Foamy Cells Associated With Platelet Phagocytosis
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Foamy Cells Associated With Platelet Phagocytosis
TOKUHIRO ISHIHARA, MD, From the First Department of Pathology, Yamaguchi University School
SHIN'ICHIRO AKIZUKI, MD, of Medicine, and the School of Allied Health Sciences,
TADAAKI YOKOTA, MD, Yamaguchi University, Ube, Japan
MUTSUO TAKAHASHI, MD,
FUMIYA UCHINO, MD, and
NOBORU MATSUMOTO, MD
In order to gain an insight into the mechanism for the cytoplasmwasof foamy cells. Although lysosomal enzyme
formation of foamy cells (macrophages with foamy activity revealed in the macrophages that con-
cytoplasm) frequently seen in spleens affectedthese
by idio- tained morphologically recognizable platelets, there
pathic thrombocytopenic purpura (ITP), cells was no demonstrable activity in the cells that contained
were experimentally induced in mice by subcutaneous myelinlike materials. From these results,
the following
injection of platelets with or without accompanied conclusion has been suggested as the mechanism for the
administration of corticosteroid. The light- and elec- formation of foamy cells. Under the state of accelerated
tron-microscopic features of experimentally repro-
duced foamy cells were essentially similar to those seen
phagocytosis of platelets by the macrophages, such as
in ITP, the amount of ingested platelet membranes is
in the spleens of ITP patients. Corticosteroid had no beyond the capacity of lysosomal digestion. Thus, the
significant effect on the formation of foamy cells. Most incompletely degraded membrane constituents, espe-
macrophages with foamy cytoplasm contained various cially membrane-derived phospholipids, remain in the
amounts of phospholipids, which were derived from macrophages, and they are most responsible for the
platelet membranes. By electron microscopy, myelin-
like materials were frequently demonstrated in the
foamy appearance of these macrophages. (Am J Pathol
1984, 114:104-111)
FOAMY CELLS (macrophages with foamy cyto- Materials and Methods
plasm) frequently observed in the spleens of patients Isolation of Murine Platelets
with idiopathic thrombocytopenic purpura (ITP)
reveal the characteristic manner of staining for phos- The blood obtained with 1% EDTA-2Na from ICR
pholipids and myelinlike materials in electron micros- mice was centrifuged at 3000 rpm for 15 minutes. We
copy.1-9 It is suggested that myelinlike materials in centrifuged the buffy coat at 1000 rpm for 10 minutes
these cells are chiefly derived from phagocytized to obtain platelet-rich plasma, which was washed by
platelets. To date, it has not been shown that such 10% EDTA-2Na saline three times. Platelet concen-
foamy cells can be experimentally induced by plate- trate obtained by centrifugation at 3000 rpm for 15
lets. In addition, it still remains unknown whether or minutes was suspended in 0.01 M phosphate-buffered
not steroid therapy, which is usually applied to ITP saline (PBS) at pH 7.2.
patients, is related to the appearance of foamy cells in
the spleen. Anti-Platelet Antibody
The main purpose of this study was to offer a satis-
factory explanation for the mechanism of the forma- Antiserums were raised in Japanese white rabbits
tion of foamy cells found in the spleens of patients by subcutaneous injection of 2 ml of platelet suspen-
with ITP. We describe the histologic, immunohisto- sion (5 x 109 platelets in 0.01 M PBS) from ICR mice
logic, and ultrastructural characteristics of foamy
cells experimentally reproduced in mice by subcu-
taneous injection of platelets with or without steroid Accepted for publication July 28, 1983.
Address reprint requests to Tokuhiro Ishihara, MD, First
administration and of the platelets sensitized with Department of Pathology, Yamaguchi University School of
platelet antibody. Medicine, Ube, 755, Japan.
104Vol. 114 · No. 1 PLATELET PHAGOCYTOSIS AND FOAMY MACROPHAGES 105
in complete Freund's adjuvant, followed by a booster Methods for Histology, Immunohistochemistry,
injection 2 weeks later. The serum collected 2 weeks and Electron Microscopy
after the booster injection was absorbed by the mu- For histologic examination, the tissue fragments
rine serum and erythrocytes. The specificity of this were fixed in 107o buffered formalin. Sections from
antiserum was confirmed by the double-immuno- paraffin blocks were stained with hematoxylin and
diffusion technique. eosin (H&E). Unlabeled antibody peroxidase-anti-
peroxidase (PAP) staining with a minor modification
of Sternberger's method'° was employed for the
Reproduction of Foamy Cells demonstration of murine platelet antigen. The tissues
A small volume of platelet suspension (8 x 108 were also fixed in formalin-calcium. Frozen sections
platelets/0.05 ml) were subcutaneously injected to were stained with Sudan III, Sudan black, Baker's
AKR and ICR mice. The tissue materials for observa- acid hematin,T" Smith-Dietrich"2 for demonstration
tions were obtained from the lesions at 3 and 6 hours of phospholipid, acid phosphatase,13 and p-glucuro-
and on Days 1, 2, 4, 5, 7, 9, 11, 13, 14, and 21 after nidase.14
the injection. Twelve mice were administered steroid For transmission electron microscopy, the subcu-
(hydrocortisone, 0.02 mg) every day after the platelet taneous tissues were fixed in 2.1%0o glutaraldehyde at 4
injection, and the material was obtained 5 and 7 days C and postfixed in 1%o osmium tetroxide. They were
later. Murine platelets were incubated with anti-ICR dehydrated in gradient ethanol and embedded in
murine platelet antibody for 30 minutes at 37 C. The Epon 812. Semithin sections stained with alkaline
mixture was centrifuged at 10,000 rpm for 15 minutes. toluidine blue were examined with a light microscope
The precipitate was suspended in 0.01 M PBS and for selection of the appropriate areas. The ultrathin
subcutaneously injected into the mice. The tissues sections cut with an LKB ultramicrotome were
from the area of the injection were removed at 5 and stained with uranyl acetate and lead citrate. They were
7 days after the injection. examined in a Hitachi H-300 electron microscope.
1 i M
..r..
Z,,.,.ii"
.t _
D.,...
. ....~~~~~~~~~~~~~o ......
Figure 1 -A large number of foamy cells are demonstrated in the peripheral area of platelet mass. (H&E) Figure 2-These foamy cells
have an eccentric nucleus and pale cytoplasm. (H&E)106 ISHIHARA ET AL AJP January 1984
*
Subcutaneous tissues fixed in cold 2.1% glutaralde- around and within the mass. After 2 and 4 days, large
hyde in cacodylate buffer were incubated in the solu- numbers of macrophages migrated around and into
tion containing lead nitrate and P-glycerophosphate the central area of the mass. Large macrophages with
for the demonstration of acid phosphatase activity. foamy cytoplasm (foamy cells) were scattered around
The selected tissues at Days 2, 5, and 7 were fixed the mass at this stage. At Day 7, a mantle of these
in cold periodate, lysine-paraformaldehyde (PLP). foamy cells surrounding the platelet mass and accom-
Frozen sections were incubated in the solution con- panied by a proliferation of capillaries was noted in
taining anti-murine antibody (Fab') conjugated with the subcutaneous tissue (Figure 1), giving a granu-
peroxidase and reacted with 3,3'-diaminobenzidine lomatous appearance to the area. The foamy cells had
(DAB). Then they were postfixed in 1%o osmium an eccentric nucleus, clear cytoplasm with irregular
tetroxide and processed for electron microscopy. stippling, and poorly defined foamy vacuolation (Fig-
ure 2). These cells were stained orange-red with
Results Sudan III, black with Sudan black, grayish black with
Baker's acid hematin, and blackish blue with Smith-
Until 2 days after the injection of platelets, miliary Dietrich stain for demonstration of phospholipid
nodules with whitish cut surface were clearly noted in (Figure 3). Moderate activity of acid phosphatase and
the subcutaneous tissue. During Days 4 and 7, these P-glucuronidase was demonstrated in the macro-
nodules were gradually dispersed and almost com- phages located in the peripheral area of the nodule.
pletely disappeared at Day 14. On the contrary, this enzyme activity was markedly
decreased in the foamy cells. Although some foamy
cells contained trace activity, the others exhibited
Microscopic Findings complete absence of this enzyme activity.
The mass of platelets injected into the subcutane- At Days 9-14, the number of foamy cells was
ous tissues was stained pinkish red with H&E. At 3 gradually decreased and replaced by macrophages
and 6 hours, a few neutrophils infiltrated around the with usual appearance and fibroblasts. Subcutaneous
mass of platelets, and after 24 hours a large number nodules induced by platelet injection were completely
of neutrophils and several macrophages were noted resolved at Day 21.
·' W ..
3
Figure 3-The foamy cells are stained blackish blue. (Smith-Dietrich reaction) Figure 4- Several macrophages and some foamy cells contain
material immunoreactive for anti-murine platelet antibody. (PAP)Vol. 114 * No. 1 PLATELET PHAGOCYTOSIS AND FOAMY MACROPHAGES 107
Immunohistochemical Findings phages that contained abundant myelinlike material.
At Day 2, a few macrophages revealed platelet- These macrophages usually had moderately devel-
associated antigens in their cytoplasm. A small oped Golgi apparatus, a few mitochondria, and oc-
amount of material positive for anti-murine platelet casional lysosomal dense bodies (Figures 8 and 9).
Based on the light-microscopic study of semithin sec-
antibody were also noted in occasional neutrophils. tions stained with toluidine blue, the foamy cells
At Day 7, some foamy cells surrounding the platelet
mass exhibited platelet antigens in the cytoplasm
demonstrable by light microscopy were estimated to
(Figure 4). correspond to the macrophages containing abundant
myelinlike material. Acid phosphatase activity was
noted around and within digested platelets (Figure 7).
Electron-Microscopic Findings However, there was no demonstrable activity in most
At Day 2, many neutrophils and several macro- of myelinlike material. At Days 9 and 11, there were
phages phagocytized the intact-appearing or partially many dense bodies and some myelinlike material in
degraded platelets. A few macrophages had various many macrophages (Figure 10). At this stage several
kinds of degraded structures of platelets. The mem- fibroblasts were demonstrated in the lesion.
brane surrounding the engulfed platelets became Histologic and ultrastructural findings of the sub-
denser, and platelets revealed some darkening of the cutaneous nodules induced by the injection of sensi-
cellular matrix and condensation of their specific tized platelets and observed at Days 5 and 7 were es-
granules (Figure 5). As the intracellular digestion ad- sentially similar to those of the lesions produced by
vanced, the phagocytized platelets were condensed untreated platelets, except for the more numerous
into coarsely granular or mottled inclusions. At this foamy cells in this group (Figure 11).
stage, structures recognizable as platelets could Administration of corticosteroid had no significant
hardly be seen. Some myelinlike material was noted in influence on the histologic and ultrastructural find-
the phagocytic vacuoles. A large amount of product ings described above. Under the experimental condi-
immunoreactive for anti-murine platelet antibody tions employed in this study, this agent had neither
was recognized in most of the phagocytic vacuoles promotive nor suppressive effects on the appearance
(Figure 6). At Days 5 and 7, there were many macro- of foamy cells.
Figure 5-A macrophage contains several platelets in various stages of digestion. (x 17,700)108 ISHIHARA ET AL AJP · January 1984
6 a 7
Figure 6- A large amount of product immunoreactive for anti-murine platelet antibody is recognized in macrophages. (x 7000) Figure 7-
Some material positive for acid phosphatase can be seen in a lysosome (arrow), but no activity is demonstrated within myelinlike material.
(x 23,000)
Discussion essentially similar to those seen in the spleen from pa-
tients with ITP. As morphologic characteristics com-
Presently, it is generally accepted that the spleen mon between the two, they contain phospholipids and
plays a central role in the pathogenesis of ITP and myelinlike material in their cytoplasm, as revealed by
that splenectomy is the therapeutic measure of choice histochemistry and electron microscopy, respectively.
in the management of patients with this disease. The In previous reports it was suggested that the phos-
light- and electron-microscopic findings of the re- pholipid content' and the cytoplasmic inclusions5'6'9
moved spleen and observations concerning platelet observed in the foamy macrophages in the spleen
phagocytosis and foamy macrophages have been de- are derived from platelets as a result of platelet
scribed by many investigators.1`9 In order to provide phagocytosis. The present observations provide more
additional evidence for the mechanism of formation direct evidence supporting this concept. Furthermore,
of foamy macrophages in ITP spleens, they were ex- this experimental investigation clarifies the role of
perimentally reproduced in mice by subcutaneous in- steroid therapy and anti-platelet antibody in the ap-
jection of platelets. In the preliminary experiment, we pearance of foamy cells and demonstrates more pre-
administered platelets into the tail vein of the mice. cisely than hitherto the sequence in which these cells
Most of the infused platelets were trapped in the pul- are produced.
monary capillaries, and the animals died within a few Several investigators have suggested that the foamy
days. Therefore, a time-course observation was im- cells in the spleen from patients with ITP are related
possible in these animals. The best way to introduce to steroid therapy.3'4 On the contrary, others insist
platelets seemed to be via the splenic artery after that the appearance of these cells are not necessarily
laparatomy. But we found it technically very difficult related to this agent.2'6'7 Because corticosteroid is
to inject a large number of platelets into the mice by known to stabilize the cell membrane, it may be ex-
this route. These were the main reasons we injected pected that the administration of corticosteroid
platelet suspensions into the subcutaneous tissue. would inhibit platelet phagocytosis and intracellular
The light- and electron-microscopic features of the degradation and thus reduce the number of foamy
foamy cells experimentally induced in this study are cells. Contrary to this expectation, administration ofVol. 114 · No. 1
...
...'-.-
~
PLATELET PHAGOCYTOSIS AND FOAMY MACROPHAGES
~ ,,.:- :~.,~- , '~,~. '
,
· .~. . :..*~'~~~~~~~~~
··*I" ·q*
::
~.'~
....'-,~
.' ,".
~..~.?.,)~?..~
·
',~ ~
~ ~
8'
,. .
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..
19.
~ ~ ~
.,-,.,.r~.i
~ ~
r~~~~~~~~~~~.,~
...."""
AV'.~'
....'"'..~·-,. .. , ... ..
~.~,,,:.~,:,x~ .,. ,.q~ ~ '.' .·...
~.,
.
. . . .'~
..,,f.
.
'~. ,
'
.c~.-~~~.:.
;.?::"~~.
.1
-,
109
t t ow -,w 1A V "If.El, .
ti-I or-
VT:
0
A0, 0, ILP_.* "IItI-. v T -4 f0,A i.I.,
,t,
., .
..4w .&;I,_ 9
9
Figure 8-There are many macrophages that contain large amounts of myelinlike material. (x 2000) Figure 9-A large amount of myelinlike
material and dense bodies is noted in a macrophage. (x7000)110 ISHIHARA ET AL AJP · January 1984
10
Figure 10-A macrophage has dense bodies and some myelinlike material. (x 15,000) Figure 11 -Myelinlike material and dense bodies
foamy cell induced by subcutaneous injection of sensitized platelets. (x8500)
are noted in aVol. 114 * No. 1 PLATELET PHAGOCYIOSIS AND FOAMY MACROPHAGES 1ll
this agent neither accelerated nor inhibited the forma- area. It is reasonable to assume that the expansion of
tion of foamy cells under the experimental conditions the cytoplasm may give a weaker reaction. At present,
employed in this study. we have no conclusive evidence to exclude this possi-
It is widely accepted that platelets in patients with bility. However, electron microscopy revealed com-
ITP are sensitized with anti-platelet antibody, and plete absence of acid phosphatase in some foamy
they are easily phagocytized by the splenic macro- cells. In addition, lysosomal enzyme activity was
phages. In this experiment, foamy cells were more demonstrated in the expanded cytoplasm of macro-
conspicuous when sensitized platelets were injected to phages that contained morphologically recognizable
the subcutaneous tissue. This suggests that the plate- platelets. Because of these findings we favor the pos-
lets coated with immunoglobulins are more suscepti- sibility of enzyme exhaustion in the foamy cells. More
ble to phagocytosis than those without antibody. experimental evidence is necessary for a definite
However, the ultrastructural findings in foamy cells answer to this question.
were essentially the same between the two groups.
Although the binding of immunoglobulin to the References
platelet seems to be a prerequisite for accelerated 1. Saltzstein SL: Phospholipid accumulation in histiocytes
phagocytosis, it is not necessarily related to the for- of splenic pulp associated with thrombocytopenic
mation of foamy cells. purpura. Blood 1961, 18:73-88
The findings obtained in this study indicate that the 2. Landing BH, Strauss L, Crocker AC, Braunstein H,
Henley WL, Will JR, Sanders M: Thrombocytopenic
incompletely degraded platelet membranes are most purpura with histiocytosis of the spleen. N Engl J Med
responsible for the phospholipid content and myelin- 1961, 265:572-577
like material of the foamy cells. Histochemical 3. Hill JM, Speer RJ, Gedikoglu H: Secondary lipidosis
of spleen associated with thrombocytopenia and other
studies revealed a weak activity of acid phosphatase blood dyscrasias treated with steroids. Am J Clin Pa-
and P-glucuronidase in occasional foamy cells. Within thol 1963, 39:607-615
the macrophages that contained myelinlike struc- 4. Dollberg L, Casper J, Djaldetti M, Klibansky C, Vries
A: Lipid-laden histiocytes in the spleen in thrombocy-
tures, a trace activity of acid phosphatase was demon- topenic purpura. Am J Clin Pathol 1965, 43:16-25
strated by electron microscopy. Although immuno- 5. Firkin BG, Wright R, Miller S, Stokes E: Splenic
reactive platelet antigen was demonstrated in the macrophages in thrombocytopenia. Blood 1969, 33:
240-245
macrophages that contained morphologically recog- 6. Ishihara T, Matsumoto N, Uchino F: Foamy histiocytes
nizable platelets, it was hardly revealed in most of in the spleen associated with idiopathic thrombocyto-
foamy cells. These findings may lead to the proposal penic purpura. Acta Pathol Jpn 1974, 24:273-284
7. Tavassoli M, McMilian R: Structure of the spleen in
that the foamy cells have already consumed their lyso- idiopathic thrombocytopenic purpura. Am J Clin Pa-
somal enzymes in the process of platelet digestion, thol 1975, 64:180-191
and that the remaining lysosomal enzymes are not 8. Takahashi K, Hakozaki H, Terashima K, Kojima M:
Two distinctive types of lipid histiocytes appearing in
sufficient to digest myelinlike material further. The the spleen of idiopathic thrombocytopenic purpura:
phospholipids within the phagocytic vacuoles seem to Sea-blue histiocyte and foam cell. Acta Pathol Jpn
be vulnerable to lysosomal digestion, because lyso- 1977, 27:447-462
9. Luk SC, Musclow E, Simo GT: Platelet phagocytosis in
somal enzymes of macrophages are shown to contain the spleen of patients with idiopathic thrombocytopen-
phospholipase.15 As a possible explanation for a ic purpura (ITP). Histopathology 1980, 4:127-136
reduced reaction for lysosomal enzymes, we suggest 10. Sternberger LA, Hardy PH Jr, Cuculis JJ, Meyer HG:
that the macrophages have enough lysosomal en- The unlabelled antibody enzyme method of immuno-
histochemistry: Preparation and properties of soluble
zymes to digest proteinous substances, but phospho- antigen-antibody complex (horseradish peroxidase-
lipase activity to digest membrane phospholipids is antihorseradish peroxidase) and its use in identification
insufficient. On the basis of these speculations, it ap- of spirochetes. J Histochem Cytochem 1970, 18:315-
333
pears possible that in case of accelerated platelet 11. Baker JR: The histochemical recognition of lipine. Q J
phagocytosis, such as in ITP, the amount of ingested Mic Sci 1946, 87:441-470
12. Smith JL, Mair W: Further observations on the bichro-
cellular membranes is beyond the capacity of lyso- mate haematoxylin method of staining lipoids. J Pathol
somal digestion. Thus, the incompletely degraded Bacteriol 1910, 15:179-181
membrane constituents, especially phospholipids, 13. Barka T, Anderson PJ: Histochemical methods for acid
remain in the macrophages for a certain period of phosphatase using hexasonium pararosanilin as coup-
ler. J Histochem Cytochem 1962, 10:741-753
time and lead them to transform into lipid-laden mac- 14. Hayashi M, Nakajima Y, Fishman WH: The cytologic
rophages with a foamy appearance. demonstration of beta-glucuronidase employing naph-
An alternative explanation for the decreased activ- thol AS-BI glucuronide and hexazonium pararosanilin:
A preliminary report. J Histochem Cytochem 1964,
ity of lysosomal enzymes in foamy cells is that the 12:293-297
enzyme may be inducible, and extensive platelet 15. Waite M, Griffin HD: The phospholipase A of lyso-
phagocytosis results in the expansion of macrophage somes, Lysosomes in Biology and Pathology. Vol 5.
Edited by JT Dingle, RT Dean, Amsterdam, North-
cytoplasm and in a dispersion of enzymes over a large Holland, 1976, p 257You can also read
