Thorax 1994;49:1146-1151

Activation of peripheral blood mononuclearcells in bronchoalveolar lavage fluid frompatients with sarcoidosis: visualisation of singlecell activation products

P Pantelidis, AM Southcott, AD Cambrey, G J Laurent, RM du Bois

AbstractBackground- Interstitial lung diseases arecharacterised by the recruitment ofmononuclear cells to disease sites wherematuration occurs and activation prod-ucts, including lysozyme (LZM), are re-leased. Analysis of in vitro cell culturesupernatants for activation productsmasks the functional heterogeneity of cellpopulations. It is therefore necessary toexamine the secretion of activation prod-ucts by single cells to assess whether theactivation ofnewly recruited mononuclearphagocytes at the sites of disease in thelung is uniform and controlled by the localmicroenvironment.Methods - The reverse haemolytic plaqueassay was used to evaluate, at a single celllevel, the ability ofbronchoalveolar lavage(BAL) fluid from seven patients with sar-coidosis to activate Ficoll-Hypaque-sep-arated peripheral blood mononuclear cellsby comparison with BAL fluid from sixnormal volunteers and nine patients withsystemic sclerosis. Monolayers of peri-pheral blood mononuclear cells and sheepred blood cells were cultured either aloneor in the presence of 20% (vlv) BAL fluidwith a polyclonal anti-LZM antibody.LZM/anti-LZM complexes bound to redblood cells surrounding the secreting cellswere disclosed following complement lysisof red blood cells and quantification ofplaque dimensions using microscopy andimage analysis.Results - Bronchoalveolar lavage fluidfrom all the patients with sarcoidosis in-creased LZM secretion by peripheralbloodmononuclear cells compared with un-stimulated mononuclear cells. By con-trast, BAL fluid from the other individualshad no effect on LZM secretion.Conclusions - Single cells activated byBAL fluid can be evaluated by the reversehaemolytic plaque assay. BAL fluid frompatients with sarcoidosis, but not frompatients with systemic sclerosis or normalindividuals, contains components capableof activating mononuclear phagocytes tosecrete lysozyme.

(Thorax 1994;49:1146-1151)

Sarcoidosis is characterised by the ac-cumulation at disease sites ofmononuclear cells

resulting in the formation of epithelioid cellgranulomas."2 Circulating monocytes whichare attracted to disease sites are subjected tothe influence of local proinflammatory signalsresulting in phenotypic and functional changesas they undergo maturation and activation.3One of the activation products of stimulatedmononuclear phagocytes is lysozyme (LZM),4and measurements ofLZM concentration havebeen used as an index of active disease.56 Theprimary function ofLZM has always been con-sidered bactericidal,47 even though previouswork has shown that it may play a part in themodulation of monocyte-lymphocyte inter-actions8 and depression of neutrophil chemo-taxis.'

If monocytes attracted to disease sites insarcoidosis are stimulated to produce increasedamounts ofLZM, we hypothesised that factorswithin the local microenvironment are re-sponsible for this change in function. To testthe hypothesis we have used a method of iden-tifying secretory products at the single cell levelusing the reverse haemolytic plaque assay toexamine the effects of bronchoalveolar lavage(BAL) fluid on LZM secretion. The ability ofthe reverse haemolytic plaque assay to visualisecytokine products by single cells allows themodulation in function of cells to be assessedin the absence of cell-cell interaction and min-imises paracrine effects ofproinflammatory me-diators from other cells in culture, thus enablingthe direct effect of added BAL fluid on cellactivation to be analysed. We therefore ex-amined BAL fluid from patients with sar-coidosis, patients with systemic sclerosis, andnormal individuals and compared the capacityof each to activate "naive" monocytes to syn-thesise and secrete LZM.

MethodsPATIENTSThree groups were studied: (1) patients withsarcoidosis (n=7; M:F 4:3; age 30-2 (1-6)years), (2) patients with fibrosing alveolitis as-sociated with systemic sclerosis (n=9; M:F4:5; age 44-6 (1-7) years), and (3) normalvolunteers (n=6; M:F 3:3; age 26-3 (046)years). Ofthe patient groups, two patients withsarcoidosis were current smokers and threewith systemic sclerosis were receiving treatment(prednisolone).A diagnosis of sarcoidosis was confirmed by

the presence of non-caseating granulomas in

Interstitial LungDisease Unit, RoyalBrompton HospitalP PantelidisA M SouthcottR M du Bois

Biochemistry Unit,Department ofThoracic MedicineA D CambreyG J Laurent

National Heart andLung Institute,Emmanuel KayeBuilding, LondonSW3 6LR, UK

Reprint requests to:Dr RM du Bois.Received 11 January 1994Returned to authors5 April 1994Revised version received18 July 1994Accepted for publication21 July 1994

1146

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from

Activation ofperipheral blood mononuclear cells in sarcoidosis

transbronchial lung biopsies or a positiveKveim reaction in a clinical setting consistentwith sarcoidosis. Patients with systemic scler-osis met the American Rheumatism As-sociation preliminary criteria for thisdiagnosis.'0 A diagnosis of pulmonary fibrosisin systemic sclerosis was made by the findingof typical changes on high resolution computedtomographic scanning."

All patients underwent bronchoalveolar lav-age (BAL) as part of their clinical evaluation.The normal volunteers chosen were non-smokers with no evidence of pulmonary dis-ease. Approval for lavage of the normal vo-lunteers was given by the Royal BromptonHospital ethics committee.

BRONCHOALVEOLAR LAVAGE FLUIDBronchoalveolar lavage fluid was collected frompatients with sarcoidosis, systemic sclerosis,and normal individuals during routine bron-choscopy and lavage.'2 Briefly, 100 ml ofwarmed normal saline was instilled in 20 mlaliquots in three subsegmental bronchi (theright middle lobe, right lower lobe, and lingula)and aspirated immediately. The cells were sep-arated from the BAL fluid by centrifugation at200 g, 4°C, seven minutes. The BAL fluid wasthen subjected to a further centrifugation at600 g for 10 minutes and stored in 5 ml aliquotsat - 80°C until further analysis. Before use inthe reverse haemolytic plaque assay the BALfluid was passed through a filter with a porediameter 02g,m.

ISOLATION OF HUMAN PERIPHERAL BLOODMONONUCLEAR CELLSPeripheral blood mononuclear cells were isol-ated from heparinised venous blood from thesame normal individual using a Ficoll-Hypaquegradient (Pharmacia, Milton Keynes, UK) ac-cording to the method of Boyum.'3 After threewashes in Hank's buffered salt solution (Gibco,Uxbridge, UK) the peripheral blood mono-nuclear cells were resuspended to a final con-centration of 2 x 106 cells/ml in RPMI 1640media (Gibco, Uxbridge, UK) containing 0-1%bovine serum albumin (BSA) and penicillin(100 U/ml)/streptomycin (100 ig/ml) (com-plete media) for use in the reverse haemolyticplaque assay.

REVERSE HAEMOLYTIC PLAQUE ASSAYThe assay was performed using a modificationof the method of Lewis et al.'4 In brief, Cun-ningham chambers'5 were prepared by coatingmicroscope slides with poly-L-lysine and con-structing a 30 gl chamber using double-edgedtape and a coverslip. 6 Human peripheral bloodmononuclear cells and protein-A coated sheeperythrocytes were mixed in equal volumes andaliquoted into Cunningham chambers. Thecells were allowed to settle on the poly-L-lysine-coated slides for 40 minutes in an atmosphereof 5% CO2 at 37°C. Excess unattached cellswere removed by rinsing each chamber withcomplete media leaving a confluent monolayer

of erythrocytes and peripheral blood mono-nuclear cells on the poly-L-lysine-coated sur-face. The chambers were then filled withcomplete media containing 1/50 dilution ofrabbit anti-human LZM (Dakopattis Ltd, HighWycombe, UK) from which sodium azide hadbeen previously removed using an AmiconConcentrator'7 (Amicon Ltd, Gloucestershire,UK), and 20% ((v/v):final dilution in culturechamber) of either normal saline (control) orBAL fluid from patients with sarcoidosis, sys-temic sclerosis, and normal subjects (the op-timum concentration ofBAL fluid determinedby preliminary experiments). The chamberswere incubated for three hours at 37°C in 5%CO2 and then washed with complete mediaand filled with complete media containing 1/30dilution of guinea pig complement (Gibco,Uxbridge, UK). After 25 minutes incubationat 37°C in 5% C02, during which periodplaques of haemolysis were formed around theLZM-secreting cells, the chambers werewashed with complete media without BSA andeither fixed with 4% glutaraldehyde in phos-phate buffer (pH 7 4) and stored at 4°C underTris-buffered saline (TBS) prior to estimationof plaque dimensions, or infused with 0 5%(v/v) Trypan Blue solution and incubated forfive minutes at 37°C to test the viability of peri-pheral blood mononuclear cells. Cell viabilitywas always >95% at the end ofthree hours incu-bation in the reverse haemolytic plaque assay.The specificity ofplaque formation for LZM

secretion was confirmed since no haemolyticplaques were observed when (a) mononuclearcells were omitted from the assay; (b) when theanti-LZM antibody was omitted and replacedwith 1:50 normal rabbit serum; (c) when com-plement was omitted; (d) when uncoated eryth-rocytes were used; (e) when the antibody waspreabsorbed at 4°C overnight with >3 ig/mlhuman LZM (Calbiochem, Nottingham, UK).To minimise the day to day variations that

exist during culturing of the peripheral bloodmononuclear cells in the reverse haemolyticplaque assay, even though culturing conditionswere stringently observed, each assay containedat least one sample of BAL fluid from eachstudy group examined and a control in whichthe volume of BAL fluid was substituted withnormal saline.

IMMUNOSTAINING OF MONOCYTESAt the end of the assay the plaque-forming cellswere identified as monocytes by immuno-labelling with KP1 mouse monoclonal anti-body (gift from Dr DY Mason, Oxford, UK)which recognises proteins of 110, 70, and40 kDa expressed in the lysosomal fraction ofmononuclear phagocytes.18 Visualisation ofpositively staining cells was achieved using thealkaline phosphatase/anti-alkaline phosphatase(APAAP) immunocytochemical method'9 em-ploying a 1/50 dilution of rabbit anti-mouseIgG, 1/50 dilution ofAPAAP complex, alkalinephosphatase substrate, and fast red withhaematoxylin counterstaining. After colour de-velopment the chambers were dried, the cover-slip removed, and a new coverslip mounted.

1147

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from

Pantelidis, Southcott, Cambrey, Laurent, du Bois

LYSOZYME CONTENT IN THE BAL FLUIDConcentrations of free LZM in the BAL fluidwere measured by radial immunodiffusion(The Binding Site, Birmingham, UK).

J,~~~,

Figure I Photomicrograph of KPI-stained, LZM secreting, monocsheep red blood cell lysis surrounding two LZM secreting cells are sh

rigure 2 lnfluence oJ increasing concentrations of bronchoalveolar lavage (BAL) fluidfrom two patients with sarcoidosis on the secretion ofLZM by peripheral blood monocytesfrom a normal individual. Cultures containing 10% and 20% (v/v) BAL fluid secretedsignificandy more LZM than control cultures (medium) (*p<0 02, **p<0.01). Barsrepresent mean and standard error in each case. ***Concentrations of BAL fluid thatresulted in the suppression ofplaque formation.

PLAQUE QUANTIFICATIONPlaque area was measured using Leitz mi-croscopy, Apple Macintosh computer, andimage analysis package (Improvision UK).Fifty plaques per chamber were measured infive replicate chambers (250 plaques in total).The following criteria were applied when

measuring plaque area: (a) lysis ofsheep erythro-cytes around the plaque-forming cell had tobe visible and well defined; (b) cells were

chosen for plaque measurement by examiningrandomly selected non-overlapping microscopefields; (c) only plaques formed by single cellswere counted; (d) plaque counting was per-

formed "blind" without knowledge of the iden-tity ofthe chambers. The coefficient ofvariationfor intraobserver variability was 4-3% for 600plaques.

ALBUMIN CONCENTRATION IN BAL FLUID

Albumin concentrations in the BAL fluidsamples were measured by bromocresol greencolorimetric assay (Sigma Chemical Co Ltd,

;;^Xs Poole, Dorset, UK) and performed in du-plicate. Briefly, a standard solution of human

serum albumin (Sigma Chemical Co Ltd),500 g/l in normal saline, was diluted to a range

covering albumin concentrations of 0-05-0 5g/l. Duplicate 1 ml aliquots of each standard

gr 9 BiandBAL fluid sample were pipetted into 3 - 5 ml

polystyrene test tubes (Sarstedt Ltd, BeaumontLeys, Leicestershire, UK) and, after the ad-dition of an equal volume ofbromocresol green

cytes (A). Plaques of solution, the reaction was allowed to proceedown. at room temperature for 10 minutes. The ab-

sorbance of each sample was measured at630 nm on a Gilford 2600 spectrophotometer(Corring Ltd, Essex, UK) and the con-

centrations ofalbumin in the BAL fluid sampleswere estimated from the linear relationshipdemonstrated between albumin concentrationand absorbance as a standard curve (r = 0 999,p<0.0001).

STATISTICAL ANALYSISResults of single replicate experiments were

analysed with the Student's t test. Group dataare distributed non-parametrically and com-

parisons between groups were made using theMann-Whitney U test. A p value of <0 05 was

considered to be significant for all analyses.The data in the text and in the figures are

shown as mean (SE).

ResultsEVALUATION OF REVERSE HAEMOLYTIC PLAQUEASSAYTo evaluate the reverse haemolytic plaque assay

as a means of identifying LZM secretion bysingle cells, preliminary experiments were con-

ducted with peripheral blood mononuclear cellscultured in the absence of stimulation to de-termine spontaneous secretion ofLZM and thespecificity of plaque formation for LZM.

Easily identifiable plaques were seen afterincubation of peripheral blood mononuclearcells for three hours, at which time point >95%of all cells were viable (fig 1). Plaque formationwas completely suppressed if antibody to LZMwas incubated with human placental LZM atconcentrations >3 ig/ml at 4°C for 12 hoursprior to being added to the culture medium(data not shown).KP 1 immunostaining for mononuclear pha-

gocytes confirmed that all plaque-forming cellswere monocytes.

INFLUENCE OF BAL FLUID FROM PATIENTS WITH

SARCOIDOSIS ON THE SECRETION OF LZMIn order to identify the concentration of BAL

4& vM*

._ ";

1148

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from

Activation ofpenipheral blood mononuclear cells in sarcoidosis

80

60

cna)

G)

a)cJ

40

L

Plaque size

Figure 3 Heterogeneity ofLZM secretion withinexamples of the effect ofBAL fluid from single inasecretion ofLZM by peripheral blood monocytes frincrease is seen in the percentage of cells forming kcwith BAL fluid from patients with sarcoidosis. Co

HETEROGENEITY OF LZM SECRETION WITHINTHE MONOCYTE POPULATION

*- *Control Because it is recognised that the mononuclearA - -- Normal phagocyte population is heterogeneous, the*....* Systemic sclerosis population of plaque-forming cells was ex-*.*- Sarcoidosis amined for heterogeneity of LZM secretion.

There was considerable variation amongst in-dividual peripheral blood mononuclear cells inthe amount of LZM secreted after three hoursof incubation. This variation of plaque sizeranged from 250 gm' to 3000 im2; fig 3illustrates the results from one such assay.

Incubation of normal peripheral bloodmononuclear cells with lavage fluid obtainedfrom patients with sarcoidosis produced anincrease in the number of cells that formedlarger size plaques, shown as a shift of thedistribution curve to the right indicating a much

._.+.--+wider distribution in plaque sizes. Plaque size2000 2500 3000 3500 distributions after incubation with lavage fluid

from normal individuals or patients with sys-3 (m2) temic sclerosis were no different from controlsthe monocyte population. Shown are using normal saline (fig 3). While 33% of(ividuals in each disease group on the plaque-forming cells stimulated by fluid fromom the same normal individual. An patients with sarcoidosis produced plaques ofirger size plaques following stimulation area greater than 1250 gm2, on average onlyontrol= 20% (v/v) normal saline. 4% of plaque-forming cells stimulated by BAL

fluid from other groups produced plaques ofthis size.

2000

N 1500 -

E

r 1000-

-L-

co

C5)

< 500-

Control

i//

Normal Systemicsclerosis

Sa rc

Figure 4 Mean (SE) average plaque size from a single assay. Shown are repli( x S) of average plaque size following stimulation with BAL fluid from a singleindividual in each group (sarcoidosis v control; p<0 01). Bars represent mean a;standard error in each case. Control= 20% (v/v) normal saline.

fluid that produced a maximum stimuJsecretion of LZM by normal periphermononuclear cells, BAL fluid was exara concentration range of 10% (v/v) to 54from two patients with sarcoidosis to detheir stimulatory capacity.Both 10% (v/v) and 20% (v/v) BP

caused a significant increase in secrLZM (fig 2; p<002, p<0-01, respewith a maximum response at 20% (v/v)formation was suppressed at conceranges of 30-50%. For all subsequperiments 20% (v/v) was used.

AUGMENTATION OF LZM SECRETION BY BALFLUID FROM PATIENTS WITH SARCOIDOSIS

* To assess the influence of BAL fluid fromnormal subjects, patients with systemic scler-osis and patients with sarcoidosis on the se-cretion of LZM by mononuclear phagocytesfrom a single normal individual, replicate assayswere performed using BAL fluid from at leasta single individual from each group in eachgroup of assays. The results from one suchassay are shown in fig 4. It can be seen thatBAL fluid from patients with sarcoidosis in-creased the secretion ofLZM by approximately50% (p<0 01).When evaluating the group data as a whole

the average plaque size measurements were

expressed as percentage increase of the averageoidosis plaque size of the control. The group data are

summarised in fig 5. There was a uniformsignificant augmentation (p<001) ofLZM se-

cates cretion by peripheral blood mononuclear cellsnd from the same normal individual cultured with

20% (v/v) BAL fluid from patients with sar-coidosis but not with BAL fluid from patientswith systemic sclerosis or normal individuals.

lation of Importantly, all seven BAL fluid samples fromal blood patients with sarcoidosis increased plaque sizesnined at by an amount greater than BAL fluid samples0% (v/v) from any of the other groups.-termine

kL fluid INFLUENCE OF LZM CONCENTRATIONS IN BALetion of FLUID ON PLAQUE FORMATIONctively), Because varying amounts of LZM are presentPlaque within BAL fluid, reflecting in vivo secretion,

ntration LZM concentrations were evaluated to de-ient ex- termine whether they influenced plaque form-

ation.

1149

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from

Pantelidis, Southcott, Cambrey, Laurent, du Bois

Figure S Group data.BAL fluid from patientswith sarcoidosissignificandy upregulated(p<O O1) monocyte LZMsecretion. Bronchoalveolarlavage fluid from patientswith systemic sclerosis andnormal individuals had nooverall influence on thespontaneous secretion ofLZM. The results areexpressed as percentageincrease of average plaquesize normalised to controlaverage plaque size. Linesrepresent mean andstandard error in each case.

100 r

4C-

0

cu

8L)

0

0)Cu

a1)

140

120

*0 a

T- O0100

80Normal Systemic

sclerosis

Bronchoalveolar lavage fluid conofLZM were similar in all three grcand average plaque size did not co:LZM concentrations in BAL fltlavage fluid samples containing I> 15 jtg/ml (final concentration in cubers 3 ,ug/ml after 20% (v/v) dilutioonly samples which abolished plaqui- that is, native LZM bound antitculture medium leaving no free antiing sites for LZM released by culttcytes. This is consistent with otexperiments which showed thatcentrations >3,ug (>15 jig/ml afterdilution in the final culture mix) tformation ofplaques. There was nobetween LZM levels in the BAL fluand the size ofthe average plaque ar

DILUTION OF RESIDENT LUNG EPITHILINING FLUIDTo determine whether the leakagfactors due to increased vascular por the volume of lavage return cotfor the observed between-group diffassayed BAL fluid for albumin, aknown to be synthesised within theparenchyma, and measured the peinstilled fluid which was returnedpared these data between grotalbumin concentrations (duplicatments) were not different between(table). Similarly, the range of fltwere comparable (data not shown)

Mean (SE) and range of lysozyme and albumin levels in the bronchoalveolfluid from patients with sarcoidosis, systemic sclerosis and normal subjects

Group Lysozyme (pg/mi) Albumin (mg/ml,

Sarcoidosis 7-7 (1-1) 0-105 (0-03)(n=7) (4- 1-11-8) (0-0197-0-243)Systemic sclerosis 10-32 (2-18) 0-0897 (0 019(n= 9) (2-8-19 4) (0-047-0-212)Normals 7-83 (0-88) 0-085 (0-018)(n = 6) (4-1-10-0) (0-036-0-161)

p < 0-01 Discussion* This study has shown that normal peripheral

blood mononuclear cells exposed to lavage fluidfrom patients with sarcoidosis, but not lavage

-9 fluid from patients with systemic sclerosis ornormal individuals, synthesised and secretedapproximately 40% more LZM than cells in-cubated in the presence of normal saline. Fur-

* thermore, the use of a reverse haemolyticplaque assay to visualise mediator production

* by single cells has shown that there is a con-siderable heterogeneity in LZM secretion bymononuclear cells, and that this heterogeneityis enhanced by BAL fluid from patients withsarcoidosis but not those with systemic sclerosisor normal individuals.

Heterogeneity among the mononuclear pha-gocyte population at disease sites has long beenrecognised202' and, in diseases like sarcoidosis

Sarcoidosis or cryptogenic fibrosing alveolitis, this hetero-geneity can be further visualised using panelsof monoclonal antibodies.2223 Furthermore, insarcoidosis diversity of cell surface phenotypeis also recognised within the mononuclearphagocyte population which make up thegranulomas.2324 Surface phenotype hetero-

icentrations geneity has not yet been related to functional)ups (table) status because of the constraints applied byrrelate with traditional cell culture methodologies. Studiesuid. Three of peripheral blood monocyte subpopulationsZM levels separated by density gradient25 do, however,

ilture cham- show a diverse secretory repertoire probablytn) were the reflecting the in vivo tight local control exertedeformation over inflammatory cells such as those of thebody in the mononuclear phagocyte series.26ibody bind- Activated cells within granulomas in sar-ired mono- coidosis are known to synthesise and secrete.ir previous LZM and serum LZM levels have been usedLZM con- in some studies as a marker of an active disease20% (v/v) process.5 6 27 In this regard, our study has shown

blocked the that BAL samples of the epithelial lining fluidFcorrelation from the lower respiratory tract of patients withaid samples sarcoidosis contain macromolecules capable of,ea (p>O0 1). enhancing LZM secretion by normal mono-

cytes. This is consistent with other studieswhich have identified stimulating factors in

ELIAL lavage fluid specific for sarcoidosis.2' Althoughthe nature of these factors is not known,

e of serum interferon gamma, a lymphokine which isermeability spontaneously released from cells obtained fromild account the lungs of patients with sarcoidosis,29 isFerences, we known to stimulate LZM secretion in culturesprotein not of macrophage cell lines.30 It remains possible,pulmonary therefore, but not proven by this study,rcentage of that interferon gamma is the stimulating factorand com- in the lavage fluid.

aps. Mean As well as a marker of biological activity ine measure- sarcoidosis, LZM may have an immuno-the groups modulatory role. Although traditionally con-uid returns sidered to be a macromolecule which has

bactericidal activity, LZM is now known toreduce neutrophil chemotaxis9 and also to

Carlavage exert an upregulatory effect on monocyte-depleted mixed lymphocyte cultures.8 Fur-thermore, LZM can reduce the in vitro re-sponse of lymphocytes to mitogens." Thisimplies a specific control of monocyte function

4) within the local microenvironment in sar-coidosis which might have a direct effect onrestoring tissue homeostasis.

1150

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from

Activation ofperipheral blood mononuclear cells in sarcoidosis

In systemic sclerosis, a condition in whichmononuclear phagocytes also traffic to diseasesites and also release potent mediators - in-cluding tumour necrosis factor alpha andinterleukin-83233 - which can modulate thedisease process, other control mechanisms are

likely to be involved in stimulating monocytesto produce a quite distinct repertoire of me-

diators which are, therefore, disease dependent.The reverse haemolytic plaque assay has a

number of advantages over traditional culturemethods. Firstly, although the assay cannotcompletely abrogate paracrine effects, the smallvolume of the incubation chamber makes thesemuch less likely, particularly as cells isolatedfrom other cells were chosen for measurement.Secondly, visualisation of products of singlecells provides the added advantage of com-

bining this with immunocytochemical stainingusing specific monoclonal antibodies which willallows changes in surface phenotype to be re-lated to plaque formation (that is, function) andthis is the subject of further studies. Thirdly,our preliminary studies using pure LZM haveconfirmed that the assay is specific for LZMsecretion. This method therefore allows a muchmore complete dissection of individual cellfunctions to be undertaken.

In conclusion, we have shown that epitheliallining fluid obtained from patients with sar-

coidosis can modulate the function of peri-pheral blood mononuclear cells which impliesthat newly arriving monocytes at disease sitesreceive specific stimulatory signals which definetheir functional role within the disease process.Furthermore, we have shown that there is het-erogeneity within the monocyte population interms of amount of LZM produced per cell,and that the stimulatory activity is found onlyin lavage fluid obtained from patients withsarcoidosis. The reverse haemolytic plaqueassay will allow the further exploration of sub-populations of mononuclear phagocytes to beundertaken with particular regard to their se-

cretory repertoire, and this may provide greaterinsights into the regulation of mononuclearphagocyte function in the lung at disease sites.

We would like to thank Mr Michael Kemp for his help withthe lysozyme measurements in the bronchoalveolar lavage fluids.

1 Mitchell DN, Scadding JG. Sarcoidosis. Am Rev Respir Dis1 974;110:774-802.

2 Mitchell DN, Scadding JG, Heard BE, Hinson KFW. Sar-coidosis: histopathological definition and clinical dia-gnosis. Clin Pathol 1977;30:395-408.

3 Hance AJ, Douches S, Wimchester RJ, Ferrans VJ, CrystalRG. Characterisation of mononuclear phagocyte sub-populations in the human lung by using monoclonalantibodies: changes in alveolar macrophage phenotypeassociated with pulmonary sarcoidosis. Jf Immunol 1985;134:284-92.

4 Gordon S, Todd J, Cohn ZA. In vitro synthesis and secretionof lysozyme by mononuclear phagocytes. Exp Med 1974;139:1228.

5 Turton CW, Grundy E, Firth G, Mitchell D, Rigden BG,Turner-Warwick M. Value ofmeasuring serum angiotensinI converting enzyme and serum lysozyme in the man-agement of sarcoidosis. Thorax 1979;34:57-62.

6 Prior C, Barbee RA, Evans PM, Townsend PJ, Primett ZS,Fyhrquist F, et al. Lavage versus serum measurementsof lysozyme, angiotensin converting enzyme and otherinflammatory markers in pulmonary sarcoidosis. EurRespirJf 1990;3:1146-54.

7 Chipman DM, Sharon N. Mechanism of lysozyme action.Science 1969;165:454-65.

8 Rinehart JJ, Cerilli JG, Jacob HS, Osserman EF. Lysozymestimulates lymphocyte proliferation in monocyte-depletedmixed lymphocyte cultures. 7 Lab Clin Med 1982;99:370-81.

9 Gordon LI, Douglas SD, Kay NE, Yamada 0, Osserman

EF, Jacob HS. Modulation of neutrophil function bylysozyme. Potential negative feedback system of in-flammation. J Clin Invest 1979;64:226-32.

10 Subcommittee for Scleroderma Criteria of the AmericanRheumatism Association Diagnostic and Therapeutic Cri-teria Committee. Preliminary criteria for the classificationof systemic sclerosis (scleroderma). Arthritis Rheum 1980;23:581-90.

11 Wells AU, Hansell DM, Rubens MB, Cullinan P, BlackCM, du Bois RM. The predictive value of appearanceson thin section computed tomography in fibrosing al-veolitis. Am Rev Respir Dis 1993;148:1076-82.

12 Saltini C, Hance AJ, Ferrans VJ, Bassett F, Bitterman PD,Crystal RG. Accurate quantification of cells recoveredby bronchoalveolar lavage. Am Rev Respir Dis 1984;130:650-6.

13 Boyum A. Isolation of lymphocytes, granulocytes and ma-crophages. Scand J Immunol 1976;5:9-15.

14 Lewis CE, McCarthy SP, Lorenzen J, McGee JO. Differ-ential effects of LPS, IFN-gamma and TNF alpha on thesecretion of lysozyme by individual human mononuclearphagocytes: relationship to cell maturity. Immunology1990;69:402-8.

15 Cunningham AJ, Szenberg A. Further improvements in theplaque technique for detecting single antibody-formingcells. Immunology 1968;14:599-600.

16 Smith PF, Luque EH, Neill JD. Detection and measurementof secretion from individual neuroendocrine cells using areverse hemolytic plaque assay. Methods Enzymol 1986;124:443-65.

17 Lewis CE, McCarthy SP, Lorenzen J, McGee JO. Het-erogeneity among human mononuclear phagocytes in theirsecretion of lysozyme, interleukin 1 and type-beta trans-forming growth factor: a quantitative analysis at the single-cell level. EurJ Immunol 1989;19:2037-43.

18 Pulford KA, Rigney EM, Micklem KJ, Jones M, Stross WP,Gatter KC, et al. KP1: a new monoclonal antibody thatdetects a monocyte/macrophage associated antigen inroutinely processed tissue sections. Jf Clin Pathol 1989;42:414-21.

19 Cordell JL, Falini B, Erber WN, Ghosh AK, AbdulazizZ, MacDonald S, et al. Immunoenzymatic labeling ofmonoclonal antibodies using immune complexes of al-kaline phosphatase and monoclonal anti-alkaline phos-phatase (APAAP complexes). Jf Histochem Cytochem 1984;32:219-29.

20 Sandron D, Raynolds HY, Laval AM, Venet A, Israel-Biet D,Chretien J. Human alveolar macrophage subpopulationsisolated on discontinuous albumin gradients. Cytologicaldata in normals and sarcoid patients. Eur J Respir Dis1986;68: 177-85.

21 Sandron D, Reynolds HY, Venet A, Laval AM, Israel-Biet C,Chretien J. Human alveolar macrophage subpopulationsisolated on discontinuous albumin gradients: functionaldata in normals and sarcoid patients. Eur Jf Respir Dis1986;69:226-34.

22 Campbell DA, Poulter LW, du Bois RM. Phenotypic analysisof alveolar macrophages in normal subjects and in patientswith interstitial lung disease. Thorax 1986;41:429-34.

23 Campbell DA, Poulter LW, du Bois RM. Immunocompetentcells in bronchoalveolar lavage reflect the cell populationsin transbronchial biopsies in pulmonary sarcoidosis intransbronchial biopsies in pulmonary sarcoidosis. Am RevRespir Dis 1985;132:1300-6.

24 Semenzato G, Chilosi M, Ossi E, Trenyin L, Pizzolo G,Cipriani A, et al. Bronchoalveolar lavage and lunghistology. Comparative analysis of inflammatory andimmunocompetent cells in patients with sarcoidosis andhypersensitivity pneumonitis. Am Rev Respir Dis 1985;132:400-4.

25 Herrmann F, Gebauer G, Lindemann A, Brach M, Mer-telsmann R. Interleukin-2 and interferon-gamma recruitdifferent subsets of human peripheral blood monocytes tosecrete interleukin-I beta and tumour necrosis factor-alpha. Clin Exp Immunol 1989;77:97-100.

26 Semenzato G. Immunology of interstitial lung diseases:cellular events taking place in the lung of sarcoidosis,hypersensitivity pneumonitis and HIV infection. EurRespirJf 1991;4:94-102.

27 Selros OB. Biochemical markers in sarcoidosis. Crit RevClin Lab Sci 1986;24:185-216.

28 Rohrbach MS, Vuk-Pavlovic Z, Martin WJ, Delaval P,Genetet N. Induction of angiotensin converting enzymein cultured human monocytes by a factor present inthe bronchoalveolar lavage fluid of sarcoidosis patients.Sarcoidosis 1988;5:17-23.

29 Robinson BWS, McLemore TL, Crystal RG. Gamma in-terferon is spontaneously released by alveolar macrophagesand lung T lymphocytes in patients with pulmonary sar-coidosis. Jf Clin Invest 1985;75:1488-95.

30 Pitzurra L, Marconi P, Bistoni F, Blasi E. Selective inhibitionof cytokine-induced lysozyme activity by tetanus toxin inthe GG2EE macrophage cell line. Infect Immun 1989;57:2452-6.

31 Varaldo PE, Valisena S, Mingari MC, Satta G. Lysozyme-induced inhibition of the lymphocyte response to mito-genic lectins. Proc Soc Exp Biol Med 1989;190:54-62.

32 Pantelidis P, Southcott AM, du Bois RM. A comparison ofsingle cell TNF-a secretion in blood and lung mono-nuclear cells in interstitial lung disease. Am Rev Respir Dis1 993;147:A756.

33 Southcott AM, Jones KP, Pantelidis P, Black CM, DaviesBH, du Bois RM. Interleukin-8 is associated with thepresence of pulmonary fibrosis in systemic sclerosis. AmRev Respir Dis 1993;147:A755.

1 151

on 18 Novem

ber 2018 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.49.11.1146 on 1 Novem

ber 1994. Dow

nloaded from