J. Cell Sci. 83, 77-87 (1986)
77
Printed in Great Britain © The Company of Biologists Limited 1986
ELEMENTAL CONTENT OF MAST CELL GRANULES
MEASURED BY X-RAY MICROANALYSIS OF RAT
THYMIC TISSUE SECTIONS
MARION D. KENDALL AND ALICE WARLEY
Division ofAnatomy, United Medical & Dental Schools, St Thomas's Hospital
Medical School Campus, Lambeth Palace Road, London SE1 7EH, UK
SUMMARY
Mast cell granules were examined by fully quantitative X-ray microanalysis of 20 cells in freezedried cryosections. The mast cells were situated mainly in the connective tissue of the thymic
capsule of five adult male Carworth Sprague Europe rats. In addition 30 red blood cells were
analysed from the same sections.
Nineteen of the mast cells had granules rich in S and K. One cell had smaller granules, and in
this cell the granules contained high [Ca] and [P] instead of high [S] and [K]. In the majority of
cells (13) the S: K ratio was highly correlated and less than 2-2, whereas in the remaining six cells
the individual granule ratios were very variable in any one cell and much higher. The mean granule
[K] (994 ± 57mmolkg~' dry wt) was about four times the mean cytoplasmic level of 227 ± 8 1
mmol kg" 1 dry wt. The existence of this difference in concentration between the granules and the
cytoplasm suggests that the K in the granules must be bound.
The relationship between the [K] and [S] is discussed with regard to the possible binding of
heparin and amines in the granules.
INTRODUCTION
The granular matrix of mast cells is a dense complex of mainly cationic proteins,
acid mucopolysaccharides and bioamines (Padawar, 1979). Heparin, a sulphated
polysaccharide, is usually greatly in excess of the amines, which can be histamine,
serotonin, dopamine or various combinations of these. Whilst the two can link
together to form a salt, the heparinate of histamine, there is usually a great excess of
heparin over histamine (West, 1959) and the binding is probably more complex,
involving a basic granule protein (Uvnas & Aborg, 1977). Granules isolated in nonionic media retain the size and the electron opacity of the granules in situ, but are said
to be unstable in the presence of mineral cations (Na + , K + , Ca2+ and Mg 2+ ), traces
of these ions causing catastrophic swelling and an explosive release of amines
(Padawar, 1979).
Mast cells are well known to be associated with thymus glands; their presence
there has been reported in normal and diseased conditions (Burnet, 1965; Goldstein,
1966). Thymic tissue is a rich source of mast cell precursors (Ginsburg & Sachs,
1963; Csaba et al. 1962; Ginsburg, 1963; Ginsburg & Lagunoff, 1967; Ishizaka
et al. 1976) and developing mast cells may be found in the thymus of many animals
Key words: mast cell granules, X-ray microanalysis, elemental content, thymus.
78
M. D. Kendall and A. Warley
(see Kendall, 1981). Increases in thymic mast cell numbers have been reported,
especially after thymic atrophy (Griss, 1970; Ruitenberg & Elgersma, 1976;
Ruitenberg & Buys, 1980; Chatamra et al. 1983). In the continuation of the latter
study (Chatamra et al. 1986), the mast cells were noted to be mainly within the
connective tissue of the capsule and septa, whilst immature cells were observed
within the cortex.
We are currently undertaking a study of different cell types in the rat thymus gland
using the technique of X-ray microanalysis. This permits the examination of the
elemental content of mast cell granules in vivo, without the disruption of cell
fractionation and granule separation techniques. Avoiding such procedures may be
important in the study of granule composition, as Jones et al. (1979) have shown that
isolating nuclei causes an alteration in their ionic content.
MATERIALS AND METHODS
Small pieces (1 mm cubes) of cortex and capsule were excised from the thymus glands of five
adult male Carworth Sprague Europe rats (approx. 250g) after anaesthetization with chloroform.
The cubes of tissue were placed on small brass stubs and frozen rapidly in just-melting liquid Freon
22 in liquid N2. The blocks were stored in liquid N2 before cutting at —65 to — 70°C in a Slee
cryoultramicrotome. Ribbons of frozen sections approx. 300 nm thick were collected onto
Formvar-coated Ni grids and freeze-dried in a stream of liquid N 2 in the cabinet at the cutting
temperature for approx. 1 h before being placed in a desiccator and left to warm up to room
temperature overnight. The dried sections were carbon coated and then analysed in an AEI
EMMA-4 fitted with a Link 860 Series 2 energy dispersive detection system. Spectra were collected
over 100 s live time with a probe area of 0-08 fim, an accelerating voltage of 60 kV and a 4 nA probe
current. A gaseous N^ cold finger was used to reduce contamination. The spectra were processed
using the Quantem program (Link Systems Ltd). This program is described by Hall & Gupta
(1982). Quantification of spectra was achieved by comparison with previously prepared gelatin
standards. Full details of the method and validation of the standards are given by Kendall et al.
(1985).
Generally 10 granules were analysed from each of 20 mast cells; the data derived from the
analysis of 30 red blood cells are also included here. Student's t tests were performed where
necessary.
RESULTS
The mast cells were mainly located in the capsule around the cortex of the thymus,
often quite close to blood capillaries. The red blood cells (RBCs) were situated in
these blood capillaries or within the adjacent cortex.
Rat thymic mast cell granules range in size from approximately 0-3 to 1-0 jitm in
diameter (Fig. 1A), so that the probe used for analysis was generally contained
within the granule except where the granules were small, when some cytoplasm was
included in the analysis. Owing to the dense packing of the granules (see Fig. IB)
only a very small number of analyses (8) could be obtained from the cytoplasm. This
number is too small for statistical purposes.
Fig. 2 is a typical spectrum from a mast cell granule; the granules in 19 of the 20
cells had a similar composition. The mean elemental concentrations from 158
granules analysed from these 19 cells are given in Table 1, along with the values for
bv
Fig. 1. A. Conventional electron micrograph of a mast cell in the capsule of the rat
thymus. Tissue fixed ( l h at room temperature) in Karnovsky, cut and stained with
uranyl acetate, bv, blood vessel nearby; co, collagen. Bar, 1 fim. B. Freeze-dried frozen
section (unstained) of a mast cell from a similar situation as in A. co, collagen. Bar, 1 /xm.
80
M. D. Kendall and A. Warley
IIIIIIIi IIIi IIIIIIi1 IiIIIiIIiIiIiIIi IIIIIIiI IIi I I
6
" i n r | 1 1 1 1 1 1 1 1 1 [i-i 1 1 1 1 [ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ( i 1 1 1 1 1 1 1 1 1 1 1
10
0
1
2
3
4
5
6
7
8
9
keV
Fig. 2 . A typical spectrum obtained from the analysis of a freeze-dried frozen section of a
mast cell granule.
mast cell cytoplasm and RBCs from the same tissues. When these are compared with
the results of the analysis of RBCs and the few mast-cell cytoplasm estimates it can be
seen that the granules contain much higher [S] and [K], and possibly raised [Na]
and [Zn].
The ratio of S: K in each granule was calculated and the mean ratio was found to be
3-0 + 3-7 but as the distribution of the frequencies is skewed (Fig. 3) the median
value is 1 -4 and 75 % of the granules had an S: K ratio of less than the mean, with the
maximum frequency at 0-8—0-9 (22%).
Table 1. Mean ^±S.E.J concentrations of elements (mmolkg ' dry wt) in red blood cells,
in mast cells and from one cell with granules high in Ca and P
Na
Mg
27 ± 5
7±3
94 ± 9
121 ±39
27 ±15
159 ±43
197 ±16
20 ± 2
142 ± 6
24 ± 5
Granules high in Ca and P (n = 10)
601 ± 3 1 73 ± 2 194 ± 7
237 ± 5
4±1
122 ± 8
Cl
K
Fe
Zn
Ca
Red blood cells (n = 30)
213 ± 8 224 ± 1 4 331 ± 1 3
57 ± 2
2±1
-4±1
Mast cell cytoplasm (n = 8)
204 ±44 351 ±152 227 ±81
2±5
4±2
3 ±0-5
10 ± 1
-8±1
6±1
1142±65
Mast cell granules (n = 158)
1418 ±56 371 ± 18 944 ±57
11 ± 3
Mast cell granules
81
The mean ratio of S: K was calculated for each cell and found to be significantly
correlated in 13 cells. The individual granule values of [S] and [K] in these cells are
indicated by dark circles in Fig. 4. In 11 cells the correlation coefficient (r) is 0-9 or
1-0 (/>< 0-001) and in two cells 0-7 (P<0-\). The ratio of S:K in all of these cells is
less than 2-2.
The cells with ratios over 2-2 were six cells in which there was no correlation
( r = < 0 - 7 ) between [S] and [K]. Individual granules from these six cells are
indicated in Fig. 4 by open circles. Most of the granules had low [K] but a few had
higher [K], similar to granules from the remaining 13 cells in which the [K] and [S]
were higher correlated (filled circles, Fig. 4). The six cells with 'high ratios' (mean
6-9 ± 4-7) came from two animals, and one cell was not situated in the capsule with
the rest, but within the cortex. This cell appeared morphologically immature, as
judged by the irregular size and variation in electron density of the granules, and the
presence of more cytoplasm between granules. This probably corresponds to a stage
II mast cell (Combs, 1966).
The mean elemental concentrations of the granules from these two populations of
mast cell (designated 'low' and 'high' ratios) are shown in Table 2. The high-ratio
granules had less Na, P, Cl and K (P< 0-001), and more Zn (P<0-01) and Ca
(P< 0-001), than the low-ratio granules. There was no difference in the [Mg], [S]
and [Fe] or in the background integrals (low-ratio mean = 6556 ± 230 compared
with 5880 ± 253 for high-ratio granules).
In one of the 20 cells, the granules (see Table 1) were rich in Ca and P, very low in
S and had low [K] (S:K ratio 0-3). The granules appeared to be slightly smaller than
those of the mast cells discussed above.
40-
S:K content
30-
E 20-
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
* * * * * * * * *
10-
*
*
*
*
* * *
11
9
Ratio
13
15
17
19
Fig. 3. The frequency distribution of the ratios of S:K in individual mast cell granules
(n = 158) from 19 cells.
M. D. Kendall and A. Warley
82
5000 H
4000-
3000-
O O
O
&
o
2000-
oo
cm o
•
•
••• m
•
m
m
•
• •• ••
«M* • •
1000- oo • • • ;
CO I
o
1000
2000
3000
4000
K (mmol kg" 1 dry wt)
Fig. 4. The individual granule [S] and [K] in 19 cells. (O) Granules from 6 cells where
the cell [S] and [K] were not correlated. (•) The others show a high correlation
(coefficient = 0-74; P < 0001; Y = 411+0-828X1).
DISCUSSION
Altogether 20 cells were studied by X-ray microanalysis, and in all but one of these
cells the spectra obtained from the granules were broadly comparable in their very
high S and K contents. A high [S] in mast cell granules is predictable, as the granules
contain much heparin, but the high K levels were unexpected. One cell was
completely different (high [Ca] and [P]). Either this cell is not a mast cell (its
Table 2. Mean (±S.E.) concentrations of elements (mmolkg~l dry wt) in mast cell
granules with low (<2-2) and high (>2-2) S:K ratios
Na
Mg
321 ±20
20 ± 3
Cl
K
Fe
Zn
Ca
3±1
9±1
—13 ± 1
4±1
14 ±2
4±1
-2-2»*
-8-3***
Mast cell granules (low ratios; n = 109)
159 ± 7
1427 ± 7 3
447 ± 2 1
1227 ± 6 5
Mast cell granules (high ratios; n = 49)
182 ±20
4-9»*»
19 ±4
104±8
1399±78
5-0***
n.s.
200±13
315 ± 3 2
Student's t test
9-9***
12-6***
•• P<001; ***P<0-001; n.s., not significant.
Mast cell granules
83
granules were smaller) or this mast cell contains little or no heparin. The significance
of the findings in this cell is not known.
Padawar (1979) reported that mast cell granules isolated in non-ionic media are
very unstable, small traces of the common mineral cations Na + , K + , Ca2+ and Mg 2+
causing swelling (up to 30 times) and an explosive release of bound amines. A
previous X-ray microanalysis study of mast cell granules did not mention K (Yarom
et al. 1975), but in that work the results referred to eight granules analysed from
conventionally fixed, dehydrated and Epon-embedded electron-microscope (EM)
material. A lack of K is not surprising as the preparative procedures would have
caused the loss of any readily diffusible elements, and because of this Na and K are
not generally detectable in conventional EM embedded specimens.
A study of the S:K ratios within individual granules revealed a much more
complicated picture than the mean values suggested. The mode of this skewed
distribution was about 1-0, but granule ratios in six cells were much higher and very
variable. Since one of these cells appeared immature, the variation may have
developmental significance, so it is important to know whether these higher and
variable ratios have a biological basis or are a consequence of the analytical technique.
The quantitative procedure used here relies upon the continuum normalization
method of Hall as described by Hall & Gupta (1982). Major factors that affect the
reliability of the method are contamination and mass loss (Hall, 1979; Hall & Gupta,
1983). Contamination could cause deposition of S, thus affecting quantification of
the elements, but there was no evidence of that on the films. Mass loss from the
organic matrix of specimens is now well documented (Hall & Gupta, 1974; Rick
et al. 1981; Andrews et al. 1981) and is generally considered to be complete well
below the beam doses that we have used. Mass loss from the organic matrix would
affect quantification of all elements, and elemental loss from the S Ktfpeak, which is
known to be sensitive to beam damage, would affect ratios. Differences in mass loss
between the mast cell granule matrix and the protein matrix of our standards would
cause errors in absolute quantification.
Mast cell granules contain basic protein, amines and heparin, a sulphated
polysaccharide from which S can be lost at the probe current used. In addition,
Shuman et al. (1976) showed that the mass loss from a sucrose specimen was greater
than that from serum albumin (a protein), so mass loss from the polysaccharide
might also add to problems in quantification. However, comparisons between highand low-ratio granules should be valid, as all mast cell granules presumably have a
similar composition in terms of organic matrix. Thus it might be expected that
behaviour under the beam would be similar in both groups. It is notable that under
the operating conditions defined here, there was no significant difference between
the background integrals from the low- and high-ratio granules, suggesting that
quantification was not affected differently in the two groups of cells. It should also be
noted that the mean S content in these two groups of granules is similar, with the
variation in the S: K ratio being caused mainly by variation in the K content, and K is
generally considered stable under the electron beam. Thus it can be concluded that
the variation in S:K ratios observed probably has a biological basis.
84
M. D. Kendall and A. Warley
The few mast cell cytoplasmic [K] estimations recorded adjacent to granules were
considerably lower than the granule values. More analyses of cytoplasm are needed
for confidence in these values. Work on dog pancreatic exocrine cells (Nakagaki et al.
1984) has shown that the water contents of cytoplasm and granules in a cell are not
the same. If this is also true for mast cells then values expressed in terms of dry
weight might be expected to differ. Nakagaki et al. (1984) found that the dry mass
fraction of pancreatic acinar cytoplasm was 22% and that of the zymogen granules
37%. Those values were similar to the results of their previous work on dog salivary
gland cells (Sasaki et al. 1983). Allowing for water contents of this order, the mean
[K] in the low-ratio granules was about 10 times the cytoplasmic level and one or two
granules had approximately 27 times more K. The mean high-ratio granule [K] was
three times that of the cytoplasm. This means that K must be bound in the granules.
In the studies reported by Nakagaki et al. (1984) the [K] in the granules was lower
than that in the surrounding cytoplasm, in contrast to our findings here. Roomans &
Wei (1985) also found that rat pancreas zymogen granules were low in K. Pancreatic
granules are largely proteinaceous, unlike mast cell granules, which are sulphated
carbohydrates. In other microprobe studies of tissues producing carbohydrate
secretions high [K] are recorded. Although the work was not fully quantitative,
Appleton et al. (1979) found elevated [K] close to the mucus surface of the mantle
collar epithelium in a snail (Otala lactea) and the microprobe studies from the
laboratory of Hall and Gupta have shown that sites rich in sulphated glycosaminoglycans sequester K (and Ca) from the surrounding medium (see references quoted
by Gupta & Hall, 1981). More recently their studies have considered the mechanisms involved in producing and maintaining such high K concentrations (Dow et
al. 1984; Gupta et al. 1985). Also, Scott (1978) considered the preferences of various
polyanions for Na and K, and found that ester sulphates strongly prefer K. Thus
there is a precedent for high [K] in some biological situations.
Although it was once thought that Zn assisted in the binding of histamine in the
granule by a chelating action, Bergqvist et al. (1971), after recording 16mmol
Znmg" 1 dried granules, considered this amount too low to explain such an effect,
and thought it more likely that the Zn originates from granule enzymes. The mean
Zn concentration recorded in this study (10 ± 1 mmolkg"1 dry wt) is slightly lower
than that of Bergqvist et al. (1971), but the group of 'high S: K ratio' granules had a
mean value of 14 ± 2 mmol kg" 1 dry wt, which is quite a close agreement.
If the cells with high and variable granule ratios are immature, then immature
granules may have a wide variation in their K content, and it appears that with
maturity the S: K ratio approaches 1-0. This narrow range of ratios in the majority of
granules implies a functional role for K in binding anions.
Until the identification of a basic granule protein (Lagunoff et al. 1964; Bergqvist
et al. 1971) it was generally assumed that heparin was bound to histamine or other
amines in the granule. The heparin-histamine complex is a stable salt, binding
efficiently at pH 2-3, but at neutrality (probably nearer to the physiological pH) the
binding capacity of heparin would account for only about 10% of the histamine.
With a low molecular weight basic protein binding to heparin the terminal COO—
Mast cell granules
85
group becomes the only group available for cation binding (Uvnas et al. 1970; Uvnas
& Aborg, 1977). Thus a ratio of S:cation should be 2:1 (or 0-5). This low ratio was
not observed in terms of S: K in any granule and although the inclusion of other
cations would improve the ratio, their contribution is not likely to be substantial.
A direct binding of K to heparin would give an S: K ratio of 1-5, but leave the roles
of the protein and histamine unaccounted for. However, although ratios around 1—2
are commonly found in these studies, there is much variation from cell to cell in the
observed granule ratios, and it is conceivable that there are a number of different
possible ways in which the heparin, granule protein and histamine (or other amines)
are arranged. Thus small changes in the relative proportions of any component - the
production of more amines, for example — might then alter the binding patterns.
This view would be in accord with the interpretation of the apparent increased
sulphation of mast cell granules with age (increased carbohydrate moiety seen with
dye reactions) as being caused by changes in dye-binding sites. A small measure of
support for this is given by the similar mean S concentrations in the granules of highand low-ratio cells.
Finally, although granules isolated in non-ionic media appear to look like granules
in vivo, this study shows that there could be differences in their composition. A loss
of K in the isolation procedures would alter the binding of granule components and
perhaps account for the instability of mast cell granules in the presence of cations
after isolation.
This study prompts a new appraisal of heparin and histamine binding in the mast
cell granules to take account of the high [K] observed. Some thoughts on the effect of
releasing high K levels at degranulation are also necessary as the membrane
permeability of surrounding cells could be greatly influenced by this event, if only for
a brief period of time.
We thank the Sir Jules Thorn Charitable Trust forfinancialsupport to work on the thymus
gland, and Mr Ian W. Morris for technical assistance.
REFERENCES
ANDREWS, S. B., KIRK, R. G. & MAZURKIEWICZ, J. E. (1981). Electron microprobe analysis of
secretory epithelia: avian salt gland. In Microprobe Analysis of Biological Systems (ed. T . E.
Hutchinson & A. P. Somlyo). London: Academic Press.
APPLETON, T . C , NEWELL, P. F. & MACHIN, J. (1979). Ionic gradients within mantle-collar
epithelial cells of the land snail Otala lactea. Cell Tiss. Res. 199, 83-97.
BERGQVIST, U., SAMUELSSON, G. & UVNAS, B. (1971). Chemical composition of basophil granules
from isolated rat mast cells. Ada physiol. scand. 83, 362-372.
BURNET, F. M. (1965). Mast cells in the thymus of NZB mice..7- Path. Bad. 89, 271-284.
CHATAMRA, K., DANIEL, P. M., KENDALL, M. D. & LAM, D. K. C. (1983). A preliminary report
on the histological changes in the thymus found in rats rendered diabetic with streptozotocin.
J. Anat. 137, 413-414.
CHATAMRA, K., DANIEL, P. M., KENDALL, M . D . & L A M , D. K. C. (1986). Atrophy of the thymus
in rats rendered diabetic by streptozotocin. Horm. Metab. Res. 17, 630-632.
COMBS, J. W. (1966). Maturation of rat mast cells. An electron microscope study. J. CellBiol. 31,
563-575.
86
M. D. Kendall and A. Warley
CSABA, G., TORO, I. & MOLD, K. (1962). Some new data concerning the functional unity of the
'lymphatic system'. Ada anat. 48, 114-121.
Dow, J. A. T., GUPTA, B. L., HALL, T . A. & HARVEY, W. R. (1984). X-ray microanalysis of
elements in frozen-hydrated sections of an electrogenic K + transport system: The posterior
midgut of Tobacco Hornworm (Manduca sexta) in vivo and in vitro. J. Membr. Biol. 77,
223-241.
GlNSBURG, H. (1963). The in vitro differentiation and culture of normal mast cells from the mouse
thymus. Ann. N.Y. Acad. Sci. 103, 20-39.
GINSBURG, H. & LAGUNOFF, D. (1967). The in vitro differentiation of mast cells. J . Cell Biol. 35,
685-697.
GINSBURG, H. & SACHS, L. (1963). Formation of pure suspensions of mast cells in tissue culture by
differentiation of lymphoid cells from the mouse thymus. .7. natn. Cancer. Inst. 31, 1-40.
GOLDSTEIN, G. (1966). Mast cells in the human thymus. Aust.jf. exp. Biol. med. Sci. 44, 593-596.
GRISS, P. (1970). Quantitative behaviour of mast cells in the normal and accidentally involuted
human thymus gland. Artzneimittel-Forsch. 24, 1-7.
GUPTA, B. L., DOW, J. A. T., HALL, T . A. & HARVEY, W. R. (1985). Electron probe X-ray
microanalysis of the effects of Bacillus thuringiensis var Kurstaki crystal protein insecticide on
ions in an electrogenic K + -transporting epithelium of the larval midgut in the lepidopteran,
Manduca sexta, in vitro. J. Cell Sci. 74, 137-152.
GUPTA, B. L. & HALL, T . A. (1981). Ions and water in cells and tissues studied by microprobe
analysis of frozen-hydrated sections in the SEM. 39th Ann. Proc. Electron Microsc. Soc. Am.,
pp. 654-656. Atlanta, Georgia: Claitors.
HALL, T . A. (1979). Problems of the continuum-normalisation method for the quantitative
analysis of sections of soft tissue. In Microbeam Analysis in Biology (ed. C. P. Lechene & R. R.
Warner), pp. 185-203. London: Academic Press.
HALL, T . A. & GUPTA, B. L. (1974). Beam-induced loss of organic mass under electronmicroprobe conditions. J. Microsc. 100, 177-188.
HALL, T . A. & GUPTA, B. L. (1982). Quantification for the X-ray microanalysis of cryosections.
J. Microsc. 126, 333-345.
HALL, T . A. & GUPTA, B. L. (1983). The localization and assay of chemical elements by
microprobe methods. Q. Rev. Biophys. 16, 279-339.
ISHIZAKA, T . , OKUDAIRA, H., MAUSER, L. E. & ISHIZAKA, K. (1976). Development of mast rat
cells in vitro. 1. Differentiation of mast cells from thymus cells. J. Immun. 116, 747-754.
JONES, R. T., JOHNSON, R. T., GUPTA, B. L. & HALL, T . A. (1979). The quantitative
measurement of electrolyte elements in nuclei of maturing erythrocytes of chick embryo using
electron-probe X-ray microanalysis. J. Cell Sci. 35, 67-85.
KENDALL, M. D. (1981). The cells of the thymus. In The Thymus Gland (ed. M. D. Kendall),
pp. 63-83. Anat. Soc. Symp. vol. 1. London: Academic Press.
KENDALL, M. D., WARLEY, A. & MORRIS, I. W. (1985). Differences in apparent elemental
composition of tissues and cells using a fully quantitative X-ray microanalysis system. J. Microsc.
138, 35-42.
LAGUNOFF, D., PHILLIPS, M. T., ISERI, O. A. & BENDm, E. P. (1964). Isolation and preliminary
characterization of rat mast cell granules. Lab. Invest. 13, 1331-1344.
NAKAGAKI, I., SASAKI, S., SHIGUMA, M. & IMAI, Y. (1984). Distribution of elements in the
pancreatic exocrine cells determined by electron probe X-ray microanalysis. Pfliigers Arch. ges.
Physiol. 401, 340-345.
PADAWAR, J. (1979). Mast cell structure: implications for normal physiology and degranulation.
In The Mast Cell (ed. J. Pepys & A. M. Edwards). London: Pitman Medical.
RICK, R., DORGE, A., ROLOFF, C. & THURAU, K. (1981). Electron microprobe analysis of Na
transporting epithelia. In Microprobe Analysis of Biological Systems (ed. T . E. Hutchinson &
A. P. Somlyo). New York: Academic Press.
ROOMANS, G. M. & WEI, X. (1985). X-ray microanalysis of resting and stimulated rat pancreas.
Ada physiol. scand. 124, 353-359.
RUITENBERG, E. J. & BUYS, J. (1980). Thymus atrophy during early pregnancy and its effect on a
Trichinella spiralis infection in mice, including intestinal pathology and blood eosinophilia. Vet.
Immun. Immunopath. 1, 199-214.
Mast cell granules
87
RUITENBERG, E. J. & ELGERSMA, A. (1976). Absence of intestinal mast cell response in
congenitally athymic mice during Trichinella spiralis infection. Nature, Land. 264, 258-260.
SASAKI, S., NAKAGAKI, J., MORI, H. &IMAI, Y. (1983). Intracellular calcium store and transport of
elements in acinar cells of the salivary gland determined by electron probe X-ray microanalysis.
Jap. J. Physiol. 33. 69-83.
SCOTT, J. E. (1978). Some comments on specific ionic interactions. In Ions in Macromolecular and
Biological Systems (ed. D. H. Everett & B. Vincent), pp. 132-133. Proc. 29th Symp. Colston
Res. Soc. Bristol: Scientechnica.
SHUMAN, H., SOMLYO, A. V. & SOMLYO, A. P. (1976). Quantitative electron probe microanalysis
of biological thin sections: methods and validity. Ultramicroscopy 1, 317-339.
UVNAS, B. & ABORG, H.-C. (1977). On the cation exchanger properties of rat mast cell granules
and their storage of histamine. Ada physiol. scand. 100, 309-314.
UVNAS, B., ABORG, H.-C. & BERGENDORFF, A. (1970). Storage of histamine in mast cells.
Evidence for an ionic binding of histamine to carboxyls in the granule heparin-protein complex.
Acta physiol. scand. 336 (Suppl.) 3-26.
WEST, G. B. (1959). Tissue mast cells and tissue amines. J . Phannac. Pharmacol. 11, 513-534.
YAROM, R., MAUNDER, C. A., HALL, T. A. &DUBOWITZ, V. (1975). Microanalysis of mast cells in
rat muscle. Experientia 31, 1339-1340.
{Received 20 August 1985 -Accepted, in revised form, 21 December 1985)