Annals of the Rheumatic Diseases, 1977, 36, 293-301

Rheumatoid synovial cells from intact jointsMorphology, growth, and polykaryocytosis

B. J. CLARRIS, J. R. E. FRASER, C. J. MORAN, AND K. D. MUIRDEN

From the University of Melbourne Department of Medicine, Royal Melbourne Hospital, Victoria, Australia

SUMMARY Synovial cell lines were isolated by instillation of trypsin or chymotrypsin into intactknee joints of patients with persistent rheumatoid effusions resistant to conventional therapy.Morphology and growth in the primary phase were compared with rheumatoid cells isolated fromexcised synovium and nonrheumatoid synovial cells obtained from intact joints of cadavers oramputated limbs. Cell populations from all sources included varying proportions ofmarcophage-likeand fibroblast-like cells, with only 1-3 % multinucleated cells. In medium supplemented with calfserum alone, rheumatoid cells from intact joints showed negligible changes in morphology. However,in the presence of nonrheumatoid, autologous rheumatoid or homologous rheumatoid serum a rapidincrease occurred in size of the macrophage-like cells and numbers of polykaryocytes, including somegiant syncytial cells. These effects were directly proportional to serum concentration and wereidentical in fresh or heat-inactivated serum. In most of these rheumatoid cell lines no multiplicationoccurred, regardless of serum type or concentration. In rheumatoid synovial cells from excisedsynovium, human serum induced both polykaryocytosis aiid rapid growth of fibroblasts. Non-rheumatoid synovial cells grew rapidly but few polykaryocytes developed, mostly with less than 6nuclei.

Evidence of viral infection in rheumatoid synovial cells was sought by electron microscopy afterstimulation of polykaryocytosis by human serum. In one of the cultures many cells were found withintranuclear particles possessing characteristics of the adenovirus group.

Some years ago it was shown that large numbers ofcells could be separated from synovial intima withoutdissection, simply by instillation of trypsin solutioninto intact joints (Fraser and Catt, 1961). Apart fromone patient with a persistent knee joint effusion(J. R. E. Fraser, 1964, unpublished), the techniquewas restricted in this laboratory to cadaver jointsand amputated limbs. However, Ford and Oh (1965)applied it to a virological study of rheumatoidsynovial cells. In this and subsequent studies nodeleterious effects were observed in the patients(D. K. Ford, personal communication, 1975). Inconsidering a possible alternative to synovectomy inthe treatment of rheumatoid joint effusions, it wasdecided to explore the therapeutic potential of thetechnique in patients with effusions resistant to localsteroid and conventional systemic therapy. The

Accepted for publication October 6, 1976Correspondence to Dr. B. J. Clarris, University of MelbourneDepartment of Medicine, The Royal Melbourne HospitalP.O., Victoria 3050, Australia

clinical results of treatment with intra-articularproteolytic enzymes will be considered separately.This paper is primarily concerned with the cyto-logical and cultural characteristics of the synovialcells isolated from intact rheumatoid joints.

Material and methods

ISOLATION OF RHEUMATOID SYNOVIAL CELLSFrom intact jointsThe procedure has so far been used only on kneejoints. A sterile 13 or 16 g catheter is inserted beneaththe patella on the medial side. The synovial fluid isaspirated with a syringe and the joint rinsedthoroughly with physiological saline to remove fibrinand free cells. Approximately 40 ml of salinecontaining 125 000 units trypsin (Tryptar, ArmourPharmaceutical Co. Ltd.) or 10 000 units of a-chymotrypsin (Chymar, Armour PharmaceuticalCo. Ltd.) are introduced and left in situ for 10-15minutes. The cell suspension is withdrawn, centri-fuged, and the pellet resuspended in culture medium

293

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

294 Clarris, Fraser, Moran, Muirden

(Eagle's basal medium with 20% fetal calf serum).The procedure is usually repeated, with an inter-vening saline rinse. The cells adhere readily to glassor polystyrene culture flasks within 24 hours. Asshown later, the behaviour of the cells depends onthe type and concentration of serum in the culturemedium.

From excised synovial tissueSynovial tissue is transported to the laboratory inEagle's medium. The synovial surface is dissectedfrom underlying tissue, then scraped gently with asterile scalpel blade into Hanks's balanced saltsolution (Ca/Mg free). Trypsin is added (finalconcentration 0 25 %), and the cell suspensionincubated at 37'C for 20-30 minutes, then centri-fuged and mixed in culture medium as above.

ISOLATION OF NONRHEUMATOID SYNOVIAL CELLSThese were obtained from intact cadaver joints usingthe procedure of Fraser and Catt (1961). Methods ofmaintaining these cells in culture have been describedpreviously (Fraser and McCall, 1965).

CULTURE MEDIAEagle's basal medium (EBM), animal sera (fetal calf,newborn calf, adult horse), trypsin, and salt solu-tions (Hanks's balanced salt solution =HBSS;Dulbecco-Vogt phosphate buffered saline = PBS)were obtained from Commonwealth Serum Labora-tories, Parkville, Victoria.Nonrheumatoid serum was obtained aseptically

from several fasting donors. Rheumatoid serum wasprepared from a patient who had received no drugsother than gold in the previous 3 months.

LIGHT MICROSCOPYLiving cultures were examined by phase contrastillumination (Olympus PBM 6). Coverslip cultureswere stained with Mayer's haematoxylin-eosin.

ELECTRON MICROSCOPYIn some isolations a portion of the trypsin-salinecell suspension direct from the joint was immediatelycentrifuged and the pellet fixed in 4% glutaralde-hyde, for 4 hours at 4°C. Cell cultures were rinsedwith warm PBS or HBSS and prepared for electronmicroscopy as described previously (Le Marshallet al., 1977).

PHAGOCYTIC ACTIVITYThis was examined by incubating cultures overnightat 37°C with washed polyvinyl chloride (Geon, B. F.Goodrich-C.S.R. Chemicals Pty. Ltd., Glen Iris,Victoria) in culture medium.

Results

ULTRASTRUCTURE OF RHEUMATOID SYNOVIALCELLS FROM INTACT JOINTS BEFORE CULTURELight and electron microscopy of joint aspiratesshowed mainly neutrophils, lymphocytes, erythro-cytes, and fibrin. Trypsin-saline cell suspensionscontained some cells of these kinds, but the dominantcell was a mononuclear cell with macrophage-likefeatures (Fig. 1), similar to the type A synoviocytein rheumatoid arthritis (Barland et al., 1962).Borders were irregular, with many filopodia whichwere often long and branched. The cytoplasmcontained a variety of vacuoles, primary lysosomes,and cytolysosomes filled with debris. Other featureswere cytoplasmic filaments, prominent Golgi zones,free ribosomes but little endoplasmic reticulum,mitochondria (sometimes swollen with disruptedcristae), and occasional myelin figures and lipiddroplets. Between the cells were granular debris,fibrin, and a few collagen fibres.

BEHAVIOUR OF SYNOVIAL CELLS IN CULTUREWithin 24 hours synovial cells from all sourcesattached firmly to the culture surface and assumedvarious shapes by flattening and extending cyto-plasmic projections (Fig. 2). At this stage some cellswere elongated, with 2 or more projections, althoughfew could be identified unequivocally as 'fibro-blasts'. Others remained more or less circular inoutline. Cytoplasmic detail in the smaller cells wasoften difficult to assess by phase contrast, butstained preparations showed the presence of 1-3%multinuclear cells (polykaryocytes = polykaryons).Growth and morphology later in culture were

related to the origin of the cultures and the type andconcentration of serum in the culture medium.

(i) Nonrheumatoid synovial cellsTypically there was a 'lag' phase of 2-10 days inwhich cell numbers did not change appreciably inany growth medium. In this period some cellsassumed the elongated forms of 'fibroblasts'; othersappeared stellate or 'tulip-shaped'. The remaindershowed morphological characteristics of macro-phages. These were either small and irregular with 2or more projections or circular cells with a relativelysmooth margin. The fibroblasts later multiplied andformed confluent monolayers which could besubcultured repeatedly. In medium containinghuman serum, some of the macrophages enlargedconsiderably and a few polykaryocytes appeared.The latter usually contained less than 6 nuclei. Fewof the macrophages could be dislodged by trypsinand only small numbers could be found in sub-cultivated cells.

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

Rheumatoid synovial cells from intact joints 295

(a)

Fig. 2 Synovial cells in culture 24 hours after isolationfrom an intact rheumatoid knee joint (Case A. V.),showing heterogeneous morphology of the cells. Phasecontrast x 240.

(b)

Fig. 1 Electron micrographs of cells washedfrom anintact rheumatoid knee joint (Case C.S.) after trypsininstillation. (a) Group ofmacrophage-like cellscharacterized by rounded outline with numerous filopodia,many vacuoles and mitochondria, andprominent Golgizones. x 4300. (b) Section ofa macrophage-like cellshowing features typical of type A synovial lining cells.x 8750.

(ii) Rheumatoid synovial cells from excised tissueGrowth and culture behaviour were similar to non-rheumatoid synovial cells except that polykaryo-cytes usually became more numerous and somesyncytial cells with many nuclei appeared (Fig. 3).

(iii) Rheumatoid synovial cells from intact jointsCell growth was rarely sufficient to permit successfulsubculture irrespective of composition of the growthmedium, although late multiplication of fibroblast-like cells occasionally occurred in widely separatedpatches. With human serum in the culture medium,the macrophage-like cells typically enlarged toseveral times their original diameter and after 3-7days cultures often consisted predominantly ofround or polygonal strongly phagocytic cells up to80 Xum in diameter. Many cells developed denseclusters of organelles around the nucleus, which wasoften partially or totally obscured. Some macro-phages appeared to be attached to the substratum byfilamentous processes radiating from the entire

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

296 Clarris, Fraser, Moran, Muirden

periphery. Large round cells were often seen with asingle long process extending from the centralcytoplasm rather than from the margin (Fig. 4).

Fig. 3 Macrophage-like cells in a rheumatoid cell lineisolatedfrom excised synovium (Case M.C.), after 21days in medium containing human andfetal calf serum(20% each). A large polykaryocyte is surrounded bysmaller macrophage-like cells with dense central clustersof organelles. Phase contrast x 240.

Polykaryocytes became common after 3-7 days andgiant syncytial cells containing large numbers ofnuclei could often be found (Fig. 5). In later culturesmultiple small vacuoles appeared in a high pro-portion of the enlarged macrophages. Cellularenlargement and polykaryocytosis were equallystriking in nonarthritic serum and homologous andautologous rheumatoid serum. In contrast, fetal calfserum (FCS) or newborn calf serum alone producedlittle increase in cell size or numbers of polykary-ocytes even after months of culture. In adult horseserum macrophages began to enlarge after severaldays in culture, and some polykaryocytes formed,most with less than 6 nuclei.The effects of human and bovine sera on poly-

karyocytosis were compared by counting polykaryonsand mononuclear cells in multiple randomly selectedmicroscope fields, using living cultures or stainedpreparations. Fig. 6 shows the typical reaction of therheumatoid cell lines isolated from intact joints. In50% FCS the polykaryon index (PI) remainedapproximately constant throughout the 13 days inboth heat-inactivated (56°C, 30 min) and unheatedserum. In contrast, changes in 50% human serumwere even more rapid than those seen previously incultures grown in mixed FCS and human serum. In10% human serum, cell numbers declined and thePI increased only slightly, although considerablenumbers of surviving cells enlarged markedly as in

Fig. 4 Macrophage-like cells in a synovial cell line Fig. 5 A giant syncytial cell in a rheumatoid synovialisolatedfrom an intact knee joint and cultivated in cell line isolatedfrom an intact knee joint (Case M.S.),medium containing nonrheumatoid human serum. Note the after 11 days in 50% heat-inactivated nonrheumatoidcytoplasmic process extending from the nuclear region. serum. Cells fixed in methanol, stained with Mayer'sPhase contrast x 300. haematoxylin-eosin x 330.

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

Rheumatoid synovial cells from intact joints 297

Lr-IbiSOFCS 1S 25HS 50HS

13DAYS

Fig. 6 Effect of50% fetal calfserum (FCS) andvarying concentrations ofnonrheumatoid human serum(HS) on the polykaryon index (percentage of the countednucleifound in polykaryons) ofa rheumatoid synovialcell line isolatedfrom an intact knee joint (Case M.G.).

the higher serum concentration. In 25% humanserum changes in cell size and numbers of poly-karyocytes were intermediate between 10% and50%. The effects of human serum were similarwhether heated or unheated.

Fig. 7 shows the distribution of polykaryons inone of the rheumatoid cell lines from an intact joint,grown in 50% FCS or human serum. In the formernearly all polykaryons contained 2 or 3 nucleithrough the l1 days of observation. However, in50% human serum the proportion of cells with 2 or3 nuclei declined as cells with much greater numbersof nuclei appeared.

loor

10

50 / HS507 FCS

DAY 0

)o -

DAY 7

eL I v10ioor

DAY 11

2 3 4 5 6-10 11-60

NUCLEI PER CELL

Fig. 7 Distribution ofnumbers ofnuclei per polykaryonin a rheumatoid cell line isolatedfrom an intact kneejoint (Case M.S.). The effect of50% FCS is comparedwith 50% nonrheumatoid HS. Vertical bars show numbersof cells with designated number of nuclei, as a percentageof total polykaryons counted. Bars at day 0 showpolykaryons before serum treatment.

Fig. 8 shows the pattern of polykaryon formationin four nonarthritic synovial cultures in the primaryphase. Polykaryons were often obscured by rapidovergrowth of fibroblasts, e.g. in 50% human serum,cell lines IN and 8N could not be counted accuratelyafter 5 and 9 days respectively. Some polykaryonsformed but were proportionately fewer than inrheumatoid cultures and rarely contained more than2 or 3 nuclei.

ELECTRON MICROSCOPY OF CULTURED RHEUMA-TOID CELLS FROM INTACT JOINTS (FIG. 9)The cells were much larger than the cells originallywashed from the joint. Although there were manypolykaryons in the cells scraped from the culturesurface, few could be recognized as such in electronmicrograph sections, probably due to thinness of thesections or orientation of cells. The cytoplasm waspleomorphic and showed the same features found inthe macrophages before culture.

Evidence of viral infection has been sought in asmall number of these cell lines and virus-likeparticles were found in one of the cultures. Theseconsisted of intranuclear structures of constant size(approximately 80 nm in diameter) occurring inmany cells. Some were closely grouped in clumps,others were scattered singly through the nucleus.Occasionally a dense core was seen and less com-monly a halo surrounded the particle. The outermargin was irregular, suggesting either 'spikes' or amany-faceted surface. The identity of the structuresis uncertain but their size, shape, and location areconsistent with virions of the adenovirus group. Thenuclei and the surrounding cytoplasm showed noother unique features. The particles were not foundin the cytoplasm or seen to bud from the nuclear orplasma membrane.

z 20

z0

' 10

0

50/1HS 50/ FCS

ION

/<3N N* *. '3N

6 10 14 2 6 10 14

TIME (DAYS)

Fig. 8 Change ofpolykaryon index (see Fig. 6) of4 nonrheumatoid synovial cell lines in 50% nonrheumatoidHS and 50% FCS. N=cell line.

[HEATED SERUM

0 UNHEATED

s-kx0 20z

z0

4o

2

5OFMS IOHS 25HS 50HS7 DAYS

(Az0

4

0.n

-

40

0w

z

LU

2

nLI

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

298 Clarris, Fraser, Moran, Muirden

Discussion

Synovial cells in primary cultures were morphologic-ally heterogeneous regardless of the source. Afterspreading on culture surfaces, cells with macrophage-like or fibroblast-like features could be recognized.The relative proportions of these cell-types variedconsiderably from one isolation to another, butrheumatoid synovial cells from intact joints wereinvariably distinguished by high proportions ofmacrophage-like cells, which is consistent with thepreponderance of type A synoviocytes in the surfaceof rheumatoid synovium (Ghadially and Roy, 1969).These cultures might also include inflammatorymacrophages of haematogenous origin.When rheumatoid synovial cells were first isolated

from an intact joint in 1964, specific differences werenot observed in the primary culture and the subse-quent life history of the cell line was typical of themany nonrheumatoid synovial cell lines grown inthisflaboratory, and of rheumatoid cells obtained by

trypsinizing cellular material scraped from thesurface of synovial tissue excised during surgery.Thus, although polykaryocytosis has long beenrecognized as a feature of human rheumatoidsynovial cells in culture (Lackington, 1959; Stanfieldand Stephens, 1963; Bartfeld, 1965), the morpho-logical peculiarities and apparently suppressedgrowth of the rheumatoid cell lines from intact jointsexamined in the present study were somewhatunexpected.Palmer (1971) examined the influence of rheuma-

toid serum on synovial cells. Giant syncytia werefound to occur in rheumatoid synovial cells fromtrypsin-dispersed tissue when cultured in heat-inactivated rheumatoid serum, but did not developin synovial cells from osteoarthrosis. Palmer alsonoted suppressed growth of fibroblasts in rheuma-toid synovial cultures, apparently related to thenumber of macrophages present.

In the present study rheumatoid synovial cellsfrom intact joints were compared with rheumatoid

a)

V

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

Rheumatoid synovial cells from intact joints 299

Fig. 9 Electron micrographs ofcells isolatedfrom an intact jointof a rheumatoid patient (Case___J.B.) and culturedfor 31 days in50% nonrheumatoid human serum.(a) Cytoplasmic features of atypical enlarged macrophage-like

~. cell. Large vacuoles (V)containing debris apparentlyphagocytosedfrom the cellenvirons, and several lysosomes

_ (L) can be seen. The nucleus (N)is surrounded by a narrow zone

- rntec(arrows) apparently containingmicrotubukes. x 32300.(b) Portion ofa nucleus showingclusters of virus-like particles

.~ (arrows) with diameter of 80 nmin the vicinity of the nucleolus(NU). x 20 000. (c) A higherpower section including virus-like

_ particles (arrows), showing densecores in some, and a granularsurface suggesting 'spikes' ormultifaceted structure. x 27500.

.3,~~~~~~~~. ~~(b)

(C)

cells from excised synovium and nonrheumatoid n- _ Icells from intact joints of cadavers or amputatedlimbs. Responses were examined in a range of serum g'stypes including rheumatoid and nonrheumatoid sera ;and serum from animals. After 24 hours in culture, [sonly 1-3% of the synovial cells were multinucleate. -No clear-cut difference could be established betweenthe proportion of polykaryocytes in rheumatoid and tk : ?t.;j:nonrheumatoid synovial cells, although such a . _difference was found by Kinsella et aL. (1970) intrypsin-dispersed synovial cells before culture.Polykaryocytes occurred as both macrophage-likeand fibroblast-like cells but the former predominatedgreatly. Occasionally, larger polykaryocytes with i.-8-12 nuclei were detected in the first 24 hours ofculture, possibly representing the giant cells identifiedin rheumatoid synovial tissue (Grimley and Sokoloff,1966).In fetal calf or newborn calf serum rheumatoid 4

synovial cells from intact joints remained virtually -3unaltered in appearance or numbers, sometimes formonths in culture. Nonrheumatoid synovial cells orrheumatoid cells from excised tissues cultured in Csimilar media were usually overgrown with fibro-blast-like cells within 6-30 days. When human

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

300 Clarris, Fraser, Moran, Muirden

serum was included in the medium, macrophages inthe rheumatoid cultures enlarged rapidly and largepolykaryocytes appeared. Polykaryocytes becameparticularly numerous in the rheumatoid culturesfrom intact joints and the process was acceleratedwhen human serum alone was used to supplementthe medium, but growth of fibroblasts usuallyremained suppressed. The rate of the morphologicalchanges in these rheumatoid synovial cells wasdirectly proportional to serum concentration.Similar changes could be produced when rheumatoidsynovial cells cultured for long periods in fetal calfserum were changed to human serum. The effects ofhuman serum were not altered by mild heating. Non-rheumatoid serum and rheumatoid serum appearedto be equally active in inducing polykaryocytosis.Interactions between rheumatoid synovial cells andserum were not species specific since some poly-karyocytosis also occurred in medium containingadult horse serum. Polykaryocytosis also occurredin some nonrheumatoid cell lines, although thechanges were comparatively slow and it was unusualto find cells with more than 6 nuclei.

These findings support other evidence (Palmer,1971; Mackay et al., 1974) that rheumatoid synovialcells are especially prone to polykaryocytosis invitro; although negligible data are available oncomparable in vitro responses of synovial cells fromconnective tissue diseases related to rheumatoidarthritis. Since blood macrophages from rheumatoidsubjects also form numerous polykaryocytes in vitro(Panayi et al., 1974), it is not clear whether theresponses of rheumatoid synovial cells to humanserum are associated with type A synoviocytes aloneor also with infiltrating macrophages from the bloodstream.Although polykaryocytes of limited size can

develop by amitotic division (Roizman and Schlue-derberg, 1962), the larger syncytia seen in rheumatoidsynovial cells in the present study are more likely toform by fusion between cells. This is supported byobservations ofnuclei in cytoplasmic bridges betweenrheumatoid synovial macrophages (Palmer, 1971)and in cytoplasm linking fusing blood macrophages(Sutton and Weiss, 1966). Further investigationwould be required to elucidate the nature of thefactor in serum responsible for induction of cellfusion.The apparent relationship between poor growth

of rheumatoid synovial cells and the presence ofmacrophage-like cells suggests that the former inhibitmitosis in neighbouring cells (Palmer, 1971). In alater study Palmer (1975) was unable to show directinhibition of fibroblast growth by rheumatoidcultures. However, soluble inhibitors secreted bymacrophages have been found in animal studies

(Nelson, 1973; Calderon et al., 1974). Work isunderway in this laboratory to investigate the effectof rheumatoid macrophages on cell growth and thereason for the difference in growth behaviour foundbetween rheumatoid synovial cells from intact jointsand excised tissue.

Viruses have been much sought in rheumatoidsynovium and cultured synovial cells. The largesyncytia seen in the rheumatoid cultures might reflectactivation of a syncytium-inducing virus (Roizman,1962). A variety of unidentified structures, whichmight be viruses, have been found in rheumatoidsynovium by electron microscopy (Highton et al.,1966; Gy6rkey et al., 1972; Serre et al., 1972;Schumacher, 1975). However, no relationship hasyet been established between viruses or otherorganisms and causation of rheumatoid arthritis. Inthe present study intranuclear inclusions were foundin synovial cells from one patient with rheumatoidarthritis, after extended culture in human serum.The structure of the particles strongly suggestedvirions ofthe adenovirus group. However, since theseviruses are extremely common, their occurrencemay have been purely by chance and investigationof the significance of this initial finding iscontinuing.

This work was supported by a grant from theNational Health and Medical Research Council ofAustralia. The authors are grateful to Dr. G. Varigosand Dr. A. Stockman for isolations of many of thecell lines from rheumatoid patients; Mr. A. Robinsonfor capable assistance in the isolations; Mr. K.Rogers for the electron microscopy; Dr. T. Mandeland Dr. I. Holmes for critical evaluation of virus-like structures in electron micrographs.

References

Barland, P., Novikoff, A. B., and Hamerman, D. (1962).Electron microscopy of the human synovial membrane.Journal of Cell Biology, 14, 207-220.

Bartfeld, H. (1965). Rheumatoid arthritic and non-rheumatoidsynovium in cell culture. Morphological observations,acridine orange and fluorescent fraction II studies. Annalsof the Rheumatic Diseases, 24, 31-39.

Calderon, J., Williams, R. T., and Unanue, E. R. (1974). Aninhibitor of cell proliferation released by cultures ofmacrophages. Proceedings of the National Academy ofSciences of the USA, 71, 4273-4277.

Ford, D. K., and Oh, J. 0. (1965). Use of 'synovial' cellcultures in the search for virus in rheumatoid arthritis.Arthritis and Rheumatism, 8, 1047-1052.

Fraser, J. R. E., and Catt, K. J. (1961). Human synovial cellculture. Use of a new method in a study of rheumatoidarthritis. Lancet, 2, 1437-1439.

Fraser, J. R. E., and McCall, J. F. (1965). Culture of synovialcells in vitro. Annals of the Rheumatic Diseases, 24,351-359.

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from

Rheumatoid synovial cellsfrom intact joints 301

Ghadially, F. N., and Roy, S. (1969). Ultrastructure ofSynovial Joints in Health and Disease, p. 28. Butterworths,London.

Grimley, P. M., and Sokoloff, L. (1966). Synovial giant cellsin rheumatoid arthritis. American Journal of Pathology,49, 931-954.

Gyorkey, F., Sinkovics, J. G., Min, K. W., and Gyorkey, P.(1972). A morphologic study on the occurrence anddistribution of structures resembling viral nucleocapsidsin collagen diseases. American Journal of Medicine, 53,148-158.

Highton, T. C., Caughey, D. E., and Ryans, D. G. (1966).A new inclusion body in rheumatoid synovia. Annals of theRheumatic Diseases, 25, 149-155.

Kinsella, T. D., Baum, J., and Ziff, M. (1970). Studies ofisolated synovial lining cells of rheumatoid and non-rheumatoid synovial membranes. Arthritis and Rheumatism,13, 734-753.

Lackington, C. (1959). Tissue culture of fibroblasts andround cells from arthritic synovial fluid. Arthritis andRheumatism, 2, 42.

Le Marshall, J., Fraser, J. R. E., and Muirden, K. D. (1977).Lysosomal activation by neutral saccharides in cellcultures of synovium. Annals of the Rheumatic Diseases,36, 130-138.

Mackay, J. M. K., Panayi, G., Neill, W. A., Robinson, A.,Smith, W., Marmion, B. P., and Duthie, J. J. R. (1974).Cytology of rheumatoid synovial cells in culture. I.Composition and sequence of cell populations in culturesof rheumatoid synovial fluid. Annals of the RheumaticDiseases, 33, 225-233.

Nelson, D. S. (1973). Production by stimulated macrophagesof factors depressing lymphocyte transformation. Nature,246, 306-307.

Palmer, D. G. (1971). Polykaryocytes and rheumatoid disease.Journal of the Royal College of Physicians of London, 6,33-40.

Palmer, D. G. (1975). Relationships between rheumatoidcells in culture. Annals of the Rheumatic Diseases, 34,235-238.

Panayi, G. S., Mackay, J. M. K., Neill, W. A., McCormick,J. N., Marmion, B. P., and Duthie, J. J. R. (1974). Cytologyof rheumatoid synovial cells in culture. II. Association ofpolykaryocytes with rheumatoid and other forms ofarthritis. Annals of the Rheumatic Diseases, 33, 234-239.

Roizman, B. (1962). Polykaryocytosis. Cold Spring HarbourSymposia on Quantitative Biology, 27, 327-342.

Roizman, B., and Schluederberg, A. E. (1962). Virus infectionof cells in mitosis. III. Cytology of mitotic and amitoticHEp-2 cells infected with measles virus. Journal of theNational Cancer Institute, 28, 35-54.

Schumacher, H. R. (1975). Synovial membrane and fluidmorphologic alteration in early rheumatoid arthritis.Microvascular injury and virus-like particles. Annals of theNew York Academy of Sciences, 256, 39-64.

Serre, H., Mandin, J., Vago, C., Sang, J., and Clot, J. (1972).Recherche et etude de virus dans la synoviale au cours dela polyarthrite chronique rhumatismale. Revue du Rhuma-tisme et des Maladies Osteioarticulaires, 39, 595-599.

Stanfield, A. B., and Stephens, C. A. L. (1963). Studies ofcells cultured from 188 rheumatoid and non-rheumatoidsynovial tissues. Texas Reports of Biology and Medicine,21, 400-411.

Sutton, J. S., and Weiss, L. (1966). Transformation ofmonocytes in tissue culture into macrophages, epithelioidcells, and multinucleated giant cells. An electron micro-scope study. Journal of Cell Biology, 28, 303-332.

copyright. on F

ebruary 1, 2021 by guest. Protected by

http://ard.bmj.com

/A

nn Rheum

Dis: first published as 10.1136/ard.36.4.293 on 1 A

ugust 1977. Dow

nloaded from