| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Diabetes Research Center, Vrije Universiteit Brussel, Brussels, Belgium
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cells. By using
gold-labeled lipoproteins, we demonstrate receptor-mediated
endocytosis of LDLs and very low-density lipoproteins in rat and human
ß cells. Specific for human ß cells is the fusion of LDL-containing
endocytotic vesicles with lipid-storing vesicles (LSVs;
diameter, 0.6–3.6 µm), which are absent in rodent
ß cells. LSVs also occur in human pancreatic and duct
cells, but these sequester little gold-labeled LDL. In humans
<25 years old, LSVs occupy 1% of the cytoplasmic surface area
in ß, , and duct cells. In humans >50 years
old, LSV surface area in ß cells (11 ± 2% of
cytoplasmic surface area) is fourfold higher than in and duct cells
and 10-fold higher than in ß cells at younger ages
(P < 0.001); the mean LSV diameter in these ß
cells (1.8 ± 0.04 µm) is larger than at younger ages (1.1
± 0.2 µm; P < 0.005). Oil red O staining on
pancreatic sections confirms that neutral lipids accumulate in ß
cells of older donors. We conclude that human ß cells can incorporate
LDL and very low-density lipoprotein material in LSVs. The marked
increase in the LSV area of aging human ß cells raises the question
whether it is caused by prolonged exposure to high lipoprotein levels
such as occurs in Western populations and whether it is causally
related to the higher risk for type 2 diabetes with
aging.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adult male Wistar rats were housed according to the guidelines of the Belgian Regulations for Animal Care. The protocol was approved by the Ethical Committee for Animal Experiments of the Vrije Universiteit Brussel. Rats were sedated and killed with CO2 followed by decapitation. Islets were isolated by collagenase digestion and dissociated in a calcium-free medium containing trypsin and DNase (both from Boehringer Mannheim, Mannheim, Germany).11 Purified ß and non-ß cells were obtained by autofluorescence-activated cell sorting as described previously.11 Cells were cultured in Ham’s-F10 medium supplemented with 2 mmol/L L-glutamine, 50 µmol/L 3-isobutyl-1-methylxanthine, 0.075 g/L penicillin, 0.1 g/L streptomycin, 10 mmol/L glucose, and 5 g/L charcoal-treated bovine serum albumin (fraction V, radioimmunoassay grade; Sigma Chemical Co., St. Louis, MO).
Human islets were isolated from donor pancreas procured by European
hospitals affiliated with Eurotransplant-Bioimplant Services (Leiden,
The Netherlands) and ß Cell Transplant, a multicenter program on
islet cell transplantation in diabetes.12
Islets were
prepared in the central unit, Medical Campus, Vrije Universiteit
Brussel (Brussels, Belgium) using collagenase digestion and Ficoll
gradient purification. The islet-enriched gradient interface was
harvested, washed, and cultured in Ham’s-F10 medium supplemented as
previously described.13
After culture, the human islet
cell preparations consisted of 55 to 65% ß cells, 6 to 12%
cells, and 20 to 30% nongranulated cells, which have been identified
as duct cells.12,13
The cultured preparations were
dissociated by the same trypsin treatment as described for rat islets.
Rat and human islet cell preparations were cultured for 48 hours in polylysine-coated (1 µg/ml, Sigma) wells (Falcon, Franklin Lakes, NJ) at a density of 105 cells/ml before exposure to gold-labeled lipoproteins.
Preparation of LDL
Human lipoproteins were prepared from serum of healthy volunteers after an overnight fast. The VLDL and LDL fractions were isolated by ultracentrifugation,14 with one additional run for LDL. The electrophoretic mobility of LDL on 75% agarose exhibited an Rf of 0.24 ± 0.02 SEM (n = 7). All lipoprotein preparations were filtered through a 22-µm filter (Millipore, Molsheim, France) before use. Their protein concentration was determined with the Pierce BCA kit using bovine serum albumin as standard. Colloidal gold particles were conjugated to LDL by electrostatic surface adsorption of their negative charge. This was accomplished by rapid mixing of 1 ml of colloidal gold (20-nm particle size, British BioCell International, Cardiff, UK) per 100 µl of LDL (concentration, 100–150 µg LDL-protein/ml).15 Conjugates were examined by negative-stain electron microscopy before use. A similar procedure was used for preparing gold-labeled VLDL.
Incubation of Cells with Lipoproteins
Experiments with gold-labeled LDL or VLDL were carried out at 4°C for studies on binding and at 37°C for studies on uptake. After incubation, cells were fixed in cacodylate-buffered glutaraldehyde (4.5%, pH 7.3), postfixed in osmium tetroxide (1%), and embedded in Spurr’s resin. Ultrathin sections were stained with uranylacetate and lead citrate and examined in a Zeiss EM 109 electron microscope.
Ultrastructural localization of acid phosphatase was performed using a modified Gomori reaction.16 After incubation with the gold-labeled lipoprotein and short fixation of the cells, the reaction medium, containing 0.067 mol/L Tris-maleate, 0.36% ß-glycerophosphate, and 0.11% lead nitrate, was added. After a 1-hour incubation at 37°C, cells were prepared for routine transmission electron microscopy, omitting the final uranylacetate-lead citrate staining step.
Characterization of Lipid-Storing Vesicles in Human ß Cells
Morphometric analysis was conducted to determine the individual diameters and surface areas of vesicles containing large (diameter > 0.6 µm) electron-lucent droplets and a peripheral rim of electron-dense material. Isolated islet tissue was examined from donors of different ages. The surface area of these vesicles was expressed as total per cell and as percentage of cellular cytoplasmic surface area (CSA). For each donor preparation, we performed measurements in 20 to 50 cells per cell type and averaged them to one value per cell type. We then averaged these values per donor age group, choosing arbitrarily a young (<25 years) and an old (>50 years) age group. The results are expressed as means ± SEM. The statistical significance of differences was calculated by unpaired Student’s t-test, assuming equal variances.
Staining for Lipids in Intact Pancreatic Tissue
Pancreatic tissue was examined for the presence of neutral lipids by an oil red O reaction on frozen material. The sections were first briefly immersed in 60% isopropanol, then incubated at room temperature for 1 hour in an oil red O solution,17 briefly discriminated in 60% isopropanol, and finally rinsed with water. The oil red O reaction was followed by hematoxylin staining. Consecutive sections were stained for insulin to localize the neutral lipids versus the endocrine tissue. Tissue was fixed for 10 minutes in 4% formaldehyde prepared in phosphate-buffered saline (PBS). After washing with PBS, the sections were incubated for 15 minutes in 90% methanol–10% H2O2, washed again with PBS, and incubated for 30 minutes with 10% normal goat serum. The insulin antibody11 was then applied at a final concentration of 1:5000 for an overnight incubation at 4°C. After washing with PBS, the biotinylated anti-guinea pig antibody (Vector Laboratories, Burlingame, CA) was added at a concentration of 1:1000 for a 30-minute incubation at room temperature. Positivity was visualized after reaction with diaminobenzidine. Insulin staining was also followed by a standard hematoxylin staining.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Incubation of rat ß cells with gold-labeled LDL at 4°C for 2 hours resulted in an association of gold particles with the plasma membrane but not with the intracellular compartment. Single as well as clustered particles occurred along both coated and noncoated regions of the membrane and, occasionally, in coated pits. No membrane-associated particles were noticed when incubation was carried out in calcium-free medium or when unconjugated colloidal gold was used. Membrane binding of gold-LDL was also negligible in the presence of excess unlabeled LDL and was markedly lower after 24 hours of preincubation with 100 to 400 µg/ml LDL at 37°C. No membrane binding was seen in islet endocrine non-ß cells that were incubated in parallel, not even after 4 hours.
Incubation at 37°C resulted in a time-dependent increase in the
number of intracellular gold particles. After 5 minutes, particles were
still associated with the plasma membrane, but, compared with 4°C
incubations, a larger proportion occurred in coated pits and coated
vesicles (Figure 1a)
. From 10 minutes on,
increasing proportions of particles were found in uncoated vesicles,
first as a rim at the inner vesicle membrane and, later on, dissociated
from the membrane. The vesicles containing gold particles varied in
size and in electron-lucent or electron-dense content (Figure 1b)
. Some
vesicles exhibited a positivity for acid phosphatase. The diameter of
vesicles containing gold-labeled LDL was always smaller than 0.6 µm.
|
Incubation of human ß cell preparations with gold-labeled
LDL led to the same observations as with rat ß cells, namely,
association of gold particles with the plasma membrane at 4°C and
their uptake in coated and uncoated vesicles at 37°C. After 3 to 6
hours, gold-containing vesicles appeared in the vicinity of larger
(diameter > 0.6 µm) vesicles that contained both
electron-lucent and electron-dense material, the latter often forming a
peripheral rim under the vesicle membrane or delineating several
spherical electron-lucent compartments within the same vesicle (Figure 2, a and b)
. With longer incubation
periods, gold-containing vesicles were found to fuse with these large
vesicles (Figure 2c)
. After incorporation, gold particles were mainly
present in the peripheral rim of electron-dense material (Figure 2c)
.
At higher magnifications, this gold-containing dense material at the
periphery was dispersed between two membranes, the outer one
surrounding the large vesicle and the inner one surrounding an
intravesicular compartment with mostly electron-lucent material (Figure 3)
. Gold particles remained sequestered
in these large vesicles during subsequent culture for 7 days. A similar
uptake and fusion process was noticed with gold-labeled VLDL.
Incubation with free gold particles did not result in their uptake and
sequestration by the ß cells. In non-ß cells, only little uptake of
gold-labeled LDL or VLDL was seen. The presence of rod-shaped granules
in secretory vesicles18
allows the identification of ß
cells from other endocrine islet cells, even when their relative
proportion is decreased as a result of culture.
|
|
. The diameter of
these membrane-bound vesicles varied widely, from 0.6 to 3.6 µm.
Smaller lipid accumulations were occasionally found in the cytoplasm,
closely associated with endoplasmic reticulum. These lipid
accumulations in vesicles or cytoplasm were not detected in ß cells
isolated and/or cultured from adult (10-week-old) rats or mice (data
not shown). Age-Dependent Accumulation of Lipid-Storing Vesicles in Human ß Cells
Electron microscopy of intact human pancreatic tissue demonstrated
the presence of the same lipid-containing vesicles in the ß cells,
indicating that they were not formed as an artifact during the
isolation or culture procedure. On routine examination of isolated
human islets, we noticed that their abundance markedly increased with
the age of the donor (Table 1)
. No
correlation was found with donor sex or body mass index. When the
surface area of the lipid-storing vesicles (LSVs) was plotted against
the age of the donors, a positive correlation was detected (Figure 4
; r = 0.938;
P < 0.001). We then arbitrarily selected the age
limits for comparing morphometric parameters in young (<25 years) and
older (>50 years) donors. In the donors of the younger age categories,
LSV surface area occupied 0.9 ± 0.5 µm2
per ß
cell, which corresponds to 1.2% of CSA (Table 2)
. These absolute and relative surface
areas were comparable to those measured in cells and in
nongranulated cells of the same preparations (Table 2)
. In ß cells
from donors of the older age category (>50 years), LSV surface area
was more than 10-fold higher (11.9 ± 1.2
µm2/ß cell; P < 0.001),
corresponding now to 11.1% of CSA (P < 0.001).
Higher surface areas were also measured in cells and in
nongranulated cells from the older donors, but they remained, in both
absolute and relative terms, 4- to 10-fold smaller than in the ß
cells (Table 2)
. In ß cells from the older donors, the mean diameter
of these vesicles was larger than in ß cells from the younger donors
(1.8 ± 0.04 versus 1.1 ± 0.2 µm;
P < 0.005). In view of their ultrastructural
characteristics, these vesicles were denoted as lipid-storing vesicles.
|
|
|
and ß cells exhibited larger CSAs than
in donors younger than 25 years (P < 0.05;
Table 2
). In ß cells, the increase (28 ± 10
µm2/cell) is 40% attributable to the
increase in LSV area (increase of 11 ± 0.9
µm2/cell; Table 2
). This is not the case in
cells, in which the increase in LSV surface area is marginal (1.9
± 0.4 µm2/cell) to the increase in CSA
(22 ± 8 µm2/cell). Higher age was not
associated with a larger CSA in nongranulated cells
(P > 0.05). Lipid Accumulation in Pancreatic Islets from Older Donors
In cryostat sections of frozen pancreatic tissue, a positive oil
red O staining was observed in the islets, as indicated by its
colocalization with the insulin-positive regions (Figure 5, a and b)
. This staining pattern was
found in all sections from donors older than 50 years, but was absent
in tissue from donors younger than 25 years (Figure 5, c and d)
. No oil
red O staining was noticed in rat or mouse pancreatic tissue,
regardless of age (data not shown). When the oil red O staining was
performed on human tissue that was embedded in paraffin before
sectioning, no positive reaction was obtained, indicating that its
positivity in frozen tissue was not caused by the presence of
lipofuscin.19
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
cells.10
We have
now used gold-labeled LDL to monitor the intracellular track of LDLs in
rat and human ß cells. This technique confirmed the presence of
LDL-binding sites on the plasma membrane of ß cells, with
characteristics that were similar to those observed in the binding
studies with I125-labeled LDL. Previous data
characterized these sites as high-affinity receptors for LDL and VLDL
but not for acetylated LDL.10
Half-maximal binding was
reached at 10 to 15 µg LDL/ml, a concentration that is comparable to
the estimated interstitial LDL levels in the rat but is fivefold lower
than those levels in humans with Western life styles.20
In
this latter condition, LDL receptors on ß cells are thus expected to
be down-regulated and saturated, mediating a steady uptake of LDL and
VLDL. Uptake of LDL by ß cells was first noticed during binding
studies with I125-LDL.10
It is now
documented ultrastructurally as a receptor-mediated endocytosis,
occurring in both rat and human ß cells but negligible in cells.
From 30 minutes on, some of the gold-LDL-containing vesicles were
positive for acid phosphatase, suggesting fusion of endocytotic
vesicles with lysosomes. This process has been described for other cell
types,21
in which the lysosomes have been found to degrade
the lipoprotein.22 In human ß cells, gold-LDL–containing vesicles were found to fuse with large (diameter >0.6 µm) lipid-containing vesicles. These large vesicles were not found in rat or mouse ß cells. Their ultrastructure does not resemble the cytoplasmic triglyceride accumulations that have been described in steatosis,19 nor do they present the features of lipofuscin accumulations.19 They contain several electron-lucent compartments, often surrounded by a membrane, as well as electron-dense material, often forming a peripheral rim under the outer vesicle membrane. This peripheral rim sometimes exhibits a positivity for acid phosphatase, suggesting prior fusion with lysosomes. Because human ß cells also presented smaller cytoplasmic lipid accumulations (data not shown), it is conceivable that the endoplasmic reticulum first envelops this cytoplasmic material and that this storage vesicle then fuses with lysosomes. A process of receptor-mediated endocytosis could deliver lipoproteins to this vesicle via secondary lysosomes. Accumulation of lipid-containing vesicles in human ß cells could be the result of a prolonged imbalance between lipid uptake, processing, and consumption. To which extent inappropriate uptake of VLDL and LDL is responsible for the development and/or increase of this subcellular compartment is not known. However, the marked postnatal increase in human LDL levels20 to concentrations that continuously saturate the LDL receptors in ß cells makes lipoproteins candidate contributors to this process.
LSVs were not only noticed in human ß cells, but also in adjacent
and duct cells. In donors under the age of 25 years, they occupied 1%
of the CSA in each of these pancreatic cell types. In donors older than
50 years, their surface area was larger, reaching 2 to 3% of the CSA
in and duct cells and 11% in ß cells. Increasing age thus causes
lipid accumulation in ß cells. It is conceivable that this process
involves receptor-mediated uptake of LDL and VLDL and therefore is less
pronounced in the adjacent non-ß cells. Further work is needed to
clarify the underlying mechanisms. This study strongly suggests that
LSVs appear after birth. In our 1-year-old donor, only 7% of the ß
cells exhibited LSVs, with a mean diameter of 0.7 ± 0.3 µm.
With increasing age, more ß cells were found to contain LSVs, and the
mean LSV diameter increased to 1.8 ± 0.04 µm in donors older
than 50 years. As discussed earlier, these LSVs can incorporate LDL and
VLDL that is taken up by receptor-mediated endocytosis. They might
accumulate lipids if their clearance is saturated, as is conceivable
with the high lipoprotein levels in Western society. Furthermore, ß
cells have a long life span, which gives them the time to accumulate
products that are inappropriately cleared. Such a process is not
necessarily specific for ß cells. LDL receptors indeed occur in many
other cell types, but a lower rate of uptake, a higher lipoprotein
clearance, and/or a shorter cellular life span might result in absent
or less pronounced accumulation of LSVs in a number of cell types.
That fat can accumulate in human islet tissue was already reported by Weichselbaum and Stangl at the beginning of this century.23 This finding was even considered to be a characteristic of diabetes because diabetic patients showed more fat in the islets of Langerhans than age-matched nondiabetic controls.23 In a subsequent paper, Weichselbaum described a vacuolization of the islets in 53% of diabetic patients.24 Later reviews have questioned the specificity and frequency of this tissue alteration25 and even warned that postmortem autolysis may be confused with the vacuolization.26 These studies have not, however, always distinguished between fatty infiltration in islet tissue and fat accumulation in the endocrine islet cells. This work demonstrates that increasing age is associated with an accumulation of neutral lipids in human ß cells but not in the surrounding exocrine cells. Analysis of the LSV content is now needed to directly demonstrate its lipid nature and identify the components. The scarcity of isolated human islet tissue will certainly be an obstacle for such subcellular fractionation and chemical characterization, but should be undertaken before speculations are made on the possible significance of these accumulations. We do not yet know whether they influence the functions of human ß cells and their ability to respond to metabolic needs. Intracellular accumulation of lipids could chronically expose the ß cells to elevated levels of free fatty acids, which, according to the lipotoxicity concept,27 can impair the cellular responsiveness to secretory stimuli and even cause cell death.9 Such a mechanism has been proposed for the development of obesity-related diabetes in rodent models, in which triglyceride deposits were formed in the pancreatic islets.8,28 It should now be examined whether the lipid accumulations in aging human ß cells bear functional consequences and could therefore explain, at least in part, the higher incidence of type 2 diabetes in aging humans.29
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by grants from the European Community (BMH-CT95–1561), the Juvenile Diabetes Foundation International (JDF 995004), the Belgian
Fonds voor Wetenschappelijk Onderzoek (F.W.O.G.0039.96 and G.0376.97), and the services of the Prime Minister (Interuniversity Attraction Pole P4/21). M. C. is Aspirant and A. G. is Postdoctoral Fellow of the Fund for Scientific Research-Flanders (F. W. O.).
Accepted for publication September 15, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. Marselli, J. Thorne, Y.-B. Ahn, A. Omer, D. C. Sgroi, T. Libermann, H. H. Otu, A. Sharma, S. Bonner-Weir, and G. C. Weir Gene Expression of Purified {beta}-Cell Tissue Obtained from Human Pancreas with Laser Capture Microdissection J. Clin. Endocrinol. Metab., March 1, 2008; 93(3): 1046 - 1053. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Ovalle and D. S.H. Bell Effect of Rosiglitazone Versus Insulin on the Pancreatic {beta}-Cell Function of Subjects With Type 2 Diabetes Diabetes Care, November 1, 2004; 27(11): 2585 - 2589. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-E. Roehrich, V. Mooser, V. Lenain, J. Herz, J. Nimpf, S. Azhar, M. Bideau, A. Capponi, P. Nicod, J.-A. Haefliger, et al. Insulin-secreting beta -Cell Dysfunction Induced by Human Lipoproteins J. Biol. Chem., May 9, 2003; 278(20): 18368 - 18375. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Cnop, J. C. Hannaert, A. Y. Grupping, and D. G. Pipeleers Low Density Lipoprotein Can Cause Death of Islet {beta}-Cells by Its Cellular Uptake and Oxidative Modification Endocrinology, September 1, 2002; 143(9): 3449 - 3453. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Andreolas, G. da Silva Xavier, F. Diraison, C. Zhao, A. Varadi, F. Lopez-Casillas, P. Ferre, F. Foufelle, and G. A. Rutter Stimulation of Acetyl-CoA Carboxylase Gene Expression by Glucose Requires Insulin Release and Sterol Regulatory Element Binding Protein 1c in Pancreatic MIN6 {beta}-Cells Diabetes, August 1, 2002; 51(8): 2536 - 2545. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. JEQUIER Leptin Signaling, Adiposity, and Energy Balance Ann. N.Y. Acad. Sci., June 1, 2002; 967(1): 379 - 388. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. G. Hanley, G. McKeown-Eyssen, S. B. Harris, R. A. Hegele, T. M. S. Wolever, J. Kwan, and B. Zinman Cross-Sectional and Prospective Associations between Abdominal Adiposity and Proinsulin Concentration J. Clin. Endocrinol. Metab., January 1, 2002; 87(1): 77 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. E. Kahn The Importance of {beta}-Cell Failure in the Development and Progression of Type 2 Diabetes J. Clin. Endocrinol. Metab., September 1, 2001; 86(9): 4047 - 4058. [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. UNGER and L. ORCI Diseases of liporegulation: new perspective on obesity and related disorders FASEB J, February 1, 2001; 15(2): 312 - 321. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
