вход по аккаунту


Nonrandom positioning of Golgi apparatus in pancreatic B cells.

код для вставкиСкачать
THE ANATOMICAL RECORD 213:182-186 (1985)
Nonrandom Positioning of Golgi Apparatus in
Pancreatic B Cells
Institute of Histology and Embryology, University Medical Center,
1211 Geneva 4, Switzerland
The Golgi apparatus has a polarized distribution in a variety of cell
types, and in certain epithelia its intracellular position correlates with the direction
of vectorial transport and release of secretory products. We have studied quantitatively the orientation of the Golgi apparatus in pancreatic islet B cells by examining
its location with respect to the juxtacapillary face of the cell. This was done a) at the
light microscope level on semithin sections of osmium-impregnated islets, and b) at
the ultrastructural level on thin sections of intact pancreas. The data obtained with
both methods show that in pancreatic B cells the Golgi apparatus is preferentially
located in the position opposite to the juxtacapillary face of the cell. These results
suggest a structural polarization of B cells, which may be important for the coordinate function of islets of Langerhans.
Polarization of cytoplasmic components and plasma
membrane domains is a n essential feature of transporting and secretory epithelial cells (De Camilli et al., 1974;
Palade, 1975; Pisam and Ripoche, 1976; Rodriguez-Boulan and Sabatini, 1978; Cereijdo et al., 1980; Reggio et
al., 1982). A typical example is provided by the pancreatic exocrine cell, in which the ordered distribution
of cytoplasmic organelles correlates with a polarized
release of secretory product into the acinar lumen (Palade, 1975). However, in endocrine cells (with the notable exception of thyroid follicular cells, which are
organized into cavitary structures) manifest evidence of
polarization has not been reported.
Islets of Langerhans are microorgans whose coqstituent cells are arranged in a well-defined topographical
pattern (for a review, see Orci, 1982). Although their
organization has been thoroughly studied in recent
years, with special regard to the structural and functional interrelationships between the various endocrine
cell types (Orci, 1982), the question of the polarity of
islet cells has so far received only limited attention.
Polarization of pancreatic B cells was previously suggested by qualitative observations indicating a clustering of secretory granules close to capillaries (Ferreira,
1957; Bencosme and Pease, 1958; Hellerstrom, 19771, as
well as a juxtacapillary thickening of the subplasmalemma1 microfilamentous cell web (Leblond et al., 1960).
Recent evidence has suggested a functional compartmentalization of the islet interstitium that limits the
diffusion of secreted hormone in that space (Kawai et
al., 1982). To study further the issue of islet cell polarization, we have examined the possibility that Golgi apparatus has a nonrandom positioning in B cells. The
Golgi apparatus plays a central role in the processing
and packaging of secretory polypeptides synthesized in
the rough endoplasmic reticulum, and in their ordered
transport to the plasma membrane for exocytotic release
(see Palade, 1975; Tartakoff, 1980; Farquhar and Palade,
0 1985 ALAN R. LISS. INC
1981, for general reviews; and Orci, 1982, for the role of
Golgi apparatus in B cell secretion). This organelle has
a polarized distribution in a variety of cell types (Weinstock and Leblond, 1974; Trelstad, 1970; Trelstad, 1977;
Garant and Cho, 1979; Holmes and Trelstad, 1980;
Icardo et al., 1982), and in certain epithelia its intracellular position correlates with the direction of vectorial
transport and release of secretory product (Weinstock
and Leblond, 1974; Palade, 1975; Leblond and Bennett,
1977). In endocrine cells a similar correlation has been
suggested by the observation of a polarized position of
the Golgi apparatus in relation to capillaries, e.g., in
parathyroid (Courrier and Reiss, 1922) and pituitary
cells (Reiss, 1922).However, observations on the location
of Golgi elements in B cells have not been in agreement
(Ferreira, 1957; Munger, 1958).
The quantitative evaluation reported in this paper
shows that in pancreatic B cells the Golgi apparatus
occupies preferentially a position opposite the capillary
face of the cells.
The orientation of Golgi apparatus in islet cells was
quantified on a) semithin sections of osmium-impregnated isolated islets and b) thin sections of intact pancreas. For reasons of clarity the two methods are
presented separately.
Isolated Islets
Islets of Langerhans were isolated from adult rat
(about 250 gm) pancreas by collagenase digestion (Lacy
.and Kostianovsky, 1967).No attempt was made to separate splenic and duodenal portions. The islets were collected in gassed (95%02,5% CO2) Krebsminger solution,
pH 7.4, containing 0.5 mg/100 ml albumin and 2.8 mM
Received December 20,1984; accepted May 23, 1985.
Osmium impregnation
Quantification of Golgi Orientation
The orientation of the Golgi apparatus was evaluated
After the collection, islets were fixed immediately in
foil-covered sealed tubes containing 2% osmium tetrox- with respect to the capillary face of the cell, which was
ide in distilled water. The incubation was continued for defined as the segment of the B cell plasma membrane
40 hours at 40°C with one change of fixative at 24 hours juxtaposed to a capillary border. The evaluation was
(Hayat, 1972). Dehydration was in graded ethanols and done according to a modification of the method described
by Trelstad (1970). Negatives of selected cells were proembedding in Epon 812.
jected onto a four-quadrant grid (see Fig. 1)which was
Three-dimensional reconstruction
positioned so that the nuclear profile fitted symmetriSerial 2-pm-thick semithin sections were photo- cally into a system of closely spaced concentric circles
graphed with a Zeiss photomicroscope using a 40 x oil- drawn about the grid center. The Golgi quadrant (GI
immersion phase-contrast objective. Negative pictures was oriented so that the largest amount of either stained
of the central portion of a n islet (i.e., > 3 cell widths Golgi (light microscopy) or Golgi cisternae (electron mifrom the periphery) were projected at a final magnifica- croscopy) was situated symmetrically inside it. The lotion of 2 , 0 0 0 ~ .Drawings of capillary, nuclear, and cation of the capillary face was described by a single
stained Golgi profiles were then transferred to perspex point on the B cell plasma membrane, perpendicular to
sheets (Goldstein and Davis, 1968; Baetens et al., 1979) the midpoint of a line joining the two extremities of the
which were superimposed and positioned using concen- capillary face. Capillary faces were thus assigned to the
quadrant in which this point fell, i.e., either the Golgi
tricity of the nuclear profiles.
quadrant (GI, or quadrants adjacent to (A1 and A21 or
Light microscopy
opposite (0)the Golgi apparatus
Semithin (1-pm-thick)sections of osmium impregnated
isolated islets were stained with 1%toluidine blue in Statistics
1%borax and photographed piecemeal in overlapping
To test whether the three samples of intact pancreas
fields with a 40x objective. The negatives were pro- studied by electron microscopy were derived from a comjected a t a final magnification of 3,000 x , and were ana- mon population, the diameters of area-equivalent circles
lysed for the orientation of Golgi apparatus (see below) were calculated from areas of ellipses using major and
in central islet cells. Cells were selected according to the minor islet axes (d = Jarea of ellipse/a) and compared
presence of a) a nuclear profile > 5 pm in diameter, b) by the Kruskal-Wallis test (Kruskal and Wallis, 1952).
a t least one capillary face (see below), and c) a n osmium- This nonparametric test is suitable for simultaneous
stained Golgi complex.
comparison of three independent samples with a n asymmetric population distribution (Siegel, 1956) a s is the
Intact Pancreas
case for rat pancreatic islets (Hellman, 1959). Cell proIn each of three experiments, a 1-2-month-old male files with one and two capillary faces were treated seprat, weighing 160-260 gm, was killed by decapitation. arately, since two faces present on the same profile do
The dorsal (splenic) portion of the pancreas was dis- not represent independent observations. Observed fresected out, chopped in 0.5-mm blocks, and fixed in 2% quencies of cells with capillary faces in various orientaglutaraldehyde in 0.1 M phosphate buffer, pH 7.4, for 2 tions (quadrants) were compared to uniform theoretical
hours. Following a wash in the same buffer, the tissue distributions. For cells with one capillary face, the x2
was postfixed in phosphate-buffered 1%Os04 solution test was used. For cells with two capillary faces, the
containing 1.5% potassium ferrocyanide (1-2 hours) to frequency of cells with faces falling in different combienhance the contrast of cell membranes (Karnovsky, nations of quadrants (i.e., G-A1, G-0, G-A2,A1-0, A1-A2,
19711, and then dehydrated in graded ethanols and and O-AZ) were compared to the theoretical uniform
embedded in Epon 812. Thin sections of selected islets distribution by the Kolmogorov-Smirnov test, since the
(see below) were cut on a n LKB ultratome, mounted on latter is more suitable for small samples than the x2 test
copper grids, and stained with aqueous uranyl acetate (Siegel, 1956).
and lead citrate.
Semithin (1-pm-thick) sections were examined every
40 pm for the occurrence of islet profiles. To obtain
sections passing through the central portion of the islets,
the major and minor islet axes were measured using a n
ocular micrometer at intervals of ten sections ( - 10 pm).
The first islet profile whose major or minor axis showed
a decrease after having previously reached a maximum
value was selected for thin sectioning provided the minor axis exceeded 40 pm. For each experiment thin
sections of six such islets were examined in a Philips
EM 300 electron microscope. Each B cell possessing a) a
nuclear profile, b) at least one capillary face (see below),
and c) at least two stacks of Golgi cisternae was photographed a t a n initial magnification of 3 , 3 0 0 ~and the
negatives were projected a t a final magnification of
12,500x for quantitative evaluation.
To obtain information on the overall distribution of
Golgi apparatus in the pancreatic B cell, we first carried
out a three-dimensional reconstruction of a central portion of a n islet (consisting mainly of B cells, cf. Orci,
1982) using plastic plates (Goldstein and Davis, 1968;
Baetens et al., 1979). This showed that Golgi elements
were grouped at one cell pole and were generally in a
juxtanuclear position. This non-uniform distribution of
Golgi elements was a prerequisite for further studies on
The quantitative evaluation of the orientation of the
Golgi apparatus with respect to the capillary face of the
cell was first carried out by the four-quadrant method
(see Materials and Methods) on semithin sections of two
osmium-impregnated isolated islets (n = 38, n = 23). To
check for the reliability of osmium staining, thin sections from all the regions quantified were examined at
Fig. 1. Thin section of a pancreatic B cell in contact with a capillary
(Cap). A four-quadrant grid was positioned so that the borders of the
nuclear profile fitted symmetrically into a system of closely spaced
concentric circles (not shown) drawn about the grid center. The Golgi
quadrant of the grid was oriented so that a maximum of thb Golgi
stacks were within this quadrant, and distributed symmetrically inside or around it. The capillary face was then assigned to the quadrant
in which its midpoint (see Methods) was found. X7,500. Abbreviations:
A,, Az, quadrants adjacent to the Golgi apparatus; G , Golgi quadrant;
0, quadrant opposite to the Golgi apparatus.
the ultrastructural level. This showed specific, consistent staining of the cis Golgi elements, thus confirming
that osmium staining observed by light microscopy represented the Golgi apparatus. The pooled data of cells
showing a single capillary face in semithin sections indicated that most cells had this face in the quadrant
opposite the Golgi apparatus, i.e., quadrant 0 (Table 1).
"he frequencies of cells with various capillary face orientations were as follows: Golgi (G) = 7, adjacent (Al) =
5, opposite (0)= 20, and adjacent (Az) = 3 (x2 = 15.8, P
< .01). In cells with two capillary faces (42% of the total
cells selected), each pair of faces was distributed most
often in the combination G-0 (P < .01; Table 1).
During this study, two potential sources of error became evident: a) at the ultrastructural level some cells
were observed which approached but did not make contact with capillaries. Such arrangements could have led
to false identification of capillary faces at the light microscopical level where cell borders were less distinct; b)
shrinkage of endothelial cells away from islet cells often
occurred in osmium-impregnated freshly isolated islets,
making identification of true capillary faces difficult.
Therefore, further studies of relative Golgdcapillary face
orientation were carried out a t the ultrastructural level
on intact pancreas.
Figure 1 shows a pancreatic B cell in contact with a
capillary. Postfixation in reduced osmium ferrocyanide
(Karnovsky, 1971) resulted in a uniformly good visualization of both plasma and intracellular membranes
even at low magnification, allowing a precise identification of B cell boundaries and a correct positioning of the
four-quadrant grid with respect to the Golgi apparatus.
Pooled data of cells showing a single capillary face
from three separate experiments indicated a preferential location of capillaries in the quadrant opposite the
Golgi apparatus (0)(Table 21, a result which agrees with
that obtained a t the light microscopical level on isolated
islets (cf. Table 1).In addition, the observed distribution
differed significantly from that expected assuming equal
distribution in the quadrants (P < .001). In electron
microscopic studies, cells with two capillary faces formed
17% of the total, a proportion less than half that ob-
TABLE 1. Distribution of capillary faces with respect to the
Golgi apparatus in central islet cells (presumptive B cells)
of isolated islets (light microscopy)'
Cells with one
Cells w i t h two
caDillarv contacts'
G A1 0
GA1 GO GA2 A10 AlA'
5 4 1 5 1
'Figures represent the number of cells having a capillary face located
in a given quadrant (G, Golgi quadrant; Al and A', quadrants adjacent
to the Golgi apparatus; 0, quadrant opposite to the Golgi apparatus).
'Cells with one capillary contact in the plane of the section formed
58%,and cells with two capillary contacts 42%of total cells examined.
3Pooled data were compared to a theoretical distribution, i.e., P =
.25/quadrant by using the'x one-sample test. 'x = 15.8, P < .01.
4Pooled data were compared to a theoretical distribution assuming
equal frequencies in each quadrant combination by the one-sample
Kolmogorov-Smirnov test, P < .01.
more capillary contacts remains unknown, but it is interesting to note that double contacts show a preferential polarization along the GolgVnuclear axis.
The main result of the present study is that Gola
apparatus has a nonrandom positioning in pancreatic B
cells. The data emerging from our quantitative analysis
show that in both methods used (semithin sections of
osmium-impregnated isolated islets and thin sections of
intact pancreas) the Golgi apparatus is preferentially
located in a position opposite to the capillary face of the
cell. The functional significance of this orientation of the
Golgi apparatus in B cells is not known. However, we
might raise the possibility that it reflects a polarized
vectorial transport and release of B cell polypeptides
into specialized compartments of the islet interstitium,
whose existence is suggested by recent evidence (Kawai
et al., 1982).
In conclusion, our results strongly suggest a structural
polarization of pancreatic B cells, which may be important for the coordinate functioning of the islet of
TABLE 2. Distribution of capillary faces with respect to the
Golgi apparatus in B cells of intact pancreas (electron
B cells w i t h o n e
G A1 0 A'
B cells with t w o
capillary contacts'
( Q u a d r a n t combination)
GA1 GO GA2 A10 AlAz OA2
1 1 3 3 2 1 5
1012 9 1 0
6 7 2 5 9 2
Totals 29 22 55 243 3
'See footnote to Table 1.
'Cells with one capillary contact formed 82.7%, and cells with two
capillary contacts 17.3%of total cells examined.
3Pooled data were compared to a theoretical distribution, i.e., P =
.25/quadrant by using the one-sample x2 test. x2 = 21.57, P < ,001.
4Pooled data were compared to a theoretical distribution assuming
equal frequencies in each quadrant combination by the one-sample
Kolmogorov-Smirnov test, P < .15.
served in light microscopy experiments on isolated islets. A likely explanation for this difference is that the
probability of sectioning more than one capillary obviously increases in parallel with the thickness of the
section. In these cases, the two capillaries fell most frequently into the combination of quadrants G-0. However, the overall distribution was more even (P < .15)
than at the light microscopical level (Table 2). Furthermore, a comparison between the frequencies in pooled
opposite (G-0, Al-AZ) and pooled adjacent quadrant combinations (G-A1,G-AZ, A1-0, A2-0) showed no significant
differences (x2 = 2.66, P > .1). These results indicate
that a n important percentage of islet cells possess at
least two capillary faces, which is in agreement with in
vivo studies (McCuskey and Chapman, 1969; O'Leary,
1930). The exact proportion of cells with none, one, or
We thank Dr. M. Amherdt and Dr. G.M. Mascherpa
for help with the statistics; Dr. R. Unger for critical
reading of the manuscript; and M. Muller for typing the
This study was supported by grant 3.460.83 from the
Swiss National Science Foundation.
Baetens, D., F. Malaisse-Lagae, A. Perrelet, and L. Orci (1979) Endocrine pancreas: Three-dimensional reconstruction shows two types
of islets of Langerhans. Science, 206:1323-1325.
Bencosme, S.A., and D.C. Pease (1958) Electron microscopy of the
pancreatic islets. Endocrinology, 63:l-3.
Cereijido, M., J. Ehrenfeld, I. Meza, and A. Martinez-Palomo (1980)
Structural and functional membrane polarity in cultured monolayers of MDCK cells. J. Membr. Biol., 52:147-159.
Courrier, R., and P. Reiss (1922) Appareil reticule de Golgi et polarite
secretoire des cellules parathyroidiennes. C.R. SOC. Biol., 86:867868.
De Camilli, P., D. Peluchetti, and J. Meldolesi (1974) Structural difference between luminal and lateral plasmalemma in pancreatic acinar cells. Nature, 248:245-246.
Farquhar, M.G., and G.E. Palade (1981) The Golgi apparatus (cornplex)-(1954-1981)-from
artifact to center stage. J. Cell Biol.,
Ferreira, D. (1957) Ultrastructure des cellules du pancreas endocrine
chez I'embryon du rat nouveau-n6. J. Ultrastruct. Res., 1:14-25.
Garant, P.R., and 1.-M. Cho (1979) Cytoplasmic polarization of periodontal ligament fibroblasts. J. Periodontal. Res., 14:95-106.
Goldstein, M.B., and E.A. Davis (1968)The three dimensional architecture of the islets of Langerhans. Acta Anat., 71:161-171.
Hayat, M.A. (1972) Basic Electron Microscopy Techniques. Van Nostrand and Reinhold Co., New York, p. 84.
Hellerstrom, C. (1977) Growth pattern of pancreatic islets in animals.
In: The Diabetic Pancreas. B.W. Volk and K.F. Wellman, eds.
Plenum Press, New York, pp. 61-97.
Hellman, B. (1959) The numerical distribution of the islets of Langerhans at different ages in the rat. Acta Endocrinol., 32:63-77.
Holmes, L.B., and R.L. Trelstad (1980) Cell polarity in precartilage
mouse limb mesenchymal cells. Dev. Biol., 78:511-520.
Icardo, J.M., J.L. Ojeda, and J.M. Hurle (1982) Endocardia1 cell polarity
during the looping of the heart in chick embryo. Dev. Biol., 90:203209.
Karnovsky, M.J. (1971)Use of ferrocyanide reduced osmium in electron
microscopy. Proc. 11th Ann. Meeting Am. SOC. Cell Biol., 146a
Kawai, K., E. Ipp, L. Orci, A. Perrelet, and R.H. Unger (19821Circulating somatostatin acts on islets of Langerhans by way of a somatos-
cules in dissociated epithelial cells. J. Cell Biol., 71:907-920.
tatin poor compartment. Science, 218:477478.
Kruskal, W.H., and W.A. Wallis (1952) Use of ranks in one-criterion Reggio, H., E. Coudrier, and D. Louvard (1982) Surface and cytoplasmic domains in polarized epithelial cells. In: Membranes in
variance analysis. J. Am. Statist. Assoc., 47:583-621.
Growth and Development. Alan R. Liss, New York, pp. 89-105.
Lacy, P.E., and M. Kostianovsky (1967) Method for isolation of intact
Reiss, P. (1922) L’appareil de Golgi dam les cellules glandulaires de
islets of Langerhans from rat pancreas. Diabetes, 16:35-39.
l’hypophyse. Polarite fonctionnelle et cycle secretoire. C.R. Soc.
Leblond, C.P., and G. Bennett (1977) Role of the Golgi apparatus in
Biol., 87:255-256.
terminal glycosylation. In: International Cell Biology ’76-’77.
Rodriguez-Boulan, E., and D.D. Sabatini (1978) Asymmetric budding
Rockefeller University Press, New York, pp. 326-336.
of viruses in epithelial monolayers: A model system for study of
Leblond, C.P., H. Putchler, and Y. Clermont (1960) Structures correepithelial polarity. Proc. Natl. Acad. Sci. U.S.A., 75:5071-5075.
sponding to terminal bars and terminal web in many types of cells.
Siegel, S. (1956)Nonparametric Statistics for the Behavioral Sciences.
Nature, 186:784-788.
McGraw-Hill Book Co., New York, pp. 47-52.
McCuskey, R.S., and T.M. Chapman (1969) Microscopy of the living
Tartakoff, A.M. (1980) The Golgi complex: Crossroads for vesicular
pancreas in situ. Am. J. Anat., 126t395-408.
traffic. Int. Rev. Exp. Pathol., 22227-251.
Munger, B.L. (1958) A light and electron microscopic study of cellular Trelstad, R.L. (1970) The Golgi apparatus in chick corneal epithelium:
differentiation in the pancreatic islets of the mouse. Am. J. Anat.,
Intracellular position during development. J. Cell Biol., 45:3442.
O’Leary, J.L. (1930) An experimental study of islet cells of the pancreas Trelstad, R.L. (1977) Mesenchymal cell polarity and morphogenesis of
chick cartilage. Dev. Biol., 59:153-163.
in vivo. Anat. Rec., 45r27-58.
Weinstock, M., and C.P. Leblond (1974) Synthesis, migration and reOrci, L. (1982) Macro- and micro-domains in the endocrine pancreas.
lease of precursor collagen by odontoblasts as visualized by raDiabetes, 31.538365.
dioautography after r3H] proline administration. J. Cell Biol.,
Palade, G.E. (1975) Intracellular aspects of the process of protein syn60:92-127.
thesis. Science, 189:347-358.
Pisam, M., and P. Ripoche (1976) Redistribution of surface macromole-
Без категории
Размер файла
661 Кб
positioning, pancreaticum, golgi, nonrandom, apparatus, cells
Пожаловаться на содержимое документа