source httpsdoiorg107892boris115444 | downloaded 27122021

631

The interaction of Trypanosoma congolense with endothelialcells

A HEMPHILL 1 I FRAME2 and C A ROSS3

1 Institute of Parasitology University of Berne Laenggass-Strasse 122 3001 Berne Switzerland2 Department of Medical Parasitology London School of Hygiene and Tropical Medicine Keppel Street London WC1E7HT UK3 Centre for Tropical Veterinary Medicine University of Edinburgh Easter Bush Roslin Midlothian Edinburgh EH259RG UK

(Received 17 January 1994 revised 11 April 1994 accepted 15 April 1994)

SUMMARY

Factors which affect adhesion of cultured Trypanosoma congolense bloodstream forms to mammalian feeder cells have beenexamined Using an in vitro binding assay the initial events following interaction of trypanosomes with bovine aortaendothelial (BAE) cells were monitored by both light- and electron microscopy Metabolic inhibitors and other bio-chemicals were incubated with either cells or parasites to test whether any inhibited the process Our findings suggest thatadhesion of the parasites is an active process requiring metabolic energy from the trypanosomes but not from endothelialcells We also provide data suggesting that T congolense bloodstream forms possess a lectin-like domain localized atdistinct sites on their flagellar surface which interacts with specific carbohydrate receptors most likely sialic acid residueson the endothelial cell plasma membrane We also suggest that the cytoskeletal protein actin is probably involved in thisinteraction

Key words African trypanosomes endothelial cells cell adhesion molecules carbohydrates actin

INTRODUCTION

Trypanosoma congolense (Broden 1904) is one ofthe most important agents of African animal trypano-somiasis and infections cause extensive losses ineconomic productivity throughout the continentThis parasite is pathogenic for domestic livestockespecially cattle and infected animals show severesigns of organ damage anaemia weight loss abor-tion and infertility (Molyneux amp Ashford 1983)Once infection has taken place after transmission bythe tsetse fly vector rapidly dividing trypanosomescirculate within the bloodstream of mammalianhosts protected from the immune response by theirvariable surface glycoprotein (VSG) coat (Vicker-man 1985 Boothroyd 1985) In contrast to othertrypanosomatids T congolense does not invadetissues like T brucei or T evansi or cells like theSouth American T cruzi and can be regarded as atrue extracellular blood parasite

In addition the flagellum of T congolense unlikeother trypanosomatids does not extensively pro-trude from the anterior end of the cell body(Vickerman 1969 Vickerman amp Tetley 1990)However this organelle appears to play a key role inthe cell adhesion events which occur at different life-cycle stages within the insect vector and in animalhosts In infected animals T congolense bloodstreamforms are found adhering to erythrocytes and toendothelial cells of the microvasculature through

their flagellum (Banks 1978 1979 1980) The directeffects of this process are not entirely understoodbut the tissue surrounding the adhesion sites isdamaged by release of a vascular permeabilizationfactor and the subsequent action of the hostsimmune system (Banks 1980)

The binding of T congolense to endothelial cellshas not been investigated at the molecular level Theaim of these studies was therefore to dissect theinteraction between cultured trypanosomes andbovine aorta endothelial cells in vitro a processwhich resembles the adhesion event between para-sites and host cells in vivo An in vitro adhesion assaywas characterized during these experiments whereobservations were made by light- and electronmicroscopy Biochemical inhibition assays wereperformed to investigate the reactive groups whichtake part in the interaction

MATERIALS AND METHODS

If not otherwise stated all reagents and tissue culturemedia were purchased from Sigma (St Louis MoUSA)

Endothelial cells

Primary cultures of bovine aorta endothelial (BAE)cells were prepared as described by Gray et al

Parasitology (1994) 109 631-641 Copyright copy 1994 Cambridge University Press

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

A Hemphill I Frame and C A Ross 632

(1985) and routinely maintained in Minimal Es-sential Medium (MEM) supplemented with 20foetal calf serum (FCS) and 2 mM glutamineCultures of passages between 11 and 20 were usedBAE cells were either used unfixed or were fixedwhen 95-100 confluent Pre-fixed BAE cells wereprepared by rinsing cover-slips 3 times with PBSand incubating them for 10 min in freshly prepared2 glutaraldehyde in PBS 1 glutaraldehyde inPBS or several concentrations of paraformaldehydeCover-slips were rinsed extensively in PBS thenstored in 0-16 M ethanolamine pH 8-3 for at least24 h at 4 degC in order to saturate free aldehydegroups After fixation in 1 paraformaldehyde for20 min at 4 degC ethanolamine treatment was notnecessary (Norgard et al 1993 data not shown)

Trypanosome culture

Cultures were prepared from T congolenseTREU1627 a cloned derivative of a Gambian isolate(primary isolation code Kantong Kunda77LUMP1794) Metacyclic derived T congolensebloodstream forms (Gray et al 1985) were grown at34 degC in volumes of 6-8 ml in T-25 flasks eitheraxenically or on BAE-feeder cells in a mediummodified from that of Majiwa et al (1993) Thisconsisted of a combination of RPMI 1640 Mediumand Iscoves modified Dulbeccos Medium (RPMIIMDM 11) supplemented with 20 heat-inacti-vated (56 degC 30 min) donor goat serum 4 mMglutamine 1 mM sodium pyruvate 0-2 mM adeno-sine 0-1 mM bathocuproine-sulphonate and 0-1 mMmonothioglycerol Trypanosomes were releasedfrom their feeder cells mechanically by removingthe medium and vigorously striking the side of theflask 2-3 times A few ml of fresh medium wereadded and trypanosomes were washed from themonolayer which remained adhered to the flask

Adhesion assay

Immediately prior to the assay BAE cells on cover-slips were rinsed 3 times with PBS and once withRPMIIMDM containing 01 bovine serum albu-min (mediumBSA) Harvested trypanosomes werecentrifuged at 800 pound for 10 min at room temperatureand resuspended at a density of 5 x 105 ml in PBSP B S 0 1 BSA PBS100 mM glucose RPMIIMDM mediumBSA or trypanosome culturemedium depending on the experiments to beperformed Trypanosomes were incubated with thecell monolayers for 1 h then the cover-slips wererinsed carefully 3 times with PBS fixed for severalhours in 1 glutaraldehyde in PBS at 4 degC andprocessed for light- and electron microscopy

The number of adherent parasites in each ex-periment was resolved light microscopically by

randomly choosing 20 different fields at 40 timesmagnification on a Polyvar light microscope count-ing the monolayer-bound trypanosomes and deter-mining the average number of adherent parasites perfield For every experiment a control incubation wasperformed The numbers of trypanosomes adheringin each experiment are expressed as a percentage inrelation to the control experiment conducted on thesame day with the same culture All experimentswere performed in duplicate and the data presentedare from 1 representative experiment out of 4-6carried out

In order to compare the number of adheringtrypanosomes to fixed and unfixed BAE cell mono-layers during a time-course trypanosomes wereresuspended in culture medium at 106 parasitesmland incubated with the monolayers After varioustime-points one cover-slip each was rinsed in PBSand fixed as described above

Light microscopy

Glass cover-slips containing glutaraldehyde-fixedT congolense bloodstream forms adhering to BAEcells were washed in distilled water and were stainedwith haemotoxylin and eosin Cover-slips werefinally dehydrated in ethanol and xylene andmounted onto glass slides

Scanning electron microscopy (SEM)

Glass cover-slips with BAE-cell monolayers andadhering T congolense were fixed in PBS plus 2 glutaraldehyde for 4 h at 4 degC were rinsed severaltimes in PBS and treated with 1 osmium tetroxidein PBS for 1 h at 4 degC Specimens were thendehydrated through a series of acetone (70mdash90mdash100) and were critical-point dried Specimenswere viewed in a Jeol JSM-255 III scanning electronmicroscope operating at 12-5mdash20 kV

Transmission electron microscopy (TEM)

Cover-slips containing bloodstream forms adheringto BAE cell monolayers were processed for thinsection EM as follows Cells were fixed in 2glutaraldehyde in 0-1 M cacodylate buffer (FlukaChemicals UK) pH 7-3 for 2 h at room tem-perature They were post-fixed with 1 osmiumtetroxide in veronal acetate buffer pH 7-4 for 1 h at4 degC followed by buffer rinses The cover-slips werethen carried through a graded series of ethanol(70-95-100) and embedded in Epon 812 resin(Fluka Chemicals UK) After polymerizing theresin at 65 degC for 24 h the glass cover-slips wereremoved by dipping the specimen into liquidnitrogen and the freshly exposed surfaces wereembedded in Epon 812 again Ultrathin sections

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

Host-parasite interactions in trypanosomiasis 633

were finally cut perpendicular to the surface ontowhich the cells had previously been adhering usinga Diatom diamond knife on a LKB-ultramicrotomeSections were loaded onto 200 mesh copper or nickelgrids and stained with uranyl acetate and lead citrate(Smith amp Croft 1991) Specimens were viewed on aJeol 100 CX II transmission electron microscopeoperating at 60-80 kV

Treatment of live endothelial cells prior to theadhesion assay

Monolayers were treated with cycloheximide(50 ^gml) in culture medium for 6 h at 37 degC priorto fixation Tunicamycin was added to the mediumat 2-5 igml for 4 h at 37 degC Glycolytic inhibition ofBAE cells was performed by incubating cells in PBScontaining 100 HIM deoxyglucose for 30 min at 37 degCThe effect of mitochondrial inhibitors was assayedby treating cells with 1-5 mM sodium azide 10mdash100 nM antimycin or 1-5 tM oligomycin (all in PBS)for 20 min prior to fixation BAE cell monolayerswere also incubated for 4 h in culture mediumcontaining either 10 figm taxol 0-5 igml noco-dazole or various concentrations of cytochalasin D(10~8-10~7 M) The effects of these drugs on themicrotubule network and the actin filament systemwere visualized by immunofluorescence using apolyclonal rabbit anti-tubulin antibody or a poly-clonal rabbit anti-skeletal muscle actin antibody (notshown) Cover-slips were then rinsed briefly in PBSand fixed for 20 min in 1 paraformaldehyde in100 mM cacodylate buffer pH 7-3 at 4 degC andrinsed several times in PBS before the adhesion assaywas carried out

(SBA) specific for iV-acetyl-glactosamine For eachlectin a control incubation was conducted in thepresence of 0-4 M of its corresponding inhibitorymonosaccharide (Benhamou 1989) After washingthe cover-slips with PBS the adhesion assay wasperformed (iii) Neuraminidase treatment Live andpre-fixed BAE cells were rinsed twice in PBS andonce in 20 mM MES-buffer 100 mM NaCl 4 mMCaCl2 pH 5-5 Neuraminidase Type V from Clostri-dium perfringes was added to the same buffer at aconcentration of 5 Uml and the cells were treatedwith this enzyme at 37 degC for 60 min before beingwashed several times with PBS As a minor impuritythe protease caseinase (0-0001 unitmg) was presentin the enzyme preparation (iv) Poly-L-lysine anddextran sulphate in order to investigate possiblecharge interactions pre-fixed and live endothelialcells were incubated with poly-L-lysine (2-20 igm)or dextran sulphate (3-30 gml) for 1 h at 4 degC

Pre-treatments of T congolense

Freshly isolated trypanosomes were given the fol-lowing treatments before performing the adhesionassay

Pre-fixation Freshly isolated trypanosomes werefixed by the addition of 1 glutaraldehyde or 2 paraformaldehyde to the medium for 5mdash10 min at4 degC They were washed in PBS and kept in 0-16 Methanolamine pH 8-3 at 4 degC overnight Afterwashing the cells 3 times in PBS they wereresuspended in mediumBSA and assayed for en-dothelial cell adhesion at 106 trypanosomesml inmedium BSA

Treatment of pre-fixed BAE cells

Monolayers were fixed in 1 paraformaldehyde in100 mM cacodylate buffer pH 7-3 at 4 degC and weregiven the following treatments before the adhesionassay was carried out (i) Periodate treatmentmonolayers were subjected to different concentra-tions (2-20 mM) of sodium periodate in 50 mMsodium acetate 100 mM NaCl pH 5 at 4 degC for30 min Alternatively incubations were performedin 2-20 mM sodium periodate in PBS (pH adjustedto 7) at 4 degC Cells were washed extensively withPBS before applying the trypanosomes (ii) Lectinbinding BAE monolayers were incubated for 45 minat 37 degC in PBS with SO^gml of the followinglectins wheat germ agglutinin (WGA) specific forJV-acetyl-glucosamine and iV-acetyl-neuraminic acidresidues succinylated WGA (Vector LaboratoriesEngland) specific for iV-acetyl-glucosamine only(Monsigny et al 1980) Tetragonolobus purpureaagglutinin (TPA) specific for fucosyl residuesConcanavalin A (Con A) specific for D-mannosyland D-glucosyl residues and soybean agglutinin

Drug treatments Trypanosomes (106ml) were incu-bated in PBS0-5 BSA in the presence of 10-100 mM deoxyglucose for 20 min at room tem-perature Sodium azide (1-20 mM) antimycin (1mdash100 nM) oligomycin (1mdash5 IM) were used as mito-chondrial inhibitors Parasite numbers were thenadjusted to five times 105ml and were immediatelyused for the adhesion assay Cycloheximide treat-ment was carried out at 30 jgml for 4 h at 34 degC inculture medium Tunicamycin was added to 106mlof trypanosomes at 0-5 tgml and cells were incu-bated for 60 min at 34 degC prior to the adhesion assayFor the cytochalasin D treatment 106 parasitesmlwere incubated in 10~7 10~6 and 105 M cytochalasinD for 20 min at 34 degC in culture medium beforeapplying them to the monolayers at 105parasitesmlTaxol was used at 1mdash10jgml and nocodazole at0-1-1 4gml in culture medium for 6 h at 34 degC

Surface protease treatment Freshly isolated trypano-somes (106ml) were spun down and resuspended inmedium without serum Trypsin protease type VIand pronase (Serva GB) were both added to

43 PAR 109

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

A Hemphill I Frame and C A Ross 634

Fig 1 SEM of Trypanosoma congolense adherent to bovine aorta endothelial (BAE) cells Monolayers were grown onglass cover-slips incubated with the parasites and processed for SEM as described in the Materials and Methodssection (A) Low magnification view demonstrating the adherence of the parasites onto the endothelial cell surface(B) Single adherent trypanosome showing that the flagellum is the only part of the parasite directly in contact withthe BAE cell surface

2-10^gml (Reinwald 1985) After an incubation at28 degC for 20 min the cells were put on ice In thecase of trypsin the reaction was stopped by theaddition of 80 igm of soybean trypsin inhibitortype I-S The cells were washed twice with ice-coldculture medium were counted and immediatelyused for the binding assay in culture medium

RESULTS

Morphology and ultrastructural observations

Scanning electron microscopy of adherent T congo-lense bloodstream forms (Fig 1A) shows that theparasites adhered to the cell surface membrane ofBAE cells and not to the cover-slip underneathTrypanosomes were bound to the endothelial cellsurface exclusively by the anterior two thirds of theflagellum (Fig 1 B) These findings were confirmed

in sections cut perpendicular to the plane of thecover-slip and viewed by TEM (Fig 2) Themicrotubule skeleton (MT) a single layer of micro-tubules forming the cage round the cell body wasclearly visible as well as the paraflagellar rod (PFR)oriented towards the trypanosome cell body and themicrotubular axoneme (Ax) facing the endothelialcell membrane The flagellum adhered at approxi-mately 45deg with respect to the medial axis laidthrough the axoneme and the PFR

Longitudinal sections through the adhering fla-gellum (Fig 3) suggest that the initial attachmentwas not mediated through an extensively wide areaon the flagellum although the two membranesseemed to be crossbridged closely to each other overlong stretches The trypanosomal surface com-ponents directly interacting with the endothelial cellmembrane appeared to be situated along a thin lineon the anterior flagellum

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

Hostmdashparasite interactions in trypanosomiasis 635

Fig 2 TEM of adherent Trypanosoma congolense (A)-(C) Micrographs of trypanosomes adhering onto live BAEcells Longitudinal section (A) and cross-section (B) at lower magnification f Flagellum (C) Higher magnificationview of (B) MT microtubules PFR paraflagellar rod Ax axoneme (D) Trypanosomes adhering onto BAE cellspre-fixed in 2 paraformaldehyde (E) BAE cells pre-fixed in 2 glutaraldehyde

The effect of pre-fixation and metabolic inhibition ofBAE-cells on their association with T congolense

Initial attachment of T congolense bloodstreamforms to feeder cells was not altered by pre-fixationof BAE cells by formaldehyde or glutaraldehyde

(data not shown) Electron microscopy showed noultrastructural alterations compared to control cul-tures (Fig 2D and E) To obtain information on thetime dependence of the adhesion process parasiteswere incubated at 34 degC with fixed and unfixedfeeder cells and the number of adherent trypano-

43-2

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

A Hemphill I Frame and C A Ross 636

Fig 3 TEM of adherent trypanosomes (A) and (B) Two longitudinal serial sections through a flagellum adherent tothe endothelial cell surface Note that the site of intimate contact between the flagellar and endothelial membraneappears to be constituted by a thin line along the anterior flagellar surface (see also Fig 2) (C) Lower magnificationview of (A)

somes counted after different time-points (Fig 4)There was no difference in the adhesion kinetics ofT congolense to fixed and unfixed cells within the

first 60 min After approximately 120 min thenumber of trypanosomes adherent to fixed mono-layers reached saturation point while the numbers

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

Host-parasite interactions in trypanosomiasis 637

Fixed BAEsLive BAEs

100 200Time (min)

300

Fig 4 Comparison of adhesion of Trypanosomacongolense bloodstream forms on live and pre-fixed (2 glutaraldehyde) BAE cell monolayers during a time-course of 4 h

carried out to determine if the adhesion processinvolved only one or both sugar residues Inhibitionwas not observed when cells were incubated withWGA in the presence of 0-4 M iV-acetylglucosamineor with succinylated WGA which has no bindingspecificity for iV-acetylneuraminic acid The possibleinvolvement of sialic acid residues in mediating thecontact between trypanosomes and BAE cells wasconfirmed when neuraminidase-treated BAE mono-layers showed a 95 reduction in trypanosomeadhesion In contrast treatment with tunicamycinan inhibitor of AT-glycosylation showed no effect

Incubation with a polycation poly-L-lysine orwith polyanionic dextran sulphate before addition oftrypanosomes had no effect showing that adherencewas not based solely on charge interactions Whenpoly-L-lysine or dextran sulphate were presentduring the adhesion assay over 60 of trypano-somes were lysed (data not shown)

counted adhering to live BAE cells showed a furtherincrease The number of parasites adhered to unfixedcells continued to rise over 48 h of incubation due totheir multiplication whereas trypanosomes on fixedcells scarcely divided during this time (data notshown)

Pre-treatment of BAE cells with glycolytic andmitochondrial inhibitors had no effect on adhesion ofthe parasites nor had inhibition of protein synthesiswith cycloheximide (data not shown) Cytoskeletalinhibitors which alter the microtubule system (taxoland nocodazole) or actin microfilaments (cyto-chalasin D) had no effect Since initial adhesion wasapparently independent of target cell metabolismsubsequent assays were carried out using pre-fixedBAE cells

The involvement of endothelial cell surfacecarbohydrates in T congolense adhesion

Both pre-treatment of BAE cells with periodate andthe pH at which this treatment was carried out hada dramatic effect on the adherence of trypanosomesto the monolayer (Fig 5) At 20 mM periodate andat pH 5 adherence of trypanosomes to the monolayerwas only 3-5 of that on control cells although witha milder treatment of 2 mM periodate adherence ontreated cells was greater than 50 of the controlThe reduction in adherence after periodate treatmentwas less marked at neutral pH

BAE cells were also incubated with differentlectins before trypanosomes were added Con ASBA and TPA had no effect on subsequent adhesionof trypanosomes but prior incubation of cells withwheat germ agglutinin (WGA) inhibited parasiteattachment by 60 (Fig 5) Since WGA recognizesboth AT-acetylglucosamine and AT-acetylneuraminicacid (sialic acid) residues further experiments were

The effects of pre-treatment of trypanosomes

Pre-fixed T congolense bloodstream forms hadcompletely lost their ability to adhere to theendothelial cell surface (Fig 6) This suggested thatadhesion depended on parasite metabolism andorthat epitopes involved in the adhesion process weredenatured or masked by fixation

In a further series of experiments trypanosomeswere resuspended in various buffers and media andwere then allowed to adhere to BAE cells Nosubstantial difference in the number of adheredtrypanosomes was found when parasites were re-suspended in culture medium (containing 20donor goat serum) mediumBSA PBS100 mMglucose or PBS05 BSA Incubations in PBSalone resulted in a 43 reduction of adherentparasites Carrying out the binding assay at 4 degC orinhibiting glycolysis by incubating trypanosomeswith 2-deoxyglucose each reduced adhesion to 50of a control culture However mitochondrial inhibi-tors such as sodium azide antimycin and oligomycinhad no effect on the number of adherent parasites

Alteration of the parasite cell surface by pre-incubating trypanosomes with cycloheximide signifi-cantly decreased adherence An even larger in-hibition was observed when the surface of blood-stream forms was altered by protease treatmentusing trypsin and pronase (Fig 6) This suggestedthe destruction of receptors necessary for parasiteadhesion to BAE cells Protease digestions at higherconcentrations and temperatures led to considerableloss of viability of parasites (data not shown)

Parasites were incubated with several drugs whichaffect the function of the cytoskeleton to investigatewhether it was involved in mammalian cell in-teraction Neither taxol which enhances and sta-bilizes microtubules nor nocodazole a microtubuledepolymerizing drug had any effect on adhesion of

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

A Hemphill I Frame and C A Ross 638

control

2mM sodium periodate pH55mM sodium periodate pH5

20mM sodium periodate pH52mM sodium periodate pH75mM sodium periodate pH7

20mM sodium periodate pH7ConASBATPA

WGAWGA+04M N-acetyleglucosamine

succ WGAneuraminidase

tunicamycin20 pgml poly-L-lysine

30ugml dextran sulphate

^^3^^ 1

WZ 2 bull

YAYA 1

yW^

^^^yyampM^^ 1^mdashI

Vy^y^^^^ 1

^ g ^ ^ ^ ^ ^ ^ ^ 1

^^^

0 20 40 60 80 100 120

of adherent T congolenseFig 5 Effects of treatments affecting endothelial cell surface carbohydrates on the efficiency of trypanosome adhesionAfter the various treatments monolayers were fixed and incubated with the parasites as described

control

1 glutaraldehyde2 paraformaldehyde

culture mediummediumBSA

PBSIOOmM glucosePBSO5 BSA

PBS4degC

10OmM deoxyglucose20mM sodium azide

lOOnM antimycin5uM oligomycin

05 jigml tunicamycin30ugml cycloheximide

10ugml trypsin10(igml pronase

01 |aM cyt DluMcyt D0 M cytD

vAmiyr^^^^

yZf^^^^vyyysyr^^^^YjPyyampgtrr^^^

YMPy^^^^ltiyyyAP^^^^

vyM^^^

VAZWA

0 20 40 60 80 100 120

of adherent T congolenseFig 6 Effects of treatments of Trypanosoma congolense bloodstream forms prior to performing the adhesion assay

trypanosomes to BAE cells (data not shown)Treatment with 105 M cytochalasin D a drug whichspecifically interacts with actin resulted in a markeddecrease in the number of adherent trypanosomes

Attempts however to label T congolense actin witha panel of several anti-actin antibodies or withrhodamine-coupled phalloidin have not been suc-cessful

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

Host-parasite interactions in trypanosomiasis 639

DISCUSSION

In this paper the in vitro interaction betweenT congolense bloodstream forms and bovine aortaendothelial cells was investigated Observationsusing TEM and SEM show that T congolenseparasites adhere to feeder cells by a specific area inthe anterior two thirds of the flagellum Only a minorpart of the flagellar surface appears to be directlyinvolved since the two membranes are in intimatecontact with each other exclusively by a small area ofthe trypanosomal flagellum These findings and thefact that parasites can be released from the feedercells mechanically indicate that the interaction israther weak Banks (1978) studied the distribution ofT congolense in infected rats and rabbits and foundthat trypanosomes attached by their anterior end tothe walls of blood vessels In addition the micro-circulation contained 4mdash1400 times as many trypano-somes than other vessels This number could reflectthe relationship between adhesion and blood press-ure since in smaller vessels the pressure generatedby the blood flow is much lower than in largervessels T congolense has also been shown to preferthe vessels of some organs over others (Losos et al1973 Kaliner 1974) and receptor-ligand inter-action during initial attachment may contribute tothis preference The EM observations also point inthis direction since the interaction is mediated solelyby the anterior part of the flagellum in direct contactwith the endothelial cell surface never by theparasite cell body and since a distinct geometricpattern of attachment is maintained in all parasites

The adhesion of T congolense in vivo to cells liningthe walls of blood vessels also occurs when parasitesare cultured with endothelial cells in vitro Otherpathogenic trypanosome species such as T bruceiand T evansi do not attach to feeder cells in thisway although they can cluster in intracellular spacesIt was possible therefore to use an in vitro assaywhereby endothelial monolayers grown on cover-slips were provided as an adhesion matrix Differentfixation regimes had no effect on the subsequent rateof attachment to cells although trypanosomes ad-hered to fixed cells did not divide and increase innumber like those on unfixed cells This suggeststhat BAE cells release certain factors essential formultiplication of the parasite Banks (1979) reportedthat T congolense adhered to glutaraldehyde-fixedand unfixed erythrocytes and to the microvascu-lature in a similar manner Fixation of target cellshas also been used in studies on the attachment ofT cruzito fibroblasts and MDCK-cells (SchenkmanRobbins amp Nussenzweig 1991a Schenkman Diazamp Nussenzweig 19916) and to rat heart myoblasts(Villata et al 1993) These papers suggested thatadherence of T cruzi trypomastigotes which pre-cedes cell invasion was mediated by specific recep-tors on target cells at restricted cell surface domains

In contrast to these findings our observations werethat T congolense did not show a preference for anyspecialized cell surface domain

Recognition of specific receptormdashligand inter-actions usually mediated by cell surface adhesivemolecules is a very early step in microbial coloniza-tion and pathogenesis In many cell types carbo-hydrates are known to be crucially involved incell-cell interactions Cell surface adhesive moleculesinclude glycolipids (Karlsson et al 1992) glycos-aminoglycans (Lander 1993) and glycoproteins(Hoffman 1992 Oebrink 1993) Modifications car-ried out on cell surface carbohydrates therefore arelikely to influence the adhesive properties of cellsurfaces Periodate oxidation at acid pH has longbeen used as a tool to cleave carbohydrate residues(Woodward Young amp Bloodgood 1985) The resultspresented here after the use of a whole range ofperiodate concentrations at low and neutral pHshow that carbohydrates are likely to be involved inT congolense binding to endothelial cells Thepossibility however that other surface membranecomponents are also affected by this treatmentcannot be ruled out Further evidence of a role forcarbohydrate receptors in adhesion has been ob-tained by the use of lectins Wheat germ agglutininspecifically inhibits trypanosome attachment in amanner which points towards a role for sialic acidresidues in this process In addition neuraminidasewhich specifically cleaves terminal sialic acid resi-dues considerably reduced the number of trypano-somes which attached to BAE cells during theadhesion assay

Modification of the trypanosome surface was alsoattempted to identify parasite components whichaffected adhesion Trypanosomes pre-fixed withaldehydes were no longer able to bind to endothelialcells indicating denaturation or masking of themolecules involved Both inhibition of proteinsynthesis prior to adhesion and protease treatmentof trypanosomes before the adhesion assay reducedthe number of adherent parasites From these resultsit can be concluded that sialic acid residues on theendothelial plasma membrane might comprise or atleast be a significant component of a receptor whichis recognized by a specific lectin-like trypanosomesurface molecule

This lectin-like surface protein located at theanterior part of the flagellum and with a specializedfunction is rather unlikely to be a part of the variablesurface glycoprotein (VSG) coat Firstly the bindingof trypanosomes is restricted exclusively to theanterior two thirds of the flagellum which suggeststhat a non-VSG component of the T congolenseplasma membrane constitutes the receptor Sec-ondly there is evidence that non-variable glyco-proteins as well as parts of the VSGs are cleavedduring the exposure of live T brucei bloodstreamforms to trypsin (Frommel Seyfang amp Balber 1988)

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

A Hemphill I Frame and C A Ross 640

The reduction in adherence after protease treatmentof trypanosomes in these studies suggests a similaroccurrence in T congolense Thirdly it is interestingto note that attachment of T congolense to red bloodcells appeared to be mediated by sialic acid residueson the erythrocyte surface and could be abolishedby protease treatment (Banks 1979) This inter-action however could be inhibited by poly-L-lysinewhich saturates the negatively charged sialic acidresidues on the erythrocyte surface In our experi-ments neither poly-L-lysine nor dextran sulphatehad any effect on trypanosome attachment to BAEcells

Initial observations indicated that trypanosomeattachment occurred independently of a functionaltarget cell metabolism (data not shown) Experi-ments were then carried out to find if the attachmentprocess required metabolic activity on the part of theparasite Inhibition of glycolysis resulted in de-creased efficiency of adhesion suggesting that try-panosomes expend metabolic energy in their in-teraction with endothelial cells Similar results wereobtained in studies of South American trypanosomesand their attachment to mammalian cells (Schenk-man et al 1991a)

The cytoskeleton in most eukaryotic cells isinvolved in cellular events such as cell motility andsurface receptor movements (Geiger 1989) Intrypanosomes the cytoskeleton is mainly micro-tubule-based (Angelopoulos 1970 HemphillLawson amp Seebeck 1991a Hemphill Seebeck ampLawson 1991 b) and the presence of actin in Africantrypanosomes has not yet been conclusively demon-strated (BenAmar et al 1988 Seebeck Hemphill ampLawson 1990) However the genes coding for actinof T brucei have been sequenced and it has beenshown that they are constitutively transcribed inboth bloodstream and insect forms (BenAmar et al1988) In this study parasites were treated withdrugs which would specifically affect microtubulesand actin filaments Taxol and nocodazole which aremicrotubule active drugs in higher eukaryotic cellshad no effect on adhesion while cytochalasin D adrug specifically destroying the organization of actinfilaments (Cooper 1987) did have an inhibitoryeffect No specialized filamentous structures how-ever could be seen by TEM There remains thepossibility that actin could form small filaments notvisible by our methodology

In conclusion our evidence indicates that theinteraction between these two cell types is mediatedby certain carbohydrate receptors present on theendothelial cell surface and one or several ligandsmost likely proteins or glycoproteins on the trypano-some whose functional activities are affected byproteases and protein synthesis inhibitors Theseligands must be present at distinct localizations nearthe anterior end of the flagellum Our investigationsprovide the necessary background for further studies

on interactions between T congolense and endothelialcells which should lead to the identification of themolecules involved The significance and the func-tional importance of this adhesion is yet to beexplained but a more thorough exploration of thisprocess will contribute to a better understanding ofthe pathogenesis of this parasite

It is a pleasure to acknowledge Simon L Croft (De-partment of Medical Parasitology LSHampTM) for hisencouraging support and George Towey and MariaMcCrossan (EM Unit LSHampTM) for their technicalhelp in electron microscopy We thank Thomas Seebeck(University of Bern Switzerland) for providing us withtaxol and with antibodies directed against tubulin andactin as well as Robert Harris (Department of MedicalParasitology LSH ampTM) and Norbert Mueller (Instituteof Parasitology University of Berne Switzerland) fortheir helpful discussions and comments on the manuscriptThis work was supported by a fellowship from the SwissNational Foundation a stipend from the Bernese CancerLeague to AH and the Overseas Development Admin-istration

REFERENCES

ANGELOPOULOS E (1970) Pellicular microtubules in thefamily Trypanosomatidae Journal of Protozoology 1739-51

BANKS K L (1978) Binding of Trypanosoma congolenseto the walls of the small blood vessels Journal ofProtozoology 25 241-5

BANKS K L (1979) In vitro binding of Trypanosomacongolense to erythrocytes Journal of Protozoology 26103-8

BANKS K L (1980) Injury induced by Trypanosomacongolense adhesion to cell membranes Journal ofParasitology 66 34mdash7

BENAMAR M F A PAYS A TEBABI P DERO B SEEBECK

T STEINERT M amp PAYS E (1988) Structure andtranscription of the actin gene in Trypanosoma bruceiMolecular and Cellular Biology 8 2166-7

BENHAMOU N (1989) Preparation and application oflectin-gold complexes In Colloidal Gold PrinciplesMethods and Applications vol 1 (ed Hayat M A)pp 95-143 London Academic Press

BOOTHROYD j c (1985) Antigenic variation in Africantrypanosomes Annual Review in Microbiology 39475-502

BRODEN A (1904) Les infections a trypanosomes auCongo chez 1homme et les animaux Bulletin de laSociete beige detudes coloniales Bruxelles

COOPER J A (1987) Effects of cytochalasin andphalloidin on actin Journal of Cell Biology 1051473-8

FROMMEL T O SEYFANG S amp BALBER A E (1988) Trypanosoma brucei ssp Cleavage of variant specificand common glycoproteins during the exposure of livecells to trypsin Experimental Parasitology 66 213mdash24

GEIGER B (1989) Cytoskeleton associated cell-contactsCurrent Opinion in Cell Biology 1 103-9

GRAY M A ROSS C A TAYLOR M A TETLEY L amp

LUCKINS A G (1985) In vitro cultivation ofTrypanosoma congolense the production of infective

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at

Host-parasite interactions in trypanosomiasis 641

forms of metacyclic trypanosomes cultured on bovineaorta endothelial cell monolayers Ada Tropica 4299-110

HEMPHILL A LAWSON D amp SEEBECK T (1991 a) T h ecytoskeletal architecture of Trypanosoma bruceiJournal of Parasitology 77 603-12

HEMPHILL A SEEBECK T amp LAWSON D (19916) T h eTrypanosoma brucei cytoskeleton ultrastructure andlocalization of microtubule associated and spectrin likeproteins using quick-freeze deep-etch immunogoldelectron microscopy Journal of Structural Biology107 211-20

HOFFMAN s (1992) Assays of cell adhesion In Cell-CellInteractions A Practical Approach (ed Rickwood Damp Hames B D) pp 1-29 Oxford IRL Press

KALINER G (1974) Trypanosoma congolense IIHistochemical findings in experimentally infectedcattle Experimental Parasitology 86 20-6

KARLSSON K A M I L H M A ANGGSTROEM J

BERGSTROEM J DEZFOOLIAN H L A N N E B

LEONARDSON I amp TENEBERG s (1992) Membraneproximity and internal binding in the microbialrecognition of host cell glycolipids a conceptualdiscussion In Molecular Recognition in HostmdashParasiteInteractions FEMS Symposium No 61 (edKorhonen T K Tapani H amp Maekelae P H) pp115-132 New York Plenum Press

LANDER A D (1993) Proteoglycans In Guidebook toExtracellular Matrix and Adhesion Proteins (ed KreisT amp Vale R) pp 12-16 Oxford Oxford UniversityPress

LOSOS G PARIS J W I L S O N A J amp DAR F K ( 1 9 7 3 )

Distribution of Trypanosoma congolense in tissues ofcattle Transactions of the Royal Society of TropicalMedicine and Hygiene 67 278

MAJIWA P A O MAINA M WAITWUBI J N MIHAK S amp

ZWEYGARTH E (1993) Trypanosoma (Nannomonas)congolense molecular characterization of a newgenotype from Tsavo Kenya Parasitology 106151-62

MOLYNEUX D H amp ASHFORD R w (1983) African animaltrypanosomiasis In The Biology of Trypanosomes andLeishmania Parasites of Man and Domestic Animals(ed Taylor and Francis) London Academic Press

MONSIGNY M ROCHE A C SENE C MAGET-DANA R amp

DELMOTTE F (1980) Sugar-Lectin interactions Howdoes wheat-germ agglutinin bindsialoglycoconjugates European Journal of Biochemistry104 147-53

NORGARD K E HAN H POWELL L KRIEGLER M

VARKI A amp VARKI N M (1993) Enhanced interactionof L-selectin with the high endothelial venule ligandvia selectively oxidized sialic acids Proceedings of theNational Academy of Sciences USA 90 1068-72

OEBRINK B (1993) Cell adhesion and cell-cell contactproteins In Guidebook to Extracellular Matrix andAdhesion Proteins (ed Kreis T amp Vale R) pp109-114 Oxford Oxford University Press

REINWALD E (1985) Role of carbohydrates withinvariant surface glycoproteins of Trypanosomacongolense European Journal of Biochemistry 151385-91

SCHENKMAN S ROBBINS E S amp NUSSENZWEIG V

(1991a) Attachment of Trypanosoma cruzi requiresparasite energy and invasion can be independent ofthe target cell cytoskeleton Infection and Immunity 59645-54

SCHENKMAN S DIAZ C amp NUSSENZWEIG V (19916)Attachment of Trypanosoma cruzi trypomastigotes toreceptors at restricted cell surface domainsExperimental Parasitology 72 76mdash86

SEEBECK T HEMPHILL A amp LAWSON D (1990) T h ecytoskeleton of trypanosomes Parasitology Today 649-52

SMITH M amp CROFT s L (1991) Embedding and thinsection preparation In Electron Microscopy in BiologyA Practical Approach (ed Rickwood D amp HarrisB D) pp 17-37 Oxford IRL Press

VICKERMAN K (1985) The developmental cycles andbiology of pathogenic trypanosomes British MedicalBulletin 41 105-14

VICKERMAN K (1969) The fine structure ofTrypanosoma congolense in its bloodstream phaseJournal of Protozoology 16 54mdash69

VICKERMAN K amp TETLEY L (1990) Flagellar surfaces ofparasitic protozoa and their role in attachment InCiliary and Flagellar Membranes (ed BloodgoodJ R) pp 267-304 New York Plenum Press

VILLATA F RUIZ-RUANO A VALENTINE A A amp LIMA

M F (1993) Purification of a 74 kD glycoprotein fromheart myoblasts that inhibits binding and entry ofT cruzi into heart cells Molecular and BiochemicalParasitology 61 217-30

WOODWARD M P YOUNG W W amp BLOODGOOD J R

(1985) Detection of monoclonal antibodies specific forcarbohydrate epitopes using periodate oxidationJournal of Immunological Methods 78 143mdash53

httpswwwcambridgeorgcoreterms httpsdoiorg101017S0031182000076514Downloaded from httpswwwcambridgeorgcore University of Basel Library on 30 May 2017 at 173902 subject to the Cambridge Core terms of use available at