Diet-induced adipocyte number increase adult rats: a new model of obesity IRVING M. FAUST, Rockefeller University,
PATRICIA R. JOHNSON, JUDITH New York, New York 10021
FAUST, IRVING M., PATRICIA AND JULES HIRSCH. Diet-induced
R. JOHNSON, JUDITH S. STERN, adipocyte number increase in ad&t rats: a new model of obesity. Am. J. Physiol. 235(3): E279-E286, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 4(3): E279-E286, 1978. -Adult rats of various strains became obese when they were fed a highly palatable diet for several months. Analysis of their adipose tissue morphology revealed increases in both adipocyte size and number in most depots. Reintroduction of an ordinary chow diet to such animals precipitated a period of v eight loss during which only mean adipocyte size returned to normal. Adipocyte number remained at the elevated level achieved during the period of weight gain. Thus, transient dietary obesity in rats results in a persistent obesity of a purely hyperplastic, nonhypertrophic form. Furthermore, the persistence of the cell number increase suggests that it is the result of proliferation or differentiation rather than of only an increase in the lipid content of a pool of very small and normally undetected adipocytes. An analysis of adipose tissue morphology changes during the course of diet-induced weight gain suggests that the achievement of some specific mean adipocyte size triggers the events that culminate in adipocyte number increase. What mechanisms may link adipocyte size to the formation of new adipocytes remains unknown. adipose tissue; hyperplasia;
Zucker rat, Osborne-Mendel
rat
OBSERVATIONS made during the past decade have supported the hypothesis that adipocyte number is determined early in life and remains constant throughout adulthood in both man and rat. In 1969, Hirsch and Han (13) reported’that adult rats do not lose adipocytes when starved nor acquire new adipocytes during the period of rapid weight gain that follows hypothalamic damage. Shortly thereafter, Hirsch and Knittle (14) and Salans et al. (27) suggested that spontaneously occurring obesity of adult onset in man could be attributed almost exclusively to enlarged adipocytes because adipocyte number in such people was found to be comparable to that of individuals of normal weight. Similarly, Salans et al. (28) induced obesity in adult volunteers and found that adipocyte numbers did not increase. The induced obesity resulted from adipocyte enlargement alone, Early studies of the development of adipose tissue cellularity in rats and mice using cell-counting techniques suggested that proliferative activity in adipose depots ceases by about the 10th wk of postnatal life because no increases in adipocyte numbers were observed afier that time (8, 13, 16, 20). Greenwood and NUMEROUS
0363-6100/78/0000-OOOO$Ol.
25 Copyright
0
1978
the American
S. STERN,
in
AND
JULES
HIRSCH
Hirsch (9), using tritiated thymidine incorporation methods, established that adipocyte proliferation in the epididymal fat pad of the chow-fed rat occurs at a rapid rate from the time of birth until 16-25 days of age, but by 35-40 days of age ceases entirely. Thus, growth of the epididymal pad in the rat afier the 5th wk of life normally occurs only by lipid filling of adipocytes already formed, rather than by the production of new adipocytes. This observation helped to explain earlier reports that undernutrition of the rat only prior ti weaning results in a reduced adipocyte number, which persists into adulthood, even afier the rat is returned to ad libitum chow feeding (20). The above studies fostered the concept of a critical period for adipocyte number determination that occurs early in the life of man, rat, and mouse. One tenet of the concept is that alterations in adipocyte number can be induced only during the critical period. However, the results of several studies suggest that under some circumstances adipocyte number may increase in adults. Mice that became obese affer treatment with gold-thioglucose (GTG) were found to have more adipocytes than untreated controls in one depot, the retroperitoneal (16). Lean Zucker (Ful- > female rats that became obese as the result of being fed a high-fat diet for 7 mo also developed a small increase in adipocyte number in the retroperitoneal depot (24). Increases in adipocyte number have even been reported to occur spontaneously with age in the male rat (2, 32). Also, in recent studies of people with clearly established adultonset obesity, some were found to be hypercellular (1, 11) These latter findings suggest that the critical period concept may need some reexamination. However, several important matters regarding the apparent cell number increases must first be addressed. First, it remains uncertain whether adipocyte number increase in adult animals is an uncommon event peculiar to one or two depots in only certain rat strains or is a general phenomenon that occurs in any rat, in any depot that enlarges with age or in response to factors such as increased diet palatability. Second, it remains uncertain whether the increases in adipocyte number in the above studies were the result of the formation of new adipocytes or merely the result of the enlargement (and thus detectability) of small adipocytes formed early in life. This latter issue was addressed in one early study (26) in which increased amounts of DNA were found in the fat depots of rats made obese as adults. However,
Physiological
Society
E279
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E280
FAUST,
because it was not determined that the extra DNA was in adipocytes rather than in cells of supporting tissues, we do not know whether the new cells that were formed were adipocytes. In the present study, we addressed the above issues by examining in some detail some of the changes in adipose tissue morphology that result from diet-induced weight gain. The major findings of the present study are as follows: I) inducible adipocyte number increase in the rat is a general phenomenon that is not strain, sex, depot, or diet specific; 2) the increased adipocyte number persists during subsequent weight loss and thus suggests new cell formation rather than lipid filling of differentiated cells normally too small to detect; 3) there is an orderly morphologic sequence of adipose depot enlargement so that initially only mean adipocyte size increases. After a certain size is reached, most further depot enlargement is due to adipocyte number increase, EXPERIMENT
1
In the first experiment, adipose tissue cellularity in several major depots was measured in males of two strains of rat and in lean and obese females of another strain. The rats were fed either a standard diet throughout life (control) or fed the diet until an age well past sexual maturity and then switched to a highly palatable high-fat diet (experimental). Rats of various strains gain moderate to large amounts of weight on such a diet (29). If some aspect of the development of obesity, such as adipocyte hypertrophy, promotes the appearance of increased numbers of adipocytes, such increased numbers should be evident in any rat, male or female, that becomes obese on a high-fat diet, and the increase should be related to degree of weight gain or adipocyte hypertrophy . Previously reported results indicate that when a rat is fed a high-fat diet, an increase in adipocyte number occurs primarily, if not exclusively in the retroperitoneal depot (24). However, if depot enlargement or cellular hypertrophy per se stimulates new adipocyte appearance, depots other than the retroperitoneal should also show cell number increases because all fat depots enlarge when a rat becomes obese. Animals and Diets. Thirteen Sprague-Dawley and 12 Osborne-Mendel male rats were switched from Purina laboratory chow to a high-fat diet, routinely used in our laboratory (5), when they were 4 mo of age. Eight 5-mo-old lean (Fal- ) and six 5-mo-old obese (falfa > female Zucker rats were similarly switched from chow to a high-fat diet. Rats of the same age, sex, strain, and mean body weight as the respective experimental rats served as controls (Sprague-Dawley, n = 8; Osborne-Mendel, n = 12; Fa/-, n = 8; fdfa, n = 8). The Sprague-Dawley and Osborne-Mendel control rats were fed chow, whereas the lean and obese Zucker control rats were fed a semipurified low-fat diet described by Lemmonier (24). Procedure. All rats were maintained on their respective diets for 5 mo and then killed. In all rats, retroperitoneal and gonadal fat depots were removed by dissec-
JOHNSON,
STERN,
AND
HIRSCH
tion, weighed, and sampled fur lipid content and cellularity determinations as previously described (4, 12). In addition, the subcutaneous inguinal depots were dissected and analyzed in the male rats, and all subcutaneous adipose tissue (axillary, buttock, scapular, and inguinal) was’dissected and analyzed for the female rats. ResuZts. High-fat feeding increased body weight and body fat in each of the strains examined. However, the amount of increase varied among the strains (see Tables 1 and 1A; 2 and ZA). The fatty Zuckers gained only a small amount of weight relative to controls (+ll%), whereas the Osborne-Mendel rats gained proportionately and absolutely the most weight and the most fat (body weight, +36%; fat in dissected depots, +258%). Adipocyte number increases were also not uniform among the strains. They were smallest in the fatty Zucker rats and greatest in the Osborne-Mendel rats. There were significant overall increases in adipocyte number in the Osborne-Mendel and Sprague-Dawley strains, whereas the increases in adipocyte number were significant in only two of the three analyzed depots of the lean Zucker rats. The failure of the Falrats to increase adipocyte number in the subcutaneous depots does not fit the pattern of cell size-cell number increase seen in all the other analyzed depots of the nonobese rats. Thus, there may be some factor or combination of factors (such as strain, sex, diet, and depot) that is not conducive to adipocyte number expansion. High-fat feeding also resulted in significant adipocyte size increases in all depots of all rats except the fatty Zucker rats in which there were no significant adipocyte size increases in any depot. Thus, excess numbers of adipocytes appeared in all depots in which significant adipocyte size increases occurred except one, (Fa/-, subcutaneous). The magnitude of the cell number increase is also related to the degree of fat That is, there is a significant depot enlargement. correlation between mean percentage of cell number increase per depot and mean percentage of weight increase per depot across all depots of all four rat groups (r = 0.845, P < 0.01). It is interesting to note that although the OsborneMendel rats gained more weight, more fat, and more adipocytes than the Sprague-Dawley or Fa/rats in response to high-fat feeding, adipocyte size increases were practically indistinguishable among the three strains. Furthermore, the mean cell size attained by the three strains in gonadal and retroperitoneal depots (0.9-1.2 pg lipid/cell) is very close to the mean cell size (1.0-l .4 pg lipid/cell) seen in all depots of the fatty Zucker rat regardless of the diet it is fed. These observations suggest that there are restraints on adipocyte size increase that operate in the same range in Sprague-Dawley, Osborne-Mendel, and Zucker rats. EXPERIMENT
2
The results of experiment I suggest that increases in adipocyte number will occur in most fat depots of most adult rats that gain weight on a high-fat diet. To insure that adipocyte number increase is a general phenome-
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED TABLE
of adult
ADIPUCYTE
INCREASE
IN
ADULT
1. Wet weight and cellularity of dissected rats fed a high-fat diet for 5 mo
Total
g
Sprague-Dawley Chow-fed (8) High-fat fed (131
adipose depots
Gonadal
wt,
Bdy
E281
RATS
lipid, g
Retroperitmeal
Cell size, pg lipid/cell
Cell no. x 1V
Total
lipid
Cell
Inguinal
size
Cell no.
Total
lipid
Cell
size
Cell no.
571 218 689 + 25*
8.03 50.72 19.99 * 1.25*
0.49 kO.04 0.98 *0.05*
16.51 20.68 20.66 k1.28*
7.09 2 1.04 22.31 *1.73*
0.67 50.04 1.20 *0.06*
10.39 kl.23 19.09 2 1.88*
10.30 20.75 25.67 22.01*
0.34 20.03 0.70 +0.04*
30.60 2 1.95 37.72 23.2s
592 +15 808 k 20*
9.63 k1.15 26.76 *1.74*
0.61 kO.05 1.10 +0.06*
15.65 21.02 24.26 -tO.86*
10.95 21.31 52.42 ?4.48*
0.69 20.07 1.14 *0.06*
16.23 21.42 48.13 k4.65*
11.73 ~1.38 36.60 +1.81*
0.35 50.03 0.75 20.04”
33.64 k2.78 50.45 23.51*
t test.
t P < 0.05,
Student
Osborne-Mendel Chow-fed
W) High-fat (12) Values
fed
are means
+ SE. Numbers
in parentheses
are numbers
lA, ANOVA
TABLE
of rats.
Sprague-Dawley Diet Depot Interaction Osborne-Mendel Diet Depot Interaction Osborne-Mendel (high-fat diet) vs. SpragueDaw ley (high-fat diet) Strain Depot Interaction < 0.01.
t
t test.
summary (F values) for Table 1 Total
*P
* P < 0.01, Student
Lipid
Cell
Size
Cell Number
df
118.32* 3.53t 0.73
135.03* 36.13* 1.72
13.54* 43.01* 0.55
l/57 2157 2157
244.88* 20.42* 17.47*
109.90* 29.25* 0.39
72.36* 32.35* 9.21*
l/66 2166 Z/66
69.92” 19.20* 15.95*
0.66 39.91* 1.59
42.08* 28.68* 10.19*
l/69 2169 2169
P < 0.05.
TABLE 2. Wet weight and cellularity of dissected adipose depots of adult obese and lean Zucker female rats fed a high-fat diet for 5 mo Gonadal Btiy
Retroperitoneal
Total
Sutxutaneous
wt, Total
8
lipid, g
Cell size, g 1ipidlcel P
Cell no., x 1v
Total
lipid
Cell size
Cell
no.
Total
lipid
Cell size
Cell no.
fQlfa Low -fat fed (8) High-fat fed
16) FQILow -fat fed (8) High-fat fed
(8)
588 k15 654 +22*
29.01 52.91 29.45 k1.62
1.42 to.08 1.33 kO.11
19.74 k2.51 20.71 21.12
34.48 k2.67 46.89 *6.08*
1.24 kO.08 1.23 20.15
27.37 52.49 34.34 24.34
143.61 k3.80 150.82 k11.34
1.12 kO.09 1.05 20.06
123.67 5 12.03 133.89 k5.05
255 +6 315 k8t
4.09 kO.22 11.02 2 1.25t
0.52 kO.05 0.94 z0.09t
7.67 20.46 11.34 20.94t
2.41 kO.25 7.96 +1*04t
0.42 kO.03 0.92 zOm06t
5.37 2 0.54 7.89 +0.60t
8.62 20.38 26.59 *2,73t
0.18 kO.03 0.48 *o.o7t
49.24 k4.54 51.40 23.17
Values are means * SE. Numbers t P < 0.01, Student t test,
in parentheses
TABLE
are numbers
2A. ANOVA Total
of rats.
Obese
rats,
falfa;
lean rats,
Fd-.
* P < 0.05,
Student
t test.
summary (F values) for Table 2 Lipid
Cell Size
Cell
Number
df
falfa Diet Depot Interaction P’u/
2.30 345.38* 0.59
0.44 4.34t 0.08
1.33 173.81* 0,27
l/36 2136 2136
71.52* 72.39* 11.19*
71.22* 26.72* 1*45
2.15 221.12* 0.06
l/42 2142 2142
-
Diet Depot Interaction *P
< 0.01.
t
P < 0.05.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E282
FAUST,
JOHNSON,
STERN, AND HIRSCH
TABLE 3. Wet weight and cellularity of dissected adipose depots of adult Osborne-Mendel rats fed a sucrose solution for 6 mo Gonad81 BUY N g
Chow-fed (7)
639 215
Retmperitoneal
Inguinal
Total lipid, B
Cell size, pg lipid/cell
Cell no., x l(P
Total lipid
Cell size
Cell no.
Total lipid
Cell size
Cell no.
8.98 21.17
0.65 kO.07
13.67 20.80
13.93 +1.65
0.63 +0.07
22.84 +2.86
15.68 k1.56
0.33 kO.03
48.45 +3.08
37.06 ?4.21*
0.59 ?0.08*
64.86 a4.64*
Sucrose-fed 882 17.20 0.95 18.50 34.74 0.96 36.74 +1.51* +0.08* +1.50* -c4.04* +0.07* zL4.11* *57* (7) Values are means 2 SE. Numbers in parentheses are numbers of rats. *P < 0.01, Student t test. TABLE
3A. ANOVA
summary
Total Lipid
Osborne-Mendel (Sucrose-fed) Diet Depot Interaction *P < 0.01.
59.37* 14.35* 3.88t tPc0.05.
non associated with weight gain rather than an unusual response to a diet high in fat content, rats in this experiment were made obese with a high-carbohydrate, low-fat diet known to induce weight gain in rats (19). Procedure. Four-month-old male Osborne-Mendel rats were either continued on their standard laboratory chow and water (n = 7) or fed chow and a 16% solution of sucrose in water ad libitum (n = 8) for 5 mo and then killed. Adipose depot cellularity determinations were made as in experiment 1. Results. The sucrose-fed Osborne-Mendel rats in this experiment gained as much weight in 5 mo and as much lipid per adipocyte relative to controls as the rats fed the high-fat diet for 5 mo in the previous experiment. Adipose depot lipid content and adipocyte number were also both significantly increased compared to control rats (Tables 3 and 3A). Thus, it can be confidently concluded from this experiment that ad libitum feeding of a 16% sucrose solution along with chow to adult rats causes increases not only in body weight and adipocyte size, but also in adipocyte number as well. Adipocyte number increases in adult rats are thus not a peculiar effect of a diet high in fat content. It is interesting to note that the mean cell size attained in gonadal and retroperitoneal depots (approximately 1.0 pg lipid/cell) is very close to the mean cell size attained in those depots by the rats in experiment 1. Thus, the same restraints or upper limits on adipocyte size increase seem to operate regardless of the composition of the diet. EXPERIMENT
3
The first two experiments simply depict the magnitude of adipocyte size and number increase in certain depots of certain rats in response to the feeding of palatable high-fat or high-carbohydrate diets. The third experiment was designed to determine whether the newly apparent adipocytes persist and thus represent a permanent morphological change in the depots
(F values) for Table 3
Cell Size
Cell Number
df
27.06* 16.51* 0.13
20.91* 86.62* 1.89
l/36 2136 2136
q
Experlmentol
0 Chowfed controls
1. Adipocyte number in retroperitoneal (RP) and epididymal (EPI) fat depots of adult male Osborne-Mendel rats fed a high-fat diet for 9 wk or fed a high-fat diet for 9 wk and refed a chow diet for 20 wk. FIG.
that is of long-term consequence to the total adiposity of the rat. Procedure. Ten 13-wk-old male Osborne-Mendel rats were fed the same high-fat diet used in experiment 1 for a period of 9 wk. They were then returned to chow feeding for 20 wk. Ten additional rats of the same age, sex, strain, and initial body weight as the experimental rats served as chow-fed controls. All rats were killed at the end of the 29-wk experimental period. Retroperitoneal and epididymal depots were dissected and analyzed as above. Results. When the rats were killed, those that had been fed a high-fat diet for 9 wk and then fed chow for 20 wk were somewhat, although not significantly, heavier than the rats fed only chow for the entire experiment (657.0 f 15.2 g vs. 635.4 f 15.2 g). However, as can be seen in Fig. 1, the difference in retroperitoneal depot cell number between the two groups of rats in this experiment is dramatic (27.2 x lo6 vs. 13.2 x lo6 cells) and is virtually identical to the difference found
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED
ADIPOCYTE
INCREASE
IN
ADULT
E283
RATS
TABLE 4. Adipose cellularity of 22-wk-old between rats fed only chow and rats fed the high-fat diet for 9 wk but not refed on chow (see experiment 4). Osborne-Mendel rats fed a high-fat diet wk Thus, 20 wk of refeeding on chow produces no measur- for3or9 able return to original retroperitoneal depot adipocyte Wet Weight, Cell Size, Group number although it does eliminate the adipocyte size pg Lipid/Cell t? increase caused by high-fat feeding (Figure 2). AdipoRetroperitoneal cyte number in the epididymal pads of the experimenA 11.55 !z 0.99 0.70 + 0.05 tal rats is also significantly greater than in control rats B 22.42 + 1.29* 1.43 + 0.08* (P < 0.011, but the difference is much smaller than that C 35.20 2 2.01* 1.42 + 0.08* which occurs in the retroperitoneal depots. As in the Epididymal retroperitoneal depot, adipocyte size in the epididymal A 10.90 + 0.82 0.55 c 0.04 pads of experimental rats is very similar to control B 19.78 r 0.94* 1.17 + 0.04* C 30.52 zk 1.42* 1.71 k 0.12* values. The Osborne-Mendel rats that had a 9-wk exposure to a high-fat diet thus have retroperitoneal Mesenteric depots that are twice the normal weight (24.30 -t- 2.20 g A 7.23 k 0.47 0.29 c 0.02 CB 19.20 -c 29.95 f 2.10* 0.80* 0.78 1.19 -r” 0.09* 0.05* vs. 11.28 f 1.22 g) because their retroperitoneal depot adipocyte number is twice normal. It is very likely that Subcutaneous inguinal the relative increase in the weight of the retroperitoA 12.68 ? 0.62 0.33 k 0.02 neal depot is permanent because the relative increase B 25.15 T 1.60* 0.79 k 0.04* in adipocyte number persists fully for at least 20 wk. C 35.31 r 1.82* 1.13 -c 0.10* EXPERIMENT
4
A
One of the above findings is that the cell number increase induced by 5 mo of high-fat feeding in the adult rat is much more dramatic in the retroperiton?al depot than in the other observed depots. In this experiment, shorter periods of high-fat feeding were employed in Osborne-Mendel rats to determine 1) whether the retroperitoneal depot is also the first to develop increases in adipocyte number, and 2) whether there is a general sequential pattern of cell size and number increase in all fat depots that would be indicative of a developmental association between the two. Procedure. Thirty-seven male Osborne-Mendel rats were fed either chow only (n = 13); switched to the high-fat diet 3 wk prior to autopsy (n = 10); or switched to the high-fat diet 9 wk prior to autopsy (n = 14). All rats were killed at 22 wk of age. Cellularity and lipid content analyses were made of epididymal, retroperitoneal, mesenteric, noninguinal subcutaneous (combined axillary, buttock, and scapular), and inguinal subcutaneous depots. Results. The results of the depot analyses are presented in Table 4. After 3 wk of high-fat feeding,
18.68 38.81 54.34
Other Ifr 0.91 zt i?.38* + 3.59*
subcutaneous 0.29 I!T 0.03 0.57 rf: 0.03* 0.85 + 0.05*
14.43 14.01 23.50
k 0.55 + 0.69t
18.12 15.56 16.82
c 1.10 + 0.83t r 1.03t
21.56 22.29 23.31
t 1.01 r 0.87t t 0.95t
27.80 25.81 25.96
2 +
45.41 45.65 49.91
-+2.62
+ 1.65*
1.13 0.97t r 1.21t
2.71t r 2.57t
Z!I
Values are means + SE. Group A, chow-fed (n = 13); I?, chow-fed until age 19 wk, high-fat diet fed between age 19 and 22 wk (n = 10); C, chow-fed until age 13 wk, high-fat diet fed between age 13 and 22 * Significantly different from chow-fed rats, (P < wk (n = 14). t Not significantly different from chow-fed 0.01, Student t test). rats.
adipocyte size is increased in all depots by about 100%. However, there is no indication of an adipocyte number increase in any of the depots. After 9 wk of high-fat feeding, further increases in cell size are evident in all depots except the retroperitoneal, in which there is instead a large increase in cell number (+63%, P < 0.01). It is as if the adipocytes of the retroperitoneal depot are the first to reach some critical size that precedes a rapid increase in adipocyte number. When cell size and number data for the Osborne-Mendel rats in experiments 1 and 4 are combined and depicted graphically, as in Fig. 3, it can be seen that in each of the three depots for which there are complete data, high-fat feeding first causes a rapid increase in mean adipocyte size. Adipocyte size reaches a peak value in each depot, but at different points in time in each depot, and then begins to decline as adipocyte number begins to increase. In each depot, the decline in cell size is significant (P < O.Ol), as is the increase in cell number (P < 0.01). Thus, even though the OsborneMendel rat continuously gains weight during the 20 wk of high-fat feeding, it significantly reduces mean cell size in each depot after some peak value is attained. DISCUSSION
t3 Expermenlol
q
Chowled ~onlrol~
2. Adipocyte size in retroperitoneal (RP) and epididymal (EPI) fat depots of adult male Osborne-Mendel rats fed a high-fat diet for 9 wk or fed a high-fat diet for 9 wk and refed a chow diet for 20 wk. FIG.
The present experiments establish that large increases in adipocyte number can be produced in adult rats of various strains by the feeding of a highly palatable diet high in fat or carbohydrate. Whereas such increases are most impressive in the retroperitoneal depots, they are evident in other depots as well,
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E284
FAUST, Ganodal
Deaats
T
:: !-/./1\_\j30 08 I "-f i/ 04-
___-- -----
+/-----
t
% \W 16 =a 12 5 08 R ;7,04 = s 0 I
_--- _--- --?O i
1
IO
1
‘
I .
lngulnal
Depots
Ttme
.’
/’
.H’
f
(weeks)
3. Adipocyte size and number in adult male Osborne-Mendel as a function of number of weeks they had been fed a high-fat (mean + SE).
FIG.
rats diet
especially after prolonged feeding of the palatable diet. The increases appear to represent genuine morphologic changes in the depots due to new adipocyte proliferation or differentiation because they are not reversed by prolonged feeding of standard laboratory chow However, the question of whether the increases are ‘due to the differentiation of preexistent precursor cells or to the proliferation of new cells can only be answered by means of direct measures of mitotic activity, e.g., via studies of tritiated thymidine incorporation into DNA. l In any case, a rat made obese by dietary means and then returned to a standard laboratory chow diet is morphologically different from a rat that has never experienced dietary obesity. It is moderately obese, and’ its obesity is a purely hyperplastic, nonhypertrophic form that is clearly not secondary to some neurologic or metabolic disorder. Furthermore, the establishment of the obesity does not require a lesion, either experimental or genetic, and the maintenance of the obesity does not require unusual treatment, dietary or otherwise. To our knowledge, there is no other comparable animal model of purely hyperplastic obesity. The availability of this model makes it possible for us to ask directly several important questions about obesity . For example, we can now ask whether there are any secondary consequences of obesity that are independent of increases in either adipocyte size or food intake (e.g., is hyperinsulinemia present in hyperplastic nonhypertrophic obesity?). The hyperplastic nonhypertrophic obese rat may also behave differently from. a normal rat when offered a palatable diet. If rats eat enough of a palatable diet to enlarge adipocytes to a specific size, the hyperplastic, nonhypertrophic rat would initially have to eat more of the diet than a normal rat because it has more adipocytes to enlarge. Sclafani and Gorman (30) found no ’ The increases tions).
results of a recent series of such are due to cellular proliferation
studies suggest (unpublished
that the observa-
JOHNSON,
STERN,
AND
HIRSCH
evidence of such a behavioral difference between rats that had or had not had a previous 60-day experience with a palatable diet. However, Peckham et al. (26) did find such an effect. Rats were fed a high-fat diet for many months and then reduced to normal body weight on chow. The diet was then reintroduced to the rats and also offered to another group of rats that had always been fed chow. The experienced rats ate more of the diet and gained weight more rapidly than the inexperienced rats. The present findings are compatible with the hypothesis, recently put forward by Faust et al. (4, 5) and by Kral (23), that adipocyte size is regulated. At the end of experiment 3, experimental rats had twice as much retroperitoneal depot fat as controls, whereas mean adipocyte size in retroperitoneal and epididymal depots did not differ between the two groups. If total body fat mass were a regulated quantity in the rat, some compensation for the retroperitoneal cell number increases (e.g., cell size decrease) should have occurred, but there is no indication that it did. That a certain degree of enlargement of existent adipocytes precedes the appearance of new adipocytes, as shown in experiment 4, suggests, as predicted previously by Kral (Zl), that when a certain large adipocyte size (e.g., a mean of 1.2-1.6 pg lipid/cell in the retroperitoneal depot of the male Osborne-Mendel rat) is attained, a stimulus for new cell production or differentiation may be produced, especially when adipocytes are maintained in the enlarged state over a long period of time. Two recent observations of adipose tissue cellularity in growing Zucker rats (7) and in growing children (10) parallel the present findings and lend support to the notion of a critical adipocyte size that promotes an increase in adipocyte number. In both cases, mean adipocyte size was found to increase with age to a certain level and then decline as adipocyte number increased. The decline in mean adipocyte size must be due in part to the introduction of new small cells into the countable depot population, but it could also represent some redistribution of caloric load among the increased numbers of cells present in the depot. The notion of a critical adipocyte size in the 1.2-1.6 ,ccgrange in the adult rat is also suggested by two other observations. First, in all depots of the adult fatty Zucker rat, mean adipocyte size is always found to be close to, but less than, 1.6 pg (17, 18, 31), and no dietary manipulation in any adult (experiments 1 and 2) or preweanling rat (17) has ever been reported to have succeeded in pushing mean adipocyte size beyond 1.6 pg. Second, when mean adipocyte size approaches the 1.6.pg level in the rat, not only do new cells begin to appear, but food intake begins to decline (5). The present observation that high-fat feeding does not promote further increase in cell size or number in the fatty Zucker rat suggests that the obesity of the fatty rat is not due to a failure of the adipose tissue to respond appropriately to the condition of adipocyte hypertrophy. New fat cells are apparently normally added as needed as the rat overeats and gains weight on an ordinary diet, whereas further increases do not occur
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED
ADIPOCYTE
INCREASE
IN
ADULT
E285
RATS
in response to the highly palatable high-fat diet. There is an experimental situation in which rats have unusually large adipocytes (over 2.0 pg lipid/cell) and no measurable increase in adipocyte number. Hirsch and collaborators (13, 18) have shown that when the ventromedial hypothalamus (VMH) of an adult rat is electrolytically lesioned, hyperphagia, weight gain, and cell size increases all occur, but cell numbers in epididymal, retroperitoneal, and subcutaneous depots remain indistinguishable from the cell numbers seen in nonlesioned contrdi rats. It thus seems possible that electrolytic ventromedial damage interrupts some pathway necessary for the production or differentiation of adipocytes, whereas at the same time it produces the hyperphagia that induces existing adipocytes to fill with lipid under the pressure of increased calorie load. When the VMH is destroyed in mice via GTG administration, rather than by electrolytic lesion, cell number does increase in the retroperitoneal depot, suggesting that chemical lesioning of the VMH produces less extensive damage and thus leaves the proliferation response intact (16). Hypophysectomy has been shown to interfere with adipocyte proliferation (15) and electrolytic VMH lesions have been shown to lower growth hormone levels in the rat (6, 25). Thus, the extensive VMH lesions in the studies by Hirsch and his collaborators may have caused a growth hormone deficit, or some other endocrine abnormality, that impaired the ability of adipocytes to proliferate or differentiate, while nevertheless allowing existing adipocytes to fill excessively with lipid. In view of the present findings, the concept of an early critical period for adipose tissue development must be reconsidered. We surely cannot now continue to assume that increases in adipocyte number can only be induced very early in the life of the rat. However, there is no question that undernutrition of the very young rat has a profound life-long impact on its adipose tissue cellularity, at least when the rat is maintained on a chow diet, whereas undernourishment of the rat at 15 wk of age or older has no such long-term impact (13, 20). The first few weeks of the life of the rat thus still appear to be critical with regard to circumstances that cause a reduction, but not necessarily with regard to factors that cause an increase in adipocyte number. There is no way to predict at this point whether a rat undernourished early in life will have the same adipocyte number increase as a normal rat in response to a palatable diet. However, if the early undernourished rat does not develop a comparable degree of hypercellularity, the first few weeks of the life of the rat would take on a new significance with regard to adipose tissue development. For example, if achieving some specific mean adipocyte size (e.g., 0.3 pg lipid/cell) normally contributes to the promotion of adipocyte proliferation in the rapidly growing rat, an underfed preweanling rat with very small cells (less than 0.3 pg lipid/cell) would produce a subnormal number of cells. Later in life, the adipocyte deficit might not be overcome because the rat may not eat enough to attain the mean cell size (e.g., 1.3 fig lipid/cell) needed to stimulate new
cell production and would thus not produce new cells. If the early nutritional experience also produces some other effect that prevents the rat from ever eating enough to attain such a large cell size, the rat would always be lean. For example, the relationship between adipocyte size and its restraining influence on food intake (5) may be developed during an early critical period so that a rat underfed prior to 3 wk of age may develop restraints on food intake that act at a smaller cell size. If so, the rat might not be able to eat enough to reach the mean cell size associated with the promotion of new adipocyte proliferation. If the present findings are at all predictive of comparable phenomena in people, there are clear implications for the treatment of the obese and for those at risk of becoming obese. Adipocytes, once differentiated, appear to remain in the adipose depots. Thus, whenever adipocyte number is increased it remains increased. Furthermore, our data suggest that once a new adipocyte is formed, it tends to store about as much lipid as previously formed adipocytes. That is, a given set of dietary conditions predisposes the rat, and perhaps man as well, to a particular mean adipocyte size, regardless of adipocyte number. Thus, once adipocyte size is established under a given dietary condition (current nutritional status), adipocyte number (a reflection of prior nutritional experience) determines total fat mass, To date, there are no known means, other than surgical removal, to reduce adipocyte number, and such surgery is not a preferred clinical approach to the problem of obesity. Most fat cannot be safely removed, and there is no assurance that surgically removed fat will not regenerate. In the rat, adipose tissue clearly regenerates in at least certain areas (3). Surgical removal of fat might also induce compensatory enlargement of other depots (22), a response that also sometimes occurs in the rat (5). It is not known at present whether adipocyte proliferation can be induced in adult human beings. Indeed, studies of experimentally induced obesity in man, in which the period of overweight was sustained for about 6 mo, failed to find an increase in adipocyte number (28). But, there are at least some people with clear adult onset obesity who have excessive numbers of adipocytes (1, 1l), a finding that suggests that induction of adipocyte production is possible in adult man. Perhaps some people are more predisposed to such induction than others, as some rat strains seem ,ti be more predisposed than others. Until these issues are properly resolved, the possibility that an initial episode of weight gain could establish an undesirable morphological alteration in adipose tissue should be a matter of general concern, We thank MS. Elizabeth Seaquist for her excellent technical assistance. This work was supported in park by National Science Foundation Grant PCM 7649324, National Institutes of Health Grant lRO1 AM 20508-01, the Howard Pack Foundation, and a Future Leaders Award to I. M. Faust from the Nutrition Foundation (Grant 5%). Present address of J. S. Stern: University of California at Davis. Received
10 February
1978; accepted
in final
form
17 April
1978.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E286
FAUST,
JOHNSON,
STERN,
AND
HIRSCH
REFERENCES 1. ASHWELL, M., M. DURRANT, AND J. S, GARROW, How a “fat cell pool” hypothesis could account for the relationship between adipose tissue cellularity and the age of onset of obesity. Proc. Nutr. Sot. 36: lllA, 1977. 2. DI GIROLAMO, M., AND S. MENDLINGER. Role of fat cell size and number in enlargement of epididymal fat pads in three species. Am. J. Physiol. 221: 859-864, 1971. 3. FAUST, I. M., P, R. JOHNSON, AND J. HIRSCH. Adipose tissue regeneration following lipectomy. Science 197: 391-393, 1977. 4. FAUST, I. M., P. R. JOHNSON, AND J. HIRSCH. Noncompensation of adipose mass in partially lipectomized mice and rats. Am. J. Physiol. 231: 538-544, 1976. 5. FAUST, I. M., P. R. JOHNSON, ~rj~ J. HIRSCH. Surgical removal of adipose tissue alters feeding behavior and the development of obesity in rats. Science 197: 393-396, 1977. 6. FROHMAN, L. A., AND L. L. BERNARDIS. Growth hormone and insulin levels in weanling rats with VMH lesions. Endocrinology 82: 1125-l 132, 1968. 7. GREENWOOD, M. R. C., AND P. R. JOHNSON. Adipose tissue cellularity and its relationship to the development of obesity in females. Nutritional Disorders of American Women, edited by M. Winick. New York: Wiley, 1977, p. 119-135. 8. GREENWOOD, M. R. C., P. R. JOHNSON, AND J. HIRSCH. Relationship of age and cellularity to metabolic activity in C57B mice. Proc. Sot. Exptl. Biol. Med. 133: 944-947, 1970. 9. GREENWOOD, M. R. C., AND J. HIRSCH. Postnatal development of adipocyte cellularity in the normal rat. J. Lipid Res. 15: 474483, 1974. 10. HAGER, A., L. SJ~STROM, B. ARUIDSSON, P. BjijRNTORP, AND U. SMITH. Body fat and adipose tissue cellularity in infants: a longitudinal study. MetaboZism 26: 607-614, 1977. 11. HIRSCH, J,, AND B. BATCHELOR. Adipose tissue cellularity and human obesity. Clin. Endocrinol. Metab. 5: 299-311, 1976. 12. HIRSCH, J., AND E. GALLIAN. Methods for the determination of adipose cell size in man and animals. J. Lipid Res. 9: 110-119, 1968. 13. HIRSCH, J., AND P. W. HAN. Cellularity of rat adipose tissue: effects of growth, starvation, and obesity. J. Lipid Res. 10: 7782, 1969. 14. HIRSCH, J., AND J. L. KNITTLE. The cellularity of obese and nonobese human adipose tissue. Federation PFOC. 29: 1516-1521, 1970. 15. HOLLENBERG, C. H., AND A. VOST. Regulation of DNA synthesis in fat cells and stromal elements from rat adipose tissue. J. Clin. Invest. 47: 2485-2498, 1968. 16. JOHN~~N, P. R., AND J. HIRSCH. Cellularity of rat adipose depots in six strains of genetically obese mice. J. Lipid Res. 13: 2-l 1, 1972.
17. JOHNSON, P. R., J. S, STERN, M, R. C. GREENWOOD, L. M. ZUCKER, AND J. HIRSCH. Effect of early nutrition on adipose cellularity and pancreatic insulin release in the Zucker rat. J. Nutr. 103: 738-743, 1973. 18. JOHNSON, P. R., L. M. ZUCKER, J. A. F. CRUCE, AND J. HIRSCH. Cellularity of adipose depots in the genetically obese Zucker rat. J. Lipid Res. 12: 706-714, 1971. 19, KANAREK, R. B., AND E. HIRSCH. Dietary-induced overeating in experimental animals. Federation Proc. 36: 154-158, 1977. 20. KNITTLE, J. L., AND J. HIRSCH. Effect of early nutrition on the development of rat epididymal fat pads: cellularity and metabolism. J. C&n. Invest. 47: 2091-2098, 1968. 21. KRAL, J. G. Surgical Reduction of Adipose Tissue. Effects of Adipectomy and Intestinal By-Pass on Adipose Tissue Cellularity (Ph.D. Dissertation). Giiteborg, Sweden: University of Gijteborg, 1976. 22. KRAL, J. G. Surgical reduction of adipose tissue hypercellularity in man. Scand. J. Plastic Reconstruction Surg. 9: 140-143, 2975. 23. KRAL, J. G. Surgical reduction of adipose tissue in the male Sprague-Dawley rat. Am. J. Physiol. 231: 1090-1096, 1976. 24. LEMMONIER, D, Effects of age, sex and weight on cellularity of the adipose tissue in mice and rats rendered obese by a high fat diet* J. Clin. Invest. 51: 2907-2915, 1972. 25. MARTIN, J. B., AND L. P. RENAUD. Pulsatile growth hormone secretion: suppression by hypothalamic ventromedial lesions and by long-acting somatostatin. Science 186: 538-540, 1974. 26. PECKHAM, S. C., C. ENTENMAN, AND H. W. CARROLL. The influence of a hypercaloric diet on gross body and adipose tissue composition in the rat. J. Nutr. 77: 187-197, 1962. 27. SALANS, L. B., S. W. CUSHMAN, AND R. E, WEISMANN. Adipose cell size and number in nonobese and obese patients. J. Clin. Invest. 52: 929-941, 1973. 28, SALANS, L. B., E, S. HORTON, AND E. A. H. SIMS. Experimental obesity in man: cellular character of the adipose tissue. J. Clin. Invest. 50: 1005-1011, 1971. 29. SCHEMMEL, R., 0. MICKELSON, AND J. L. GILL. Dietary obesity in rats: body weight and body fat accretion in seven strains of rats. J. Nuts. 100: 1041-1048, 1970. 30. SCLAFANI, A., AND A. N. GORMAN. Effects of age, sex, and prior body weight on the development of dietary obesity in adult rats. Physiol. Behavior 18: 10214026, 1977. 31. STERN, J. S., P. R. JOHNSON, B. R. BATCHELOR, L. M. ZUCKER, AND J. HIRSCH. Pancreatic insulin release and peripheral tissue resistance in Zucker obese rats fed high- and low-carbohydrate diets. Am. J. Physiol. 228: 543-548, 1975. 32. STILES, J. W., A. A. FRANCENDESE, AND J. MASORO. Influence of age on size and number of fat cells in the epididymal depot. Am. J. Physiol. 229: 1561-1568, 1975.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
PATRICIA R. JOHNSON, JUDITH New York, New York 10021
FAUST, IRVING M., PATRICIA AND JULES HIRSCH. Diet-induced
R. JOHNSON, JUDITH S. STERN, adipocyte number increase in ad&t rats: a new model of obesity. Am. J. Physiol. 235(3): E279-E286, 1978 or Am. J. Physiol.: Endocrinol. Metab. Gastrointest. Physiol. 4(3): E279-E286, 1978. -Adult rats of various strains became obese when they were fed a highly palatable diet for several months. Analysis of their adipose tissue morphology revealed increases in both adipocyte size and number in most depots. Reintroduction of an ordinary chow diet to such animals precipitated a period of v eight loss during which only mean adipocyte size returned to normal. Adipocyte number remained at the elevated level achieved during the period of weight gain. Thus, transient dietary obesity in rats results in a persistent obesity of a purely hyperplastic, nonhypertrophic form. Furthermore, the persistence of the cell number increase suggests that it is the result of proliferation or differentiation rather than of only an increase in the lipid content of a pool of very small and normally undetected adipocytes. An analysis of adipose tissue morphology changes during the course of diet-induced weight gain suggests that the achievement of some specific mean adipocyte size triggers the events that culminate in adipocyte number increase. What mechanisms may link adipocyte size to the formation of new adipocytes remains unknown. adipose tissue; hyperplasia;
Zucker rat, Osborne-Mendel
rat
OBSERVATIONS made during the past decade have supported the hypothesis that adipocyte number is determined early in life and remains constant throughout adulthood in both man and rat. In 1969, Hirsch and Han (13) reported’that adult rats do not lose adipocytes when starved nor acquire new adipocytes during the period of rapid weight gain that follows hypothalamic damage. Shortly thereafter, Hirsch and Knittle (14) and Salans et al. (27) suggested that spontaneously occurring obesity of adult onset in man could be attributed almost exclusively to enlarged adipocytes because adipocyte number in such people was found to be comparable to that of individuals of normal weight. Similarly, Salans et al. (28) induced obesity in adult volunteers and found that adipocyte numbers did not increase. The induced obesity resulted from adipocyte enlargement alone, Early studies of the development of adipose tissue cellularity in rats and mice using cell-counting techniques suggested that proliferative activity in adipose depots ceases by about the 10th wk of postnatal life because no increases in adipocyte numbers were observed afier that time (8, 13, 16, 20). Greenwood and NUMEROUS
0363-6100/78/0000-OOOO$Ol.
25 Copyright
0
1978
the American
S. STERN,
in
AND
JULES
HIRSCH
Hirsch (9), using tritiated thymidine incorporation methods, established that adipocyte proliferation in the epididymal fat pad of the chow-fed rat occurs at a rapid rate from the time of birth until 16-25 days of age, but by 35-40 days of age ceases entirely. Thus, growth of the epididymal pad in the rat afier the 5th wk of life normally occurs only by lipid filling of adipocytes already formed, rather than by the production of new adipocytes. This observation helped to explain earlier reports that undernutrition of the rat only prior ti weaning results in a reduced adipocyte number, which persists into adulthood, even afier the rat is returned to ad libitum chow feeding (20). The above studies fostered the concept of a critical period for adipocyte number determination that occurs early in the life of man, rat, and mouse. One tenet of the concept is that alterations in adipocyte number can be induced only during the critical period. However, the results of several studies suggest that under some circumstances adipocyte number may increase in adults. Mice that became obese affer treatment with gold-thioglucose (GTG) were found to have more adipocytes than untreated controls in one depot, the retroperitoneal (16). Lean Zucker (Ful- > female rats that became obese as the result of being fed a high-fat diet for 7 mo also developed a small increase in adipocyte number in the retroperitoneal depot (24). Increases in adipocyte number have even been reported to occur spontaneously with age in the male rat (2, 32). Also, in recent studies of people with clearly established adultonset obesity, some were found to be hypercellular (1, 11) These latter findings suggest that the critical period concept may need some reexamination. However, several important matters regarding the apparent cell number increases must first be addressed. First, it remains uncertain whether adipocyte number increase in adult animals is an uncommon event peculiar to one or two depots in only certain rat strains or is a general phenomenon that occurs in any rat, in any depot that enlarges with age or in response to factors such as increased diet palatability. Second, it remains uncertain whether the increases in adipocyte number in the above studies were the result of the formation of new adipocytes or merely the result of the enlargement (and thus detectability) of small adipocytes formed early in life. This latter issue was addressed in one early study (26) in which increased amounts of DNA were found in the fat depots of rats made obese as adults. However,
Physiological
Society
E279
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E280
FAUST,
because it was not determined that the extra DNA was in adipocytes rather than in cells of supporting tissues, we do not know whether the new cells that were formed were adipocytes. In the present study, we addressed the above issues by examining in some detail some of the changes in adipose tissue morphology that result from diet-induced weight gain. The major findings of the present study are as follows: I) inducible adipocyte number increase in the rat is a general phenomenon that is not strain, sex, depot, or diet specific; 2) the increased adipocyte number persists during subsequent weight loss and thus suggests new cell formation rather than lipid filling of differentiated cells normally too small to detect; 3) there is an orderly morphologic sequence of adipose depot enlargement so that initially only mean adipocyte size increases. After a certain size is reached, most further depot enlargement is due to adipocyte number increase, EXPERIMENT
1
In the first experiment, adipose tissue cellularity in several major depots was measured in males of two strains of rat and in lean and obese females of another strain. The rats were fed either a standard diet throughout life (control) or fed the diet until an age well past sexual maturity and then switched to a highly palatable high-fat diet (experimental). Rats of various strains gain moderate to large amounts of weight on such a diet (29). If some aspect of the development of obesity, such as adipocyte hypertrophy, promotes the appearance of increased numbers of adipocytes, such increased numbers should be evident in any rat, male or female, that becomes obese on a high-fat diet, and the increase should be related to degree of weight gain or adipocyte hypertrophy . Previously reported results indicate that when a rat is fed a high-fat diet, an increase in adipocyte number occurs primarily, if not exclusively in the retroperitoneal depot (24). However, if depot enlargement or cellular hypertrophy per se stimulates new adipocyte appearance, depots other than the retroperitoneal should also show cell number increases because all fat depots enlarge when a rat becomes obese. Animals and Diets. Thirteen Sprague-Dawley and 12 Osborne-Mendel male rats were switched from Purina laboratory chow to a high-fat diet, routinely used in our laboratory (5), when they were 4 mo of age. Eight 5-mo-old lean (Fal- ) and six 5-mo-old obese (falfa > female Zucker rats were similarly switched from chow to a high-fat diet. Rats of the same age, sex, strain, and mean body weight as the respective experimental rats served as controls (Sprague-Dawley, n = 8; Osborne-Mendel, n = 12; Fa/-, n = 8; fdfa, n = 8). The Sprague-Dawley and Osborne-Mendel control rats were fed chow, whereas the lean and obese Zucker control rats were fed a semipurified low-fat diet described by Lemmonier (24). Procedure. All rats were maintained on their respective diets for 5 mo and then killed. In all rats, retroperitoneal and gonadal fat depots were removed by dissec-
JOHNSON,
STERN,
AND
HIRSCH
tion, weighed, and sampled fur lipid content and cellularity determinations as previously described (4, 12). In addition, the subcutaneous inguinal depots were dissected and analyzed in the male rats, and all subcutaneous adipose tissue (axillary, buttock, scapular, and inguinal) was’dissected and analyzed for the female rats. ResuZts. High-fat feeding increased body weight and body fat in each of the strains examined. However, the amount of increase varied among the strains (see Tables 1 and 1A; 2 and ZA). The fatty Zuckers gained only a small amount of weight relative to controls (+ll%), whereas the Osborne-Mendel rats gained proportionately and absolutely the most weight and the most fat (body weight, +36%; fat in dissected depots, +258%). Adipocyte number increases were also not uniform among the strains. They were smallest in the fatty Zucker rats and greatest in the Osborne-Mendel rats. There were significant overall increases in adipocyte number in the Osborne-Mendel and Sprague-Dawley strains, whereas the increases in adipocyte number were significant in only two of the three analyzed depots of the lean Zucker rats. The failure of the Falrats to increase adipocyte number in the subcutaneous depots does not fit the pattern of cell size-cell number increase seen in all the other analyzed depots of the nonobese rats. Thus, there may be some factor or combination of factors (such as strain, sex, diet, and depot) that is not conducive to adipocyte number expansion. High-fat feeding also resulted in significant adipocyte size increases in all depots of all rats except the fatty Zucker rats in which there were no significant adipocyte size increases in any depot. Thus, excess numbers of adipocytes appeared in all depots in which significant adipocyte size increases occurred except one, (Fa/-, subcutaneous). The magnitude of the cell number increase is also related to the degree of fat That is, there is a significant depot enlargement. correlation between mean percentage of cell number increase per depot and mean percentage of weight increase per depot across all depots of all four rat groups (r = 0.845, P < 0.01). It is interesting to note that although the OsborneMendel rats gained more weight, more fat, and more adipocytes than the Sprague-Dawley or Fa/rats in response to high-fat feeding, adipocyte size increases were practically indistinguishable among the three strains. Furthermore, the mean cell size attained by the three strains in gonadal and retroperitoneal depots (0.9-1.2 pg lipid/cell) is very close to the mean cell size (1.0-l .4 pg lipid/cell) seen in all depots of the fatty Zucker rat regardless of the diet it is fed. These observations suggest that there are restraints on adipocyte size increase that operate in the same range in Sprague-Dawley, Osborne-Mendel, and Zucker rats. EXPERIMENT
2
The results of experiment I suggest that increases in adipocyte number will occur in most fat depots of most adult rats that gain weight on a high-fat diet. To insure that adipocyte number increase is a general phenome-
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED TABLE
of adult
ADIPUCYTE
INCREASE
IN
ADULT
1. Wet weight and cellularity of dissected rats fed a high-fat diet for 5 mo
Total
g
Sprague-Dawley Chow-fed (8) High-fat fed (131
adipose depots
Gonadal
wt,
Bdy
E281
RATS
lipid, g
Retroperitmeal
Cell size, pg lipid/cell
Cell no. x 1V
Total
lipid
Cell
Inguinal
size
Cell no.
Total
lipid
Cell
size
Cell no.
571 218 689 + 25*
8.03 50.72 19.99 * 1.25*
0.49 kO.04 0.98 *0.05*
16.51 20.68 20.66 k1.28*
7.09 2 1.04 22.31 *1.73*
0.67 50.04 1.20 *0.06*
10.39 kl.23 19.09 2 1.88*
10.30 20.75 25.67 22.01*
0.34 20.03 0.70 +0.04*
30.60 2 1.95 37.72 23.2s
592 +15 808 k 20*
9.63 k1.15 26.76 *1.74*
0.61 kO.05 1.10 +0.06*
15.65 21.02 24.26 -tO.86*
10.95 21.31 52.42 ?4.48*
0.69 20.07 1.14 *0.06*
16.23 21.42 48.13 k4.65*
11.73 ~1.38 36.60 +1.81*
0.35 50.03 0.75 20.04”
33.64 k2.78 50.45 23.51*
t test.
t P < 0.05,
Student
Osborne-Mendel Chow-fed
W) High-fat (12) Values
fed
are means
+ SE. Numbers
in parentheses
are numbers
lA, ANOVA
TABLE
of rats.
Sprague-Dawley Diet Depot Interaction Osborne-Mendel Diet Depot Interaction Osborne-Mendel (high-fat diet) vs. SpragueDaw ley (high-fat diet) Strain Depot Interaction < 0.01.
t
t test.
summary (F values) for Table 1 Total
*P
* P < 0.01, Student
Lipid
Cell
Size
Cell Number
df
118.32* 3.53t 0.73
135.03* 36.13* 1.72
13.54* 43.01* 0.55
l/57 2157 2157
244.88* 20.42* 17.47*
109.90* 29.25* 0.39
72.36* 32.35* 9.21*
l/66 2166 Z/66
69.92” 19.20* 15.95*
0.66 39.91* 1.59
42.08* 28.68* 10.19*
l/69 2169 2169
P < 0.05.
TABLE 2. Wet weight and cellularity of dissected adipose depots of adult obese and lean Zucker female rats fed a high-fat diet for 5 mo Gonadal Btiy
Retroperitoneal
Total
Sutxutaneous
wt, Total
8
lipid, g
Cell size, g 1ipidlcel P
Cell no., x 1v
Total
lipid
Cell size
Cell
no.
Total
lipid
Cell size
Cell no.
fQlfa Low -fat fed (8) High-fat fed
16) FQILow -fat fed (8) High-fat fed
(8)
588 k15 654 +22*
29.01 52.91 29.45 k1.62
1.42 to.08 1.33 kO.11
19.74 k2.51 20.71 21.12
34.48 k2.67 46.89 *6.08*
1.24 kO.08 1.23 20.15
27.37 52.49 34.34 24.34
143.61 k3.80 150.82 k11.34
1.12 kO.09 1.05 20.06
123.67 5 12.03 133.89 k5.05
255 +6 315 k8t
4.09 kO.22 11.02 2 1.25t
0.52 kO.05 0.94 z0.09t
7.67 20.46 11.34 20.94t
2.41 kO.25 7.96 +1*04t
0.42 kO.03 0.92 zOm06t
5.37 2 0.54 7.89 +0.60t
8.62 20.38 26.59 *2,73t
0.18 kO.03 0.48 *o.o7t
49.24 k4.54 51.40 23.17
Values are means * SE. Numbers t P < 0.01, Student t test,
in parentheses
TABLE
are numbers
2A. ANOVA Total
of rats.
Obese
rats,
falfa;
lean rats,
Fd-.
* P < 0.05,
Student
t test.
summary (F values) for Table 2 Lipid
Cell Size
Cell
Number
df
falfa Diet Depot Interaction P’u/
2.30 345.38* 0.59
0.44 4.34t 0.08
1.33 173.81* 0,27
l/36 2136 2136
71.52* 72.39* 11.19*
71.22* 26.72* 1*45
2.15 221.12* 0.06
l/42 2142 2142
-
Diet Depot Interaction *P
< 0.01.
t
P < 0.05.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E282
FAUST,
JOHNSON,
STERN, AND HIRSCH
TABLE 3. Wet weight and cellularity of dissected adipose depots of adult Osborne-Mendel rats fed a sucrose solution for 6 mo Gonad81 BUY N g
Chow-fed (7)
639 215
Retmperitoneal
Inguinal
Total lipid, B
Cell size, pg lipid/cell
Cell no., x l(P
Total lipid
Cell size
Cell no.
Total lipid
Cell size
Cell no.
8.98 21.17
0.65 kO.07
13.67 20.80
13.93 +1.65
0.63 +0.07
22.84 +2.86
15.68 k1.56
0.33 kO.03
48.45 +3.08
37.06 ?4.21*
0.59 ?0.08*
64.86 a4.64*
Sucrose-fed 882 17.20 0.95 18.50 34.74 0.96 36.74 +1.51* +0.08* +1.50* -c4.04* +0.07* zL4.11* *57* (7) Values are means 2 SE. Numbers in parentheses are numbers of rats. *P < 0.01, Student t test. TABLE
3A. ANOVA
summary
Total Lipid
Osborne-Mendel (Sucrose-fed) Diet Depot Interaction *P < 0.01.
59.37* 14.35* 3.88t tPc0.05.
non associated with weight gain rather than an unusual response to a diet high in fat content, rats in this experiment were made obese with a high-carbohydrate, low-fat diet known to induce weight gain in rats (19). Procedure. Four-month-old male Osborne-Mendel rats were either continued on their standard laboratory chow and water (n = 7) or fed chow and a 16% solution of sucrose in water ad libitum (n = 8) for 5 mo and then killed. Adipose depot cellularity determinations were made as in experiment 1. Results. The sucrose-fed Osborne-Mendel rats in this experiment gained as much weight in 5 mo and as much lipid per adipocyte relative to controls as the rats fed the high-fat diet for 5 mo in the previous experiment. Adipose depot lipid content and adipocyte number were also both significantly increased compared to control rats (Tables 3 and 3A). Thus, it can be confidently concluded from this experiment that ad libitum feeding of a 16% sucrose solution along with chow to adult rats causes increases not only in body weight and adipocyte size, but also in adipocyte number as well. Adipocyte number increases in adult rats are thus not a peculiar effect of a diet high in fat content. It is interesting to note that the mean cell size attained in gonadal and retroperitoneal depots (approximately 1.0 pg lipid/cell) is very close to the mean cell size attained in those depots by the rats in experiment 1. Thus, the same restraints or upper limits on adipocyte size increase seem to operate regardless of the composition of the diet. EXPERIMENT
3
The first two experiments simply depict the magnitude of adipocyte size and number increase in certain depots of certain rats in response to the feeding of palatable high-fat or high-carbohydrate diets. The third experiment was designed to determine whether the newly apparent adipocytes persist and thus represent a permanent morphological change in the depots
(F values) for Table 3
Cell Size
Cell Number
df
27.06* 16.51* 0.13
20.91* 86.62* 1.89
l/36 2136 2136
q
Experlmentol
0 Chowfed controls
1. Adipocyte number in retroperitoneal (RP) and epididymal (EPI) fat depots of adult male Osborne-Mendel rats fed a high-fat diet for 9 wk or fed a high-fat diet for 9 wk and refed a chow diet for 20 wk. FIG.
that is of long-term consequence to the total adiposity of the rat. Procedure. Ten 13-wk-old male Osborne-Mendel rats were fed the same high-fat diet used in experiment 1 for a period of 9 wk. They were then returned to chow feeding for 20 wk. Ten additional rats of the same age, sex, strain, and initial body weight as the experimental rats served as chow-fed controls. All rats were killed at the end of the 29-wk experimental period. Retroperitoneal and epididymal depots were dissected and analyzed as above. Results. When the rats were killed, those that had been fed a high-fat diet for 9 wk and then fed chow for 20 wk were somewhat, although not significantly, heavier than the rats fed only chow for the entire experiment (657.0 f 15.2 g vs. 635.4 f 15.2 g). However, as can be seen in Fig. 1, the difference in retroperitoneal depot cell number between the two groups of rats in this experiment is dramatic (27.2 x lo6 vs. 13.2 x lo6 cells) and is virtually identical to the difference found
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED
ADIPOCYTE
INCREASE
IN
ADULT
E283
RATS
TABLE 4. Adipose cellularity of 22-wk-old between rats fed only chow and rats fed the high-fat diet for 9 wk but not refed on chow (see experiment 4). Osborne-Mendel rats fed a high-fat diet wk Thus, 20 wk of refeeding on chow produces no measur- for3or9 able return to original retroperitoneal depot adipocyte Wet Weight, Cell Size, Group number although it does eliminate the adipocyte size pg Lipid/Cell t? increase caused by high-fat feeding (Figure 2). AdipoRetroperitoneal cyte number in the epididymal pads of the experimenA 11.55 !z 0.99 0.70 + 0.05 tal rats is also significantly greater than in control rats B 22.42 + 1.29* 1.43 + 0.08* (P < 0.011, but the difference is much smaller than that C 35.20 2 2.01* 1.42 + 0.08* which occurs in the retroperitoneal depots. As in the Epididymal retroperitoneal depot, adipocyte size in the epididymal A 10.90 + 0.82 0.55 c 0.04 pads of experimental rats is very similar to control B 19.78 r 0.94* 1.17 + 0.04* C 30.52 zk 1.42* 1.71 k 0.12* values. The Osborne-Mendel rats that had a 9-wk exposure to a high-fat diet thus have retroperitoneal Mesenteric depots that are twice the normal weight (24.30 -t- 2.20 g A 7.23 k 0.47 0.29 c 0.02 CB 19.20 -c 29.95 f 2.10* 0.80* 0.78 1.19 -r” 0.09* 0.05* vs. 11.28 f 1.22 g) because their retroperitoneal depot adipocyte number is twice normal. It is very likely that Subcutaneous inguinal the relative increase in the weight of the retroperitoA 12.68 ? 0.62 0.33 k 0.02 neal depot is permanent because the relative increase B 25.15 T 1.60* 0.79 k 0.04* in adipocyte number persists fully for at least 20 wk. C 35.31 r 1.82* 1.13 -c 0.10* EXPERIMENT
4
A
One of the above findings is that the cell number increase induced by 5 mo of high-fat feeding in the adult rat is much more dramatic in the retroperiton?al depot than in the other observed depots. In this experiment, shorter periods of high-fat feeding were employed in Osborne-Mendel rats to determine 1) whether the retroperitoneal depot is also the first to develop increases in adipocyte number, and 2) whether there is a general sequential pattern of cell size and number increase in all fat depots that would be indicative of a developmental association between the two. Procedure. Thirty-seven male Osborne-Mendel rats were fed either chow only (n = 13); switched to the high-fat diet 3 wk prior to autopsy (n = 10); or switched to the high-fat diet 9 wk prior to autopsy (n = 14). All rats were killed at 22 wk of age. Cellularity and lipid content analyses were made of epididymal, retroperitoneal, mesenteric, noninguinal subcutaneous (combined axillary, buttock, and scapular), and inguinal subcutaneous depots. Results. The results of the depot analyses are presented in Table 4. After 3 wk of high-fat feeding,
18.68 38.81 54.34
Other Ifr 0.91 zt i?.38* + 3.59*
subcutaneous 0.29 I!T 0.03 0.57 rf: 0.03* 0.85 + 0.05*
14.43 14.01 23.50
k 0.55 + 0.69t
18.12 15.56 16.82
c 1.10 + 0.83t r 1.03t
21.56 22.29 23.31
t 1.01 r 0.87t t 0.95t
27.80 25.81 25.96
2 +
45.41 45.65 49.91
-+2.62
+ 1.65*
1.13 0.97t r 1.21t
2.71t r 2.57t
Z!I
Values are means + SE. Group A, chow-fed (n = 13); I?, chow-fed until age 19 wk, high-fat diet fed between age 19 and 22 wk (n = 10); C, chow-fed until age 13 wk, high-fat diet fed between age 13 and 22 * Significantly different from chow-fed rats, (P < wk (n = 14). t Not significantly different from chow-fed 0.01, Student t test). rats.
adipocyte size is increased in all depots by about 100%. However, there is no indication of an adipocyte number increase in any of the depots. After 9 wk of high-fat feeding, further increases in cell size are evident in all depots except the retroperitoneal, in which there is instead a large increase in cell number (+63%, P < 0.01). It is as if the adipocytes of the retroperitoneal depot are the first to reach some critical size that precedes a rapid increase in adipocyte number. When cell size and number data for the Osborne-Mendel rats in experiments 1 and 4 are combined and depicted graphically, as in Fig. 3, it can be seen that in each of the three depots for which there are complete data, high-fat feeding first causes a rapid increase in mean adipocyte size. Adipocyte size reaches a peak value in each depot, but at different points in time in each depot, and then begins to decline as adipocyte number begins to increase. In each depot, the decline in cell size is significant (P < O.Ol), as is the increase in cell number (P < 0.01). Thus, even though the OsborneMendel rat continuously gains weight during the 20 wk of high-fat feeding, it significantly reduces mean cell size in each depot after some peak value is attained. DISCUSSION
t3 Expermenlol
q
Chowled ~onlrol~
2. Adipocyte size in retroperitoneal (RP) and epididymal (EPI) fat depots of adult male Osborne-Mendel rats fed a high-fat diet for 9 wk or fed a high-fat diet for 9 wk and refed a chow diet for 20 wk. FIG.
The present experiments establish that large increases in adipocyte number can be produced in adult rats of various strains by the feeding of a highly palatable diet high in fat or carbohydrate. Whereas such increases are most impressive in the retroperitoneal depots, they are evident in other depots as well,
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E284
FAUST, Ganodal
Deaats
T
:: !-/./1\_\j30 08 I "-f i/ 04-
___-- -----
+/-----
t
% \W 16 =a 12 5 08 R ;7,04 = s 0 I
_--- _--- --?O i
1
IO
1
‘
I .
lngulnal
Depots
Ttme
.’
/’
.H’
f
(weeks)
3. Adipocyte size and number in adult male Osborne-Mendel as a function of number of weeks they had been fed a high-fat (mean + SE).
FIG.
rats diet
especially after prolonged feeding of the palatable diet. The increases appear to represent genuine morphologic changes in the depots due to new adipocyte proliferation or differentiation because they are not reversed by prolonged feeding of standard laboratory chow However, the question of whether the increases are ‘due to the differentiation of preexistent precursor cells or to the proliferation of new cells can only be answered by means of direct measures of mitotic activity, e.g., via studies of tritiated thymidine incorporation into DNA. l In any case, a rat made obese by dietary means and then returned to a standard laboratory chow diet is morphologically different from a rat that has never experienced dietary obesity. It is moderately obese, and’ its obesity is a purely hyperplastic, nonhypertrophic form that is clearly not secondary to some neurologic or metabolic disorder. Furthermore, the establishment of the obesity does not require a lesion, either experimental or genetic, and the maintenance of the obesity does not require unusual treatment, dietary or otherwise. To our knowledge, there is no other comparable animal model of purely hyperplastic obesity. The availability of this model makes it possible for us to ask directly several important questions about obesity . For example, we can now ask whether there are any secondary consequences of obesity that are independent of increases in either adipocyte size or food intake (e.g., is hyperinsulinemia present in hyperplastic nonhypertrophic obesity?). The hyperplastic nonhypertrophic obese rat may also behave differently from. a normal rat when offered a palatable diet. If rats eat enough of a palatable diet to enlarge adipocytes to a specific size, the hyperplastic, nonhypertrophic rat would initially have to eat more of the diet than a normal rat because it has more adipocytes to enlarge. Sclafani and Gorman (30) found no ’ The increases tions).
results of a recent series of such are due to cellular proliferation
studies suggest (unpublished
that the observa-
JOHNSON,
STERN,
AND
HIRSCH
evidence of such a behavioral difference between rats that had or had not had a previous 60-day experience with a palatable diet. However, Peckham et al. (26) did find such an effect. Rats were fed a high-fat diet for many months and then reduced to normal body weight on chow. The diet was then reintroduced to the rats and also offered to another group of rats that had always been fed chow. The experienced rats ate more of the diet and gained weight more rapidly than the inexperienced rats. The present findings are compatible with the hypothesis, recently put forward by Faust et al. (4, 5) and by Kral (23), that adipocyte size is regulated. At the end of experiment 3, experimental rats had twice as much retroperitoneal depot fat as controls, whereas mean adipocyte size in retroperitoneal and epididymal depots did not differ between the two groups. If total body fat mass were a regulated quantity in the rat, some compensation for the retroperitoneal cell number increases (e.g., cell size decrease) should have occurred, but there is no indication that it did. That a certain degree of enlargement of existent adipocytes precedes the appearance of new adipocytes, as shown in experiment 4, suggests, as predicted previously by Kral (Zl), that when a certain large adipocyte size (e.g., a mean of 1.2-1.6 pg lipid/cell in the retroperitoneal depot of the male Osborne-Mendel rat) is attained, a stimulus for new cell production or differentiation may be produced, especially when adipocytes are maintained in the enlarged state over a long period of time. Two recent observations of adipose tissue cellularity in growing Zucker rats (7) and in growing children (10) parallel the present findings and lend support to the notion of a critical adipocyte size that promotes an increase in adipocyte number. In both cases, mean adipocyte size was found to increase with age to a certain level and then decline as adipocyte number increased. The decline in mean adipocyte size must be due in part to the introduction of new small cells into the countable depot population, but it could also represent some redistribution of caloric load among the increased numbers of cells present in the depot. The notion of a critical adipocyte size in the 1.2-1.6 ,ccgrange in the adult rat is also suggested by two other observations. First, in all depots of the adult fatty Zucker rat, mean adipocyte size is always found to be close to, but less than, 1.6 pg (17, 18, 31), and no dietary manipulation in any adult (experiments 1 and 2) or preweanling rat (17) has ever been reported to have succeeded in pushing mean adipocyte size beyond 1.6 pg. Second, when mean adipocyte size approaches the 1.6.pg level in the rat, not only do new cells begin to appear, but food intake begins to decline (5). The present observation that high-fat feeding does not promote further increase in cell size or number in the fatty Zucker rat suggests that the obesity of the fatty rat is not due to a failure of the adipose tissue to respond appropriately to the condition of adipocyte hypertrophy. New fat cells are apparently normally added as needed as the rat overeats and gains weight on an ordinary diet, whereas further increases do not occur
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
DIET-INDUCED
ADIPOCYTE
INCREASE
IN
ADULT
E285
RATS
in response to the highly palatable high-fat diet. There is an experimental situation in which rats have unusually large adipocytes (over 2.0 pg lipid/cell) and no measurable increase in adipocyte number. Hirsch and collaborators (13, 18) have shown that when the ventromedial hypothalamus (VMH) of an adult rat is electrolytically lesioned, hyperphagia, weight gain, and cell size increases all occur, but cell numbers in epididymal, retroperitoneal, and subcutaneous depots remain indistinguishable from the cell numbers seen in nonlesioned contrdi rats. It thus seems possible that electrolytic ventromedial damage interrupts some pathway necessary for the production or differentiation of adipocytes, whereas at the same time it produces the hyperphagia that induces existing adipocytes to fill with lipid under the pressure of increased calorie load. When the VMH is destroyed in mice via GTG administration, rather than by electrolytic lesion, cell number does increase in the retroperitoneal depot, suggesting that chemical lesioning of the VMH produces less extensive damage and thus leaves the proliferation response intact (16). Hypophysectomy has been shown to interfere with adipocyte proliferation (15) and electrolytic VMH lesions have been shown to lower growth hormone levels in the rat (6, 25). Thus, the extensive VMH lesions in the studies by Hirsch and his collaborators may have caused a growth hormone deficit, or some other endocrine abnormality, that impaired the ability of adipocytes to proliferate or differentiate, while nevertheless allowing existing adipocytes to fill excessively with lipid. In view of the present findings, the concept of an early critical period for adipose tissue development must be reconsidered. We surely cannot now continue to assume that increases in adipocyte number can only be induced very early in the life of the rat. However, there is no question that undernutrition of the very young rat has a profound life-long impact on its adipose tissue cellularity, at least when the rat is maintained on a chow diet, whereas undernourishment of the rat at 15 wk of age or older has no such long-term impact (13, 20). The first few weeks of the life of the rat thus still appear to be critical with regard to circumstances that cause a reduction, but not necessarily with regard to factors that cause an increase in adipocyte number. There is no way to predict at this point whether a rat undernourished early in life will have the same adipocyte number increase as a normal rat in response to a palatable diet. However, if the early undernourished rat does not develop a comparable degree of hypercellularity, the first few weeks of the life of the rat would take on a new significance with regard to adipose tissue development. For example, if achieving some specific mean adipocyte size (e.g., 0.3 pg lipid/cell) normally contributes to the promotion of adipocyte proliferation in the rapidly growing rat, an underfed preweanling rat with very small cells (less than 0.3 pg lipid/cell) would produce a subnormal number of cells. Later in life, the adipocyte deficit might not be overcome because the rat may not eat enough to attain the mean cell size (e.g., 1.3 fig lipid/cell) needed to stimulate new
cell production and would thus not produce new cells. If the early nutritional experience also produces some other effect that prevents the rat from ever eating enough to attain such a large cell size, the rat would always be lean. For example, the relationship between adipocyte size and its restraining influence on food intake (5) may be developed during an early critical period so that a rat underfed prior to 3 wk of age may develop restraints on food intake that act at a smaller cell size. If so, the rat might not be able to eat enough to reach the mean cell size associated with the promotion of new adipocyte proliferation. If the present findings are at all predictive of comparable phenomena in people, there are clear implications for the treatment of the obese and for those at risk of becoming obese. Adipocytes, once differentiated, appear to remain in the adipose depots. Thus, whenever adipocyte number is increased it remains increased. Furthermore, our data suggest that once a new adipocyte is formed, it tends to store about as much lipid as previously formed adipocytes. That is, a given set of dietary conditions predisposes the rat, and perhaps man as well, to a particular mean adipocyte size, regardless of adipocyte number. Thus, once adipocyte size is established under a given dietary condition (current nutritional status), adipocyte number (a reflection of prior nutritional experience) determines total fat mass, To date, there are no known means, other than surgical removal, to reduce adipocyte number, and such surgery is not a preferred clinical approach to the problem of obesity. Most fat cannot be safely removed, and there is no assurance that surgically removed fat will not regenerate. In the rat, adipose tissue clearly regenerates in at least certain areas (3). Surgical removal of fat might also induce compensatory enlargement of other depots (22), a response that also sometimes occurs in the rat (5). It is not known at present whether adipocyte proliferation can be induced in adult human beings. Indeed, studies of experimentally induced obesity in man, in which the period of overweight was sustained for about 6 mo, failed to find an increase in adipocyte number (28). But, there are at least some people with clear adult onset obesity who have excessive numbers of adipocytes (1, 1l), a finding that suggests that induction of adipocyte production is possible in adult man. Perhaps some people are more predisposed to such induction than others, as some rat strains seem ,ti be more predisposed than others. Until these issues are properly resolved, the possibility that an initial episode of weight gain could establish an undesirable morphological alteration in adipose tissue should be a matter of general concern, We thank MS. Elizabeth Seaquist for her excellent technical assistance. This work was supported in park by National Science Foundation Grant PCM 7649324, National Institutes of Health Grant lRO1 AM 20508-01, the Howard Pack Foundation, and a Future Leaders Award to I. M. Faust from the Nutrition Foundation (Grant 5%). Present address of J. S. Stern: University of California at Davis. Received
10 February
1978; accepted
in final
form
17 April
1978.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
E286
FAUST,
JOHNSON,
STERN,
AND
HIRSCH
REFERENCES 1. ASHWELL, M., M. DURRANT, AND J. S, GARROW, How a “fat cell pool” hypothesis could account for the relationship between adipose tissue cellularity and the age of onset of obesity. Proc. Nutr. Sot. 36: lllA, 1977. 2. DI GIROLAMO, M., AND S. MENDLINGER. Role of fat cell size and number in enlargement of epididymal fat pads in three species. Am. J. Physiol. 221: 859-864, 1971. 3. FAUST, I. M., P, R. JOHNSON, AND J. HIRSCH. Adipose tissue regeneration following lipectomy. Science 197: 391-393, 1977. 4. FAUST, I. M., P. R. JOHNSON, AND J. HIRSCH. Noncompensation of adipose mass in partially lipectomized mice and rats. Am. J. Physiol. 231: 538-544, 1976. 5. FAUST, I. M., P. R. JOHNSON, ~rj~ J. HIRSCH. Surgical removal of adipose tissue alters feeding behavior and the development of obesity in rats. Science 197: 393-396, 1977. 6. FROHMAN, L. A., AND L. L. BERNARDIS. Growth hormone and insulin levels in weanling rats with VMH lesions. Endocrinology 82: 1125-l 132, 1968. 7. GREENWOOD, M. R. C., AND P. R. JOHNSON. Adipose tissue cellularity and its relationship to the development of obesity in females. Nutritional Disorders of American Women, edited by M. Winick. New York: Wiley, 1977, p. 119-135. 8. GREENWOOD, M. R. C., P. R. JOHNSON, AND J. HIRSCH. Relationship of age and cellularity to metabolic activity in C57B mice. Proc. Sot. Exptl. Biol. Med. 133: 944-947, 1970. 9. GREENWOOD, M. R. C., AND J. HIRSCH. Postnatal development of adipocyte cellularity in the normal rat. J. Lipid Res. 15: 474483, 1974. 10. HAGER, A., L. SJ~STROM, B. ARUIDSSON, P. BjijRNTORP, AND U. SMITH. Body fat and adipose tissue cellularity in infants: a longitudinal study. MetaboZism 26: 607-614, 1977. 11. HIRSCH, J,, AND B. BATCHELOR. Adipose tissue cellularity and human obesity. Clin. Endocrinol. Metab. 5: 299-311, 1976. 12. HIRSCH, J., AND E. GALLIAN. Methods for the determination of adipose cell size in man and animals. J. Lipid Res. 9: 110-119, 1968. 13. HIRSCH, J., AND P. W. HAN. Cellularity of rat adipose tissue: effects of growth, starvation, and obesity. J. Lipid Res. 10: 7782, 1969. 14. HIRSCH, J., AND J. L. KNITTLE. The cellularity of obese and nonobese human adipose tissue. Federation PFOC. 29: 1516-1521, 1970. 15. HOLLENBERG, C. H., AND A. VOST. Regulation of DNA synthesis in fat cells and stromal elements from rat adipose tissue. J. Clin. Invest. 47: 2485-2498, 1968. 16. JOHN~~N, P. R., AND J. HIRSCH. Cellularity of rat adipose depots in six strains of genetically obese mice. J. Lipid Res. 13: 2-l 1, 1972.
17. JOHNSON, P. R., J. S, STERN, M, R. C. GREENWOOD, L. M. ZUCKER, AND J. HIRSCH. Effect of early nutrition on adipose cellularity and pancreatic insulin release in the Zucker rat. J. Nutr. 103: 738-743, 1973. 18. JOHNSON, P. R., L. M. ZUCKER, J. A. F. CRUCE, AND J. HIRSCH. Cellularity of adipose depots in the genetically obese Zucker rat. J. Lipid Res. 12: 706-714, 1971. 19, KANAREK, R. B., AND E. HIRSCH. Dietary-induced overeating in experimental animals. Federation Proc. 36: 154-158, 1977. 20. KNITTLE, J. L., AND J. HIRSCH. Effect of early nutrition on the development of rat epididymal fat pads: cellularity and metabolism. J. C&n. Invest. 47: 2091-2098, 1968. 21. KRAL, J. G. Surgical Reduction of Adipose Tissue. Effects of Adipectomy and Intestinal By-Pass on Adipose Tissue Cellularity (Ph.D. Dissertation). Giiteborg, Sweden: University of Gijteborg, 1976. 22. KRAL, J. G. Surgical reduction of adipose tissue hypercellularity in man. Scand. J. Plastic Reconstruction Surg. 9: 140-143, 2975. 23. KRAL, J. G. Surgical reduction of adipose tissue in the male Sprague-Dawley rat. Am. J. Physiol. 231: 1090-1096, 1976. 24. LEMMONIER, D, Effects of age, sex and weight on cellularity of the adipose tissue in mice and rats rendered obese by a high fat diet* J. Clin. Invest. 51: 2907-2915, 1972. 25. MARTIN, J. B., AND L. P. RENAUD. Pulsatile growth hormone secretion: suppression by hypothalamic ventromedial lesions and by long-acting somatostatin. Science 186: 538-540, 1974. 26. PECKHAM, S. C., C. ENTENMAN, AND H. W. CARROLL. The influence of a hypercaloric diet on gross body and adipose tissue composition in the rat. J. Nutr. 77: 187-197, 1962. 27. SALANS, L. B., S. W. CUSHMAN, AND R. E, WEISMANN. Adipose cell size and number in nonobese and obese patients. J. Clin. Invest. 52: 929-941, 1973. 28, SALANS, L. B., E, S. HORTON, AND E. A. H. SIMS. Experimental obesity in man: cellular character of the adipose tissue. J. Clin. Invest. 50: 1005-1011, 1971. 29. SCHEMMEL, R., 0. MICKELSON, AND J. L. GILL. Dietary obesity in rats: body weight and body fat accretion in seven strains of rats. J. Nuts. 100: 1041-1048, 1970. 30. SCLAFANI, A., AND A. N. GORMAN. Effects of age, sex, and prior body weight on the development of dietary obesity in adult rats. Physiol. Behavior 18: 10214026, 1977. 31. STERN, J. S., P. R. JOHNSON, B. R. BATCHELOR, L. M. ZUCKER, AND J. HIRSCH. Pancreatic insulin release and peripheral tissue resistance in Zucker obese rats fed high- and low-carbohydrate diets. Am. J. Physiol. 228: 543-548, 1975. 32. STILES, J. W., A. A. FRANCENDESE, AND J. MASORO. Influence of age on size and number of fat cells in the epididymal depot. Am. J. Physiol. 229: 1561-1568, 1975.
Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (005.189.202.082) on August 24, 2018. Copyright © 1978 American Physiological Society. All rights reserved.
