ORIGINAL ARTICLE

Differences in adipocyte long chain fatty acid uptakein Osborne–Mendel and S5B/Pl rats in response tohigh-fat diets

O Petrescu1, AF Cheema2, X Fan1, MW Bradbury1 and PD Berk1,2

1Division of Liver Diseases, Department of Medicine, Mount Sinai School of Medicine, New York, NY, USA and 2Division ofDigestive and Liver Disease, Department of Medicine, Columbia University College of Physicians and Surgeons, New York,NY, USA

Objective: To determine whether strain differences in adipocyte uptake of long chain fatty acids (LCFAs) contribute todifferences in weight gain by Osborne–Mendel (OM) and S5B/Pl rats (S) fed a high-fat diet (HFD).Subjects: Ninety-four adult (12–14-week old) male OM and S rats.Measurements: Body weight; epididymal fat pad weight; adipocyte size, number, LCFA uptake kinetics; and plasma insulin andleptin during administration of HFD or chow diets (CDs).Results: In both strains, rate of weight gain (RWG) was greater on an HFD than a CD; RWG on an HFD was greater, overall, inOM than S. A significant RWG increase occurred on days 1 and 2 in both strains. It was normalized in S by days 6–9 but persistedat least till day 14 in OM. RWGs were significantly correlated (Po0.001) with the Vmax for saturable adipocyte LCFA uptake(Vmax). In S, an increase in Vmax on day 1 returned to baseline by day 7 and was correlated with both plasma insulin and leptinlevels throughout. In OM, a greater increase in Vmax was evident by day 2, and persisted for at least 14 days, during which bothinsulin and leptin levels remained elevated. Growth in epididymal fat pads on the HFD correlated with body weight, reflectinghypertrophy in OM and both hypertrophy and hyperplasia in S.Conclusions: (a) Changes in Vmax contribute significantly to changes in RWG on HFDs. (b) There are important strain differencesin circulating insulin and leptin responses to an HFD. (c) Both insulin and leptin responses to an HFD are closely correlated withVmax of adipocyte fatty acid uptake in S animals, but suggest early onset of insulin resistance in OM. Thus, differences inhormonal regulation of adipocyte LCFA uptake may underlie the different responses of OM and S to HFD.

International Journal of Obesity (2008) 32, 853–862; doi:10.1038/sj.ijo.0803792; published online 15 January 2008

Keywords: adipose tissue; uptake kinetics; insulin; leptin

Introduction

Obesity is the accumulation of excess fat in the form of

triglycerides (TGs) in adipose tissue. However, since adipo-

cytes are not major sites of long chain fatty acid (LCFA)

synthesis,1 and the insulin resistance associated with obesity

leads to increased lipolysis and cellular fatty acid release,2

this excessive TG accumulation must ultimately reflect, at

least in part, increased LCFA uptake, irrespective of the

kinetics for esterification of LCFA to TG. Uptake occurs via

two discrete processes: regulatable, protein-mediated, facili-

tated transport and passive flip-flop (diffusion) across the

plasma membrane.3–5 We have demonstrated several fold

upregulation of saturable LCFA uptake in adipocytes from

single-gene obesity models such as the Zucker rat,6 ob, db, fat

and tubby mouse,7 and in dietary models of obesity in both

rats and mice.7 A similar upregulation of LCFA uptake occurs

in omental adipocytes from obese patients undergoing

bariatric surgery.8 Upregulation of adipocyte LCFA uptake

precedes onset of overt obesity in weanling Zucker rats,6 and

downregulation of adipocyte LCFA uptake precedes weight

loss during leptin administration to ob mice,9 suggesting that

regulation of adipocyte LCFA uptake is a key control point

for body adiposity. While these data suggest that dysregula-

tion of adipocyte LCFA uptake might contribute to the

pathogenesis of obesity, it is alternatively possible that the

ability of adipocytes to upregulate LCFA uptake duringReceived 31 May 2007; revised 18 October 2007; accepted 19 November

2007; published online 15 January 2008

Correspondence: Professor PD Berk, Department of Medicine, Columbia

Presbyterian Medical Center, Russ Berrie Medical Science Pavillion, 1150

St Nicholas Avenue, Room 412, New York, NY 10032, USA.

E-mail: pb2158@columbia.edu

International Journal of Obesity (2008) 32, 853–862& 2008 Nature Publishing Group All rights reserved 0307-0565/08 $30.00

www.nature.com/ijo

caloric excess represents a buffer that protects hepatocytes,

pancreatic b-cells, cardiac and skeletal muscle from LCFA-

mediated lipotoxicity.10–16

Comparisons of obesity-prone and -resistant rodent strains

are useful in studying the pathogenesis of obesity. C57BL/6J

but not A/J mice,17,18 and Osborne–Mendel (OM) but not

S5B/Pl rats develop obesity when fed high-fat diets

(HFDs).19,20 The reasons for these differences are not fully

understood. We therefore postulated that the differences

between OM and S rats in susceptibility to obesity while

consuming an HFD might reflect differences in regulation of

facilitated LCFA uptake. The studies described below test this

hypothesis.

Methods

Animals

Male Osborne–Mendel (OM) and S5B/Pl (S) rats were

purchased at 6–8 weeks of age from Pennington Biomedical

Research Institute (Baton Rouge, LA, USA). Animals were

maintained in a temperature-controlled facility with a 12 h

light/dark cycle with free access to water and laboratory

chow until the start of the studies. Weights were taken

regularly to determine when the growth rate decreased at

maturity. For OM, this occurred by B12 weeks and in S at

12–14 weeks of age.

Protocols

The impact of an HFD on body weight, adipocyte cell size

and adipocyte uptake of LCFA in OM and S was compared in

three protocols. In protocol 1, 10 12-week-old male rats

of each strain were weighed on day 0 and randomized to

receive either an HFD containing 35% lard (55% of calories

from fat) (Bio-Serv, Frenchtown, NJ, USA) (OMfat, Sfat), or a

chow diet (CD) (Diet no. 5053, 13% of calories from fat; WF

Fisher, Somerville, NJ, USA) (OMchow, Schow). Animals were

weighed at 4 and 8 weeks and killed after 10–11 weeks on

the designated diet. Separate groups of nine OM and six S

animals, obtained subsequently, were designated as controls

(OMcontrol, Scontrol) and were fed the same CD until killed

for study on reaching ages equivalent to those in the four

experimental groups on day 0. At the time of killing,

adipocytes were isolated for cell sizing and [3H]-oleic acid

uptake kinetics were measured using standard methods (see

below). For protocol 2, designed to determine the effects of

time on diet on the rate of weight gain (RWG), two

additional groups of 12-week-old male OM rats (n¼4 per

group; OMfat, OMchow) were given the same HFD or CD as in

protocol 1. Corresponding groups of S received the same

diets Sfat, Schow). Animals were kept on their assigned diet

and weighed weekly for 10–11 weeks. In protocol 3, groups

of OM and S comparable to those in protocol 1 were placed

on the HFD and killed on days 0, 1, 2, 6–9, 14 and 70 for

studies of adipocyte size, LCFA uptake, and insulin and

leptin responses. Animals were weighed 1 day prior to and

just before killing, and RWG (g per day) was estimated as the

24-h increment between weighings. There were no chow-fed

controls in this study.

Ethics statement

All applicable institutional and governmental regulations on

ethical use of animals were followed in this research. All

protocols were approved by the Mt Sinai IACUC Committee.

Adipocyte isolation and measurements

In protocols 1 and 3, epididymal fat pads were removed after

killing, weighed and collected in DMEM. Adipocytes were

isolated by collagenase digestion of weighed samples,3,6

suspended in Krebs–Ringer bicarbonate/Hepes buffer con-

taining 5% bovine albumin21 and counted. The total number

of epididymal adipocytes was calculated from the weights

of the epididymal fat pads in grams and the measured cell

counts per gram of tissue. The distribution of adipocyte

diameters (in mm) in each suspension was determined by

direct microscopy using a graduated eyepiece reticle.22 The

mean cell surface area (mm2) and volume (mm3) of these

spherical cells were calculated from the diameters.22

Fatty acid uptake studies

Long chain fatty acid uptake kinetics by adipocyte suspen-

sions were measured by rapid filtration using tracer amounts

of [3H]-oleic acid (OA).3,6 As previously,3,6–9,23 data for OA

uptake were best fitted by the sum of a saturable plus a non-

saturable function of the unbound OA concentration in the

incubation medium, of the form:

UTð½Ou�Þ ¼ ðVmax � ½Ou�=ðKm þ ½Ou�ÞÞ þ k � ½Ou� ð1Þ

where UT([Ou]) is OA uptake velocity at unbound OA

concentration [Ou] (nM), Vmax (pmol per sec per 50 000 cells)

and Km (nM) are, respectively, the maximal OA uptake

velocity of the saturable uptake process and the unbound OA

concentration at which 1/2 of maximal uptake velocity is

achieved, and k (ml per sec per 50 000 cells) is the rate

constant for non-saturable uptake. Curve fitting and deter-

mination of the kinetic constants of the uptake process were

accomplished as described3 with the Simulation, Analysis

and Modeling II (SAAMII) program.24,25

Other measurements

Whole blood glucose was measured in tail vein samples with

a glucose meter (One-Touch; LifeScan Inc., Milpitas,

CA, USA). Additional blood was collected after killing into

EDTA and promptly separated by centrifugation. Plasma was

stored at �20 1C for leptin and insulin measurements by

RIA (Linco Research, St Louis, MO, USA) and free fatty acids

(FFA) using a commercial kit (Wako Chemicals USA Inc.,

Richmond, VA, USA).

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

854

International Journal of Obesity

Statistical analyses

The key hypothesis being tested in these studies was that

RWG in the rat strains studied would be correlated with the

Vmax for adipocyte LCFA uptake. The choice of statistical

methods, therefore, is focused on those most appropriate to

test this hypothesis. Measurements at each time point within

each treatment group were initially summarized in terms of

mean, standard deviation and standard error, using conven-

tional methods of descriptive statistics.26 Measured Differ-

ences between strains at specific times, or between time

points within a strain, were determined by Student’s two-

tailed t-tests. Relationships between parameters were assessed

by both linear and nonlinear correlations.

The weights recorded in protocol 2 were influenced by

multiple factors, those most important being rat strain,

diet and time on diet. Rigorous analysis, therefore, involved

simultaneous consideration of four variables measured

repeatedly at succeeding times, requiring a novel statistical

approach.27 Specifically, differences in the experimental

growth curves in OM and S rats on both HFD and CD

(protocol 2) were analyzed by a model27 in which the natural

log (ln) of weight at week i was expressed as a function of diet

(b1), strain (b2), a term that changes with i (b3, to capture the

fact that the animals of both strains continue to gain some

weight independent of dietary fat content), a term that

changes with i2 (b4, to capture the fact that the relative

weight gain changes with time), and two interaction terms

(b5 and b6), which were included because the pattern of

weight gain for the high-fat animals (of both strains) differs

from that of the chow animals. The model can also include

terms for differences in the pattern of weight gain between

the two strains, but those were not statistically significant

in the present study. The equation describing the model is

presented in Results.

Growth curve data were fitted to the model by methods of

hierarchical mixed models for repeated measures,27 assum-

ing a first order autoregressive covariance structure (that is,

the correlation between weights decreases as time between

measurements increases) with SAS software (SAS Institute

Inc., Cary, NC, USA), generating estimates for model

parameters, designated b0–b6. The consequent interpreta-

tions are presented in Results. As illustrated there, the model

permits identification and separation of both strain and

dietary factors contributing to differences in weight gain,

and would be widely applicable to other similar studies.

Data are presented as mean±s.e.

Results

Protocol 1: preliminary investigations

Weight gain. At day 0, OMfat and OMchow animals were

significantly heavier than their Sfat and Schow counterparts.

For each strain, the initial weight of the two experimental

groups did not differ from that of the corresponding controls

(OMcontrol and Scontrol). After killing at 10–11 weeks on their

assigned diets, all four experimental groups weighed

significantly more than initial weight (Table 1). Moreover,

OMfat animals were significantly heavier than OMchow, while

the differences in weight between Sfat and Schow were not

significant. Thus, over the 70 days of the study, OMfat rats

gained 55% of their initial weight at an average rate of

3.1 g day–1, while the OMchow group gained 30% of their

initial weight at an average of 1.7 g day–1. Over the same

period, Sfat animals gained 39% of their initial weight at an

average of 1.6 g day–1, while Schow animals gained 28% of

their initial weight at a mean of 1.1 g day–1. There were,

however, appreciable changes in RWGs over time in all

groups. On the basis of weights at 4 and 8 weeks, and at the

time of killing, RWGs in groups OMfat, OMchow, Sfat and

Schow, respectively, averaged 5.0, 2.6, 2.3 and 1.3 g day–1 over

the first 4 weeks of the study, 2.0, 0.9, 1.3 and 1.1 g day–1

during weeks 4–8, and 0.9, 0.9, 0.6 and 0.5 g day–1 from 8

weeks to killing. These changes were explored in more detail

in protocol 2 (Table 1).

Adipocyte cell size. On day 0, adipocyte diameters in

OMcontrol rats were significantly greater than in Scontrols

(Table 1). Thereafter, diameters in both OM and S adipocytes

grew significantly between days 0 and 70 on the HFD, but

only minimally on the CD. However, the increases in

diameter (48 vs 26%) and surface area (113 vs 59%) were

much greater in the OMfat than in the Sfat animals.

Fatty acid uptake. There was no difference in LCFA uptake

kinetics between OMcontrol and Scontrol animals on day 0

(Table 1). Likewise, there were no differences in LCFA

uptake Vmax between days 0 and 70 for OMchow, Schow or Sfat

animals. However, the Vmax for LCFA uptake increased

significantly over this period in the OMfat animals, so that

the value at day 70 in OMfat animals was 31% greater than in

Sfat. Thus, the effect of the HFD in preferentially promoting

weight gain in OM animals is accompanied by a selective

upregulation of saturable LCFA uptake. In this protocol,

when Vmax (pmol per sec per 50 000 cells) was divided by the

surface area (in mm2) of 50 000 cells, expressing the results

as pmol per sec per mm2 of adipocyte surface area (Vmax0),

there was no difference in the values of Vmax0 between the

OMcontrol and Scontrol groups (Table 1). Despite the modest

increase in Vmax after 70 days of the HFD, Vmax0 decreased

significantly in OMfat animals, because the increase in cell

size and surface area was much greater than that in Vmax.

Vmax0 fell to a lesser degree in the OMchow, Sfat and Schow

groups, reflecting the lack of change in Vmax, and smaller

increases in cell size and surface area.

Other measurements. At killing, none of the animals was

hyperglycemic; FFA were higher in the OM (OMfat, OMchow)

than in the corresponding S animals (Sfat, Schow) (395±28 vs

307±21 mM, Po0.05). The HFD produced no difference in

FFA between Sfat and Schow animals, and the increase in FFA

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

855

International Journal of Obesity

to a mean of 425 mM in OMfat vs 365 mM in OMchow was not

significant.

Protocol 2: time course study

Many studies have compared the effects of HFDs on weight

gain in OM and S over a specific period, such as 14 days (for

example, 19, 20). Protocol 1 was designed assuming that

prolonged HFD administration would produce a sustained

increase in RWG and saturable LCFA uptake in OM rats.

However, the results above suggested that the effects of the

HFD were not sustained, but rather decreased with time.

Accordingly, additional groups of 12-week-old male OM and

S animals (n¼4 per group) were given either the HFD or CD

and weighed weekly for 10–11 weeks. The weight curves are

shown in Figure 1a. In all four groups, the RWGs appear

greatest by inspection over the initial 3–4 weeks. From week

4 to approximately week 8, there is a nearly linear RWG,

which is slower than the initial rate for that strain and diet.

Finally, in CD-fed animals, weight gain slows further during

the final 2–3 weeks. These observations are amplified by

analysis of daily RWGs (Figure 1b). The RWGs in the OMchow

and Schow animals were virtually identical during week 1, and

little changed during weeks 1–3. They then fell compared

Tab

le1

Effect

sof

lon

g-t

erm

ad

min

istr

ation

of

ah

igh

-fat

die

ton

bod

yw

eig

ht

an

dad

ipocy

teup

take

of

ole

icaci

din

Osb

orn

eF

Men

del

(OM

)an

dS5B/P

l(S

)ra

tsa,b

Str

ain

Gro

up

Die

tIn

itia

lw

eight

(g)

Finalw

eight

(g)

Cel

ldia

m(m

m)

Cel

lsu

rface

(mm

2)

Vm

axc

Vm

ax/s

urf

ace

d(p

m/s

/mm

2)

10

8K

me

kf

OM C

on

trol

OM

con

trol

Ch

ow

374±

10a

NA

77±

5f

21

000±

1601

5.5

±0.6

70.5

24

±0.0

75

23.1

±9.4

0.0

058

±0.0

030

Exp

eri

men

tal

OM

fat

(n¼

5)

Hig

h-f

at

392±

12b

606±

19c,

d114±

14g

44

781±

5042

8.5

±0.2

7h

,i0.3

80

±0.0

44

7.4

±2

0.0

064

±0.0

017

OM

ch

ow

(n¼

5)

Ch

ow

509±

12c

83±

622

911±

1636

6.5

±0.1

60.5

67

±0.0

43

21±

2.2

0.0

046

±0.0

011

S5

B/P

l

Con

trol

Scon

trol

Ch

ow

308±

11a

NA

66±

114

792±

478

6.6

±1.2

0.8

92

±0.2

07

35±

17.9

0.0

054

±0.0

024

Exp

eri

men

tal

Sfa

t(n

¼5)

Hig

h-f

at

290±

5b

404±

7c,

e83±

7g

23

538±

3284

6.3

±0.1

30.5

35

±0.0

75

15±

1.7

0.0

13±

0.0

008

Sch

ow

(n¼

5)

Ch

ow

370±

14c

76±

219

197±

1306

6.1

±0.2

20.6

36

±0.0

49

22±

0.5

0.0

03±

0.0

009

Ab

bre

viati

on

s:N

A,

not

ap

plic

ab

le;

NS,

not

sig

nific

an

t.aA

llva

lues

are

mean

±s.

e.m

.bD

ata

for

con

trolg

roup

sare

from

day

0.

Th

ose

from

exp

eri

men

tal

gro

up

sw

ere

ob

tain

ed

at

killi

ng

at

day

70.

Un

its:

cV

max¼

pm

olp

er

sec

per

50

000

cells

;dV

max/s

urf

ace

are

a¼

pm

olp

er

sec

per

mic

ron

2;

eK

m¼

nM

;f k¼

mlp

er

sec

per

50

000

cells

.Sta

tist

icalco

mp

ari

son

s:a:P¼

NS

vsco

rresp

on

din

gexp

eri

men

talg

roup

;b

:

Po

0.0

01

vsS5BP/l

exp

eri

men

tal

gro

up

;c:

Po

0.0

01

vsin

itia

lw

eig

ht

of

corr

esp

on

din

gexp

eri

men

tal

gro

up

;d

:Po

0.0

05

vsO

Mch

ow;

e:

0.1

4P4

0.0

5vs

Sch

ow

;f:

Po

0.0

5vs

Scon

trol;

g:

Po

0.0

5vs

corr

esp

on

din

gco

ntr

olg

roup

;h

:Po

0.0

05

vsO

Mcon

trol;

i:Po

0.0

1vs

Sfa

t.

650

600

550

500

450

400

350

300

250

Wei

ght (

g)

OM Hi-FatS5B/Pl Hi-FatOM ChowS5B/Pl Chow

OM Hi-FatS5B/Pl Hi-FatOM ChowS5B/Pl Chow

0 1 2 3 4 5 6 7 8 9 10Weeks

8

7

6

5

4

3

2

1

0Rat

e of

Wei

ght G

ain

(g/d

ay)

to1 2 3 4 5 6 7 10Weeks on High Fat Diet

Figure 1 (a) Weights of adult male Osborne–Mendel (OM) and S5B/Pl

rats fed either standard laboratory chow or a high-fat diet for 10–11 weeks

(protocol 2). Results are mean±s.e. (n¼4 per group). (b) Average rates of

weight gain (g per day) in fat- and chow-fed OM and S5B/Pl rats in protocol 2,

during each week of the study. Results are mean±s.e.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

856

International Journal of Obesity

with baseline in both strains during week 4 and remained

below baseline thereafter. In contrast, RWG in OMfat rats was

greatest during week 1 (6.2±0.6 g day–1), which is signifi-

cantly greater than the basal rate preceding onset of the

study (1.6±0.5 g day–1, Po0.001). It then declined over the

next 4 weeks to 1.4±0.5 g day–1 at week 5, a value similar to

baseline, which persisted thereafter. While the peak value

(4.0±0.3 g day–1) in Sfat animals, also observed during week

1, was significantly less than that in OM (Po0.025), and

RWG in OMfat consistently exceeded that in Sfat, the pattern

over time was similar (Figure 1b): increases in RWG in both

OMfat and Sfat animals were greatest during weeks 1–3 and

least during week 9 and thereafter (Figure 1).

As noted above, the weights recorded in Figure 1a were

influenced by multiple factors, of which those most

important to this study were strain, diet and time on diet.

To examine simultaneously the influence of each of these

factors on the observed weights, the data were analyzed by a

model in which the natural logarithm of weight at week i is

given by:

Inðweight � iÞ ¼b0 þ b1ðHFÞ þ b2ðOMÞ þ b3ðiÞ þ b4ði2Þþ b5ðHF�iÞ þ b6ðHF�i2Þ þ error

ð2Þwhere HF is assigned a value of 1 for HFD, 0 for CD. Similarly,

OM¼1 for Osborne–Mendel rats, 0 for S5B rats; HF*i¼ i for

HFD, 0 for CD; and HF*i2¼ i2 for HFD, 0 for CD. Data were

fitted to Equation (2) as described in Methods. The solution

yielded numerical values for parameters b0–b6, giving

the following results: b0¼5.70 (Po0.0001), b1¼0.065

(P¼0.077), b2¼0.23 (Po0.0001), b3¼0.04 (Po0.0001),

b4¼�0.0019 (Po0.0001), b5¼0.023 (P¼0.0001), and

b6¼�0.0012 (P¼0.029). The meaning of these results is

most readily illustrated by examples. Substituting 1 or 0, as

appropriate, for HF and OM and using the estimated

b-coefficients, the expected ln of the weight of an HFD-fed

OM rat at week 3 is estimated as:

LnðWeight � 3Þ ¼5:70 þ 0:065 þ 0:23 þ 0:04ð3Þ � 0:0019ð9Þþ 0:023ð3Þ � 0:0012ð9Þ ¼ 6:156;

and Weight � 3 ¼ expð6:156Þ ¼ 471:6

The relative weight gain from one week to the next can be

calculated by subtracting the ln of weight at week i from that

at week (iþ1) and exponentiating. This involves dividing

weight at week (iþ1) by that at week i. The relative change

in weight from week i to week (iþ1) is:

High Fat :1:066 expf�0:00308ð2i þ 1Þgand Chow : 1:042 expf�0:00193ð2i þ 1Þg

From this, we get the following model estimates of relative

changes in body weight over time on the two diets: day 0 to

week 1: HFD: 1.066 (that is, þ6.6%) vs CD:1.042 (þ4.2%);

week 3–4: HFD: 1.043 (þ4.3%) vs CD: 1.029 (þ2.9%); and

week 9–10: HFD: 1.005 (þ0.5%) vs CD: 1.004 (þ0.4%).

These model-dependent results support experimental

observations that (1) initial weights of OM rats were

significantly higher than those of the S rats of the same

age (b2), and (2) the HFD-fed rats of both strains started out

marginally heavier than the Chow (b1). (3) Weights of both

strains increased significantly with time (b3 for S), but (4) the

rate of increase from week to week was maximal during week

1 and thereafter slowed down with time (b4 for S). (5) The fat-

fed rats of both strains gained weight faster than the chow

(b5 for the HFD excess), but (6) their weekly rate of increase

diminished faster (b6 for the HFD excess fall-off). A more

detailed explanation of the approach will be found in Brown

and Prescott.27

Protocol 3: short-term response study

Protocols 1 and 2 suggested that the RWG produced by an

HFD in both OM and S animals was greatest soon after

introduction of the diet. Since protocol 1 studies were

conducted mainly during week 10, they might have missed

key strain differences during the initial diet period. This was

addressed in protocol 3, which tested our basic hypothesis by

examining RWG, LCFA uptake kinetics, and plasma levels

of insulin and leptinFboth are putative regulators of LCFA

uptakeFat frequent intervals over the first 14 days and on

day 70 (Figures 2–5).

Weight gain. Individual S rats gained 4–6 g on day 1, but the

average, which was 3.2±0.2 g day–1 prior to starting the

HFD, was stable at B3.0 g day–1 over the initial 14 days

(Figure 2). In contrast, mean RWG in OM animals increased

to 5.5 g day–1 over the initial 14 days of fat feeding, although

with much day-to-day variability. Thus, RWG in OM peaked

at 8.9±0.7 g day–1 on day 2, with individual rats gaining

p10.5 g on that day. It then fell by 50% by days 6–9 and to

baseline on day 70 (Figure 2). All values except day 70 were

significantly greater than baseline.

12

10

8

6

4

2

0

Rat

e of

Wei

ght G

ain

(g/d

ay)

Days on Diet

OM

S5B/PI

0 1 2 6-9 14 70

Figure 2 Rates of weight gain (g per day) in fat-fed Osborne–Mendel (OM)

and S5B/Pl rats on specific days during protocol 3. Data represent differences

in weight measurements made on the day stipulated and 24 h earlier. Results

are mean±s.e. Values in OM rats on days 1 (Po0.025), 2 (Po0.001), 6–9

(Po0.005) and 14 (Po0.01) are all significantly greater than those at baseline.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

857

International Journal of Obesity

Fatty acid uptake. The Vmax for OA uptake on day 0 was

similar in OM and S animals (5.5 vs 6.6 pmol per sec per

50 000 cells, respectively). The rate of OA uptake rose briefly

in S on day 1 (14±2 pmol per sec per 50 000 cells; Po0.05 vs

day 0), but then declined by nearly one-third by day 2 and

to baseline by days 6–9 (Figure 3a). In contrast, Vmax in OM

rats was unchanged on day 1, but increased fourfold, to

21.5±4.6 pmol per sec per 50 000 cells, on day 2 (Po0.025 vs

day 0) before declining by 30% on days 6–9 and thereafter to

baseline. Vmax0 also peaked within 24 h in S and within

48 h in OM rats, before drifting downward to, or below

baseline, by day 14 (Figure 3b). The increased Vmax0 strongly

suggests upregulation of a transport system,8,9 while the

subsequent downslope of Vmax0 over time presumably

reflects both a decline in membrane transport activity and

Figure 5 Plasma hormone responses to an HFD in Osborne–Mendel (OM)

and S5B/Pl rats in protocol 3. (a) Insulin. (b) Leptin. Results are mean±s.e.

Both insulin (Po0.005) and leptin (Po0.05) concentrations were significantly

increased in S5B/Pl rats on days 1 and 2 of the HFD, but had declined to

baseline by days 6–9. In contrast, in OM rats, although leptin was increased by

day 1 and remained significantly increased on days 2–14 (Po0.001), insulin

levels did not increase until day 2. Thereafter, they, too, remained significantly

increased through day 14 (Po0.01).

OMS5B/Pl

OMS5B/Pl

30

25

20

15

10

5

0

Vm

ax (

pmol

/sec

/50,

000

cells

)

0 1 2 6-9 14 70

Days on High Fat Diet

3.5

3

2.5

2

1.5

1

0.5

00 1 2 6-9 14 70

Days on High Fat Diet

Vm

ax/S

A (

pmol

/sec

/µm

2 x

108 )

Figure 3 Saturable oleic acid uptake by epididymal adipocytes from fat-fed

Osborne–Mendel (OM) and S5B/Pl rats on specific days during protocol 3. (a)

Saturable uptake capacity as measured by the Vmax for oleic acid uptake (pmol

per sec per 50 000 cells). In S5B/Pl animals, Vmax is significantly increased only

on day 1. In OM rats, Vmax is significantly greater (Po0.05 or better) on days 2

and 6–9 than on day 0. (b) Saturable uptake as measured by the Vmax0 for oleic

acid uptake (pmol per sec per mm2 of adipocyte surface area). Increases in

Vmax0 are more specifically indicative of increased activity of one or more

transport processes. In both OM and S rats, the observed pattern is similar to

that illustrated for Vmax in panel a. Results are mean±s.e.

Figure 4 Relationship of Vmax to the rate of weight gain in the combined

series fat-fed Osborne–Mendel (OM) and S5B/Pl rats in protocol 3. The

correlations between these variables were significant in all rats of both strains

combined (r¼0.57, Po0.001), as illustrated, and in both the OM animals

(r¼0.59, Po0.001) and S animals (r¼ 0.45, Po0.05) separately.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

858

International Journal of Obesity

a progressive increase in cell surface area. These results are

in contrast to those in protocol 1, in which Vmax0 in OM

declined on day 70 at which a small increase in Vmax was

overshadowed by a much larger increase in surface area.

Vmax was linearly correlated with RWG throughout the

study (Figure 4).

Insulin and leptin responses. Both plasma insulin and leptin

levels were similar in the two strains on day 0 (Figure 5). In S,

insulin subsequently doubled on day 1, declined somewhat

on day 2 and returned to baseline by days 6–9 (Figure 5a);

leptin was significantly elevated on days 1 and 2 (Po0.001)

and declined thereafter, although it still exceeded baseline

on day 14 (Po0.025) (Figure 5b). The parallel rise and

subsequent fall in Vmax, insulin and leptin concentrations in

S animals (Figures 3 and 5) resulted in highly significant

linear correlations between Vmax and insulin, and Vmax and

leptin during days 0–2 (insulin: r¼0.46, Po0.05; leptin:

r¼0.64, Po0.01), days 2–14 (insulin: r¼0.73, Po0.01;

leptin: r¼0.79; Po0.01) and over the entire period from

day 0 through day 14 (insulin: r¼0.52, Po0.01; leptin

r¼0.69, Po0.01). In contrast, in OM animals, Vmax, insulin

and leptin concentrations moved roughly in parallel only

during days 0–2. Both insulin and Vmax were essentially

unchanged on day 1, before increasing strikingly on day 2.

Thereafter Vmax declined toward baseline, while both insulin

and leptin remained elevated (Figures 3 and 5). Conse-

quently, correlations in OM between Vmax and insulin

concentration, and Vmax and leptin from day 0 through

day 2 were both statistically significant (r¼0.53, Po0.05 and

r¼0.75, Po0.01, respectively). Subsequently, from days

2–14, correlations between Vmax and insulin (r¼0.28), and

Vmax and leptin (r¼0.08) were no longer significant.

Adipocyte measurements. On day 0, combined weights of

the two epididymal fat pads were similar in OM and S

(Figure 6a), but consistently more cells were recovered in

OM (12.7±2.5106) than S (9.1±1.0106). After the

initiation of the HFD, cell counts remained constant over

the next 14 days, but adipocytes increased in size (data not

shown), and fat pad weights increased in both strains

(Figure 6a) and were highly correlated with changes in body

weight (r¼0.9790, Po0.001) (Figure 6b) and with the total

volume of adipocytes (the product of the number of isolated

cells and their mean cell volume) (r¼0.838, Po0.001).

While the numbers of cells recovered during days 0–14

remained unchanged in both strains, the weight of the fat

pads increased by 49% from baseline in the S group

(Po0.005) but nearly threefold in OM (Po0.001) during

this period (Figure 6a). From days 14–70, epididymal fat-pad

growth remained much greater in OM than S (Figure 6a). The

number of cells recovered in OM between days 14 and 70

remained essentially constant (12.1±1.7106), but recovery

of fat-pad cells in S increased 26% to 11.5±2.8106. While

adipocyte numbers were thus equivalent at day 70, the

average volume of an OM adipocyte, 812 pl, was 2.75 times

that in S (295 pl).

Discussion

Although obesity can be defined as the excessive accumula-

tion of LCFA in the form of TG within adipocytes, the

importance of fatty acid disposition to obesity and its

comorbidities, including the entire metabolic syndrome,

was long underestimated. This largely reflected the long-held

view that the movement of LCFA into and out of cells was an

entirely passive, unregulatable and therefore uninteresting

process. Multiple laboratories have now shown that LCFAs

enter cells by both a facilitated, regulatable, protein-

mediated transport process and diffusion.3–5,28–39 At the

unbound LCFA concentrations that commonly exist in most

mammals, X90% of LCFA uptake into key tissues occurs by

Figure 6 (a) Weights of epididymal fat pads over time in fat-fed Osborne–

Mendel (OM) and S5B/Pl rats in protocol 3. Results are mean±s.e. (b)

Correlation between body weight and total epididymal fat pad weights over

time in fat-fed OM and S5B/Pl rats in protocol 3. The illustrated correlation

with a third-order polynomial has a correlation coefficient of r¼0.9790

(Po0.001). It provides a visually better fit than a simple linear regression,

which nevertheless has a highly significant correlation coefficient of r¼ 0.876

(Po0.001).

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

859

International Journal of Obesity

the facilitated process.23,37–40 At least four proteins have

been proposed as LCFA transporters: plasma membrane fatty

acid-binding protein (FABPpm),32,37–39 fatty acid translocase

(FAT),30 the fatty acid transport polypeptide family

(FATP)41–43 and caveolin-1.44,45 While there is good evidence

supporting a role in LCFA transport for these proteins, there

are also unresolved issues that have led to persistent doubts

about each. While some or all of them probably do

participate in LCFA uptake in different tissues under

different conditions, it would not be surprising if entirely

new classes of LCFA transporters were discovered. Although

the nature of LCFA transporters is still debated, the existence

of LCFA uptake mechanisms besides diffusion seems well

established.

Osborne–Mendel and S rats are known to gain weight at

different rates on an HFD.19,20 The central hypothesis tested

in the current studies is that differences in the regulation of

saturable adipocyte LCFA uptake could explain these

differences in RWGs. The core data of this study suggest

that changes in the Vmax for adipocyte fatty acid uptake,

which in other settings has been shown to be tightly

correlated with expression of one or more putative LCFA

transporters,31,46 contribute significantly to changes in the

RWG in response to an HFD. It remains to explain what

underlies the observed differences in Vmax. More preliminary

data suggest (a) that differences in circulating insulin and

leptin responses to an HFD may explain the strain differ-

ences in saturable adipocyte LCFA uptake; and (b) that

different mechanisms of adipose tissue growth may con-

tribute to the differences in hormonal responses. Both issues

require more definitive study.

On the basis of studies in ob mice,7,9 which lack leptin,47

and Zucker fatty rats,6 with a defective leptin receptor,48,49

we proposed that insulin was an upregulator and leptin a

downregulator of adipocyte LCFA uptake.6,7,9 The former

suggestion was supported by strong correlations between

plasma insulin and the Vmax for adipocyte LCFA uptake in

obese and nonobese humans.8 While leptin is an insulin

counter-regulatory hormone in some settings,50 in others its

effects parallel those of insulin. Its precise role in LCFA

uptake is unclear.

In the present studies, Vmax and circulating insulin and

leptin in S animals rose and fell in parallel during the first 2

weeks of HFD feeding; Vmax was highly correlated with each

of the hormones. In contrast, in OM rats, insulin and leptin

concentrations correlated significantly with Vmax only for

the first 2 days, after which elevated insulin and leptin levels

persisted as Vmax returned to baseline. This suggests the

development of insulin resistance, and possibly also of leptin

resistance, in OM animals, within the first week of HFD

administration. It is also noteworthy that, in contrast to

their parallel increase in S animals on day 1 of the HFD, Vmax

and plasma insulin concentration in OM animals were both

unchanged on day 1, but increased together 24 h later. Thus,

we speculate that differences in hormonal regulation of

adipocyte LCFA uptake may underlie the different body

weight responses of OM and S rats to an HFD. Moreover,

these differences in hormonal responses, themselves, may be

related to different mechanisms of fat pad growth. While

adipocytes increased in size in both strains, the constancy of

adipocyte numbers in OM animals suggests that the increase

in fat pad weight was largely a consequence of cellular

hypertrophy, while the late increase in adipocytes in S

animals suggested, in addition, an appreciable degree of

hyperplasia. The resulting differences in adipocyte size

between strains could contribute to differences in produc-

tion of leptin and other adipokines.51

Another important observation is that responses to the

initiation of an HFD are not fixed, but rather evolve over

time despite continuation of the diet. Most papers on HFD

responses in these strains report data at a single time after

starting the diet. Our data suggest that the responses

observed depend critically on time at which measurements

are made, and that regulatory responses may obliterate

important early effects by day 14.

Hormonal responses to HFDs have been studied exten-

sively in these strains, although their effects were thought

to be mediated mainly by changes in caloric intake or

energy expenditure. Our hypothesis, supported by the

current studies, is that these responses impact on adiposity

by altering LCFA partitioning,8 with OM rats storing an

increased proportion of available LCFA as TG in adipose

tissue. While many studies of the differences between OM

and S rats during fat feeding focused on leptin18,52–55 and

other hormones, neuropeptides and uncoupling proteins

(for example, 52, 55), they have not yet led to a clear

explanation for the differences in HFD responses.

We have proposed that the diverse neuroendocrine

signaling resulting from food-related stimuli converge on a

final common pathway that regulates adipocyte LCFA

uptake and sequestration, which we believe to be a major

control point for body adiposity.8 The present studies are

consistent with, but certainly do not prove, that hypothesis.

Detailed studies of the role of hormones and fat depot

growth in regulating LCFA uptake, of the transporters and

genes involved, and of changes in gene expression associated

with HFD responses, are planned.

Acknowledgements

This work was supported, in part, by Grants DK-23718,

DK-52401 and DK-72526 from the National Institute of

Diabetes, Digestive and Kidney Diseases of the National

Institutes of Health. Dr Carol Bodian developed the mixed

model used in our analysis of experimental growth curves.

References

1 Shrago E, Spennetta A, Gordon E. Fatty acid synthesis in humanadipose tissue. J Biol Chem 1969; 244: 2761–2766.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

860

International Journal of Obesity

2 Delarue J, Magnan C. Free fatty acids and insulin resistance.Curr Opin Clin Nutr Metab Care 2007; 10: 142–148.

3 Stump DD, Fan X, Berk PD. Oleic acid uptake and binding by ratadipocytes define dual pathways for cellular fatty acid uptake.J Lipid Res 2001; 42: 509–520.

4 Kampf JP, Kleinfeld AM. Fatty acid transport in adipocytesmonitored by imaging intracellular free fatty acid levels. J BiolChem 2004; 279: 35775–35780.

5 Kleinfeld AM, Kampf JP, Lechene C. Transport of 13C-oleate inadipocytes measured using multi imaging mass spectrometry.J Am Soc Mass Spectrom 2004; 11: 1572–1580.

6 Berk PD, Zhou SL, Kiang CL, Stump D, Bradbury M, Isola LM.Uptake of long chain free fatty acids is selectively up-regulated inadipocytes of Zucker rats with genetic obesity and non-insulin-dependent diabetes mellitus. J Biol Chem 1997; 272: 8830–8835.

7 Berk PD, Zhou S, Kiang C, Stump DD, Fan X, Bradbury MW.Selective up-regulation of fatty acid uptake by adipocytescharacterizes both genetic and diet-induced obesity in rodents.J Biol Chem 1999; 274: 28626–28631.

8 Petrescu O, Fan X, Gentileschi P, Hossain S, Bradbury M, GagnerM et al. Long-chain fatty acid uptake is upregulated in omentaladipocytes from patients undergoing bariatric surgery for obesity.Int J Obes (Lond) 2005; 29: 196–203.

9 Fan X, Bradbury MW, Berk PD. Leptin and insulin modulatenutrient partitioning and weight loss in ob/ob mice throughregulation of long-chain fatty acid uptake by adipocytes. J Nutr2003; 133: 2707–2715.

10 Unger RH, Zhou YT. Lipotoxicity of beta-cells in obesity and inother causes of fatty acid spillover. Diabetes 2001; 50 (Suppl 1):S118–S121.

11 Listenberger LL, Ory DS, Schaffer JE. Palmitate-induced apoptosiscan occur through a ceramide-independent pathway. J Biol Chem2001; 276: 14890–14895.

12 Unger RH. Lipotoxic diseases. Annu Rev Med 2002; 53: 319–336.13 Unger RH. Minireview: weapons of lean body mass destruction:

the role of ectopic lipids in the metabolic syndrome. Endocrinology2003; 144: 5159–5165.

14 Borradaile NM, Schaffer JE. Lipotoxicity in the heart. CurrHypertens Rep 2005; 7: 412–417.

15 Lee Y, Naseem RH, Park BH, Garry DJ, Richardson JA, Schaffer JEet al. Alpha-lipoic acid prevents lipotoxic cardiomyopathy in acylCoA-synthase transgenic mice. Biochem Biophys Res Commun2006; 344: 446–452.

16 Borradaile NM, Han X, Harp JD, Gale SE, Ory DS, Schaffer JE.Disruption of endoplasmic reticulum structure and integrity inlipotoxic cell death. J Lipid Res 2006; 47: 2726–2737.

17 Surwit RS, Feinglos MN, Rodin J, Sutherland A, Petro AE, OparaEC et al. Differential effects of fat and sucrose on the developmentof obesity and diabetes in C57BL/6J and A/J mice. Metabolism1995; 44: 645–651.

18 Bergen HT, Mizuno T, Taylor J, Mobbs CV. Resistance todiet-induced obesity is associated with increased proopio-melanocortin mRNA and decreased neuropeptide Y mRNA inthe hypothalamus. Brain Res 1999; 851: 198–203.

19 Osborne TB, Mendel LB. The growth of rats upon diets of isolatedfood substances. Biochem J 1916; 10: 534–538.

20 Schemmel R, Mickelsen O, Gill JL. Dietary obesity in rats: bodyweight and body fat accretion in seven strains of rats. J Nutr 1970;100: 1041–1048.

21 Cushman SW, Wardzala LJ. Potential mechanism of insulinaction on glucose transport in the isolated rat adipose cell.Apparent translocation of intracellular transport systems to theplasma membrane. J Biol Chem 1980; 255: 4758–4762.

22 Di Girolamo M, Mendlinger S, Fertig JW. A simple method todetermine fat cell size and number in four mammalian species.Am J Physiol 1971; 221: 850–858.

23 Schwieterman W, Sorrentino D, Potter BJ, Rand J, Kiang CL,Stump D et al. Uptake of oleate by isolated rat adipocytes ismediated by a 40-kDa plasma membrane fatty acid binding

protein closely related to that in liver and gut. Proc Natl Acad SciUSA 1988; 85: 359–363.

24 Berman M, Weiss MF. User’s Manual for SAAM. US GovernmentPrinting Office: Washington, DC, 1967.

25 SAAM II User’s Guide. SAAM Institute: Seattle, WA, 1997, 1998.26 Snedecor GW, Cochran WG. Statistical Methods. Iowa State

University Press: Ames Iowa, 1967.27 Brown H, Prescott R. Applied Mixed Models in Medicine. John Wiley

& Sons Ltd: Chichester, 1999.28 Abumrad NA, Perkins RC, Park JH, Park CR. Mechanism of long

chain fatty acid permeation in the isolated adipocyte. J Biol Chem1981; 256: 9183–9191.

29 Abumrad NA, Park JH, Park CR. Permeation of long-chain fattyacid into adipocytes. Kinetics, specificity, and evidence forinvolvement of a membrane protein. J Biol Chem 1984; 259:8945–8953.

30 Abumrad NA, el-Maghrabi MR, Amri EZ, Lopez E, Grimaldi PA.Cloning of a rat adipocyte membrane protein implicated inbinding or transport of long-chain fatty acids that is inducedduring preadipocyte differentiation. Homology with humanCD36. J Biol Chem 1993; 268: 17665–17668.

31 Berk PD, Stump DD. Mechanisms of cellular uptake of long chainfree fatty acids. Mol Cell Biochem 1999; 192: 17–31.

32 Bradbury MW, Berk PD. Cellular uptake of long chain free fattyacids: the structure and function of plasma membrane fatty acidbinding protein. Adv Mol Cell Biol 2004; 33: 47–81.

33 Cupp D, Kampf JP, Kleinfeld AM. Fatty acid–albumin complexesand the determination of the transport of long chain free fattyacids across membranes. Biochemistry 2004; 43: 4473–4481.

34 Glatz JF, van Nieuwenhoven FA, Luiken JJ, Schaap FG, van derVusse GJ. Role of membrane-associated and cytoplasmic fattyacid-binding proteins in cellular fatty acid metabolism. Prosta-glandins Leukot Essent Fatty Acids 1997; 57: 373–378.

35 Luiken JJ, Turcotte LP, Bonen A. Protein-mediated palmitateuptake and expression of fatty acid transport proteins in heartgiant vesicles. J Lipid Res 1999; 40: 1007–1016.

36 Luiken JJ, Glatz JF, Bonen A. Fatty acid transport proteinsfacilitate fatty acid uptake in skeletal muscle. Can J Appl Physiol2000; 25: 333–352.

37 Stremmel W, Berk PD. Hepatocellular influx of [14C]oleate reflectsmembrane transport rather than intracellular metabolism orbinding. Proc Natl Acad Sci USA 1986; 83: 3086–3090.

38 Stremmel W, Strohmeyer G, Berk PD. Hepatocellular uptake ofoleate is energy dependent, sodium linked, and inhibited by anantibody to a hepatocyte plasma membrane fatty acid bindingprotein. Proc Natl Acad Sci USA 1986; 83: 3584–3588.

39 Sorrentino D, Stump D, Potter BJ, Robinson RB, White R, KiangCL et al. Oleate uptake by cardiac myocytes is carrier mediatedand involves a 40-kD plasma membrane fatty acid bindingprotein similar to that in liver, adipose tissue, and gut. J ClinInvest 1988; 82: 928–935.

40 Sorrentino D, Robinson RB, Kiang CL, Berk PD. At physiologicalbumin/oleate concentrations oleate uptake by isolatedhepatocytes, cardiac myocytes, and adipocytes is a saturablefunction of the unbound oleate concentration. Uptake kineticsare consistent with the conventional theory. J Clin Invest 1989;84: 1325–1333.

41 Schaffer JE, Lodish HF. Expression cloning and characterization ofa novel adipocyte long chain fatty acid transport protein. Cell1994; 79: 427–436.

42 Hirsch D, Stahl A, Lodish HF. A family of fatty acid transportersconserved from mycobacterium to man. Proc Natl Acad Sci USA1998; 95: 8625–8629.

43 Stahl A, Gimeno RE, Tartaglia LA, Lodish HF. Fatty acid transportproteins: a current view of a growing family. Trends EndocrinolMetab 2001; 12: 266–273.

44 Pohl J, Ring A, Stremmel W. Uptake of long-chain fatty acids inHepG2 cells involves caveolae: analysis of a novel pathway. J LipidRes 2002; 43: 1390–1399.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

861

International Journal of Obesity

45 Pohl J, Ring A, Ehehalt R, Herrmann T, Stremmel W. Newconcepts of cellular fatty acid uptake: role of fatty acid transportproteins and of caveolae. Proc Nutr Soc 2004; 63: 259–262.

46 Berk PD, Zhou SL, Bradbury M, Stump D, Kiang CL, Isola LM.Regulated membrane transport of free fatty acids in adipocytes:role in obesity and non-insulin dependent diabetes mellitus.Trans Am Clin Climatol Assoc 1996; 108: 26–43.

47 Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, FriedmanJM. Positional cloning of the mouse obese gene and its humanhomologue. Nature 1994; 372: 425–432.

48 Lee GH, Proenca R, Montez JM, Carroll KM, Darvishzadeh JG,Lee JI et al. Abnormal splicing of the leptin receptor in diabeticmice. Nature 1996; 379: 632–635.

49 Chua Jr SC, Chung WK, Wu-Peng XS, Zhang Y, Liu SM, Tartaglia Let al. Phenotypes of mouse diabetes and rat fatty due tomutations in the OB (leptin) receptor. Science 1996; 271: 994–996.

50 Remesar X, Rafecas I, Fernandez-Lopez JA, Alemany M. Is leptinan insulin counter-regulatory hormone? FEBS Lett 1997; 402:9–11.

51 Skurk T, Alberti-Huber C, Herder C, Hauner H. Relationshipbetween adipocyte size and adipokine expression and secretion.J Clin Endocrinol Metab 2007; 92: 1023–1033.

52 Schaffhauser AO, Madiehe AM, Braymer HD, Bray GA, York DA.Effects of a high-fat diet and strain on hypothalamic geneexpression in rats. Obes Res 2002; 10: 1188–1196.

53 Madiehe AM, Schaffhauser AO, Braymer DH, Bray GA,York DA. Differential expression of leptin receptor in high- andlow-fat-fed Osborne–Mendel and S5B/Pl rats. Obes Res 2000; 8:467–474.

54 Lin X, Chavez MR, Bruch RC, Kilroy GE, Simmons LA, Lin Let al. The effects of a high fat diet on leptin mRNA, serum leptinand the response to leptin are not altered in a rat strainsusceptible to high fat diet-induced obesity. J Nutr 1998; 128:1606–1613.

55 York DA, Singer L, Thomas S, Bray GA. Effect of topiramate onbody weight and body composition of Osborne–Mendel rats fed ahigh-fat diet: alterations in hormones, neuropeptide, anduncoupling-protein mRNAs. Nutrition 2000; 16: 967–975.

Adipocyte fatty acid uptake in fat-fed ratsO Petrescu et al

862

International Journal of Obesity