A Quantitative Ultrastructural Evaluation of the Cell Organelle
Specificity of the Uranaffin Reaction
in Normal Human Platelets
CLAIRE M. PAYNE, PH.D.
The purpose of this study was to determine the normal variation
in the number of uranaffin-positive organelles in normal human
platelets and to study the effect of pH and fixation on the uranaffin reaction. The normal variation in uranaffin organelles
was determined by studying platelets from nine normal subjects.
At pH 3.9, the mean number of reactive sites/platelet profile
was 0.43 ± 0.10 when platelets were fixed with lower glutaraldehyde concentrations and 0.56 ± 0.24 with higher glutaraldehyde concentrations.
Fixed platelets were reacted at pH 2.8, 3.9, 5.0, and 7.0 with
four different uranaffin procedures that varied in the extent of
fixation and rinse steps (isotonic saline vs. cacodylate buffer).
The number of uranaffin-positive sites in 200 platelet profiles
was scored under the electron microscope. There was a progressive increase in the number of reactive sites/platelet profile
as the pH increased from 2.8 to 7.0. In general, higher pH
favored granule matrix and core staining, whereas lowpH favored
both the staining of granule membranes and their contents.
The uranaffin reaction showed organelle specificity when run
under certain experimental conditions. At low pH and using
isotonic saline in the rinse steps, only the dense bodies and
ribosomes stained. The biochemical content of the dense body
responsible for uranaffin reactivity is discussed. (Key words:
Uranaffin; Ultrastructure; Platelets) Am J Clin Pathol 1984;
81:62-70
THE NORMAL HUMAN PLATELET is ultrastructurally complex29 and contains several cell organelles and
structural elements shared with muscle cells and neurons.18 Platelet cell organelles include mitochondria, microfilaments, microtubules, a-granules, dense bodies, lysosomes, 116 peroxisomes, a dense tubular system, and an
open canalicular system.28 Electron microscopy has made
a significant contribution to our understanding of the
normal and defective human platelet.6-22'24'28'2931'32 Ultrastructural defects involving platelet granules have been
well documented and include a deficiency or absence of
a-granules (gray platelet syndrome3,30) and dense bodies
(Hermansky-Pudlak syndrome [HPS],14 storage pool disease [SPD]9'25'26), and the presence of giant or fusion-
Received April 22, 1983; received revised manuscript and accepted
for publication July 6, 1983.
Address reprint requests to Dr. Payne: Department of Pathology,
University of Arizona, Arizona Health Sciences Center, 1501 North
Campbell Avenue, Tucson, Arizona 85724.
62
Department of Pathology, University of Arizona, AHSC,
Tucson, Arizona
type granules10 in various neoplastic myeloproliferative
disorders.12 The normal human platelet is now believed
to contain at least five distinct granule types, each of
which has characteristic biochemical, ultrastructural, or
ultrastructural cytochemical markers. The a-granules
contain an eccentric nucleoid by electron microscopy28
and contain fibrinogen, platelet factor 4, and growth
factor26; the dense body contains serotonin, nonmetabolic
ADP and ATP, calcium," 17 and pyrophosphate. There
are two distinct lysosome types,5 one of which contains
/8-N-acetylglucosaminidase and /3-glucuronidase, and the
other, /?-glycerophosphatase. Microperoxisomes have been
identified at the ultrastructural level and shown to have
catalase activity.2
Tranzer23 in 1971 first described en bloc staining of
platelets with uranyl ions. Six years later, Richards and
DaPrada 15 reported in detail what they called the "uranaffin reaction," discussed the specificity of this reaction
for organelles containing nucleotides and storing biogenic
amines and showed an absence of uranaffin-positive organelles in patients with HPS. Lorenz and associates9
(using the method of Richards and DaPrada 15 ) went on
to quantitate the number of uranaffin-positive organelles
in two control subjects and six patients with storage pool
disease. Although the uranaffin reaction now has been
used in the study of two related platelet disorders, very
little is known about the specificity and variability of this
reaction in normal human platelets.
The purpose of this study was to determine the normal
variation in the number of uranaffin-positive organelles
in normal human platelets and to study the effect of pH
and fixation on the uranaffin reaction. The parameters
studied included the extent of heavy metal binding, cell
organelle specificity, and uranaffin organelle substructure
resulting from the uranaffin reaction. The significance of
these nucleotide-containing cell organelles to the physiology of the blood platelet is discussed.
Vol. 81 • No. I
63
URANAFFIN REACTION IN HUMAN PLATELETS
Materials and Methods
Preparation and Fixation of Platelets
for Electron Microscopy
Human blood was obtained from healthy volunteers
by venipuncture using a 20-gauge butterfly needle attached
to a 12-inch plastic tubing (Abbott Hospital Products,
Chicago, IL). The first 2-3 mL of blood were discarded
and the next 4.5 mL of blood then were added to a plastic
tube containing 0.5 mL of 3.8% sodium citrate as an
anticoagulant. Multiple specimens were drawn on each
subject. Platelet-rich plasma (PRP) was obtained by gentle
centrifugation at 800 rpm for 10 minutes in a Sorvall
GLC-1 swinging bucket table-top centrifuge. The PRP
was drawn off using one piece disposable plastic transfer
pipettes (#B-75-100, Bernal Enterprises, Inc., Arleta, CA)
and placed in small round-bottomed plastic centrifuge
tubes (No. 2054 tube; Falcon, Oxnard, CA). The PRP
was mixed with an equal volume of 0.1% glutaraldehyde27
made up in 0.1 M cacodylate buffer (pH 7.2) and fixed
at room temperature for 15 minutes. The suspension was
then centrifuged at 3,000 rpm for 10 minutes and fixed
at 4°C with either 1% glutaraldehyde in 0.1 M cacodylate
buffer (pH 7.2) for 1 hour or 3% glutaraldehyde in the
same buffer for 90 minutes. The cell pellets then were
rinsed with either 0.9% NaCl or 0.1 M cacodylate buffer
(pH 7.2) and reacted with uranyl acetate as described
below.
Uranaffin Procedure (A)
The cell pellets that werefixedwith 1% glutaraldehyde
for 1 hour were rinsed in three changes of 0.9% NaCl
(15 minutes each), placed in the refrigerator, and reacted
overnight with a 4% aqueous solution of uranyl acetate
(pH 2.8*, 3.9, 5.0, and 7.0, respectively). The cells then
were rinsed in three changes of 0.9% NaCl (15 minutes
each), dehydrated in a graded series of ethanols, and
embedded in Spurr's epoxy resin.20 Selected blocks were
thin-sectioned, picked up on uncoated 200-mesh copper
grids, and examined unstained in an Hitachi HU-12 electron microscope.
Uranaffin Procedure (B)
The cell pellets were prepared as in Uranaffin Procedure
(A) with the exception that the pellets were fixed with
•Uranyl acetate was titrated to pH 2.8 with glacial acetic acid and 1
N hydrochloric acid (HC1), respectively. A 4% aqueous solution of uranyl
acetate has a pH of 3.9; therefore, no additional ions are present. Uranyl
acetate was titrated to pH 5.0 and pH 7.0 using 1 N NaOH. The physiochemical properties of the uranyl acetate (UA) solution changed with
pH as evidenced by a change in color intensity and optical density. At
pH 3.9, the UA solution was clear and had a distinct yellow color. At
pH 2.8, the UA solution remained clear but changed to a light yellow
color, and at pH 7.0 the solution was cloudy and finely particulate.
3% glutaraldehyde for 90 minutes (uranaffin procedure
[UP] of Richards and Da Prada15).
Uranaffin Procedure (C)
The cell pellets were prepared as in UP(A) with the
exception that the pellets were rinsed in 0.1 M cacodylate
buffer (pH 7.2) instead of 0.9% NaCl.
Uranaffin Procedure (D)
The cell pellets were prepared as in UP(B) with the
exception that the pellets were rinsed in 0.1 M cacodylate
buffer (pH 7.2) instead of 0.9% NaCl.
Method of Quantitation
The platelets were viewed at a scope magnification of
X 10,000 using the attached X10 ocular system. To insure
a random scoring of platelet profiles, the viewing was
started in the upperrightcorner of a grid square progressed
to the upper left corner, down to the lower left corner,
and over to the lower right corner. The viewing commenced to successive grid squares until 100 platelet profiles were scored. Platelet profiles partly obscured by grid
bars and profiles consisting only of very narrow pseudopodal-type extensions were not counted. The number
of reactive sites in each platelet profile was scored. A
reactive site was defined as any electron-dense precipitate
and consisted of granule membrane staining, core staining,
membrane and core staining, and crystals, depending on
the specific UP used (see "Results").
Two serial sections were scored (100 platelet profiles
per section) for each experimental procedure and the average thereby computed. Student's /-test was used to
compare means with an n = 200.
Results
Ultrastructure of Uranaffin Organelles
The substructure of uranaffin organelles was best visualized using UP(A) and UP(B) and consisted of three
basic parts: a surrounding membrane, a core, and a matrix,
which consists of the area between the core and membrane
(Figs. IA-H). Uranaffin organelles showed great variability in morphology but could be arranged in a sequence
to represent different stages in granule maturation (Figs.
IA-H). Granules with unstained cores first may exhibit
membrane staining, followed by peripheral core staining,
then progressive staining of the core interior, and finally
matrix staining. Most of the uranaffin organelles appear
to fit somewhere in this maturation sequence. Variants
64
PAYNE
A.J.C.P. -January 1984
FIG. 1. Electron micrographs of platelet dense bodies arranged in order of increasing uranaffin reactivity to illustrate a possible maturation
sequence of the granules. Granules with unstained cores (A) (upper left) may first exhibit membrane staining, followed by peripheral core staining
(B) (upper, left center), then progressive staining of the core interior (C-F) (upper right center; upper right; lower left; lower left center), and finally
matrix staining (G. H) (lower right center, lower right). The core of granule H (arrow) is almost obscured by the heavy matrix staining. All electron
micrographs shown have no counterstain. Granule (A) was printed darker than the others to reveal an unstained "core." The magnifications of the
individual granules are as follows: A—X79,800; B— X68,000; C— X68.000; D— X65.200; £—X (05,700; F— X65.200; G—X 105,700; H— X 105.7QO.
that show only core staining or matrix and core staining
with only weak membrane reactivity are exceptions.
In order to quantitatively determine if different pH and
fixation conditions selected for certain uranaffin organelle
characteristics, the uranaffin organelles were classified into
three basic categories. Type I uranaffin organelles showed
membrane staining but no definitive staining of the core
or matrix (Fig. \A, Fig. 2). The unstained core appears
as an electron-lucent area usually located in an eccentric
position (Figs. \A and B) and better visualized when outlined by a fine reaction product. Type II uranaffin organelles showed core (Fig. 2) or matrix staining but no
membrane staining. Type III uranaffin organelles showed
reactivity of both the membranes and granule contents
(matrix and/or core staining [Figs. \C-H, Fig. 2]). The
Type III uranaffin organelles were quite heterogeneous
in nature; some contained eccentric cores (Figs. \C-H,
Fig. 2), centralized cores (Fig. 2), smooth contoured cores
(Fig. 1F, Fig. 2), and crystalline-type cores (Fig. 2); others
contained no definitive cores (stained or unstained) but
exhibited extensive matrix staining; some showed core
and matrix staining to different degrees (Figs. IE, G, H,
Fig. 2) and infrequently contained double cores. Some
granule variants showed microvesicles or unstained microcores and appeared in Type I and III organelles (Fig.
ID, Fig. 2). The membranes of all uranaffin organelles
observed in this study showed staining throughout the
membrane leaflet.
Effect ofpH and Fixation on the Uranaffin Reaction
This part of the study was only performed on platelets
prepared by UP(A) and UP(B).
1. Extent of Heavy Metal Binding. There was a progressive increase in the number of reactive sites/platelet
profile as the pH increased from 2.8 (titrated with HC1)
to pH 7.0 (Fig. 3). This represents overall a threefold
increase in the number of reactive sites using UP(A) compared with almost a sevenfold increase using UP(B); the
latter procedure represents a greater extent of fixation.
There was no statistically significant difference in the
number of reactive sites at pH 2.8 (titrated with HC1)
compared with pH 3.9 using UP(A) or UP(B). There was,
however, a statistically significant increase in the number
of reactive sites between pH 3.9 and 5.0 using UP(B) (P
< 0.001) but not using UP( A). There was also a statistically
significant increase in the number of reactive sites between
pH 5.0 and pH 7.0 using UP(A) (P < 0.01) and UP(B)
(P< 0.001).
A difference in the number of reactive sites was noted
depending upon what acid was used to titrate the uranyl
acetate solution. There was a twofold increase (using both
URANAFFIN REACTION IN HUMAN
Vol. 81 • No. I
0.5
PLATELETS
65
im
FlG. 2. Electron micrograph of a normal human platelet showing the variation in uranaffin reactivity of several dense bodies. The Type I uranaffin
organelle shown here (1) exhibits membrane staining but no definitive core or matrix staining. The Type II uranaffin organelle (2) shows core
staining but no membrane staining. The Type III uranaffin organelles (3) show stained cores (one of which is eccentrically located) and membrane
staining. Note the lack of reactivity of the a-granules(a-G). Uranaffin reaction, no counterstain (X47,900).
uranaffin procedures) in the number of reactive sites/
platelet profile when acetic acid was used instead of HC1
(Fig. 3). This increase was statistically significant using
UP(A) (P < 0.05) and UP(B) (P < 0.01). The number
of reactive sites/platelet profile at pH 2.8 (using acetic
acid for titration) also was increased over the number of
reactive sites at pH 3.9. This increase was statistically
significant using UP(B) (P < 0.05) but not using UP(A).
2. Substructure of Uranaffin Organelles. pH had a
marked effect on the selective staining of the different
types of uranaffin organelles. In general, higher pH favored
matrix and core staining (Type II uranaffin organelles
[Fig. 4]), whereas low pH favored both the staining of
granule membranes and their contents (Type III uranaffin
organelles) (Fig. 5). The staining of Type II organelles
was markedly inhibited below pH 3.9 and showed a progressive increase from pH 3.9 to 7.0 using UP(A) (Fig.
4). On the other hand, Type III uranaffin organelles
showed a progressive inhibition of staining from pH 2.8
to 7.0 (Fig. 5); this was particularly evident using UP(A).
The decrease in reactivity of Type III uranaffin organelles
and the increase in reactivity of Type II uranaffin organelles can be explained by the general inhibition of mem-
brane staining at higher pH. This was evidenced by the
progressive inhibition of staining of Type I uranaffin organelles from /?H 2.8 to 5.0 using UP(B) (Fig. 6). There
was, however, a significant increase in the number of
Type I uranaffin organelles stained at pH 7.0 using more
lengthy fixation (Fig. 6). The Type I organelles scored at
pH 7.0 were probably a combination of true granule
membranes (as evidenced by the presence of unstained
cores) and the x-sectional profiles of elements of either
the dense tubular system (DTS) or the open canalicular
system (OCS) (as evidenced by the presence of stained
anastomosing and tortuous membrane profiles). This
predominance of membrane staining (65% of all uranaffin
organelles) contributes the most to the overall increase
in the number of reactive sites/platelet profile as shown
in Fig. 3.
3. Cell Organelle Specificity. The only cell organelles
that were stained by the uranaffin reaction under all pH
conditions tested were membrane profiles, dense bodies
(uranaffin organelle Types I—III), and ribosomes. The
membranes that were stained appeared to be granule associated when the uranaffin reaction was run at pH 2.85.0. At pH 7.0, other membrane profiles associated with
PAYNE
66
Quantitation of Reactive Sites in Platelets
Prepared by Uranaffin Procedures (A) and (B)
.
•
UP (A)
•
UP (B)
1.1
r°
J
0.9
|
0.8 H
|
0.7 H
|
0.6 H
Comparison of Two Different Rinse Steps in the
Uranaffin Reaction
When fixed platelets were rinsed in cacodylate buffer
(UP[C] and UP[D]) instead of isotonic saline before and
after exposure to uranyl acetate, a striking change in organelle specificity was observed. An electron-dense material filled some areas of the cytosol, the cisternae of the
DTS, and mitochondria, in addition to staining the dense
bodies and ribosomes. The extent of each organelle involvement varied from one platelet profile to another.
Only an occasional a-granule was stained. The reaction
product in some areas had a distinct crystalline appearance
(Fig. 7) and in others afineto coarse granular appearance.
The number of reactive sites using the cacodylate rinse
2 0.5
|
0.4 H
z
0.3
0.2 H
0.1
0
prepared by UP(B). The difference in the means was not
statistically significant (P < 0.2).
There were statistically more Type I and III uranaffin
organelles represented than Type II organelles under both
conditions of fixation (Table 1). The Type I and III organelles were represented equally. There was no statistical
difference in the percentage of Type I, II, and III uranaffin
organelles when platelets were prepared by both fixation
methods.
1.3
1.2
AJ.C.P. .January 1984
2.8
(HCI)
2.8
(acetic
acid)
7.0
Quantitation of Type II Uranaffin Organelles in
Platelets Prepared by Uranaffin Procedures (A) and (B)
PH
IUU
FIG. 3. Bar graph illustrating the effect of pH on the number of
reactive sites in platelets prepared by uranaffin procedures (A) and (B).
-
•
•
90U>
the DTS and/or OCS became stained. The granule-associated membranes appeared to be specific for dense
bodies. No surrounding membrane from any typical agranule (showing a distinct eccentric nucleoid) (Fig. 2)
appeared to be stained in its entirety by the uranaffin
reaction. An occasional a-granule membrane appeared
stained on the inner aspect or granule side of the membrane leaflet. The few a-granules showing this particular
staining appeared larger than the vast majority of other
normal-sized a-granules.
4. Quantitation of Uranqffin-positive Organelles in
Platelets from Normal Subjects. The normal variation in
uranaffin organelles was determined by studying platelets
obtained from nine normal subjects and prepared according to UP(A) and UP(B). A pR of 3.9 was chosen
for this part of the study as it corresponds to the commonly
used method of Richards and DaPrada15 using the fixation
protocol of UP(B). The mean number of reactive sites/
platelet profile was 0.43 ±0.10 when platelets were prepared by UP(A) and 0.56 ± 0.24 when platelets were
e
c
(0
UP (A)
UP (B)
80-
CJ>
6c ™-
_,
% 60c
a
=> 50•
>: 400
2,
a
§
0
30-
1
2v0 -
CO
°-
b
~b
•
10-
1 ri
2.8
(HCI)
2.8
(acetic
acid)
3.9
5.0
1
7.0
PH
FIG. 4. Bar graph illustrating the effect of pH on the percentage of
Type II uranaffin organelles scored in platelets prepared by uranaffin
procedures (A) and (B).
URANAFFIN REACTION IN HUMAN PLATELETS
Vol. 81 • No. I
was increased greatly over the number of reactive sites
using isotonic saline. At pH 3.9, there were 8.7 reactive
sites/platelet profile using UP(C) and 4.6 reactive sites/
platelet profile using UP(D) (P < 0.001), compared with
0.22 and 0.25, respectively, using UP(A) and UP(B).
There was no appreciable change in the number of
reactive sites or percentage of reactive platelets when
platelets were reacted with uranyl acetate at pH 2.8 or
3.9. There was, however, a 91% and 94% inhibition in
the number of reactive sites at pH 5.0 compared with
3.9, and a 97% and 92% inhibition at/?H 7.0 using UP(C)
and UP(D), respectively (Fig. 8).
Quantitation of Type III Uranaffin Organelles
in Platelets Prepared by Uranaffin Procedures (A) and (B)
-
IUU
90-
£
1 1 UP (A)
•
UP (B)
80-
CO
O
c
g
7060-
3
=
50-
•
Q.
£
40-
Discussion
8, 30-
o o
o
CO
Percen
We have determined the normal variation in the number of uranaffin-positive organelles in normal human
platelets and the specificity of the uranaffin reaction under
different conditions of pH, fixation, and rinsing.
The mean number of uranaffin-positive organelles/
platelet profile, determined from an analysis of nine
healthy volunteers was 0.56 ± 0.24 using the method of
Richards and DaPrada 15 and 0.43 ±0.10 using a shorter
fixation time and a lower percentage of glutaraldehyde
[UP(A)]. The difference in these means was not statistically significant. Lorez and associates9 analyzed two human controls and found values of 0.27 and 0.49, respectively. Their values fall within our range of 0.250.94 uranaffin-positive organelles/platelet profile using the
same experimental conditions.
The hydrogen ion concentration had a marked effect
on both the number of reactive sites and the proportion
of uranaffin-positive organelle types. Although the number
of reactive sites increased with pH, there was a progressive
decrease in the number of uranaffin-positive organelles
that showed both matrix/core and granule membrane
staining (Type III organelles) as the pH increased from
2.8 to 7.0. Type III uranaffin organelles represent the
granule morphology that has been associated most commonly with the typical platelet dense bodies. On the other
hand, a progressive increase in the staining of granule
cores and/or matrix without granule membrane staining
(Type II organelles) occurred as the pH increased from
2.8 to 7.0. This progressive increase was most evident
using lower glutaraldehyde concentrations. The affinity
of uranyl ions for both matrix/core and membrane staining at low pH and reduced membrane staining at high
pH can be explained by the change in size and charge of
the uranyl ion and its complexes that are known to occur
at different pHs. In general, as the pH increases, the positive charge decreases and the size of the ionic complexes
increases. Zobel and Beer33 reported that at low pH (2.3)
only uncomplexed uranyl ions are present and the uranyl
ion has a +3 charge (based on the stoichiometric ratio
67
1 r
111 1 .
2.8
(HCI)
2.8
(acetic
acid)
3.9
5.0
7.0
PH
FlG. 5. Bar graph illustrating the effect of pH on the percentage of
Type III uranaffin organelles scored in platelets prepared by uranaffin
procedures (A) and (B).
Quantitation of Type I Uranaffin Organelles in Platelets
Prepared by Uranaffin Procedures (A) and (B)
100
!
o
a
>•
o
f
2.8
(HCI)
2.8
(acetic
acid)
3.9
5.0
7.0
PH
FlG. 6. Bar graph illustrating the effect of pH on the percentage of
Type I uranaffin organelles scored in platelets prepared by uranaffin
procedures (A) and (B).
PAYNE
68
Table 1. The Percentage of Uranaffin Organelle Types
in Normal Human Platelets Prepared
by UP(A) and UP(B)
Fixation Procedure
Type I
Type II
Type III
UP (A)
UP(B)
P values
34.9 + 6.9
41.2 ± 6 . 4
<0.1
26.1 ± 8 . 0
21.7 ± 9 . 3
<0.5
39.0 ± 5.2
37.1 ± 9 . 9
<0.9
between the number of uranyl ions bound for every phosphate group). The small size of the complex and its large
positive charge probably allows for greater accessibility
and binding to polyphosphates associated with the platelet
granule membrane matrix and core. Stoeckenius21 determined that at p\i 3.5 the uranyl ions are still uncomplexed (U0 2 ++ ) but has a +2 charge. At pH 4.8, Huxley
and Zubay7 report that monovalent uranyl acetate complexes (U02Ac+) form. Zobel and Beer33 state that at/?Hs
higher than 3.9 polynuclear complexes of uranyl and hy-
A.J.C.P. • January 1984
droxyl ions, and ionic complexes such as U02Ac2 and
UO2AC3" occur. The large size and change in charge of
these ionic species most probably prohibit access of the
uranyl complexes not only to the platelet dense bodies
but also inhibit their transport across the cell membrane.
The staining of granule membranes (Type I organelles)
and other cytoplasmic membranes (at higher pHs) with
uranyl ions was favored by higher glutaraldehyde concentrations, whereas the staining of Type II and III organelles was favored by lower glutaraldehyde concentrations. Perhaps the cross-linking of membrane proteins
allowed the uranaffin-reactive material to be better retained in the Type I organelles, whereas in the core and
matrix of the granule, cross-linking may lead to the blockage of reactive sites. Spicer and co-workers19 reported a
similar effect of fixation in the "nucleoid" of platelet agranules. The nucleoid lacked affinity for colloidal iron
in glutaraldehyde-fixed specimens but showed reactivity
after formalin or osmium tetroxide fixation.
The uranaffin reaction as reported by Richards and
r%
FlG. 7. Low-power electron micrograph of human platelets reacted with uranaffin procedure (C). Small and large deposits of an electron-dense
crystalline material fill areas of the cytosol in addition to some granules. Uranaffin reaction, no counterstain (XI 4,500).
vol. 81-No. I
URANAFFIN REACTION IN HUMAN PLATELETS
DaPrada15 is run at pH 3.9. We have found a significant
increase in the number of Type III organelles at pU 2.8.
The highest number of typical Type III organelles was
found using acetic acid for titration with more extensive
fixation (UP[B]), and using HC1 for titration with less
fixation (UP[A]). Low pH appears to favor dense body
staining with uranyl ions. The effect of fixation and the
presence of other ions on this reaction requires further
study.
We previously have studied the organelle specificity of
the uranaffin reaction in two human neoplasms, a pheochromocytoma and islet cell carcinoma.13 The neurosecretory granules present in these paraneuromas4 are similar to platelet dense bodies in that they both store neurotransmitters or hormones that are associated with a
high concentration of nucleotides. It was determined from
that study that the uranaffin reaction was specific for
organelles known to contain a high concentration of nucleotides: the nucleus, ribosomes, and cytoplasmic neurosecretory-like granules. In the present study, a similar
specificity was observed with platelets. Since human
platelets lack a nucleus, the only organelles found to stain
were the ribosomes and the dense bodies. Platelet dense
bodies are similar to the neurosecretory granules found
in the normal adrenal medulla or pheochromocytomas
in that both can display an eccentric core. One notable
difference is that in the platelet, 65% of all dense bodies
showed a surrounding uranaffin-positive membrane,
whereas no staining was observed in the membranes surrounding the neurosecretory granules in both the pheochromocytoma and the islet cell tumor.13 Although Richards and DaPrada15 discuss the possibility that the uranaffin reaction might represent a phosphonucleotide
stain, the evidence to date indicates that its specificity
might be only for polyphosphates. Zobel and Beer33 have
determined that uranyl-stained DNA show shifts in the
bands of the infrared spectra assigned to vibrations of
the phosphate and sugar groups. Although the core and
matrix staining of the platelet dense bodies most probably
is a result of the polyphosphates associated with the high
nucleotide content (ADP, ATP), there is a strong possibility that the surrounding membrane may be staining
because of a high concentration of phosphoproteins,33
phospholipids,8 nucleotides in transit across the membrane, other polyphosphate complexes, or a combination
thereof.
A nonspecific staining with uranyl acetate was found
in the present study if the platelets were rinsed with cacodylate buffer instead of isotonic saline. An electron-dense
crystalline material was deposited in the cytosol and other
cell organelles in addition to the dense bodies. We observed this same phenomenon with human tissue13 and
suggested that sodium chloride may facilitate the transport
69
Quantitation of Reactive Sites in Platelets
Prepared by Uranaffin Procedures (C) and (D)
10.0
2.8
2.8
(HCI)
(acetic
acid)
3.9
5.0
7.0
pH
FIG. 8. Bar graph illustrating the effect of pH on the number of
reactive sites in platelets prepared by uranaffin procedures (C) and (D).
of uranyl ions into membrane-bound compartments and
effectively washes the unbound uranyl ions out of nonnucleotide-storing organelles. An unexpected finding from
the present study was the marked inhibition (>90%) in
the number of reactive sites when the reaction was run
at a pYL > 3.9. Perhaps even larger complexes, resulting
from an association of the arsenate ion (As02 (CH3)2+)
in the cacodylate buffer with U02Ac3" species, form at
these higher pHs. The large size of these complexes may
prohibit their transport across the cell membrane into
the cytosol.
It has become evident from this study that the cytochemistry of uranyl staining is complex but appears to
have organelle specificity when run under certain experimental conditions. The specificity of the uranaffin reaction for platelet dense bodies and its accessibility for
quantitation makes it a useful cytochemical stain for the
evaluation of patients with storage pool disease.
Acknowledgments. The author wishes to thank Dr. Jack Layton (Head,
Department of Pathology) without whose support this work would not
be possible. Much gratitude goes to Dr. Lewis Glasser for his continuing
intellectual stimulation, to Mrs. Virginia Borduin and Ms. Alison Kim
for excellent technical assistance, and Mrs. Minde Juric for superb secretarial assistance.
PAYNE
70
References
1. Bentfeld ME, Bainton DF: Cytochemical localization of lysosomal
enzymes in rat megakaryocytes and platelets. J Clin Invest 1975;
56:1635-1649
2. Breton-Gorius J, Guichard J: Two different types of granules in
megakaryocytes and platelets as revealed by the diaminobenzidine
reaction. J Microscop Biol Cell 1975; 23:197-202
3. Breton-Gorius J, Vainchenker W, Nurden A, Levy-Toledano S,
Caen J: Defective a-granule production in megakaryocytes from
gray platelet syndrome. Ultrastructural studies of bone marrow
cells and megakaryocytes growing in culture from blood precursors. Am J Pathol 1981; 102:10-19
4. Fujita T: Concept of paraneurons. Arch Histol Jpn 1977; 40
(Suppl):l-12
5. Fukami MH, Salganicoff L: Human platelet organelles: A review.
Thrombos Haemostas (Stuttg) 1977; 38:963-970
6. Hovig T: The ultrastructure of blood platelets in normal and abnormal states. Ser Haematol 1968; 1:3-64
7. Huxley HE, Zubay G: Preferential staining of nucleic acid-containing
structures for electron microscopy. J Biophys Biochem Cytol
1961; 11:273-296
8. Jones CJ: Perivascular nerves in the rat submandibular salivary
gland. Neuroscience Letters 1979; 13:19-23
9. Lorez HP, Richards JG, Da Prada M, et al: Storage pool disease:
comparative fluorescence microscopical, cytochemical and biochemical studies on amine-storing organelles of human blood
platelets. Br J Haematol 1979; 43:297-305
10. Maldonado JE: Platelet granulopathy. A new morphologic feature
in preleukemia and myelomonocytic leukemia: Light microscopy
and ultrastructural morphology and cytochemistry. Mayo Clin
Proc 1976;51:452-462
11. Martin JH, Carson FL, Race GJ: Calcium-containing platelet granules. J Cell Biol 1974; 60:775-777 '
12. Payne CM, Glasser L: The diagnostic significance of platelet giant
and fusion granules: An ultrastructural morphometric analysis.
Lab Invest 1983; 48:67A
13. Payne CM, Nagle RB, Borduin VF, Kim A: An ultrastructural
evaluation of the cell organelle specificity of the uranaffin reaction
in two human endocrine neoplasms. J Submicrosc Cytol
1983;15:833-841
14. Rendu F, Breton-Gorius J, Trugnan G, et al: Studies on a new
variant of the Hermansky-Pudlak syndrome: qualitative, ultrastructural, and functional abnormalities of the platelet-dense
bodies associated with a phospholipase A defect. Am J Hematol
1978; 4:387-399
15. Richards JG, DaPrada M: Uranaffin reaction: A new cytochemical
technique for the localization of adenine nucleotides in organelles
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
A.J.C.P. -January 1984
storing biogenic amines. J Histochem Cytochem 1977; 25:13221336
Siegel A, Luscher EF: Non-identity of the a-granules of human
blood platelets with typical lysosomes. Nature 1967; 215:745747
Skaer RJ, Peters PD, Emmines JP: The localization of calcium and
phosphorus in human platelets. J Cell Sci 1974; 15:679-692
Sneddon JM: Blood platelets as a model for monoamine-containing
neurones. Prog Neurobiol 1973; 1:151-198
Spicer SS, Greene WB, Hardin JH: Ultrastructural localization of
acid mucosubstance and antimonate-precipitable cation in human
and rabbit platelets and megakaryocytes. J Histochem Cytochem
1969; 17:781-792
Spurr AR: A low viscosity epoxy resin embedding medium for
electron microscopy. J Ultrastruct Res 1969; 26:31-43
Stoeckenius W: Electron microscopy of DNA molecules "stained"
with heavy metal salts. J Biophys Biochem Cytol 1961; 11:297310
Stuart MJ: Inherited defects of platelet function. Sem Hematol 1975;
12:233-253
Tranzer JP: Fixation et apport de contraste par les ions U 0 2 + + et
Pb ++ en l'absence de Os0 4 . Experientia 1971; 27:741-742
Weiss HJ: Congenital disorders of platelet function. Sem Hematol
1980; 17:228-241
Weiss HJ, Ames RP: Ultrastructural findings in storage pool disease
and aspirin-like defects of platelets. Am J Pathol 1973; 71:447466
Weiss HJ, Witte LD, Kaplan KL, et al: Heterogeneity in storage
pool deficiency: Studies on granule-bound substances in 18 patients including variants deficient in a-granules, platelet factor
4, /3-thromboglobulin, and platelet-derived growth factor. Blood
1979; 54:1296-1319
White JG: Fine structural alterations induced in platelets by adenosine diphosphate. Blood 1968; 31:604-622
White JG: Ultrastructural defects in congenital disorders of platelet
function. Ann NY Acad Sci 1972; 201:205-233
White JG: Current concepts of platelet structure. Am J Clin Pathol
1979; 71:363-378
White JG: Ultrastructural studies of the gray platelet syndrome.
Am J Pathol 1979; 95:445-462
White JG, Gerrard JM: Ultrastructural features of abnormal blood
platelets. Am J Pathol 1976; 83:590-632
White JG, Gerrard JM: The ultrastructure of defective human platelets. Mol Cell Biochem 1978; 21:109-128
Zobel CR, Beer M: Electron stains: I. Chemical studies on the
interaction of DNA with uranyl salts. J Biophys Biochem Cytol
1961; 10:335-346