вход по аккаунту


Interaction of polymorphonuclear leukocytes with immune complexes trapped in joint collagenous tissues.

код для вставкиСкачать
The present experiments were designed to investigate in vitro interactions between polymorphonuclear
leukocytes (PMN) and rabbit joint collagenous tissues
containing trapped immune complexes. Articular cartilages and menisci from antigen-injected and control
joints were incubated with normal PMN isolated from
rabbit peritoneal exudates or blood. After incubation of
cartilage and menisci from antigen-injected joints with
PMN, large numbers of PMN became attached to the
articular surface. In areas of superficial erosion, the
PMN invaded the tissue several cell diameters below
the articular surface. Through immunoelectron microscopy, degranulated PMN were observed in scattered
areas to phagocytose amorphous material containing
rabbit Ig. Following addition of PMN to control tissues,
only a few PMN became attached to the articular surface. When tissues from monosodium urate-injected
joints were incubated with PMN, these cells were found
From the Rheumatic Diseases Unit, Department of internal
Medicine, University of Texas Southwestern Medical School, Dallas,
Supported by U.S. Public Health Service Research Grants
#AM16209 and #AM09989.
Kazuhiro Ugai, MD: Fellow in Internal Medicine; Morris
Zif. MD: Professor of Internal Medicine, Recipient. USPHS Research Career Award, Southwestern Medical School; Hugo E. Jasin,
MD: Professor of internal Medicine, Southwestern Medical School.
Address reprints to Hugo E. Jasin, MD. Department of lnternal Medicine, University of Texas Health Science Center, 5323 Harry
Hines Blvd., Dallas, Texas 75235.
Submitted for publication August 24, 1978; accepted in revised form October 9, 1978.
Arthritis and Rheumatism, Vol. 22, No. 4 (April 1979)
attached to the surface in moderate numbers, but invasion into the tissues was not seen. These studies indicate that immune complexes trapped in joint collagenous tissues may lead to enhanced release of
lysosomal hydrolases.
In rabbits previously immunized to a soluble antigen, intraarticular injection of the antigen results in
the development of a chronic arthritis and the formation of immune complexes (IC) that are trapped within
the joint collagenous tissues and persist in the inflamed
joint for long periods of time (1-3). IC have also been
found in the articular cartilage and menisci obtained
from a majority of patients with rheumatoid arthritis
(4.5). The role of the trapped IC in the maintenance of
acute and chronic inflammation both in the animal
model and the rheumatoid joint is unknown, but recent
work from this laboratory (6) suggests that the trapped
complexes are able to activate the complement system
and generate mediators of inflammation. Moreover, the
trapped complexes induced the formation of pannuslike inflammatory capsules in immune complex-containing menisci transferred to the suprapatellar pouches
of previously immunized rabbits ( 6 ) .
Henson (7,8) and Hawkins (9) have previously
shown that 1C trapped on artificial membranes were
able to induce the release of lysosomal hydrolases from
polymorphonuclear leukocytes (PMN) permitted to interact with the 1C irreversibly trapped on such surfaces.
The release of lysosomal enzymes was attributed to “re-
3 54
verse endocytosis” secondary to t h e inability o f t h e
phagocytic cells t o remove t h e immune complexes attached t o t h e membranes, i.e. “frustrated phagocytosis.”
Lysosomal enzyme release was o b t a i n e d a t IC concentrations 40-fold less t h a n t h e concentrations needed t o
elicit enzyme release with IC complexes in solution (7).
The presence of IC irreversibly trapped near t h e articular surfaces in experimental arthritis and rheumatoid
joint collagenous tissues suggested that these tissues
could bring a b o u t a similar release of lysosomal en-
zymes from PMN.
Accordingly, t h e interaction of rabbit PMN with
j o i n t collagenous tissues containing irreversibly trapped
IC has been investigated. The results obtained suggest
that t h e trapped IC induce t h e phenomenon o f frustrated phagocytosis, which may play a role in t h e pathologic changes seen on t h e articular surface of animals
with antigen-induced arthritis and in patients with rheum a t o i d arthritis.
New Zealand white rabbits of either sex weighing 2.53 kg were used. Immunization and induction of arthritis with
bovine serum albumin (BSA) were performed as previously
described (10). Acute monosodium urate crystal induced arthritis was generated by intraarticular injection of 15 mg of
monosodium urate crystals (MSU) prepared by the method of
Seegmiller et a1 ( I I). PMN were obtained from normal rabbits
from two different sources: I ) PMN were collected from the
peritoneal cavity of rabbits injected intraperitoneally with 100
ml of sterile saline solution containing 0. I% oyster glycogen
(Eastman Kudak Co., Rochester, New York) 12 hours previously. The animals were killed and the peritoneal cavities
washed with 150 ml of sterile saline solution that contained
preservative-free heparin at a concentration of 10 units/ml.
The harvested cells were washed three times with sterile saline
solution and resuspended in RPMI 1640 tissue culture medium (Grand Island Biological Co., Grand Island, New York)
containing 10% fetal calf serum or normal rabbit serum. 2)
PMN were also obtained from rabbit peripheral blood by the
method described by De Shazo et al (12). Briefly, blood obtained from intracardiac puncture was mixed with 1/7 volume
of acid citrate dextrose solution and centrifuged at 500 x g for
20 minutes at room temperature. The plasma containing
mononuclear cells was discarded, and the erythrocytes and
P M N were resuspended in a 2.5% gelatin solution. The
erythrocytes were allowed to settle at 37°C for 1.5 hours and
the supernatant containing PMN was obtained. The cells were
washed several times and resuspended in tissue culture medium as described above. Differential counts showed consistently over 95% PMN when either method of purification was
Incubation of joint collagenous tissues with PMN. Immunized rabbits were injected intraarticularly with 2.5 mg
BSA in I ml of sterile saline 1-2 weeks prior to the experi-
ments. At the same time, the contralateral knees were injected
with either I ml sterile saline solution or 15 mg monosodium
urate crystal suspension and were used as a source of control
tissues. Articular cartilage and menisci obtained from such
animals were cut into 2 x 2 mm pieces. A few pieces were immediately fixed in buffered formalin for routine histologic
studies and in 5% glutaraldehyde-phosphate buffered saline
(PBS). pH 7.3, for electron microscopy studies. The remaining
pieces were incubated in I ml volumes with the PMN suspension containing between 5-10 X lo7 cells per mm3. Most experiments were performed with medium that contained 10%
fresh or heat-inactivated normal rabbit serum. Incubation was
performed in polystyrene capped tissue culture tubes (Falcon
Plastics, Division of Bioquest, Oxnard, California) at 37°C for
60 minutes with continuous end-to-end rotation at a rate of 12
rpm. After the incubation period, the tissues were washed
briefly in Hanks’ solution and fixed for routine histologic and
electron microscopic studies.
Electron microscopy studies. The joint collagenous
tissues were fixed in 5% glutaraldehyde for 2 hours and subsequently fixed in 1% OsO,. For immunoelectron microscopy,
the glutaraldehyde-fixed tissues were washed with PBS for at
least 24 hours, and tissue sections of 40-60 pm thickness were
reacted with horseradish peroxidase (HRP) conjugated goat
antirabbit immunoglobulin (Cappel Laboratories, Downingtown, Pennsylvania) for 3 hours at room temperature with
constant agitation and kept at 4°C overnight. Controls consisted of blocking tests with unconjugated goat antirabbit immunoglobulin prior to reacting the sections with the HRPconjugated reagent or HRP-conjugated normal goat IgG. After washing with PBS for at least 24 hours, the sections were
incubated at pH 7.6 in the diaminobenzidine medium with
H , 0 2 as described by Graham and Karnovsky (13) for 30
minutes at room temperature. All the sections were fixed in
1% OsO,. After dehydration and embedding in Epon, thin
sections were cut perpendicularly to the articular surface on
an LKB microtome and examined in a Phillips 300 electron
microscope with or without uranyl acetate and lead citrate
Assessment of results. Histologic sections stained with
hematoxylin and eosin were used to quantify the number of
PMN attached to the joint collagenous tissues. Three or more
areas of articular surface that showed accumulation of PMN
were selected, 100 pm lengths were scanned, and the number
of PMN attached to the surface were counted. For the purpose of graphic representation, the results were also expressed
as follows: 0 = 0-1 PMN/IOO pm; I + = 2-10 PMN/IOO pm;
2+ = 11-20 PMN/IOO pm; 3+ = more than 20 PMN/100 pm
length. A typical example of each grade is shown in Figure 1.
Statistical analysis was carried out by the paired sample Student’s t test for the numerical values (PMN/100 pm) and the
Wilcoxon matched pairs signed rank test for the discontinuous values.
Quantitation of PMN attached to articular surfaces. Cartilage and menisci from 11 BSA-immunized
rabbits were examined. They were all injected in one
Figure 1. Quantitation of polymorphonuclear leukocyte (PMN) attachment to joint collagenous tissues. 0 = 0-1 PMN/100 pm;
1+ = 2-10 PMN/100 pm: 2+ = 11-20 PMN/100 pm; 3+ = >20 PMN/100 pm ( X 250).
knee with BSA, and 7 were injected in the contralateral
knees with saline solution and 4 with MSU crystal suspension. The magnitude of PMN attachment to the articular surfaces is shown in Table I and Figure 2. PMN
were seen attached in greater numbers to cartilage and
menisci obtained from BSA-injected joints. Quantitation of PMN attachment yielded values ranging
from 0 to 3+, averaging 2+ or 19.7 for cartilage and
20.9 PMN per 100 pm for meniscus. The joint tissues
obtained from saline- and MSU-injected joints showed
a significantly lower degree of attachment of PMN; only
one specimen of articular cartilage obtained from MSUinduced arthritis showed a 2+ degree of attachment,
whereas the remaining tissues showed very small numbers or no attached PMN. Statistical analysis showed
significant differences between cartilage and menisci
from BSA-injected arthritic joints and the corresponding tissues from saline-injected joints (P < 0.01), and
between cartilage and menisci from BSA-injected joints
and the corresponding tissues from MSU-injected joints
Table 1. Magnitude of polyrnorphonuclear leukocyte attachment to
articular surfaces
Mean PMN/100pm f SEM*
19.7 4.3
20.9 f 3.2
4.2 f 2.7
3.2 2 1.5
2.7 f 0.9
0.9 f 0.3
* PMN = polymorphonuclear leukocytes; BSA
bumin: MSU = monosodium urate crystals.
bovine serum al-
3 56
Figure 2. Polymorphonuclear leukocyte attachment to experimental
and control joint collagenous tissues. ( 1 = cartilage: 0 = meniscus.
( P < 0.05 and 0.005, respectively). The numbers of
PMN attached to tissues from saline and MSU-injected
joints were not statistically different. Significant differences were not seen when the incubation medium contained either fresh or heat-inactivated normal rabbit
serum, so for the purpose of this work the results are
grouped together.
Electronmicroscopy studies. Prior to incubation
with PMN, the joint tissues, even from BSA-injected
joints, only rarely showed the presence of PMN on the
surface Qr within the tissues. After 1 hour of incubation
with PMN, articular cartilage pieces obtained from
BSA-arthritic joints showed many leukocytes attached
tightly to the articular surfaces (Figure 3). Many of the
PMN were seen to extend foot processes several micrometers into the cartilage tissues (Figure 4), and actual invasion of the cells into the depth of the eroded
cartilage was frequently seen (Figures 5 and 6). The
PMN were commonly degranulated, particularly the
cells within the substance of the cartilage tissue (Figures
5 and 7). The cartilage surfaces appeared markedly
eroded, and there was thick fibrin-like material on the
surface (Figure 6) which was in close apposition to the
superficial collagen fibers. In other areas examined at
higher magnifications, the collagen fibers appeared
freely exposed to the surface. The superficial chondrocytes were damaged, and cell debris and lipid bodies
were present within the matrix. Amorphous electrondense bodies were scattered within the matrix in the superficial portion of the cartilage, and degranulated
PMN were commonly seen engulfing this material (Figure 7).
The findings in menisci were similar to those described for the articular cartilage. However, PMN invading the substance of this tissue were seen in greater
Figure 3. Attachment of polymorphonuclear leukocytes to articular cartilage obtained from a BSA-injected joint after one hour incubation at 37°C ( X 2900).
Figure 4. Higher power view of a polymorphonuclear leukocyte attached to cartilage obtained from a BSA-injected joint. A foot process
extends several micrometers into the cartilage ( X 9300).
numbers and to a greater depth than in cartilage extending to about 40 pm from the articular surface. Electron dense amorphous material was also found in the
matrix of menisci, and this material was detected to a
greater depth than that seen in cartilage (Figure 8). As
in cartilage, degranulated PMN were seen engulfing the
amorphous material as shown in Figure 9.
In saline-injected control tissues, the findings
were quite different. Rarely were PMN in close contact
with the articular surface. The few PMN seen in contact
with the surface were not degranulated (Figure lo), nor
were PMN seen to invade the substance of the tissue in
any of the joint tissues examined. There was no evidence of amorphous electron-dense material near the
articular surface.
The articular surface of the tissues obtained from
MSU-injected joints showed some damage, but not to
the same extent as that seen in tissues from BSA-injected animals. There was some fibrin-like material covering the surface in a patchy or irregular fashion (Figure 1 I ) . In some areas, a few PMN were attached, and
occasionally very short foot processes just penetrated
the articular surface, but only to one of two micrometers
deep (Figures I I and 12). No amorphous electron-dense
material was seen and the PMN appeared not to be
degranulated. No MSU crystals were seen attached to
the tissues or within PMN.
Immunoelectron microscopy studies. The joint
tissues were stained with HRP-conjugated antirabbit Ig
Figure 5. Invading polymorphonuclear leukocytes in the depth of eroded cartilage from a BSA-injected
joint. Note the degranulated PMN ( X 3200).
Figure 6. Invading polymorphonuclear leukocytes in the depth of eroded cartilage from BSA-injected
joint. The surface is covered with thick fibrin-like material ( X 3200).
Figure 7. Polyrnorphonuclear leukocyte within the substance of articular cartilage obtained from a BSAinjected joint showing degranulation of the cell and engulfment of amorphous electron dense material (x
in order to determine whether the electron-dense
amorphous material phagocytosed by the PMN contained immunoglobulin. In cartilage obtained from
BSA-injected joints, there were large amounts of peroxidase-positive material beneath the articular surfaces,
particularly within 5 pm of the surface (Figure 13). The
peroxidase-positive material had the same appearance
and distribution as the amorphous electron-dense material seen by conventional electron microscopy. Many
polymorphonuclear leukocyte processes were observed
in close apposition to HRP-positive material. That the
HRP-positive staining was specific for immunoglobulin
was shown by blocking studies using nonconjugated
goat antirabbit immunoglobulin (Figure 14), and HRPconjugated normal goat immunoglobulin. Both types of
controls showed almost complete disappearance of
HRP-positive material within the cartilage matrix,
while the PMN granules still showed positive staining
demonstrating the presence of myeloperoxidase. These
findings indicate that the amorphous material was rich
in rabbit immunoglobulin. Rabbit immunoglobulin was
not detectable in saline-injected articular tissues (Figure
Figure 8. Polymorphonuclear leukocytes invading a meniscus obtained from a BSA-injected joint, The deepest penetration in menisci
was about 40 pm from the articular surface (X 2900).
In the experiments described, the interaction of
rabbit PMN with joint Collagenous tissues Containing irreversibly trapped IC has been investigated. When
Figure 9. High power view of degranulated polymorphonuclear leukocyte engulfing amorphous electron
dense material in the depth of a meniscus obtained from a BSA-injected joint ( X 16,700).
Figure 10. Polymorphonuclear leukocytes close to cartilage obtained from a normal saline injected joint.
No contac! is seen between the cells and the articular surface. The polymorphonuclear leukocytes are not
degranulated ( X 8000).
Figure 11. Polymorphonuclear leukocyte interaction with articular cartilage obtained from a monosodium
crystal injected joint. Cells are occasionally attached to the cartilage and very short foot processes are
seen penetrating the articular surface to about 1 or 2 pm deep (X 3400).
Figure 12. Polymorphonuclear leukocytes attached to cartilage obtained from monosodium urate crystal
injected joint. Three cells are attached to the surface, but no invasion is seen ( X 9800).
Figure 13. Horseradish peroxidase-conjugated antirabbit Ig staining of articular cartilage incubated with
polymorphonuclear leukocytes. Large amounts of peroxidase positive material are seen beneath the articular surface. Many polymorphonuclear leukocyte (PMN) processes (arrows) are seen in close apposition to peroxidase positive material. Figures 13 to 15 have not been counterstained ( X 26,800).
36 1
Figure 14. Blocking study of articular cartilage from a BSA injected joint by use of nonconjugated goat
antirabbit immunoglobulin prior to incubation with peroxidase-conjugated anti-Ig. Note the disappearance of peroxidase positive staining within the substance of the cartilage in the lower half of the
field while some positive myeloperoxidase staining of polymorphonuclear leukocyte granules persists (arrow) (X 15,900).
PMN were incubated in vitro with articular tissues obtained from BSA-arthritic joints, large numbers of leukocytes were seen attached to the articular surfaces and
a significant number of cells penetrated within the substance of the collagenous tissues to a variable depth.
These findings were specific for antigen-injected joint
tissues, since neither saline- nor MSU-injected joint tissues demonstrated significant attachment or invasion by
the PMN. The difference between the numbers of PMN
attached to cartilage and menisci in BSA-injected arthritic joints and saline- or MSU-injected joints was statistically significant. No differences in PMN attachment
were seen between saline- and MSU-injected tissues.
We have previously shown that the immune
complexes trapped within the joint collagenous tissues
are capable of activating complement and generating
mediators of inflammation (6). Hollister has also shown
recently that in antigen-induced arthritis, the cartilaginous tissues incubated with fresh serum are able to generate chemotactic factors for PMN (14). Thus, it was not
surprising that a short incubation of such tissues with
normal PMN resulted in the extensive cellular attach-
ment and invasion seen in our studies. That the PMN
seemed to be actively attracted toward the tissues was
supported by the morphology of many of the cells attached to the tissues obtained from experimental joints.
Malech et a1 (15) have shown that in PMN oriented in a
chemotactic gradient, the nucleus tends to be situated
away from the attractant, as seemed to be the case in
our experiments (Figure 3).
No apparent differences were seen in the attachment and invasion of PMN when either fresh or heatinactivated normal rabbit serum was added to the medium. However, on the basis of the available data, it
was not possible to arrive at any definite conclusions
with regard to the role of complement in this phenomenon. Invasion of PMN in the absence of complement
could be due to chemotactic factors generated in vivo
within the antigen-injected joints and still remaining in
the tissues after the washing procedure.
The PMN invading the tissues were significantly
degranulated, and cells engulfing amorphous electrondense material that contained rabbit immunoglobulin
were seen in the depth of the articular tissues. These
Figure 15. Horseradish peroxidase conjugated anti-lg staining of cartilage obtained from a saline injected joint. Note the absence of peroxidase positive material ( X 30,000).
findings are similar to those by Henson (7,8) and Hawkins (9) who showed increased lysosomal enzyme release when PMN were allowed to interact with insoluble immune complexes attached to artificial
membranes. In experiments not presented in this report
(16), lysosomal enzyme release by PMN has been measured in our laboratory in experiments similar to the
ones reported here, but no detectable differences were
found between experimental and control tissues. This
was thought to be due to the small amount of IC available in the joint. In the experiments of Henson (7), at
least 25 p g of IC per membrane were needed to elicit the
release of lysosomal hydrolases, while in the experimental joint tissues used in this study, the greatest amount
of complexes trapped did not exceed 5 p g of antigen per
joint ( 1,2).
The pathologic significance of our findings is further supported by the studies of Oronsky et al (17) and
Ignarro (18). These workers demonstrated that in rabbit
auricular cartilage previously opsonized with aggregated IgG, PMN were able to digest the matrix proteoglycan whereas nonopsonized cartilage remained intact.
Careful examination of articular surfaces of car-
tilage samples obtained from rheumatoid joints and of
cartilage and meniscus samples from rabbits with antigen-induced arthritis has failed to reveal significant
numbers of PMN attached to the surface in a fashion
similar to that seen following in vitro incubation (19).
This failure to detect cells attached to the surfaces in
vivo may not be surprising if one considers that inflammatory synovial fluids are rich in chemotactic factors
(20), and these would tend to inhibit the chemotaxis of
PMN toward the tissue and its trapped immune complexes in vivo. In the in vitro situation, the tissues would
presumably provide the only source of chemotactic factor, either through entrapment of synovial fluid or
through generation of chemotactic factors at the surface
of the cartilage, thus resulting in the attachment and invasion of PMN observed. However, any decrease in
concentration of chemotactic factors in the synovial
fluid might allow PMN invasion into the tissues.
Since IC trapped in joint tissues obtained from
patients with rheumatoid arthritis have been previously
demonstrated in this laboratory (4,5), it would be of interest to determine whether the phenomenon demonstrated here with the rabbit joint collagenous tissues
would take place with the human counterparts. If the
3 64
phenomenon of frustrated phagocytosis and consequent
increase in lysosomal hydrolase release takes place in
vivo, these observations would suggest t h a t this mechanism may play a role i n t h e destructive changes seen in
t h e articular tissues of patients with rheumatoid arthritis.
We wish to thank Mr. Ellis Lightfoot for expert technical assistance, and Ms Jean Freeman and Monica Cassano
for secretarial help.
I . Cooke TD. Jasin HE: The pathogenesis of chronic inflammation in experimental antigen-induced arthritis. I. The
role of antigen on the local immune response. Arthritis
Rheum 15:327-337, 1972
2. Cooke TD. Hurd ER. Ziff M. Jasin HE: The pathogenesis
of chronic inflammation in experimental antigen-induced
arthritis. 11. Preferential localization of antigen-antibody
complexes to collagenous tissues. J Exp Med 135:323-338.
3. Hollister JR. Mannik M: Antigen retention in joint tissues
in antigen-induced synovitis. Clin Exp lmmunol 16:615627, 1974
4. Cooke TD, Hurd ER, Jasin HE, Bienenstock J. Ziff M:
Identification of immunoglobulins and complement in
rheumatoid articular collagenous tissues. Arthritis Rheum
18:541-551, 1975
5. lshikawa H, Smiley JD, Ziff M: Electron microscopic
demonstration of immunoglobulin deposition in rheumatoid cartilage. Arthritis Rheum 18563-576, 1975
6. Jasin HE, Cooke TD: The inflammatory role of immune
complexes trapped in joint collagenous tissues. Clin Exp
Immunol 33:416-424. 1978
7. Henson PM: The immunologic release of constituents
from neutrophilic leukocytes. I. The role of antibody and
complement on non-phagocytosable surfaces or phagocytosable particles. J lmmunol 107: 1535-1546, 1971
8. Henson PM: The immunologic release of constituents
from neutrophilic leukocytes. 11. Mechanism of release
d u r i n g phagocytosis a n d adherence to non-phagocytosable surfaces. J Immunol 107:1547-1557, 197 I
9. Hawkins D: Biopolymer membrane: a model system for
the study of the neutrophilic leukocyte response to immune complexes. J lmmunol 107:344-352, 1971
10. Jasin HE: Mechanism of trapping to immune complexes
in joint collagenous tissues. Clin Exp Immunol 22:473485. 1975
I I. Seegmiller JE, Howell RR, Malawista SE: The inflammatory reaction to sodium urate. JAMA 180:469-475, 1962
12. DeShazo CV. Henson PM, Cochrane CG: Acute immunologic arthritis in rabbits. J Clin Invest 5150-57, 1972
13. Graham RC. Karnovsky MJ: The early stages of absorption o f injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a
new technique. J Histochem Cytochem 14:29 1-302, 1966
14. Hollister JR: Inflammatory potential of tissue-bound immune complexes (abstract). XIV International Congress
of Rheumatology 97, 1977
15. Malech HL. Root RK. Gallin JI: Structural analysis of
human neutrophil migration: centriole, microtubule and
microfilament orientation and function during chemotaxis. J Cell Biol 75:666-693, 1977
16. Jasin HE: Unpublished observations
17. Oronsky A. lgnarro L. Perper R: Release of cartilage
mucopolysaccharide degrading neutral protease from human leukocytes. J Exp Med 138:461-472, 1973
1X. Ignarro LJ: Release of neutral protease and P-glucuronidase from human neutrophils in the presence of cartilage
treated with various immunologic reactants. J Immunol
I 13:298-308. 1974
19. Ugai K: Unpublished observations
20. Ward PA. Zvaifler NJ: Complement-derived leukotactic
factors in inflammatory synovial fluids of humans. J Clin
Invest 50:606-616. 1971
Без категории
Размер файла
4 998 Кб
interactions, immune, trappes, polymorphonuclears, collagenous, joint, leukocytes, complexes, tissue
Пожаловаться на содержимое документа