Altered fat differentiation and adipocytokine expression are inter-related and linked to morphological changes and insulin resistance in HIV-1-infected lipodystrophic patients.

Abstract

To achieve a better understand of the pathophysiology of HIV-related lipoatrophy, we compared the mRNA expression of adipocytokines in fat samples from patients and healthy HIV-seronegative controls together with fat morphology and we studied the relationship between changes in fat morphology, adipocytokine expression, markers of adipose tissue differentiation and whole body insulin sensitivity.
Cross-sectional analytical study.
The mRNA expression of adipocytokines and transcriptional factors in fat samples from 26 patients with peripheral lipoatrophy (all under anti-retroviral therapy associating protease inhibitor and nucleoside-analogue reverse transcriptase inhibitors) and from 16 non-HIV-infected controls was measured by real time quantitative RT-PCR. Fat morphology was assessed histologically on a subgroup of 10 patients and six controls: collagen fibres by Sirius Red staining, apoptosis by the TUNEL technique, vessels by smooth muscle alpha-actin staining and macrophages by CD68 staining. Insulin resistance was assessed by using the homeostasis model assessment.
The patients' fat showed higher values of apoptosis (P=0.005), fibrosis (P<0.05), vessel density (P=0.001) and macrophage infiltration (P<0.05) than the controls' fat, together with lower adiponectin and leptin mRNA levels and higher interleukin (IL)-6 and tumour necrosis factor (TNF)alpha mRNA levels. TNFa and IL-6 expression correlated positively with the level of apoptosis (P=0.05 and P<0.05, respectively) and negatively with CCAAT-enhancer binding protein (C/EBP)alpha (P<0.001 and P<0.05, respectively). Apoptosis correlated negatively with the expression level of sterol-regulatory-element-binding-protein-1c (SREBP1c) (P=0.01) and C/EBPalpha (P=0.01) whilst the vessel density correlated negatively with SREBP1c (P<0.005), C/EBPalpha (P=0.001) and beta (P=0.001). Adiponectin and leptin expression correlated positively with each other, and also with adipogenic marker expression and overall insulin sensitivity. These relationships were also present when the patient group was studied separately. Finally, fat morphological abnormalities correlated positively with whole body insulin resistance.
Adipose tissue from patients with HIV-1-related lipoatrophy shows increased apoptosis, together with decreased adipocyte differentiation. Increased TNFalpha and IL-6 expression could be a major phenomenon linking these alterations. Decreased adiponectin and leptin expression, which may result from decreased adipocyte differentiation, could be involved in the observed whole body insulin resistance.

Full text

Page 1

Objective: To achieve a better understand of the patho-
physiology of HIV-related lipoatrophy, we compared the
mRNA expression of adipocytokines in fat samples from
patients and healthy HIV-seronegative controls together
with fat morphology and we studied the relationship
between changes in fat morphology, adipocytokine
expression, markers of adipose tissue differentiation and
whole body insulin sensitivity.
Design: Cross-sectional analytical study.
Subjects and methods: The mRNA expression of adipocy-
tokines and transcriptional factors in fat samples from
26 patients with peripheral lipoatrophy (all under anti-
retroviral therapy associating protease inhibitor and
nucleoside-analogue reverse transcriptase inhibitors) and
from 16 non-HIV-infected controls was measured by real
time quantitative RT-PCR. Fat morphology was assessed
histologically on a subgroup of 10 patients and six
controls: collagen fibres by Sirius Red staining, apoptosis
by the TUNEL technique, vessels by smooth muscle
α-actin staining and macrophages by CD68 staining.
Insulin resistance was assessed by using the homeostasis
model assessment.
Results: The patients’ fat showed higher values of apop-
tosis (P=0.005), fibrosis (P<0.05), vessel density (P=0.001)
and macrophage infiltration (P<0.05) than the controls’
fat, together with lower adiponectin and leptin mRNA
levels and higher interleukin (IL)-6 and tumour necrosis
factor (TNF)α mRNA levels. TNFα and IL-6 expression
correlated positively with the level of apoptosis (P=0.05
and P<0.05, respectively) and negatively with CCAAT-
enhancer binding protein (C/EBP)α (P<0.001 and P<0.05,
respectively). Apoptosis correlated negatively with the
expression level of sterol-regulatory-element-binding-
protein-1c (SREBP1c) (P=0.01) and C/EBPα (P=0.01)
whilst the vessel density correlated negatively with
SREBP1c (P<0.005), C/EBPα (P=0.001) and β (P=0.001).
Adiponectin and leptin expression correlated positively
with each other, and also with adipogenic marker expres-
sion and overall insulin sensitivity. These relationships
were also present when the patient group was studied
separately. Finally, fat morphological abnormalities corre-
lated positively with whole body insulin resistance.
Conclusions: Adipose tissue from patients with HIV-1-
related lipoatrophy shows increased apoptosis, together
with decreased adipocyte differentiation. Increased TNFα
and IL-6 expression could be a major phenomenon
linking these alterations. Decreased adiponectin and
leptin expression, which may result from decreased
adipocyte differentiation, could be involved in the
observed whole body insulin resistance.
Altered fat differentiation and adipocytokine
expression are inter-related and linked to
morphological changes and insulin resistance in
HIV-1-infected lipodystrophic patients
Véronique Jan1,2†, Pascale Cervera3†, Mustapha Maachi1,2, Marielle Baudrimont4, Minji Kim1, Hubert
Vidal5, Pierre-Marie Girard6, Philippe Levan7, Willy Rozenbaum8, Anne Lombès9, Jacqueline Capeau1,2
and Jean-Philippe Bastard1,2*
1INSERM Research Unit 402, Faculty of Medicine Saint-Antoine, Paris, France
2Department of Biochemistry, 8Department of Infectious Diseases, Tenon Hospital, Paris, France
3Department of Pathology, 6Department of Infectious Diseases, Saint-Antoine Hospital, Paris, France
4Department of Pathology, Saint-Anne Hospital, Paris, France
5INSERM Research Unit 449 and Human Nutrition Research Center of Lyon, Faculty of Medicine R Laennec, Lyon, France
7Department of Plastic Surgery, Rothschild Hospital, Paris, France
9INSERM Research Unit 582, Myology Institute, Salpêtrière Hospital, Paris, France
†These authors contributed equally to this work
*Corresponding author: Tel: +33 1 56 01 66 76; Fax: +33 1 56 01 78 40; E-mail: jean-philippe.bastard@tnn.ap-hop-paris.fr
Antiviral Therapy 9:555–564
©2004 International Medical Press 1359-6535/02/$17.00555

Page 2

A high proportion of HIV-1-infected patients on anti-
retroviral therapy develop lipodystrophy, characterized
by peripheral fat wasting, visceral fat redistribution
and metabolic alterations with dyslipidemia and
insulin resistance [1–4]. Although the lipodystrophy
syndrome is multifactorial, protease inhibitors (PIs)
and nucleoside reverse transcriptase inhibitors
(NRTIs) have been clearly implicated in cohort studies
and in vitro experiments. NRTIs, and especially
thymidine analogues, have been implicated in periph-
eral lipoatrophy. Their combination with PIs results in
an increased incidence and severity of this phenotype
[5]. In vitro, PIs alter adipocyte differentiation and
insulin sensitivity and induce apoptosis [6,7]. NRTIs
reduce adipocyte lipid content and increase apoptosis
[8] and, when combined with PIs, affect several
adipocyte functions [8,9]. Therefore, the two classes
of molecules have deleterious effects at the adipocyte
level. They could act in synergy and play a role in clin-
ical lipoatrophy.
An increasing number of studies report data on fat
tissue from HIV-1-infected patients [10–17]. An altered
fat morphology has been reported by two groups, with
modified mitochondrial structure and numbers together
with increased apoptosis [10,11,13]. We have previ-
ously reported that subcutaneous adipose tissue from
patients with HIV-1-related lipoatrophy exhibits
marked changes in adipogenesis, with decreased mRNA
concentrations of the main adipogenic transcription
factors [sterol-regulatory-element-binding-protein-1
(SREBP-1), peroxisome proliferators-activated receptor
(PPAR)γ
(C/EBP)α], as well as lower adipocyte insulin sensitivity
[12]. However, these alterations had not been linked to
morphological changes in adipose tissue.
In addition to its metabolic functions, adipose tissue
secretes a variety of adipocytokines. Leptin and
adiponectin act through endocrine mechanisms in the
liver and muscles, controlling insulin sensitivity
[18,19]. Interleukin (IL)-6 and tumour necrosis factor
(TNF)α probably act mainly at the local level. TNFα
induces adipocyte dedifferentiation, insulin resistance
and apoptosis, while IL-6 induces insulin resistance in
adipocytes and reduces their differentiation in vitro
[20–22]. Altered expression of these cytokines might
thus be involved in some of the observed adipose tissue
alterations in HIV-infected patients. We have previ-
ously observed lower leptin expression and higher
TNFα expression in fat from HIV-infected patients
with lipodystrophy than in healthy controls [12]. Very
recently, the increased TNFα
confirmed in lipodystrophic compared with non-
lipodystrophic HIV-infected patients [15,16] and an
increased expression of IL-6 together with a decreased
expression of adiponectin was reported [15,17].
Here we compared histological adipose tissue
abnormalities and adipocytokine expression in HIV-1-
infected subjects and studied the relationship between
morphological alterations in adipose tissue, adipocy-
tokine expression, adipocyte differentiation and whole
body insulin sensitivity. Our findings allow us to
propose this hypothetical scheme (Figure 1).
and CCAAT-enhancer binding protein
expression was
V Jan et al.
©2004 International Medical Press
556
Introduction
Protease
inhibitors
NRTI,
thymidine
analogues
Decreased
SREBP-1,
C/EBPα
Altered adipocyte
phenotype or
differentiation
Increased
TNFα, IL-6
Mitochondrial
dysfunction
Decreased
adiponectin
Adipose tissue
insulin resistance
Apoptosis
Lipoatrophy
Whole body
insulin
resistance
Figure 1. Pathogenesis of lipodystrophy

Page 3

Materials and methods
Subjects
All the subjects studied had been previously presented
[12]. Briefly, the HIV-1-infected group consisted of 26
patients treated with both PI and NRTI (80% with
stavudine) who had developed peripheral lipoatrophy
and who had undergone subcutaneous abdominal lipo-
suction with fat re-injection into the cheeks, using
Coleman’s technique [23]. An additional abdominal fat
sample was stored in liquid nitrogen or fixed in formol
or Carson solution (Millonig’s phosphate-buffered
formalin). All the patients gave their written consent to
participate in the study.
The patients were comprised of 21 men and five
women, with a mean age of 45 years (range 36–62)
and a mean body mass index (BMI) of 23.3 kg/m2
(16.5–28.3). CD4+ cell counts were above 200/µl in all
but three of the patients (145–165/µl). Nineteen
patients had no detectable HIV-1 RNA at the time of
the study and the remaining seven had viral load values
between 730 and 135000 copies/ml. The mean known
duration of HIV infection was 11 years (2–19). Fifteen
patients were on indinavir and nine were on nelfinavir;
21 were on stavudine and 19 on lamivudine. The mean
duration of antiretroviral combination therapy was
37 months (15–48).
The control group consisted of 16 healthy HIV-1-
seronegative non-diabetic and non-obese subjects (eight
women and eight men). The mean age of the control
subjects was 34 years (24–69) and their mean BMI was
24.1 kg/m2(19.0–31.3). Fat from control subjects was
obtained during abdominal plastic surgery (liposuction)
or from normal volunteers involved in a nutritional
study with subcutaneous fat biopsies (needle fat aspira-
tion) and immediately frozen.
Fat morphology studies involved a subgroup of 16
subjects (10 patients and six controls). The mean age
and BMI of the HIV-1-infected patients was 43 years
(37–53) and 24.0 kg/m2(21.3–26.7). All the patients
were treated with stavudine plus either lamivudine
(n=8), didanosine (n=1) or abacavir (n=1) and plus
indinavir (n=7), nelfinavir (n=3) or ritonavir (n=1). The
mean age and BMI of the controls was 49 years (28–69)
and 24.8 kg/m2(19.8–30.0), with no statistical differ-
ence between the two groups.
Adipose tissue histology
Light microscopy was performed by a histopathologist
on 10% zinc-formol-fixed paraffin-embedded 5-µm
tissue sections stained with hemalum-phloxine and
Sirius Red to detect collagen fibres. For immunocyto-
chemical studies, we used an automated immunostainer,
Optimax plus (BioGenex Laboratories, Inc., Calif., USA
and an avidin–biotin immunoperoxidase method with
3-amino-9-ethylcarbazole chromogene (AEC) (Vector
Laboratories, Inc., Calif., USA) or peroxidase-labelled
secondary antibody followed by a peroxidase-labelled
tertiary antibody before AEC revelation. A prestaining
heat-based epitope-retrieval technique was used prior
to staining for several antibodies, with sodium citrate
pH 6.0 or water as buffer. Appropriate positive and
negative control tissues were stained using the same
methods. The following monoclonal and polyclonal
antibodies were used: CD45 (1/10; Dako Corp, Calif.,
USA) a lymphocyte and neutrophil marker; CD68
(1/300; Dako) a macrophage marker; Ki67, which
stains replicative nuclei; anti-smooth-muscle α actin
(1/50; Dako), which labels vascular smooth muscle
cells; and anti-mitochondria (1/50, BioGenex).
Apoptosis was measured by the terminal deoxynu-
cleotidyl transferase dUTP-digoxigenin nick-end
labelling method (TUNEL) using the Apoptag kit
(Oncor, Gaithersburg, Md., USA).
Each fat specimen was examined with a semi-
automatic image analysis system (Mercator, La
Rochelle, France). Firstly, the surface area of three
regions chosen at random was measured. Then, in each
region, adipocytes, apotag-positive nuclei and total
nuclei (except vessels) were mapped and counted. The
density of CD45-, CD68- and Ki67-positive cells and
anti-smooth muscle actin-positive vessels was evalu-
ated as the average of 10 fields (×40).
Fat samples for ultrastructural studies were fixed in
Carson solution, rinsed in 0.1 M cacodylate buffer and
then post-fixed in 1% osmium tetroxide for 1 h. After
rinsing in cacodylate buffer, fragments were dehy-
drated in graded alcohol series and embedded in epoxy
resin. Semi-fine sections (0.5 µm) were stained with
toluidine blue. Ultrastructure sections (60 nm) were
contrasted with uranyl acetate and lead citrate and
examined on Geol 100, CX-2 electron microscope.
mRNA assays
Fat tissue RNA was extracted using the RNeasy total
RNA kit (Qiagen, Courtaboeuf, France). The yield of
total RNA was 2.1 µg (0.7–4.8) per 100 mg of tissue.
The mRNA concentrations of PPARγ, C/EBPα,
C/EBPβ, SREBP-1c, glucose transporter 4 (GLUT4),
hormone-sensitive lipase (HSL), leptin, TNFα and β2-
microglobulin were determined by reverse transcription
followed by competitive polymerase chain reaction
amplification (RT-cPCR); adiponectin and IL-6 mRNAs
were quantified by real-time PCR. Real-time PCR was
performed with a Light cycler®device (Roche-
Boehringer, Meylan, France) using a SYBR Green®
(Finnzymes, Espoo, Finland)
Quantification was obtained with reference to a
plasmid dilution range that contained the human
mRNA target sequence. Each quantification was
detection signal.
Antiviral Therapy 9:4
557
Fat disorders in HIV-related lipoatrophy

Page 4

performed twice and each run was validated by the
specificity of the PCR (Tm evaluation) and by the slope
and error obtained for the different plasmid dilutions.
The primer sequences were as follows: IL-6: forward
primer 5′-AGCCCTGAGAAAGGAGACATGTAACA
AG-3′ and backward primer 5′-TTCTGCAGGAACTG
GATCAGGACTTT-3′; adiponectin: forward primer 5′-
CAGAGATGGCACCCCTGGTG-3′
primer 5′-TTCACCGATGTCTCCCTTAG-3′. The results
were expressed as attomoles (amol/µg) of total RNA.
β2-microglobulin was used as an internal standard for
mRNA expression and all the mRNA results were
expressed as amol/µg of total RNA and as ratios rela-
tive to β2-microglobulin mRNA expression.
and backward
Insulin resistance
Insulin resistance was assessed by calculating the
homeostasis model assessment (HOMA) index defined
as fasting plasma glucose (mmol/l) × fasting plasma
insulin (µU/ml)/22.5 [24].
Statistical analysis
All results are presented as means ±SE or mean with
range. Differences between groups were determined
using the non-parametric Mann–Whitney U test. The
significance of correlations was determined using the
non-parametric Spearman’s rank correlation test. The
threshold of significance was set at P=0.05.
Results
Adipose tissue shows major morphological alterations
The population of small adipocytes is higher in
patients than in controls [12]. In addition, we observed
an increased number of apoptotic nuclei in fat from
patients: 17.5% (4.3–30.2) versus 4.5% in controls
[(0.7–7.2); P=0.005]. Blood vessels were more
numerous in fat from patients [30 per field (5–47)]
than from controls [three per field (1–4); P=0.001]
(Figure 2A). The proportion of fibrosis was higher in
fat from patients [10% (1–30)] than from controls
[2% (0–8); P<0.05] (Figure 2B). Marked mitochon-
drial labelling was present in the cytoplasm of some
adipocytes in the patients’ fat (Figure
Interestingly, these labelled adipocytes were small, with
an increased cytoplasm often containing small fat
droplets adjacent to the large unilocular fat vacuole or
small residual fat vacuoles. These cells (Figure 2C;
inset) were adipocytes
macrophages since labelling by CD68 was negative
(Figures 2D; inset). Control fat contained large
adipocytes that showed no mitochondrial staining. The
number of macrophages was higher in fat from
patients [14.8 per field (1.6–25.0)] than from controls
[3.2 per field (1.1–8.0), P<0.02]. Lipogranulomas were
2C).
and not lipid-laden
often seen, with macrophages surrounding a fat
droplet lying free in connective tissue (Figure 2D). No
difference was found between the two groups as
regards the number of CD45 positive cells [10.5 per
field (4.3–18.0) and 7.4 per field (2.0–16.1), P=0.23]
in fat from patients and controls, respectively. Cell
proliferation was evaluated by labelling replicative
nuclei with Ki67 – few nuclei were labelled in the fat
from both patients and controls (not shown), arguing
against a process of active regeneration.
The proportion of apoptotic nuclei correlated posi-
tively with blood vessel density (r=0.63, P=0.01) and
fibrosis (r=0.51, P=0.05). The proportion of small
adipocytes (less than 70 µm) correlated strongly with
blood vessel density (r=0.70, P<0.01). Overall, the
altered fat morphology in the HIV-1-infected group
was characterized by fields of small adipocytes with
increased cytoplasm enriched in mitochondria and
surrounded by increased stroma with fibrosis, blood
vessel density and apoptosis.
Ultra-structural studies performed in 14 patients
revealed small adipocytes in 13, and stromal fat
droplets in all the patients. Small adipocytes were
present in the fat of two out of six controls, while
stromal lipid droplets were found in only one control.
Adipocytokine expression
All the results are presented as absolute levels in the
patient and control groups and as ratios relative to β2-
microglobulin, which was used as an internal control
(Table 1). These ratios were used for all the data
presented below. The mRNA concentration of leptin
was markedly lower in fat from patients than controls
as previously reported [12]. We found here that the
mRNA concentration of adiponectin was 2.1-fold
lower in fat from patients than from controls (Table 1).
The mRNA concentrations of leptin and adiponectin
correlated positively with each other (Table 2). We
found that the mRNA concentration of both
adiponectin and leptin correlated positively with the
expression of all the adipogenic transcription factors
tested previously [12], the strongest correlation being
with C/EBPα (Table 2).
The mRNA concentration of IL-6 was fourfold
higher in fat from patients than in fat from controls
(Table 1), as previously observed for TNFα [12]. The
expression of TNFα correlated negatively with that of
SREBP1c, PPARγ, C/EBPα and β while adiponectin and
leptin were positively correlated with all these
adipogenic transcriptional factors. Otherwise IL-6 was
only inversely and significantly correlated with C/EBPα
(Table 2).
When we studied the HIV-infected group separately,
we found a significant positive correlation between
adiponectin and C/EBPα (r=0.522, P<0.05), C/EBPβ
V Jan et al.
©2004 International Medical Press
558

Page 5

Antiviral Therapy 9:4
559
Fat disorders in HIV-related lipoatrophy
(A) Blood vessels (anti-smooth-muscle α actin labelling); (B) fibrosis (Sirius Red staining); (C) mitochondria (anti-mitochondria labelling), inset: an adipocyte with
multiple fat droplets; (D) macrophages with a lipogranuloma structure (CD68 labelling), inset: absence of labelling with CD68 of the adipocyte shown in (C) inset.
The magnification was the same for all pictures including insets except for Figure 1C left which magnification was twofold lower.
Figure 2. Typical histology of subcutaneous abdominal adipose tissue from an HIV-1-infected patient and a healthy control
A
B
C
D
Patients with HIV-1 infection
Healthy control
100µm
50µm

Page 6

(r=0.511, P<0.05), GLUT4 (r=0.443, P<0.05) and LPL
(r=0.490, P<0.05) mRNA expression while TNFα was
inversely correlated with PPARγ and C/EBPα (r=
–0.407, P<0.05 and r=–0.412, P<0.05, respectively).
There was no difference in adipose tissue gene
expression between male and female subjects in both
controls and patients groups for all the genes studied
except for leptin. Indeed, female patients exhibited a
higher mRNA expression than male patients (0.289
±0.082 vs 0.126 ±0.034, P<0.05). However, leptin
expression remained markedly decreased as compared
with controls in both male (P<0.0001) and female
groups (P=0.005).
When we compared male controls and patients or
female controls and patients, the same significant vari-
ations as those observed in the whole group were
found for SREBP1c, PPARγ, C/EBPα and β, HSL,
GLUT4, adiponectin, leptin and TNFα.
V Jan et al.
©2004 International Medical Press
560
Table 2. Correlation between adipose tissue parameters and adipocytokine mRNA
Adiponectin Leptin IL-6 TNFα
SREBP1c r=0.576
P<0.001
r=0.693
P<0.0001
r=0.583
P<0.001
r=0.484
P<0.005
r=0.521
P<0.005
r=0.601
P=0.0005
r=0.696
P<0.0001
–
r=0.647
P<0.0001
r=0.660
P<0.0001
r=0.344
P=0.05
r=0.562
P<0.001
r=0.478
P<0.005
r=0.432
P=0.01
r=0.632
P<0.0001
r=0.559
P<0.001
–
r=-0.305
P=0.10
r=–0.396
P=0.05
r=–0.174
P=0.86
r=–0.379
P=0.09
r=–0.387
P=0.05
r=–0.219
P=0.24
r=–0.234
P=0.20
r=–0.036
P=0.84
r=–0.326
P=0.08
–
r=–0.527
P<0.001
r=–0.548
P<0.001
r=–0.354
P<0.05
r=–0.572
P<0.001
r=–0.234
P=0.13
r=–0.102
P=0.56
r=–0.549
P<0.001
r=–0.407
P<0.05
r=–0.318
P<0.05
r=–0.257
P=0.16
C/EBPα
C/EBPβ
PPARγ
HSL
LPL
GLUT4
Adiponectin
Leptin –
IL-6– –
The study was performed in 42 subjects except for IL-6, LPL and C/EBPβ (n=30–34). All the gene mRNA expression was related to that of β2-microglobulin.
Table 1. Comparison of mRNA gene expression from adipose tissue in controls (n=16) and HIV-infected patients (n=26)
expressed as attomol/µg RNA or per β2-microglobulin mRNA gene expression
ControlsHIV patientsControls HIV patients
mRNA (amol/µg RNA) mRNA (amol/µg RNA) mRNA /β2 mRNA
expression × 100
mRNA /β2 mRNA
expression × 100
SREBP1c
C/EBPα
C/EBPβ
PPARγ
HSL
LPL
GLUT4
Adiponectin
Leptin
IL-6
TNFα
26.0 ±3.3
82.1 ±11.7
29.7 ±3.4
16.1 ±2.4
212.0 ±26.1
72.2 ±11.2
33.7 ±6.2
287.0 ±22.5
4.32 ±0.73
0.64 ±0.18
0.048 ±0.007
2.8 ±0.7¶
26.6 ±3.7¶
15.4 ±1.9∫
5.1 ±1.0∫
145.0 ±16.2*
45.7 ±5.6*
6.6 ±0.9¶
113.1 ±17.1†
0.28 ±0.07¶
2.36 ±0.55*
0.125 ±0.014§
18.6 ±2.4
53.9 ±6.5
17.5 ±1.4
12.8 ±2.8
149.8 ±19.2
44.2 ±7.1
26.7 ±6.1
206.3 ±20.5
3.34 ±0.71
0.40 ±0.11
0.033 ±0.004
2.0 ±0.6¶
17.2 ±3.1¶
10.0 ±1.7‡
2.9 ±0.6¶
87.9 ±10.6‡
29.4 ±5.0*
4.0 ±0.6¶
96.1 ±14.0∫
0.16 ±0.03¶
1.61 ±0.58*
0.070 ±0.008∫
The study was performed in 42 subjects except for IL-6, LPL and C/EBPβ (n=30–34).
*P<0.05; †P<0.01; ‡P<0.005; §P<0.001; ∫P<0.0005; ¶P<0.0001.

Page 7

Correlations between morphological alterations of
adipose tissue and adipocytokine expression
We found numerous tight correlations between the
diverse aspects of fat morphological changes (apop-
tosis, fibrosis and blood vessel density) and the
different components of the underlying molecular
events, namely adipocytokine expression and factors
involved in fat differentiation.
The adipocytokines appeared to be involved in fat
morphological changes, as the mRNA expression of
IL-6 and TNFα correlated positively with apoptosis
(Table 3). TNFα transcript levels also correlated with
vessel density (Table 3). Conversely, leptin transcript
expression correlated negatively with apoptosis,
fibrosis and blood vessel density (Table 3), while
adiponectin correlated only and negatively with blood
vessel density.
Finally, a strong negative correlation was observed
between the expression of several adipogenic transcrip-
tion factors, including C/EBPα and SREBP1c, and both
apoptosis and blood vessel density (Table 3).
Correlations between altered adipocytokine
expression and insulin resistance
HIV-infected patients had a higher HOMA score than
controls (4.9 ±0.7 vs 1.4 ±0.20, P<0.0001) confirming
that HIV-infected patients presented a state of whole
body insulin resistance. We have previously found a
positive correlation between TNFα mRNA concentra-
tions in fat samples and insulin resistance and a
positive correlation between leptin mRNA expression
and insulin sensitivity [12]. We observed that the rate
of TNFα and leptin relative to β2-microglobulin were
highly correlated with HOMA (r=0.531, P<0.005 and
r=–0.670, P<0.0001 respectively) in accordance with
previous results [12]. Accordingly, we found a positive
correlation between IL-6 mRNA expression and both
HOMA (r=0.529, P<0.01) and fasting plasma insulin
(r=0.476, P<0.05). By contrast, we found a negative
correlation between adiponectin mRNA concentrations
and both HOMA (r=–0.535, P<0.01) and insulin levels
(r=–0.507, P=0.01). This emphasizes the involvement
of adipocytokines in insulin sensitivity in humans, and
particularly in HIV-1-infected patients. Moreover, we
found a significant correlation between HOMA and
apoptosis (r=0.782, P<0.005), fibrosis (r=0.632,
P<0.05) and the number of vessels (r=0.880, P<0.001).
Discussion
This is the first comprehensive study of morphological
and molecular alterations in subcutaneous adipose
tissue from HIV-1-infected lipodystrophic patients
treated with NRTI–PI combinations. However, an
important limitation of its cross-sectional design
together with the lack of a treated HIV-1-positive
control group without lipodystrophy, is that it does not
allow us to distinguish between potential effects of HIV
or its treatment. In addition, no conclusions can be
drawn concerning the role of specific antiretroviral
drug classes. We observed striking morphological alter-
ations in fat from patients as compared with healthy
controls. We confirmed the increased proportion of
small adipocytes [10–13,25], the higher rate of apop-
tosis [10] and the presence of stroma containing fibrosis
and vessels [11,25]. Interestingly, these morphological
alterations correlated with one another and also with
abnormal adipocytokine expression, markers of altered
adipocyte differentiation and insulin resistance.
We report here that the mRNA concentrations of
C/EBPα and SREBP1c, two key transcription factors
involved in adipocyte differentiation and insulin sensi-
tivity, correlate strongly and negatively with apoptosis
and increased blood vessel density. Altered cytokine
expression could represent a possible link between
differentiation and apoptosis. In vitro and animal
studies have shown that TNFα, which mainly acts
through autocrine/paracrine mechanisms, can promote
adipocyte dedifferentiation, resistance to insulin and
apoptosis [20]. While TNFα production by white
adipose tissue is low in physiological conditions, it can
be markedly increased in animal models of obesity
[20]. A strong set of arguments suggests the involve-
ment of TNFα in the adipose tissue changes associated
with HIV-1-related lipodystrophy. We have previously
observed higher mRNA concentrations of TNFα in fat
from lipodystrophic patients [12]. This result has been
Antiviral Therapy 9:4
561
Fat disorders in HIV-related lipoatrophy
Table 3. Correlation between adipose tissue alterations and
adipocytokine and transcription factor gene expression in
controls (n=6) and HIV-infected patients (n=10)
ApoptosisFibrosis Vessels
IL-6r=0.647
P<0.05
r=0.517
P=0.05
r=–0.671
P<0.01
r=–0.649
P=0.01
r=–0.648
P=0.01
r=–0.261
P=0.31
r=– 0.648
P=0.01
r=–0.480
P=0.06
r=0.197
P=0.46
r=0.126
P=0.48
r=–0.688
P<0.01
r=–0.693
P=0.01
r=–0.371
P=0.15
r=–0.068
P=0.79
r=–0.769
P<0.005
r=–0.629
P<0.05
r=0.482
P=0.07
r=0.515
P=0.05
r=–0.873
P<0.001
r=–0.771
P<0.005
r=–0.739
P<0.005
r=–0.507
P=0.05
r=–0.878
P=0.001
r=–0.823
P=0.001
TNFα
Adiponectin
Leptin
SREBP1c
PPARγ
C/EBPα
C/EBPβ
All the gene mRNA expression was related to that of β2-microglobulin.

Page 8

recently confirmed when HIV-infected lipodystrophic
patients were compared with non-lipodystrophic HIV
infected patients [15,16]. Likewise, higher circulating
levels of TNFα and its soluble receptors have been
reported by us and others in lipodystrophic patients as
compared with controls [26–28]. An accumulation of
circulating T cells primed for TNFα synthesis has been
reported in patients with HIV-1-related lipodystrophy,
and was linked to circulating lipid abnormalities [29].
Interestingly, we found that TNFα expression corre-
lated negatively with the expression of adipogenic
factors and positively with apoptosis. Together with
other published data, this suggests that TNFα could be
involved in adipocyte dedifferentiation and in apop-
tosis in fat of lipodystrophic patients.
We report an increased IL-6 expression in fat from
lipodystrophic patients as recently reported by two
groups [15,17]. In addition, we observed a negative
correlation between IL-6 and C/EBPα expression and a
strong positive correlation between IL-6 expression and
apoptosis. This suggests that IL-6 could also play a
major role at the local level, through paracrine/autocrine
mechanisms, as previously reported in insulin-resistant
obese non-diabetic and type 2 diabetic subjects [30]. In
vitro studies suggest that this cytokine could alter
adipocyte differentiation [21,22]. However, data are
lacking to link increased IL-6 expression to adipocyte
dedifferentiation or apoptosis.
Antiretroviral drugs could be responsible for the
increased expression and secretion of these cytokines.
Recent studies indicate that not only PIs but also thymi-
dine analogues can induce the expression of TNFα and
IL-6 by cultured 3T3F442A adipocytes ([31,32] and
Lagathu et al., personal results). It is therefore conceiv-
able that PIs and NRTIs can increase the expression of
TNFα and IL-6 in adipocytes, resulting in decreased
differentiation and increased apoptosis.
We found that patients’ adipose tissue contained an
increased number of macrophages, which surrounded
adipocytes in lipogranuloma-like structures. These
macrophages are probably activated, expressing IL-6
and TNFα. Therefore, the increased IL-6 and TNFα
expression observed in patients’ fat could also derive
partly from macrophages. As IL-6 and TNFα act
through autocrine and paracrine mechanisms, these
cytokines could be responsible for adipocyte dysfunc-
tion, whatever their cellular origin.
We observed an increased level of fibrosis and a
markedly increased blood vessel density in patients’
fat. This could reflect a process of tissue reparation
after elimination of apoptotic adipocytes by
macrophage phagocytosis. This tissue is enriched in
blood vessels, suggesting that angiogenesis could also
be involved in the remodelling process. We found that
the mRNA concentration of leptin correlated strongly
and negatively with blood vessel density. This is in line
with a recent study showing that leptin can ablate
adipose tissue by inducing a loss of adipose vasculature
[33]. Likewise, leptin has been shown to induce the
expression of the negative angiogenesis signal angiopoi-
etin-2 [33]. Further studies are required to explain the
increased angiogenesis in lipodystrophic patients’ fat.
Altered mitochondrial biogenesis could play a role
in increased apoptosis and could result from altered
differentiation. The ability of NRTIs, and particularly
thymidine analogues, to decrease mtDNA has been
reported and linked to peripheral lipoatrophy [11,13].
Morphological studies have shown the presence of
increased numbers of mitochondria with an altered
ultrastructure in lipoatrophic adipose tissue [10,11,25].
We found that adipocytes with mitochondria-rich cyto-
plasm were small, and often contained multiple fat
droplets or a very small fat vacuole, in accordance with
previous studies [11,25]. As we found only scarce repli-
cating nuclei in patients’ fat, this aspect is probably due
to toxicity rather than regeneration. A prominent role
of NRTIs in this process was recently reported [11,13].
The responsibility for NRTIs and histological alter-
ations in some patients is suggested by Nolan et al.
who observed detectable improvement in tissue toxi-
city in one of three cases after switching d4T for
zidovudine [13].
Metabolic alterations in lipodystrophic HAART-
treated patients are frequently associated with insulin
resistance, but the underlying mechanisms are
unclear. However, in patients with HAART-related
lipodystrophy, insulin resistance at the level of
adipose tissue was revealed by the increased level of
circulating free fatty acids [34,35], which have been
shown to induce insulin resistance in human liver and
muscle [36]. This could result from the effects of
TNFα or IL-6 on adipocytes.
The role of leptin and adiponectin in insulin sensi-
tivity has been reported several times. Serum
adiponectin and leptin levels are decreased in human
genetic syndromes with generalized lipoatrophy and
insulin resistance: the metabolic disorders in these
patients are markedly improved by leptin replacement
therapy [37]. More recently, we and others showed a
reduction in serum adiponectin and leptin levels in
HIV-infected patients on HAART [26,27]. In addition,
adiponectin and sTNF-R1 levels were related to meta-
bolic alterations and insulin resistance, suggesting a
role of these cytokines in insulin sensitivity.
A decreased mRNA concentration of adiponectin
in fat from patients with HIV-related lipodystrophy
was recently described, and was found to correlate
with insulin resistance [14,15]. Accordingly, we found
that adiponectin mRNA expression correlated nega-
tively with insulin resistance and insulin levels in our
V Jan et al.
©2004 International Medical Press
562

Page 9

population. These data argue for a role of adipocy-
tokines in insulin resistance in this setting. In addition,
insulin resistance was related to morphological alter-
ations suggesting a close relationship between adipose
tissue dysfunction and whole body insulin resistance in
lipodystrophic HIV patients. The decreased expression
of adiponectin could result from drug toxicity: PIs and
NRTIs were both found to decrease adiponectin
mRNA expression in 3T3F442A adipocytes ([32] and
Lagathu et al., personal results).
In conclusion, we show here that some morpholog-
ical and molecular alterations in adipose tissue from
lipodystrophic HIV-1-infected patients are inter-
related. We had previously obtained evidence that
antiretroviral treatment including PIs and NRTIs can
inhibit adipocyte differentiation and induce insulin
resistance, in part by decreasing SREBP-1 expression
[12]. We recently found that, in vitro, some PIs alter
lamin A/C maturation, resulting in abnormal nuclear
lamina stability which, in turn, is probably responsible
for altered SREBP-1 nuclear location and function
[7,38]. The observed increase in TNFα and IL-6
expression could result from exposure to both PIs and
NRTIs and lead to altered adipocyte differentiation
and insulin sensitivity, as well as increased apoptosis,
ultimately leading to lipoatrophy. Drug-induced mito-
chondrial dysfunction might also contribute to this
phenotype. Finally, a drug-induced decrease in
adiponectin secretion, together with an increase in free
fatty acid release by insulin-resistant adipose tissue,
could also be involved in whole body insulin resis-
tance and metabolic disorders (Figure 1).
Acknowledgements
We thank Dr Martine Caron and Dr Corinne
Vigouroux for helpful discussion and critical evalua-
tion of the manuscript. We are grateful to Dr Sabine Le
Gouvello (Immunology Department, Henri-Mondor
Hospital, Créteil, France) for providing the plasmid for
IL-6 mRNA RT-PCR assay.
This work was supported by grants from Agence
Nationale de Recherche sur le SIDA (ANRS), Institut
National de la Santé et de la Recherche Médicale
(INSERM) and Association de Recherche sur le VIH
(ARVIH).
References
1. Carr A, Samaras K, Thorisdottir A, Kaufmann GR,
Chisholm DJ & Cooper DA. Diagnosis, prediction, and
natural course of HIV-1 protease inhibitor-associated
lipodystrophy, hyperlipidaemia, and diabetes mellitus: a
cohort study. Lancet 1999; 353:2093–2099.
Jain RG, Furfine ES, Pedneault L, White AJ & Lenhard
JM. Metabolic complications associated with antiretroviral
therapy. Antiviral Research 2001; 51:151–177.
2.
3. Nolan D, John M & Mallal S. Antiretroviral therapy and
the lipodystrophy syndrome, part 2: concepts in
aetiopathogenesis. Antiviral Therapy 2001; 6:145–160.
Schambelan M, Benson CA, Carr A, Currier JS, Dube MP,
Gerber JG, Grinspoon SK, Grunfeld C, Kotler DP,
Mulligan K, Powderly WG & Saag MS. International AIDS
Society-USA: management of metabolic complications asso-
ciated with antiretroviral therapy for HIV-1 infection:
recommendations of an international AIDS society-USA
panel. Journal of Acquired Immune Deficiency Syndromes
2002; 31:257–275.
Mallal SA, John M, Moore CB, James IR & McKinnon EJ.
Contribution of nucleoside analogue reverse transcriptase
inhibitors to subcutaneous fat wasting in patients with HIV
infection. AIDS 2000; 14:1309–1316.
Dowell P, Flexner C, Kwiterovich PO & Lane MD.
Suppression of preadipocyte differentiation and promotion
of adipocyte death by HIV protease inhibitors. Journal of
Biological Chemistry 2000; 275:41325–41326.
Caron M, Auclair M, Vigouroux C, Glorian M, Forest C
& Capeau J. The HIV protease inhibitor indinavir
impairs sterol regulatory element-binding protein-1
intranuclear localization, inhibits preadipocyte differenti-
ation, and induces insulin resistance. Diabetes 2001;
50:1378–1388.
Caron M, Auclair M, Kornprobst M & Capeau J.
Stavudine and didanosine partly reversed the adverse
effects of indinavir on cell differentiation and response to
insulin but enhanced apoptosis of cultured adipocytes.
Antiviral Therapy 2002; 7(Suppl. 3):16.
Roche R, Poizot-Martin I, Yazidi CM, Compe E, Gastaut
JA, Torresani J & Planells R. Effects of antiretroviral drug
combinations on the differentiation of adipocytes. AIDS
2002; 16:13–20.
10. Domingo P, Matias-Guiu X, Pujol RM, Francia E, Lagarda
E, Sambeat MA & Vazquez G. Subcutaneous adipocyte
apoptosis in HIV-1 protease inhibitor-associated lipodys-
trophy. AIDS 1999; 13:2261–2267.
11. Nolan D, Hammond E, Martin A, Taylor L, Herrmann S,
McKinnon E, Metcalf C, Latham B & Mallal S.
Mitochondrial DNA depletion and morphologic changes in
adipocytes associated with nucleoside reverse transcriptase
inhibitor therapy. AIDS 2003; 17:1329–1338.
12. Bastard JP, M Caron, H Vidal, V Jan, M Auclair, C
Vigouroux, J Luboinski, M Maachi, PM Girard, W
Rozenbaum, P Levan & J Capeau. Association between
altered expression of adipogenic factor SREBP1 in lipoat-
rophic adipose tissue from HIV-1-infected patients and
abnormal adipocyte differentiation and insulin resistance.
Lancet 2002; 359:1026–1031.
13. Nolan D, Hammond E, James I, McKinnon E & Mallal S.
Contribution of nucleoside-analogue reverse transcriptase
inhibitor therapy to lipoatrophy from the population to the
cellular level. Antiviral Therapy 2003; 8:617–626.
14. Sutinen J, Korsheninnikova E, Funahashi T, Matsuzawa Y,
Nyman T & Yki-Jarvinen H. Circulating concentration of
adiponectin and its expression in subcutaneous adipose
tissue in patients with highly active antiretroviral therapy-
associated lipodystrophy. Journal of Clinical
Endocrinology & Metabolism 2003; 88:1907–1910.
15. Lihn AS, Richelsen B, Pedersen SB, Haugaard SB, Rathje
GS, Madsbad S & Andersen O. Increased expression of
TNF-α, IL-6, and IL-8 in HIV-associated lipodystrophy.
Implications for the reduced expression and plasma
levels of adiponectin. American Journal of Physiology,
Endocrinology & Metabolism 2003; 285:E1072–E1080.
16. Johnson JA, Albu JB, Engelson ES, Fried SK, Inada Y,
Ionescu G & Kotler DP. Increased systemic and adipose
tissue cytokines in patients with HIV-associated lipodys-
trophy. American Journal of Physiology, Endocrinology &
Metabolism 2004; 286:E261–E271.
17. Kannisto K, Sutinen J, Korsheninnikova E, Fisher RM,
Ehrenborg E, Gertow K, Virkamäki A, Nyman T, Vidal H,
Hamsten A & Yki-Järvinen H. Expression of adipogenic
4.
5.
6.
7.
8.
9.
Antiviral Therapy 9:4
563
Fat disorders in HIV-related lipoatrophy

Page 10

transcription factors, peroxisome proliferator-activated
receptor gamma co-activator 1, IL-6 and CD45 in subcuta-
neous adipose tissue in lipodystrophy associated with
highly active antiretroviral therapy. AIDS 2003;
17:1753–1762.
18. Shimomura I, Hammer RE, Ikemoto S, Brown MS &
Goldstein JL. Leptin reverses insulin resistance and diabetes
mellitus in mice with congenital lipodystrophy. Nature
1999; 401:73–76.
19. Yamauchi T, Kamon J, Waki H, Terauchi Y, Kubota N,
Hara K, Mori Y, Ide T, Murakami K, Tsuboyama-Kasaoka
N, Ezaki O, Akanuma Y, Gavrilova O, Vinson C, Reitman
ML, Kagechika H, Shudo K, Yoda M, Nakano Y, Tobe K,
Nagai R, Kimura S, Tomita M, Froguel P & Kadowaki T.
The fat-derived hormone adiponectin reverses insulin resis-
tance associated with both lipoatrophy and obesity. Nature
Medicine 2001; 7:941–946.
20. Coppack SW. Pro-inflammatory cytokines and adipose
tissue. Proceedings of the Nutrition Society 2001;
60:349–356.
21. Päth G, Bornstein SR, Gurniak M, Chrousos GP,
Scherbaum WA & Hauner H. Human breast adipocytes
express interleukin-6 (IL-6) and its receptor system:
increased IL-6 production by β-adrenergic activation and
effects of IL-6 on adipocyte function. Journal of Clinical
Endocrinology & Metabolism 2001; 86:2281–2288.
22. C Lagathu, JP Bastard, M Auclair, M Maachi, J Capeau &
M Caron. Chronic interleukin-6 (IL-6) treatment increased
IL-6 secretion and induced insulin resistance in adipocyte:
prevention by rosiglitazone. Biochemical & Biophysical
Research Communications 2003; 311:372–379.
23. Coleman WP 3rd. Autologous fat transplantation. Plastic
Reconstructive Surgery 1991; 13:2261–2267.
24. Matthews DR, Hosker JP, Redenski AS, Naylor BA,
Treacher DF & Turner RC. Homeostasis model assess-
ment: insulin resistance and beta-cell function from fasting
plasma glucose and insulin concentration in man.
Diabetologia 1985; 28:412–419.
25. Lloreta J, Domingo P, Pujol RM, Arroyo JA, Baixeras N,
Matias-Guiu X, Gilaberte M, Sambeat MA & Serrano S.
Ultrastructural features of highly active antiretroviral
therapy-associated partial lipodystrophy. Virchows Archive
2002; 441:599–604.
26. Mynarcik DC, Combs T, McNurlan MA, Scherer PE,
Komaroff E & Gelato MC. Adiponectin and leptin levels in
HIV-infected subjects with insulin resistance and body fat
redistribution. Journal of Acquired Immune Deficiency
Syndromes 2002; 31:514–520.
27. Vigouroux C, Maachi M, Nguyen TH, Coussieu C,
Gharakhanian S, Funahashi T, Matsuzawa Y, Shimomura I,
Rozenbaum W, Capeau J & Bastard JP. Serum adipocy-
tokines are related to lipodystrophy and metabolic
disorders in HIV-infected men under antiretroviral therapy.
AIDS 2003; 17:1503–1511.
28. Christeff N, Melchior JC, de Truchis P, Perronne C &
Gougeon ML. Increased serum interferon alpha in HIV-1
associated lipodystrophy syndrome. European Journal of
Clinical Investigation 2002; 32:43–50.
29. Ledru E, Christeff N, Patey O, de Truchis P, Melchior JC
& Gougeon ML. Alteration of tumor necrosis factor-alpha
T-cell homeostasis following potent antiretroviral therapy:
contribution to the development of human
immunodeficiency virus-associated lipodystrophy
syndrome. Blood 2000; 95:3191–3198.
30. Bastard JP, Maachi M, Tran Van Nhieu J, Jardel C,
Bruckert E, Grimaldi A, Robert JJ, Capeau J & Hainque B.
Adipose tissue IL-6 content correlates with resistance to
insulin activation of glucose uptake both in vivo and in
vitro. Journal of Clinical Endocrinology & Metabolism
2002; 87:2084–2089.
31. Jones SP, Janneh O, Maher B, Khoo S, Back DJ &
Pirmohamed N. Altered TNFα and IL-6 levels and the
adipogenic effect of antiretrovirals on cultured adipocytes:
possible mechanism for their role in lipodystrophy in HIV
infected patients. Antiviral Therapy 2002; 7(Suppl. 3):12.
32. Jones SP, Janneh O, Back DJ & Pirmohamed N. Altered
adipocytokine gene expression in murine 3T3-F442A
adipocytes treated with protease inhibitors and nucleoside
reverse transcriptase inhibitors (NRTI). Antiviral Therapy
2003; 8:L11.
33. Cohen B, Barkan D, Levy Y, Goldberg I, Fridman E,
Kopolovic J & Rubinstein M. Leptin induces angiopoietin-
2 expression in adipose tissues. Journal of Biological
Chemistry 2001; 276:7697–7700.
34. Van der Valk M, Bisschop PH, Romijn JA, Ackermans MT,
Lange JM, Endert E, Reiss P & Sauerwein HP.
Lipodystrophy in HIV-1-positive patients is associated with
insulin resistance in multiple metabolic pathways. AIDS
2001; 15:2093–2100.
35. Hadigan C, Borgonha S, Rabe J, Young V & Grinspoon S.
Increased rates of lipolysis among human
immunodeficiency virus-infected men receiving highly
active antiretroviral therapy. Metabolism 2002;
51:1143–1147.
36. Shulman GI. Cellular mechanisms of insulin resistance.
Journal of Clinical Investigation 2000; 106:171–176.
37. Oral EA, Simha V, Ruiz E, Andewelt A, Premkumar A,
Snell P, Wagner AJ, DePaoli AM, Reitman ML, Taylor SI,
Gorden P & Garg A. Leptin-replacement therapy for
lipodystrophy. New England Journal of Medicine 2002;
346:570–578.
38. Caron M, Auclair M, Sterlingot H, Kronprobst M &
Capeau J. Some HIV protease inhibitors alter lamin A/C
maturation and stability, SREBP1 nuclear localization and
adipocyte differentiation. AIDS 2003; 17:2437–2444.
V Jan et al.
©2004 International Medical Press
564
Received 14 November 2003, accepted 5 April 2004