis [25]
- increased in the medium, this being concomitant with
an increase in the level of CAMPin the cells (Fig. 8).
When culturing was continued for 1 or 2weeks, the
morphology of the unilocular fat cells changed, so that they
assumed the appearance of multilocular fat cells or fibroblast-like cells containing many small lipid droplets (Fig. 5);
in other words, dedifferentiation occurred. These dedifferentiated cells began to divide (for an example of mitosis,
see the inset in Fig. lob), and the lipid droplets were transferred to the daughter cells. The growth curves of these
cells are shown in Fig. 9. The doubling time for cells obtained from 1-week-old rats was 2.5 days. For cells obtained
from 3-week-old rats as well as from children and adolescents, it took longer for the cell divisions to begin, and
the doubling time was also longer. To confirm the proliferative ability of these fat cells in culture, we used autoradiography to study 3H-thymidine incorporation in cells obtained from 1-week-old rats on day 4 of culture. The label~
ing indices were 2% for unilocular fat cells, 20% for multilocular fat cells, and 40% for fibroblast-like fat cells
(Fig. 10a, b). After DNA-ase treatment of saponin-lysed
cells, the number of grains observed sharply decreased, and
cells containing more than five silver grains could no longer
be found. No grains were observed over cultured cells
treated with hydroxyurea.
Dedifferentiated fat cells such as the multilocular and
fibroblast-like cells formed typical monolayers on the surface of the flasks (Fig. 11). After these cells had grown
to confluence, the number and size of the cytoplasmic lipid
droplets increased and unilocular fat cells reappeared
among these cells (Fig. 11). The addition of 0.1-1 mIU/ml
insulin enhanced the synthesis of lipid droplets. Lipolysis
induced by norepinephrine or db-CAMP was also observed
in the fat cells which had formed monolayers.
Discussion
Using collagenase treatment, Rodbell was the first to succeed in separating unilocular fat cells from mature white
fat tissue of rat epididymal fat pads [21]. Since then, most
biochemical studies of fat cells have been performed using
such dispersed cells. However, culturing unilocular fat cells
in vitro has proved to be technically very difficult, and
probably for this reason, most investigators have preferred
to culture putative precursor cells, so-called preadipocytes,
which are also obtained from fat tissue [2, 4, 8, 14, 15,
18, 19, 22, 24, 261. In the present study, we showed that
46
Fig. 5. Cultured fat cells with a fibroblast-like appearance containing many lipid droplets. These cells were derived from a 5-year-old
child; 7th day of culture. They can be referred to as dedifferentiated
fat cells. Oil-red-O/hematoxylin stain. Bar, 10 pm
Fig. 6. Scanning clectron micrograph of a cultured mature white
Tat cell treated with l o - ' M norepinephrine for 3 h. The cellular
surface is covered with a network of many short microvilli. This
cell is undergoing lipolysis, as demonstrated by an increase in the
glycerol level in the culture medium. Bar, 5 pm
it is possible to maintain unilocular fat cells derived from
mature white fat tissue in culture for long periods of time,
and we also found that such cells can proliferate. The novelty of our method is that the fat cells are provided with
a surface on which they can attach and multiply; this is
achieved by taking advantage of a feature that has been
a hindrance to previous methods for their culture, i.e., their
buoyancy in medium. We have termed this system of cell
culturing ceiling culture. Under these culture conditions,
rat and human unilocular fat cells remain alive and function
as fat cells. We hope that our method will allow white
fat cells to be investigated in vitro in biochemical as well
as in cell-biological studies. We are currently investigating
the potential of our ceiling-culture method for the study
of other sorts of cells.
Under the present culture conditions, the morphology
of unilocular fat cells changed, so that they had the appearance of multilocular or fibroblast-like fat cells. Such a
change has previously been suggested by other researchers
[23]. In terms of cellular morphology, this phenomenon
may be called dedifferentiation, but these dedifferentiated
cells were able to maintain the biochemical functional ability of fat cells. It is widely believed that, once a fat cell
has differentiated enough to accumulate a large unilocular
lipid droplet consisting of triglycerides, it can no longer
dedifferentiate or divide. In our study, the dedifferentiative
and proliferative ability of this particular cell type was observed in vitro only in cells adhering to the surface of culture
flasks and seeded at a h g h density, i.e., over lo5 cells in
40 ml medium. In autoradiographic studies of the incorporation of H-thymidine into cultured fat cells after 6 h, the
labeling indices were 2% in unilocular fat cells, 20% in
multilocular fat cells, and 40% in fibroblast-like fat cells.
The labeling index of unilocular fat cells was low. However,
the obtained data do not indicate the lack of proliferative
ability of the mature fat cells obtained using our culture
system. Unilocular fat cells dedifferentiated into multiocular fat cells and fibroblast-like fat cells, and these dedifferentiated cells proliferated actively. Miller et al. [I31 and Klyde
and Hirsch [lo, 111 have described the incorporation of
3H-thymidine into mature fat tissue in vivo, and they have
also suggested the possibility that mature fat cells have the
ability to proliferate. Even though there are some differences between our autoradiographic procedures and those
used by Miller et al., (i.e. in vitro vs. in vivo, and the materials used) the low but positive labeling index (2%) of unilocular fat cells observed in our culture system is compatible
with the in vivo results obtained by Miller et al. Not all
of the unilocular fat cells divided. In cells obtained from
young rats, those with a diameter of 60 pm or less dedifferentiated, and about 50% of the cells proliferated. Larger
cells exhibited adherence, but they did not extend their cyto-
47
Fig. 7. Cultured fat cells treated with norepinephrine and cultured under the same conditions as the cells shown in Fig. 4. These fat
cells were formerly multilocular fat cells. They are now undergoing lipolysis and have numerous finely granular lipid droplets. Oil-redO/hematoxylin stain. Bar, 10 pm
30
DJ
>
:
10
::
a'
Z
E'
Fig. 8. Lipolysis of cultured fat cells on the 3rd day of culture as
revealed by changes in glycerol released into the medium and
intracelluar CAMP contents. The cells were derived from young rats and
were treated with
M norepinephrine. 0-0, glycerol content;
A-----A, CAMP level. Data show the results of six experiments
1
a
I
4,
0
60
15 30
120
/-*-
*---.a
**.-=-
/*
A/<-
-\
0
1
2
3
4
5
6
QayS
7
8
9
1 0 2 0 Z l 2 2
Fig. 9. Growth curve of mature white fat cells in
culture. Cells were derived from 1-week-old rats
(0-o),
3-week-old rats (0-----o), and a 5-year-old
child (A-A)
48
Fig. IOa, b. Autoradiograph showing 'H-thymidine incorporation. Note the grains located over the nuclei of unilocular fat cells (a)
and multilocular fat cells (b). 3H-Thymidine (2 pCi/ml) was applied to the cultured cells for 6 h. After fixation, the cells were coatcd
with Sakura NR-M2 emulsion, exposed for 2 weeks at 4"C and developed using Konidol X. Arrowheads, grains. Inset. mitosis (arrowheud)
in a multilocular fat cell. Bars, 10 pm
Fig. 11. Phase-contrast micrograph of a monolayer culture of mature white fat cells at confluence. Most of the cells have unilocular
or multilocular lipid droplets; €3indicates approximately the size of one cell. Bar, 40 pm
plasmic tentacles widely, nor did they divide. In cells obtained from children, those with diameters of 5&120 pm
dedifferentiated, and about 20% of these fat cells did divide.
Huge unilocular fat cells of adult human origin were not
included in the present study.
It has been proposed that an in vivo increase in the
number of mature white fat cells is due to the proliferation
of preadipocytes that are present among the mature uni-
locular fat cclls in fat tissue [6, 1G121. On the basis of
our findings, it seems justifiable to assume that unilocular
fat cells themselves may also proliferate under some conditions in vivo.
The factors affecting mature fat cells can be studied
using our system for culturing unilocular fat cells. The culturing of the precursor cells of fat cells - preadipocytes
- has elucidated some factors specific to fat cells, e.g., insu-
49
lin [4, 8, 141 and heparin [4] as lipogenetic factors, epinephrine and ACTH [4, 81 as lipolytic factors, and human serum
[24] and estrogen [22] as growth factors. Some of these
factors have been already examined in cultures of unilocular
fat cells, and as already mentioned, some similarities were
found. The information accumulated from studies of cultures of preadipocytes should be useful in further studies
of unilocular fat cells in culture. We would like to add
that, in preliminary experiments, we have found that newborn calf serum and fibroblast growth factor (FGF) promote the proliferation of cultured unilocular fat cells obtained from rats.
Acknowledgements. We wish to thank Prof. K. Tohkaichi for correcting the English text, and Mr. M. Esaki and Mr. H. Ideguchi
for their technical assistance. This study was supported by a grantin-aid for Scientific Research in Japan (59570153).
References
1. Adebonojo FO (1975) Studies on human adipose cells in cul-
ture: Relation of cell size and cell multiplication to donor age.
Yale J Biol Med 48:9-16
2. Adebonojo FO (1975) Monolayer culture of disaggregated human adipocytes. In Vitro 11:50-54
3. Bierman EL (1981) Obesity. In: Williams RH (ed) Textbook
of endocrinology. W.B. Saunders, Philadelphia, pp 907-921
4. Bjorntorp P, Karlsson M, Petterson P, Sypniewska G (1980)
Differentiation and function of rat adipocyte precursor cells
in primary culture. J Lipid Res 21 :714-723
5. Eggstein M, Kuhlmann E (1974) Triglycerides and glycerol.
In: Bergmeyer HU (ed) Methods of enzymatic analysis, vol4.
Verlag Chemie, Weinheim/Academic Press, New York London,
pp 1825-1831
6. Hager A, Sjostrom L, Arvidsson B, Bjorntorp P, Smith U
(1977) Body fat and adipose tissue cellularity in infants: a longitudinal study. Metabolism 26: 607-614
7. Honma M, Satoh T, Takezawa T, Ui M (1977) An ultrasensitive method for the simultaneous determination of cyclic AMP
and cyclic GMP in small volume samples from blood and tissue.
Biochem Med 18 :257-273
8. Johnson PR, Goldstein AL (1983) Cell culture systems in obesity research. In: Greenwood MRC (ed) Obesity, Churchill Livingstone, Edinburgh Harlow, pp 159-191
9. Kather H (1981) Hormonal regulation of adipose tissue lipolysis in man: Implications for the pathogenesis of obesity. Triangle 20:131-143
10. Klyde BJ, Hirsch J (1979) Isotonic labeling of DNA in rat
adipose tissue : Evidence for proliferating cells associated with
mature adipocytes. J Lipid Res 20: 691-704
11. Klyde BJ, Hirsch J (1979) Increased cellular proliferation in
adipose tissue of adult rat fed a high-fat diet. J Lipid Res
20: 705715
12. Knittle J, Timmers K, Ginsberg-Fellner F (1979) The growth
of adipose tissue in children and adolescents. J Clin Invest
63;239-246
13. Miller WH, Faust IM Jr, Hirsch J (1984) Demonstration of
de novo production of adipocytes in adult rats by biochemical
and radioautographic techniques. J Lipid Res 25 :336-347
14. Nkgrel R, Grimaldi P, Ailhaud G (1978) Establishment of a
preadipocyte clonal line from epididymal fat pad of objob
mouse that responds to lipolytic hormones. Proc Natl Acad
Sci USA 75 :6054-6058
15. Ng CW, Pornanski WJ, Borowiecki M, Reimer G (1971) Differences in growth in vitro of adipose cells from normal and
obese patients. Nature 231 :445
16. Pearse AGE (1972) The Tween method for lipase after Gomori.
In: Histochemistry, theoretical and applied. Churchill Livingstone, Edinburgh Harlow, p 1309
17. Pearse AGE (1972) NAD-linked glycerophosphate dehydrogenase. In : Histochemistry, theoretical and applied. Churchill
Livingstone, Edinburgh Harlow, p 929
18. Pettersson P, Cigolini M, Sjostrom L, Smith U, Bjorntorp P
(1981) Cells in human adipose tissue developing into adipocytes. Acta Med Scand 21 5 :447451
19. Poznanski WJ, Waheed I, Van R (1973) Human fat cell precursors; morphologic and metabolic differentiation in culture. Lab
Invest 29: 57C-576
20. Pomanski WJ, Ruston J, Waheed 1(1974) The origin and metamorphosis of the human fat cell in the regulation of the adipose
tissue mass. In: Vague J, Boyer J (eds) Excerpta Medica, Elsevier Science Publishers, Amsterdam, pp 145-1 50
21. Rodbell M (1964) Metabolism of isolated fat cells. 1. Effects
of hormones on glucose metabolism and lipolysis. J Biol Chem
239:375-380
22. Roncari DAK, Van R (1978) Promotion of human adipocyte
precursor replication by 17P-estrddiol in culture. J Clin Invest
62: 503-508
23. Roncari DAK, Van R (1978) Adipose tissue cellularity and
obesity : New perspectives. Clin Invest Med 1 :71-79
24. Smith U (1971) Morphologic studies of human subcutaneous
adipose tissue in vitro. Anat Rec 169:97-1 04
25. Sugihara H, Miyabara S, Yonemitsu N, Ohta K (1983) Hormonal sensitivity of brown fat cells of fetal rats in monolayer
culture. Exp Clin Endocrinol82: 233-241
26. Tavassoli M (1982) In vitro development of adipose tissue following implantation of lipid-depleted cultured adipocytes. Exp
Cell Res 137 : 55-62
Received July 1985 / Accepted in revised form December I I, 1985