| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Pediatrics, University of Ulm (P.F., K.M.D., M.W.), D-89075 Ulm, Germany; Department of Pathology, University of Ulm (P.M., I.M.), D-89070 Ulm, Germany; Department of Pediatrics, University of Bonn (L.B.), D-53113 Bonn, Germany; and Department of Diabetes Biology, Novo Nordisk A/S (H.T.), DK-2880 Bagsvaerd, Denmark
Address all correspondence and requests for reprints to: Martin Wabitsch, M.D., Department of Pediatrics, University of Ulm, Prittwitzstrasse 43, 89075 Ulm, Germany. E-mail: . martin.wabitsch{at}medizin.uni-ulm.de
Abstract
Congenital generalized lipodystrophy (CGL) is characterized by the absence of adipose tissue from birth due to a hypothetical differentiation block. The genetic causes of CGL are still not completely understood.
Subepidermal, fibroblast-like cells were prepared from the sc tissue of an infant with CGL. Preadipocytes from sc adipose tissue and foreskin fibroblasts from three healthy patients, respectively, were used as controls. Adipose differentiation was induced in cultured cells by exposure to 10 nM insulin, 200 pM T3, 1 µM cortisol, and 2 µM rosiglitazone. Under these conditions 42% of the subepidermal, fibroblast-like CGL cells developed into mature adipocytes. Adipogenic differentiation was dependent on rosiglitazone. The differentiation rate was comparable in cultures of preadipocytes from control patients maintained under the same conditions (53%, 38%, and 20%). In contrast, foreskin fibroblasts did not differentiate into adipocytes. Morphological changes in CGL cells during differentiation were associated with the expression of fat cell-specific mRNAs (PPAR
, leptin, and glut-4). In addition, these cells revealed characteristic features of mature adipocytes, such as lipogenesis or leptin secretion.
Taken together, we show that adipocyte precursor cells were present in subepidermal tissue of a patient with CGL and were able to differentiate into adipocytes in the presence of a thiazolidinedione. These findings strongly support clinical trials with thiazolidinediones in patients with CGL.
CONGENITAL GENERALIZED lipodystrophy (CGL) (OMIM 269700) is an extremely rare disorder reported initially by Berardinelli (1) and Seip (2). Affected patients have a nearly complete absence of metabolically active adipose tissue from birth (3) and usually have a voracious appetite. Metabolic consequences, such as high serum triglyceride concentrations, severe hyperinsulinemia, and insulin resistance, are present during infancy. Secondary comorbidities develop early in life, including diabetes and its complications, hypertrophic cardiomyopathy, fatty infiltration of the liver, and liver cirrhosis, finally leading to a poor life expectancy (3). As no cure of patients with CGL is possible, treatment consists of a fat-reduced diet and management of the developing complications, including the use of hypoglycemic drugs.
The genetic basis of CGL, which has an autosomal recessive inheritance, is still not completely understood. Several candidate genes have been excluded (4, 5, 6). Linkage analysis studies localized a gene responsible for the disorder in some patients to human chromosome 9q34 (7). A new disease locus, called bscl2, has been identified recently (8) on chromosome 11q13. The encoded protein of unknown function was called Seipin. Most variants are null mutations that probably lead to severe disruption of Seipin expression. In addition to these genetic studies in large human pedigrees, a new candidate gene for lipodystrophy was identified in a murine model. Fld mice have features of human lipoatrophy and carry mutations in the fatty liver dystrophy gene, which encodes a nuclear protein called lipin (9).
Adipocytes originate from mesenchymal multipotent stem cells that develop into adipocyte precursor cells (usually termed preadipocytes) by largely unknown mechanisms. Preadipocytes are found in large numbers in human adipose tissue. In vitro, isolated preadipocytes can be stimulated to differentiate into mature adipocytes by treatment with specific adipogenic factors that finally activate PPAR/RXR receptors (10, 11).
We had the opportunity to study subepidermal fibroblast-like cells obtained by skin biopsy of an infant with CGL. We report here that in vitro these cells are able to differentiate into adipocytes upon stimulation with rosiglitazone, a PPAR/RXR receptor agonist.
Materials and Methods
Materials
Culture media were obtained from Life Technologies, Inc. (Karlsruhe, Germany). 14C-Labeled D-glucose was purchased from Amersham Pharmacia Biotech (Freiburg, Germany). Other chemicals and reagents were purchased from Sigma (Taufkirchen, Germany). Rosiglitazone (BRL 49653) was a gift from SmithKline Beecham (London, UK). Recombinant human insulin was provided by Novo Nordisk (Gentofte, Denmark).
Subjects and adipose tissue samples
Patient with CGL. The male infant was born by cesarian section due to progressive hydrops. Echocardiography revealed hypertrophic cardiomyopathy with left ventricular outflow tract obstruction developing later and pulmonary valve stenosis. In the first days recurrent episodes of hyperglycemia and lipemic serum were observed. Further blood analysis showed hyperinsulinemia and dyslipoproteinemia with extremely elevated concentrations of very low density lipoprotein up to 1.04 g/liter (normal range, 0–0.18 g/liter) and triglyceride levels up to 21 g/liter after induction of mother’s milk. After resolution of the edema, the complete absence of sc adipose tissue and pronounced muscularity became apparent. An umbilical hernia was present. The diagnosis of CGL was suspected. A whole body magnetic resonance imaging study (T1W COR SPIR) on d 63 demonstrated the absence of sc adipose tissue, sparing the palms, soles, scalp, and the periarticular regions. Perirenal fat tissue was preserved. Thus, the definitive diagnosis of CGL was confirmed. Two measurements of serum leptin showed levels of 0.02 and 0.01 ng/ml (less than first age-related percentile).
A skin biopsy of 2 x 2 mm2 was taken from the subscapular area by stamp comprising the complete skin above the muscle fascia. No sc adipose tissue was macroscopically visible.
Control subjects. To compare the differentiation capacity of subepidermal fibroblast-like cells of the CGL patient with that of subepidermal fibroblast-like cells from normal patients we obtained tissue samples from the foreskin from three healthy boys of 3, 3, and 5 yr of age undergoing circumcision. In addition, preadipocytes were isolated from subepidermal adipose tissue from two male and one female healthy infants of 1, 6, and 7 months of age undergoing herniotomia. Informed consent was obtained from the parents of these patients. The study was approved by the ethical committee of the University of Ulm.
Histology
The biopsy material was routinely fixed in 4% neutral buffered formalin. After routine tissue processing, 2-µm-thick paraffin sections were made, which were subsequently stained with hematoxylin/eosin.
Cell isolation and culture
Subepidermal cells. The skin tissue of the patient with CGL and samples of the three foreskins equal in size were placed with the subepidermal site on the surface of a culture well. After 2 h, DMEM/Ham’s F-12 (1:1) containing 10% FCS and antibiotics was added. After 7 d, the tissue was removed, and between 200–300 fibroblast-like cells were visible on the culture well surface. Cells were detached with HBSS containing 0.05% trypsin and 0.02% EDTA and multiplied in subcultures until enough cells were available to perform the experiments described below. Cells of the 8th, 12th, and 16th generations were used to induce adipose differentiation.
Preadipocytes. Preadipocytes were prepared from adipose tissue samples by collagenase digestion and cultured according to an established protocol (12, 13). Preconfluent cells were repeatedly subcultured in DMEM/Ham’s F-12 (1:1) containing 10% FCS and antibiotics (serum-containing medium) until cells of the 8th, 12th, and 16th generations were obtained, which were then used in the experiments described below for comparison. In the experiments performed, either serum-containing medium or serum-free basal medium [DMEM/F-12 (1: 1) supplemented with 33 mM biotin, 17 mM pantothenate, and antibiotics] was used.
The subepidermal, fibroblast-like cells from the patient with CGL are subsequently called subepidermal CGL cells, the subepidermal, fibroblast-like cells from the foreskin of the control patients are called subepidermal foreskin cells, and the preadipocytes obtained from sc adipose tissue from control patients are called preadipocytes from control patients.
Induction of adipose differentiation
After reaching near confluence, subepidermal cells and preadipocytes were repeatedly washed with PBS buffer and cultured thereafter in serum-free, basal medium supplemented with 10 µg/ml transferrin, 10 nM insulin, 200 pM T3, and 1 µM cortisol (adipogenic medium) with or without addition of 2 µM rosiglitazone (BRL 49653) for the first 4 d. The medium was changed every 4 d. Morphologically differentiated adipocytes were obtained after 20 d. The number of differentiated cells in the monolayers was estimated by direct counting using a net micrometer.
Nile Red and 4',6-Diamidino-2-phenylindole (DAPI) staining
Subepidermal CGL cells, subepidermal foreskin cells, and preadipocytes from control patients were seeded on culture slides (Becton Dickinson and Co., Heidelberg, Germany). After reaching near confluence, cells were cultured in either adipogenic or serum-free, basal medium. After 20 d, monolayers were washed twice with PBS and then fixed for 10 min at room temperature with 2% paraformaldehyde. Nile Red (Sigma, Taufkirchen, Germany) was added to a final concentration of 0.5 µg/ml. After removing the staining solution, slides were mounted in Mowiol (Calbiochem, Bad Soden, Germany) containing 0.2 µg/ml DAPI (Sigma). Slides were viewed with an Olympus Corp. fluorescence microscope (Melville, NY), and pictures were taken using Analysis software.
Measurement of lipogenesis and leptin concentrations in culture medium
Lipogenesis was assessed as described previously (14) by incubating cells with 3-D-[14C]glucose for 48 h, followed by determination of the radioactivity incorporated into cellular lipids using cells of the eighth generation. Leptin concentrations in medium of cultures of generation 8 were measured using a specific RIA (15).
RT and duplex PCR
RT reaction and duplex PCR were performed as described previously with some modifications (16, 17). RNA was extracted from cells at time points as indicated. Total RNA was digested with ribonuclease-free deoxyribonuclease, and 5 µg of each were reverse transcribed using Superscript RT kit and random hexamers (both Life Technologies, Inc.) according to the instruction of the manufacturer. PCR primer pairs were picked from human cDNA libraries (Wisconsin Package version 9.1, Genetic Computer Group, Madison, WI; PPAR2: accession no. U63415; forward primer position, 23–44; reverse primer position, 320–340; leptin: accession no. U18915; forward primer position, 39–58; reverse primer position, 410–428; Glut-4: accession no. M20747; forward primer position, 1186–1203; reverse primer position, 1643–1660; aP2: accession no. J02874; forward primer position, 71–90; reverse primer position, 431–450; Sp1: accession no. J03133; forward primer position, 757–777; reverse primer position, 977–997). Semiquantitative duplex PCR with the transcription factor Sp1 as internal standard was performed on 0.1 µg cDNA using the Taq PCR Core Kit (QIAGEN, Hilden, Germany). Thermocycling was as follows: denaturation at 95 C for 5 min; 25 cycles: 95 C for 1 min, 56 C for 1 min, and 72 C for 1 min; and prolongation for 10 min at 72 C. Primers were used at a final concentration of 0,5 µM each, dNTPs at 10 µM, 2.5 U Taq DNA polymerase in a total of 50 µl, 10x buffer, and 5x Q-solution according to the protocol. Omitting RT reaction resulted in no detectable PCR products.
Electrophoretically separated PCR products were ethidium bromide stained, and the fluorescence images were analyzed by an ImageMaster VDS (Pharmacia Biotec, San Francisco, CA). The ratio of the background corrected integrated optical densities of the DNA bands related to SP1 expression were calculated.
Results
Histological examinations
Histological examination of the skin biopsy of the patient with CGL showed complete absence of sc adipose tissue. The dermal structure, the dermal appendices, and the epidermis were normal (Fig. 1A
).
|
When subepidermal CGL cells were transferred to serum-free, adipogenic medium supplemented with 2 µM rosiglitazone for the first 4 d a large number of the cells developed within 20 d the typical morphological criteria of in vitro differentiated adipocytes such as a round shape and a cytoplasm filled with lipid droplets (Fig. 2, A and B). As suggested from results of earlier studies preadipocytes from control patients (eighth generation) cultured under the same conditions also partly differentiated into adipocytes. Subepidermal fibroblast-like cells from foreskin (Fig. 1B
), however, did not change their morphology and did not differentiate into adipocytes when cultured in adipogenic medium supplemented with rosiglitazone (Fig. 2, E and F). Neither subepidermal CGL cells nor subepidermal fibroblast-like cells from foreskin developed any morphological signs of mature adipocytes when maintained in serum-free, basal medium (Fig. 2, C and D, and G and H).
|
). The percentage of differentiated adipocytes in cultures of subepidermal cells of the patient with CGL was as high as the percentage of newly developed adipocytes (Fig. 2, I and J) in the cultured preadipocytes derived from sc adipose tissue of control patients (Fig. 1C). No differentiation could be observed after culture in serum-free, basal medium (Fig. 2, K and L).
|
Expression of specific genes in in vitro differentiated adipocytes from the CGL patient
The phenotypical changes during adipose differentiation of subepidermal CGL cells were associated with the expression of adipocyte-specific genes. The expression of mRNAs encoding Glut-4 (Fig. 3B), leptin (Fig. 3C), and aP2 (Fig. 3D) were similar to those seen in differentiating cells of control patients. PPAR mRNA (Fig. 3A), however, is obviously expressed to a lesser extent in in vitro differentiated adipocytes derived from the CGL patient in comparison to in vitro differentiated adipocytes from healthy patients.
|
To demonstrate two functional characteristics of mature fat cells we investigated the capacity of the cells to synthesize lipids from [14C]glucose and to produce and secrete leptin. Insulin at 1 nM was able to stimulate lipogenesis in differentiated adipocytes of subepidermal CGL cells 10-fold (4934 ± 384 vs. 460 ± 128 cpm), which was comparable to results obtained in in vitro differentiated human preadipocytes (14) and which was not found in undifferentiated subepidermal CGL cells, in undifferentiated preadipocytes from control patients, and in subepidermal foreskin cells (data not shown).
In accordance with the expression of leptin mRNA in differentiated adipocytes of subepidermal CGL cells, leptin production in these cultures could be demonstrated that increased steadily during adipose differentiation. When cells were cultured in adipogenic medium with rosiglitazone present for the first 4 d leptin could be firstly detected in culture medium between d 8 and d 12 and reached thereafter maximal levels up to 3.9 ng/dl·105 adipocytes/72 h which is comparable to results found in in vitro differentiated preadipocytes (14). Leptin was not detectable by the sensitive RIA (sensitivity, 0.02 ng/ml) used in cultures of undifferentiated subepidermal CGL-cells, in undifferentiated preadipocytes from control patients, and in subepidermal foreskin cells.
Discussion
The molecular basis for the finding that adipose tissue is almost completely absent in patients with CGL is still not completely understood. Possible pathogenetic defects have been hypothesized (3). Theoretically, the absence of adipose tissue in patients with CGL could result from agenesis of adipocyte precursor cells, failure of adipocyte precursor cells to differentiate into adipocytes, or failure of mature adipocytes to synthesize or store triglycerides.
In the present study we have investigated subepidermal cells obtained from a patient with CGL. We could show that these subepidermal mesenchymal derivatives differentiate in vitro into adipocytes after stimulation with the thiazolidinedione, rosiglitazone. Furthermore, we have demonstrated that subepidermal cells from the foreskin of healthy patients cultured under the same conditions do not differentiate into adipocytes. From these findings it can be concluded that in subepidermal tissue of the patient with CGL, preadipocytes or committed precursor cells at an earlier stage of development were present. We could show furthermore that the differentiation rate in cultures of subepidermal cells of the patient with CGL was comparable with the one found in cultures of preadipocytes isolated from human adipose tissue from healthy control patients. These findings support the hypothesis of a block in the adipose differentiation process of precursor cells in patients with CGL.
The concentrations reached in plasma of patients treated with rosiglitazone are comparable to the concentrations that could induce adipose differentiation in our in vitro experiments. After an oral dose of 2 mg rosiglitazone (18) a maximal plasma concentration of 150 ng/ml is achieved (3 x 10-7 M). After 12 h, the plasma concentration is approximately 20 ng/ml (5 x 10-8 M). In all of our in vitro experiments shown here, rosiglitazone was used for the first 4 d of differentiation at a concentration of 2 x 10-6 M. However, adipose differentiation could be also induced using concentrations of 10-7, 10-8, and 10-9 M, comparable to the plasma concentrations achieved after oral administration of 2 mg rosiglitazone.
Whole body magnetic resonance imaging as well as autopsy findings revealed a near-complete absence of adipose tissue from sc areas, intraabdominal and intrathoracic regions, and bone marrow, whereas normal amounts of adipose tissue were present in the orbits, crista galli, buccal region, tongue, palms, soles, scalp, perineum, vulva, periarticular regions, epidural area, and pericalyceal regions of the kidney (3). On the basis of this observation, one group of scientists has introduced a hypothetical classification of human white adipose tissue into metabolically active and mechanical adipose tissue (3). The theoretical reasons for the different behavior and the molecular basis for it are not known. It has to be suggested, however, that a possible block in the adipose differentiation process is present only in precursor cells of the affected tissue sites.
During periods of weight gain in man resulting in an increase in adipose tissue new adipocytes can be formed from precursor cells throughout life (10). It is unknown whether in postnatal life human sc tissue contains mesenchymal stem cells that are able to differentiate into adipocytes, if already committed preadipocytes are present, or both. Although there are several recent reports on markers identifying adipocyte precursor cells derived from rats (19), there is to our knowledge at present no specific marker for human preadipocytes available. It was therefore not possible to characterize the investigated cells more precisely.
Surgical implantation of adipose tissue into A-ZIP/F-1, mice which have a severe form of lipoatrophic diabetes, reversed diabetes and dramatically lowered insulin levels, demonstrating that diabetes was caused by the lack of adipose tissue. Furthermore, all aspects of the A-ZIP/F-1 phenotype, including hyperphagia, were reversed (20). Treatment of these mice with leptin resulted in a moderate improvement of glucose and insulin levels and a slight reduction in food intake (21), suggesting that leptin as a hormone produced in adipose tissue might be at least in part responsible for these positive effects. However, overexpression of leptin rescued these mice from insulin resistance and diabetes (22).
The clinical course of patients with CGL is complicated by early-onset diabetes, accelerated atherosclerosis, and hypertrophic cardiomyopathy leading to death at early ages (3). Rigid energy and fat restriction is the mainstay of treatment supported by the use of anorexiant drugs in patients with CGL (23). However, the outcome of such dietary interventions is rather poor, and there exists no evidence that the complications of the disease can be slowed in their appearance. Clinical trials of leptin in patients with CGL have been recommended on the basis of findings in aP2-SREBP-1c mice, a transgenic mouse model with moderate fat deficiency (24), and in A-ZIP/F-1 mice, as mentioned above. An open-labeled study to investigate the efficacy of leptin replacement in patients with lipoatrophy was started at two centers (25), and the first results suggest that patients seem to benefit from this treatment (Arioglu-Oral, E., unpublished data).
The severe insulin resistance seen in patients with CGL with regard to hepatic glucose production as well as muscle glycogen synthesis and lipid oxidation may be a late consequence of the initially present lack of storage capacity for fatty acids (26).
Besides their differentiation-promoting capacity, thiazolidinediones have a strong insulin-sensitizing effect. Recently, they have been introduced in the treatment of type 2 diabetes mellitus. Clinical studies have shown that the drug also has beneficial effects on other aspects of the metabolic syndrome (27).
On the basis of their mode of action, thiazolidinediones should be an effective drug for managing the metabolic disturbances occurring in patients with CGL. Recently, data from a first study were published showing troglitazone to improve metabolic control and to increase body fat in patients with lipoatrophic diabetes (28). Patients with CGL were included in this study. The doses of troglitazone used were about 100-fold higher than the trosiglitazone doses usually given.
In conclusion, our findings suggest that in patients with CGL adipocyte precursor cells are present in subepidermal tissue. Under in vitro conditions these cells differentiate into adipocytes upon stimulation with rosiglitazone. With respect to possible other beneficial metabolic effects of this drug, our findings support clinical trials with thiazolidinediones in patients with CGL.
Acknowledgments
Footnotes
Abbreviations: CGL, Congenital generalized lipodystrophy; DAPI, 4',6-Diamidino-2-phenylindole.
Received August 8, 2001.
Accepted January 16, 2002.
References
2: genetic mapping, identification of a variant in the coding sequence, and exclusion as the gene responsible for lipoatrophic diabetes. Diabetes 47:490–492[Medline]
This article has been cited by other articles:
 |
|
 |
|
  H. Koutnikova, T.-A. Cock, M. Watanabe, S. M. Houten, M.-F. Champy, A. Dierich, and J. Auwerx Compensation by the muscle limits the metabolic consequences of lipodystrophy in PPAR{gamma} hypomorphic mice PNAS, November 25, 2003; 100(24): 14457 - 14462. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
