Low density lipoproteins bound to Schistosoma mansoni do not alter the

rapid lateral diffusion or shedding of lipids in the outer surface membrane

JOHN P. CAULFIELD1'5*, CHUN-PIN CHIANG5, PATRICK W. YACONO2, LAUREL A. SMITH3'4 and

DAVID E. GOLAN2'3-6

Departments of1 Pathology, ^Biological Chemistry and Molecular Pharmacology, and ^Medicine, Harvard Medical School and*Channing Laboratories, the Departments of b Rheumatology/Immunology and 6 Medicine (Hematology Division), Brigham andWomen's Hospital, Boston, MA 02115, USA

* Author for correspondence at Harvard Medical School, The Seeley G. Mudd Building, Room 517, 250 Longwood Avenue, Boston,MA 02115, USA

Summary

Schistosomula of Schistosoma mansoni bind humanlow density lipoproteins (LDL) in a concentration-dependent and saturable manner. The bound LDLcould provide phospholipids and sterol to the worm,which cannot synthesize sterol de novo and lacksacyl chain-modifying capability. Here we have usedthree phospholipid analogues to explore the effect ofLDL binding on the parasite's outer tegumentalmembrane, i.e. the outer of the two membranes thatcover its surface syncytium. Fluorescein phosphati-dylethanolamine (Fl-PE) and rhodamine phosphatd-dylethanolamine (Rh-PE) bound to the parasite in asaturable manner and, as shown by fluorescencemicroscopy, were confined to the surface. Fl-PEfluorescence was completely quenched by TrypanBlue and Fl-PE was lost from the surface followingsingle exponential decay kinetics (tj=12h), furthersuggesting that the probe was confined to the outermembrane. 1,1' -Dioctadecy 1-3,3,3' ,3' -tetramethyl-indocarbocyanine perchlorate (Dil-C18(3); Dil) didnot bind saturably and was seen in both the surfaceand the internal parasite membranes. Fluorescencephotobleaching recovery was used to measure the

lateral mobility of Fl-PE in the outer membrane. Thelateral diffusion coefficient of Fl-PE was approxi-mately 10"7cm2s~1. The fractional mobility of Fl-PEwas 85 % when measured using a laser beam of radius1.8 fan and 45% using a beam of radius 4.3/an. Thesemeasurements suggest that the outer membrane iscomposed of /an-scale liquid crystalline-phase lipiddomains that lack significant amounts of transmem-brane proteins. LDL binding to the parasite surfacedid not alter the lateral mobility of Fl-PE or the rateof loss of either Fl-PE or Rh-PE. These studiessuggest that the binding of LDL to the outertegumental membrane of schistosomula does notchange either the organization of the membranelipids or their rate of loss from the membrane. If LDLdo provide lipids to schistosomula, then directinsertion of lipid molecules into the outer tegumentalmembrane is an unlikely mechanism for lipid trans-fer.

Key words: trematode, fluorescence photobleaching recovery,Schistosoma mansoni.

Introduction

Schistosomes are trematodes that infect more than 200million people worldwide. The adult parasites of Schisto-soma mansoni survive for years in the portal vasculaturewhere they escape killing by the immune system despite ademonstrable humoral response to the infection. Larvalparasites, called schistosomula, can be killed in vitro byisolated human effector cells, i.e. granulocytes andmacrophages, in antibody-dependent, cell-mediated cyto-toxic reactions. The parasite can defend itself in vitro andpresumably in vivo against these effector cells by severalmechanisms localized to the parasite surface, including areduction in antigenicity (Samuelson et al. 1980) and theelaboration of molecules toxic to host cells (Golan et al.1986).

A schistosomulum is covered by a syncytium, called thetegument, which is in turn covered by two lipid bilayers.Journal of Cell Science 99, 167-173 (1991)Printed in Great Britain © The Company of Biologiste Limited 1991

The inner bilayer is rich in intramembrane particles butthe outer bilayer is nearly devoid of particles, suggestingthat the outer bilayer is composed almost entirely of lipidswith little associated protein. The recovery of labeledligands and surface moieties from the culture medium(Samuelson and Caulfield, 1982; Samuelson et al. 1982)and the absolute loss of iodinated surface antigens fromthe parasite (Pearce et al. 1986; Saunders et al. 1987)suggest that the worm continually sheds its outermembrane. Further, the parasite does not endocytoseligands bound to its surface. Thus, a schistosome is coveredby a large sheet of membrane that is in turn covered by asheet of phospholipids and sterol. The latter sheet movescontinuously outward, shedding antigens and boundantibody.

The parasite also binds host molecules that may furtherreduce its antigenicity. Recently, human low densitylipoproteins (LDL) have been shown to bind to the surface

167

of schistosomula. LDL bound to the parasite partiallyblock the binding of antibodies from the sera of scbistoso-miasis patients (Chiang and Caulfield, 1989o,6). BoundLDL are partially displaced by treatment with suramin orother polyanions. Others have suggested that Schistosomamansoni binds LDL with a 45x KrAfr protein doublet thatis expressed on schistosomula after transformation fromcercariae (Rumjanek et al. 1983, 1985, 1988). An LDL-binding protein of similar molecular weight has beenidentified in Schistosoma japonicum (Rogers et al. 1989,1990).

The binding of human LDL is also important becauseschistosomes do not synthesize fatty acids or cholesterol denovo. Bound LDL could provide these lipids to the parasite.In particular, LDL could alter the lipid composition of theworm surface membranes by inserting sterols or phospha-tidylcholine. Fatty acids derived from LDL phospholipidscould be enzymatically incorporated into the parasite'slipids. If LDL were broken down on the surface of theworm, lipids could be incorporated directly into the surfacebilayer. Rumjanek and McLaren (1981) showed thatserum causes worms to release iodinated surface materialbut delipidated serum does not. Although these authorsspeculated that serum-derived LDL was the cause of thisrelease, they did not examine the role of LDL directly.Alternatively, LDL could be ingested into the parasite gutand degraded, and the lipids incorporated metabolicallyinto parasite membranes. This pathway has been sugges-ted by recent work in which the worm ingested carbocya-nine-labeled LDL and the fluorophore was subsequentlyredistributed throughout the parasite (Bennett and Caul-field, 1991).

Here, we have examined the effect of bound LDL on theproperties of lipid probes inserted into the surfacemembranes of the parasite. First, three probes, fluoresceinphosphatidylethanolamine (Fl-PE), rhodamine phosphati-dylethanolamine (Rh-PE), and l,l'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine perchlorate (Dil-C18(3);DiT), were tested for their ability to label the parasitesurface selectively. Second, schistosomula were labeledwith Fl-PE or Rh-PE and the effect of bound LDL on theloss of these labels from the surface was measured. LDLwas also displaced with suramin before measurement totest whether the fluorescent label is partitioned into thebound LDL. Finally, fluorescence photobleaching recoveryexperiments were performed on worms labeled with Fl-PEor Dil to examine the effect of LDL binding on lipid probelateral mobility.

Materials and methods

MaterialsA Puerto Rican strain of Schistosoma mansoni was carried inBiomphalaria glabrata snails and CBA/J mice (Jackson Labora-tories, Bar Harbor, Maine). Schistosomula were prepared fromcercariae by vortexing (Ramalho-Pinto et al. 1974). The larvaewere separated from the tails by centrifugation on a Percollgradient (Lazdins et al. 1982). The parasites were approximately3 h post-transformation at the beginning of culture.

Culture conditionsAll cultures were performed in 1.5 ml polypropylene centrifugetubes at 37 °C under 5% CO2/95% air. In general, 1000schistosomula were cultured overnight in a tube containing 600 /ilRPMI 1640, lOOi.u.mr1 penicillin, lOO/fgrnP1 streptomycin,2 miu L-glutamine, and 20 HIM Hepes buffer (RPMI 1640-PS) plus1 % BSA (RPMI/BSA). The worms were labeled with fluorescent

lipid probes and incubated in RPMI/BSA with or without LDL,suramin, or Trypan Blue, as described below. The relativefluorescence of the bound fluorophore was measured with aphotometer attached to a Leitz Orthoplan microscope, aspreviously described (Chiang and Caulfield, 1989a), or lipid probelateral mobility was measured as described below. Worms wereimmobilized with 10 mM eserine sulfate immediately prior tophotometry or lateral mobility measurements.

Fluorescent labelingDil, Fl-PE and Rh-PE (Molecular Probes, Eugene, OR) were usedto label the schistosomula. To determine whether probe binding tothe parasites was saturable, 1000 cultured worms were washedonce in phosphate-buffered saline (PBS), pH7.4, incubated for30 min at room temperature (RT) with 500 /il of PBS containing0.25-12 fig ml"1 probe and washed three times with RPMI/BSA.Relative fluorescence was measured as described above.

Effect of LDL on rate of probe loss from the membrane1000 schistosomula per tube were labeled for 30 min at RT with500 /J of PBS containing 4 /<g ml"1 Fl-PE or 6 /<g ml"] Rh-PE, andthen cultured for 3, 6,12 or 24 h at 37 °C in 300 /d RPMI/BSA withor without 20 ^g LDL. At each time point 3 tubes of worms, 1without and 2 with LDL, were removed from the 37 °C bath andwashed 3 times with RPMI/BSA. Relative fluorescence of wormsfrom the tube cultured without LDL and one of the tubes culturedwith LDL were measured directly after washing. To removebound LDL from the worm, the other tube cultured with LDL wastreated with 2mM suramin for 15 min (Chiang and Caulfield,1989a,b) and washed, and the relative fluorescence was measured.

Quenching of Fl-PE by Trypan BlueAbout 1000 worms were labeled for 30 min at RT with 4/<gml~1

Fl-PE in PBS, pH7.4. The worms were washed 3 times,resuspended in 130/d RPMI containing 0.25% Trypan Blue,incubated for either 5 min or 20 min, and washed and the relativefluorescence was measured. Positive controls were labeled withFl-PE and incubated in RPMI without Trypan Blue; negativecontrols were incubated in Trypan Blue without prior exposure toFl-PE.

Fluorescence photobleaching recovery (FPR)Schistosomula were labeled with 4 /<g ml"1 of either Dil-C18(3) orFl-PE in PBS and cultured for 0,3,6 or 12 h at 37°C in RPMI/BSAwith or without LDL. The lateral mobility of the lipid analoguesin the labeled schistosomula was measured by FPR (Axelrod et al.1976). Briefly, the worm surface was observed in a fluorescencemicroscope using a focused laser beam as the excitation source. Asmall area of membrane was exposed to a brief, intense laserpulse, causing irreversible bleaching of the fluorophore in thatarea. Fluorescence recovery, resulting from the lateral diffusionof unbleached fluorophore into the bleached area, was measured.Analysis of the fluorescence recovery curves yielded the fractionof Dil or Fl-PE that was free to diffuse in the plane of themembrane (the mobile fraction, /), as well as the diffusioncoefficient (D) of the mobile fraction.

Our FPR apparatus and analytical methods have beendescribed in detail (Golan et al. 1986), with the followingmodifications. Two computer-interfaced acousto-optic modulators(N35085-3; Newport Electro-Optics Systems, Fountain Valley,CA) were used to produce the monitoring and photobleachinglaser pulses. A multi-channel sealer (370; Nicolet InstrumentCorp., Madison, WI) was used to collect fluorescence data from theamplifier/discriminator. The experiment was controlled by acomputer (386i; Sun Microsystems Inc., Mountain View, CA) anda custom-built timing board (Corbett, J.D., V.G. Bose and D.E.Golan, unpublished data). The Gaussian beam radius at thesample plane was determined daily by a two-dimensionalemission scan technique (Stolpen et al. 1988). Photobleachingpower at the sample was approximately 2 mW. Sample tempera-tures were controlled to 23.0(±0.1)°C with a thermal stage.

168 J. P. Caulfield et al.

Statistical analysisAnalysis of the loss of Fl-PE and Rh-PE from the parasite surfaceover time was carried out using PC-SAS software, version 6.3(SAS Institute Inc., Cary, NC).

Results

Fluorescein phosphatidylethanolamineParasites exposed to Fl-PE acquired fluorescence on theirsurface in a concentration-dependent and saturable man-ner (Fig. 1). Fl-PE binding increased monotonically atconcentrations between 0.25 and 4/tgml~1 Fl-PE andsaturated at concentrations greater than 4^gml~1. Byfluorescence microscopy the Fl-PE was seen to be presenton the parasite surface and in the esophagus and cecum(see Fig. 2). Fl-PE fluorescence was largely, approximately95 %, quenched by exposure to Trypan Blue (Table 1),indicating that the bound fluorophore was confined to theouter membrane of the worm surface. By comparing thefluorescence intensity in the labeled parasite membranewith that in the Fl-PE labeled human red cell membrane(Golan et al. 1986), the Fl-PE concentration in the outermembrane at saturation was estimated to be 0.2 mole %.

Worms were exposed to saturating concentrations of Fl-PE and cultured with or without LDL and surfacefluorescence was measured. To examine the effect of boundLDL on the loss of label from the membrane, fluorescencewas also measured before and after displacement of thebound LDL with suramin. Under all three conditions theFl-PE fluorescence decreased with time, following singleexponential decay kinetics (Fig. 3). Analysis of covariance,using the log of the percentage of Fl-PE bound as thedependent variable, showed no significant differencebetween lines fitted to the data in the presence and

2 4 6 8 10Concentration of Fl-PE or Rh-PE

12

Fig. 1. Fl-PE ( ) and Rh-PE ( ) bind saturably toschistosomula. About 1000 worms were incubated for 30min atRT in 0.5 ml RPMI containing the concentration of Fl-PE orRh-PE expressed on the x-axis. After washing, the relativefluorescence of the parasite surface was measured with aphotometer attached to a light microscope. Saturation ofbinding occurred at a concentration of about 4-6//g ml"1 forboth fluorescent lipid analogues. No internal fluorescence wasobserved, indicating that the loading of the surface membraneswith the lipid analogues was saturated.

absence of LDL (P=0.65). Similarly, there was nosignificant difference between curves obtained fromworms cultured in the absence of LDL and worms exposedto suramin after culture in LDL (P=0.09). Single exponen-tial decay accounted for almost all the variability in thepercentage of Fl-PE bound (r2=0.94). The half-time ((,) ofloss for worms cultured in the absence of LDL was 12.4 h(95% confidence interval=ll.l to 13.4 h); for wormscultured in the presence of LDL, 11.9 h (10.8 to 13.3 h); and

Fig. 2. Schistosomula incubatedwith 6 jig ml"1 Rh-PE for 3h. In(A) the objective is focused onthe surface and its spines. Thefluorescence is uniformlydistributed except for the brightareas on the head at the right,which are probably glandularsecretions, and the tail socketand openings of the excretorysystem on the left. The bright,out-of-focus fluorescence in thecenter is due to fluorophore inthe esophagus (e) and the cecum(c). In (B) a higher-power viewshows the surface in cross-section. The fluorescence isconfined to a line that is thesurface membranes. Thedistribution of Fl-PE was similarto that shown here for Rh-PE.Bars: A, 20 jnn; B, 10 fan.

LDL binding to schistosome membranes 169

10

Fig. 3. Clearance of Fl-PE from thesurface of schistosomula in thepresence and absence of bound LDL.Worms were labeled with Fl-PE,washed, and cultured in RPMI/BSA( ) or incubated with Fl-PE,washed, and cultured in RPMI/BSAwith LDL ( ). After the timeindicated on the x-axis, either therelative fluorescence on the wormsurface was measured or the boundLDL was partially displaced with 2 mMsuramin ( ) and the fluorescencewas then measured. The curves arefirst-order regressions. Error barsrepresent S.D. Bound LDL did not alterthe rate of loss of Fl-PE from thesurface (i,=11.7h). Fl-PE did nottransfer from the parasite membranesinto the bound LDL, since partialdisplacement of bound LDL by suramindid not alter surface fluorescence.

Table 1. Trypan Blue quenching of fluorescence ofschistosomula labeled with Fl-PE

TreatmentRelative

fluorescence (%)

Fl-PE+Trypan Blue, 5min+Trypan Blue, 20 mm

Trypan Blue only

1006.6±0.64 3±1.1

0.5±0.2

Values represent mean±standard deviation (S.D ) of 3 experiments inwhich worms were labeled with Fl-PE, washed, and then exposed toTrypan Blue for 5 or 20 minutes before a final wash. Relativefluorescence was measured by photometry on at least 16 worms in eachexperiment. Values were normalized to the Fl-PE fluorescence beforeexposure to Trypan Blue in each experiment and the means of theexperiments were then averaged. The relative fluorescence of the wormsincubated with Trypan Blue alone was measured twice in parallel withthe Fl-PE quenching experiments; the value represents the mean±s.D.of 2 sets of measurements on 16 worms each.

Table 2. Fl-PE lateral mobility in the surfacemembranes of schistosomula before and after exposure to

LDLTime (h) LDL

66

1212

1.9±1.6

2.9±1.62.7±1.1

3.0±1.32.8±1.9

45±22

60±2748±12

43±947±25

22

158

1524

Schistosomula were labeled with Fl-PE and cultured for 0, 6 or 12 hat 37 °C in RPMI/BSA with or without LDL. Lateral mobility wasmeasured by FPR using a Gaussian beam radius of 4.3 /on. Values areexpressed as mean±s.D. D is the diffusion coefficient, /'is the mobilefraction, and n the total number of worms that were measured tocalculate D and f. Data are pooled from measurements performed on 2or 3 different days on different sets of worms.

for worms exposed to suramin after culture in LDL, 11.2 h(10.2 to 12.4 h). Because there were no significantdifferences among the three sets of data, they were pooledand analyzed. The t± of the pooled data was 11.7 h (95%confidence interval=10.6 to 13.0 h). These experimentsshow that LDL binding to the surface of the parasite didnot alter the rate of loss of the membrane probe Fl-PE fromthe parasite surface. Further, Fl-PE did not appear topartition into the surface-bound LDL, because the fluor-escence intensity at the parasite surface did not decreaseafter exposure of the parasite to suramin at concentrationsthat removed approximately 40 % of the bound LDL.

The rate of lateral diffusion of Fl-PE in the outermembrane was extremely high, approximately2xlO~7cm2s~1, and did not change as the intensity ofsurface fluorescence decreased over 12 h (Table 2). Themobile fraction of Fl-PE remained constant, at approxi-mately 40-60 %, over the same time period. LDL bindingto the parasite surface did not alter the lateral diffusioncoefficient or the fractional mobility of Fl-PE over 12 h ofincubation (Table 2). Thus, LDL bound to the surface ofschistosomula did not affect either the lateral mobility ofFl-PE or the loss of this lipid probe from the surface.

Laser beams of different sizes can be used in FPRmeasurements of lipid probe mobility to examine the long-

Table 3. Effect of laser beam radius on lateral mobilityof fluorescently labeled lipid probes in schistosomular

membranesProbe Beam radius (/an) D (xl0ecmas"1) f(%) n

Fl-PE

Dil

1.8±0.044.3±0.1

1.8±0.044.3±0.1

6.6±3.919±16*

0.23±0.060.30±0.13

22*85±1146±22'

83±10 1662±10 9

The beam radius at the sample plane was determined by a two-dimensional emission scan technique (Stolpen et al 1988) and isexpressed as the mean±s.D. The values for D and /'represent themean±s.D. of measurements taken on 2 or 3 different days on differentsets of worms; n is number of worms.

* Data reproduced from Table 2.

range organization of membrane lipid. Specifically, thepresence of /an-scale domains can be inferred from theobservation that the fractional mobility of a lipid probedecreases with increasing beam size (Yechiel and Edidin,1987). To investigate whether or not such domains couldbe present in the outer surface membrane of schistoso-mula, we used Gaussian laser beams of two different sizesto measure the lateral mobility of Fl-PE. Fractionalmobilities of 45 % and 85 % were found using beams ofradius 4.3 and 1.8 jan, respectively (Table 3), suggesting

170 J. P. Caulfield et al.

100 j90 •onOU

7060

50

40

30

20

10 i 1 1 1 112

Time (h)18 24

Fig. 4. Clearance of Rh-PE from thesurface of schistosomula in thepresence and absence of bound LDL.Worms were labeled with Rh-PE,washed, and cultured in RPMI/BSA( ) or incubated with Rh-PE,washed, and cultured in RPMl/BSAwith LDL ( ). After the timeindicated on the x-axis, either therelative fluorescence on the wormsurface was measured or the boundLDL was partially displaced with 2 DIMsuramin ( ) and the fluorescencewas then measured. Error barsrepresent S.D. The curves are second-order regressions. The rate of loss offluorescence was similar to that seenfor Fl-PE for the first 12 h. At 24 h,however, less Rh-PE was lost than Fl-PE. Bound LDL did not alter the rateof loss of Rh-PE from the surface.Suramin treatment did not displacefluorophore from the parasite surface,suggesting that Rh-PE was notpartitioned into the bound LDL.

that jan-scale lipid domains are present in the outermembrane. Beams of radius < 1.8 /an could not be used,since Fl-PE diffusion in the outer membrane was so rapidthat the duration of our most intense photobleaching pulsewould have approached the half-time for fluorescencerecovery.

Rhodamine phosphatidylethanolamineLike Fl-PE, Rh-PE bound to schistosomula in a concen-tration-dependent and saturable manner (Fig. 1). Bindingincreased monotonically between 0.5 and 6.0/igml"1 Rh-PE and saturated at concentrations greater than6/igm]"1. By fluorescence microscopy Rh-PE was visibleon the parasite surface and in the gut (Fig. 2). Trypan Bluequenching of Rh-PE could not be performed because thefluorescence emission of Trypan Blue is in the samewavelength region as that of rhodamine. Both thesaturation of the binding curve and direct microscopicobservation suggested, however, that Rh-PE was confinedto the worm surface.

Parasites labeled with 6 /^gml"1 Rh-PE lost fluorescenceover time (Fig. 4). Over the first 12 h the rate of loss of Rh-PE followed single exponential decay kinetics with*j=11.6h (95% confidence limits, 10.3-13.1 h; r2=0.91).After 24 h, however, more fluorescence was retained on theRh-PE-labeled worms than on the Fl-PE-labeled parasites,37.6±3.0% versus 20.9±2.8% (mean±standard error),respectively. These values were significantly different byStudent's t-test (P=0.001) and by Wilcoxon's test (P=0.05).When the 24 h point was included, the Rh-PE data fit aquadratic model slightly better, r =0.94, than a singleexponential model, P=0.88, although too few data pointswere available to determine whether a quadratic modelprovides an adequate description over the entire 24 hperiod. Thus, the initial rates of loss of Fl-PE and Rh-PEfrom the parasite surface were the same but there was adecrease in the rate of loss of Rh-PE after 12 h. Thisdecrease was most likely due to gradual partitioning ofRh-PE into the inner surface membrane. FPR studiescould not be performed on Rh-PE-labeled worms becausethe laser intensity required to bleach the rhodamine alsodamaged the parasite membrane.

Saturating concentrations of LDL bound to the parasite

surface did not cause an alteration in the rate of loss of Rh-PE from the worm (Fig. 4). Further, displacement of 40%of the bound LDL with 2mM suramin did not cause adecrease in fluorescence on the worm surface, suggestingthat Rh-PE did not partition into the bound LDL. The rateof loss of Rh-PE in the presence of LDL, both before andafter suramin treatment, was not significantly differentfrom the rate of loss observed in medium alone (P=0.76and P=0.44, respectively). Thus, although the rate of lossof Fl-PE and Rh-PE differed after long times in culture,LDL binding did not alter the kinetics of loss of eitherphospholipid analog.

Dil-C18(3)Incubation of schistosomula with Dil, 0.25-12j/gml"1,resulted in a concentration-dependent increase in fluor-escence associated with the schistosomula. Unlike Fl-PEand Rh-PE binding, Dil binding did not become saturatedat high probe concentration. Further, Dil fluorescence wasobserved in the internal membranes of the parasite as wellas on the surface. Both of these observations indicate thatDil was not confined to the parasite surface. These resultsare in agreement with those of Foley et al. (1986), whocould quench only 15 % of the Dil fluorescence on labeledadult worms.

The rate of lateral diffusion of Dil in schistosomularmembranes was approximately 2xlO~9cm2s~1, or 30- to100-fold slower than that of Fl-PE. The fractional mobilityof Dil, like that of Fl-PE, was dependent on the radius ofthe laser beam, varying from 52 % using a 4.3 [an beam to82% using a 1.8 fan beam (Table 3). Because Dil fluor-escence was observed internally as well as on the surface,these diffusion coefficients and fractional mobilitiesrepresent averages for Dil mobility among the variousmembranes into which Dil was partitioned.

Discussion

These studies have examined the surface of schistosomulausing three fluorescent lipid probes. Fl-PE appears to be aspecific surface label, since it binds saturably, is quenched

LDL binding to schistosome membranes 171

by Trypan Blue, and is observed only on the surface. Theother two probes enter the parasite, Dil rapidly during thecourse of labeling and Rh-PE slowly over 12-24 h. Fl-PEhas an extremely high rate of lateral diffusion on theparasite surface, suggesting that it is inserted into amembrane with a unique composition that approximatesthat of an artificial lipid bilayer rather than that of atypical biological membrane. Both Fl-PE and Rh-PE arelost from the parasite surface with a half-time of about12 h, probably as a consequence of the shedding of thesurface membrane into the culture medium (Samuelsonand Caulfield, 1982a). Finally, saturating concentrationsof human LDL bound to the parasite surface do not alterthe rapid lateral diffusion and significant immobilefraction of Fl-PE or the rate of loss of Fl-PE and Rh-PE,suggesting that lipids from the bound LDL do not disruptthe native composition or organization of the parasitesurface membrane.

Of the three lipid probes used here, Fl-PE is mostdemonstrably confined to the parasite surface. Our findingthat 95 % of Fl-PE fluorescence is quenched by TrypanBlue differs from the results of others, who observed a 40 %reduction in fluorescence after incubation with TrypanBlue (Foley et al. 1986). These investigators used TrypanBlue in medium containing 10 % fetal calf serum, however,in which albumin could bind Trypan Blue and reduce theeffective concentration of this quencher. The distributionof Fl-PE between the inner and outer surface membranesremains an open question. The extremely rapid rate ofprobe diffusion, the maintainence of this rate over 12 h andthe first-order rate of probe loss from the parasite suggest,however, that Fl-PE is inserted into and remains in aunique environment, namely the outer tegumental mem-brane. In contrast, Rh-PE probably localizes initially tothe outer membrane and is partitioned over 12-24 h intothe inner membrane, whereas Dil is partitioned withinminutes into the internal parasite membranes.

The rapid rate of diffusion of lipid probes (10~8 to10"7cm2s~1, Tables 2 and 3; Levi-Schaffer et al. 1986)indicates that the composition of the outer tegumentalmembrane is unique. Lateral diffusion coefficients of5xlO~8 to 20xl0~8cm2s"1 have been measured forfluorescently labeled phospholipid analogs at room tem-perature in fluid-phase artificial bilayers composed ofdimyristoylphosphatidylcholine or egg phosphatidylcho-line (Wu et al. 1977; Rubenstein et al. 1979; Smith et al.1979; Derzko and Jacobson, 1980; Alecio et al. 1982; Golanet al. 1984). Addition of cholesterol to these artificialsystems reduces the rate of probe diffusion by two- tothreefold (Wu et al. 1977; Rubenstein et al. 1979; Alecio etal. 1982; Golan et al. 1984). Further, in at least twobiological systems there is a fourfold increase in thelateral diffusion rates of lipid probes upon extraction andreconstitution of plasma membrane lipid, but not protein,into multilamellar liposomes (Jacobson et al. 1981; Golanet al. 1984). The rapid diffusion of Fl-PE may indicate thatthe outer schistosomular membrane is composed primarilyof fluid-phase phospholipid, with little transmembraneprotein or cholesterol.

This inference is consistent with current knowledge ofthe chemical composition of the tegumental membranes.The outer membrane appears to contain some sterol aswell as phospholipid (Allan et al. 1987; Young and Podesta,1986; Torpier and Capron, 1980). Phosphatidylcholine isthe major phospholipid in the outer membrane (Allan et al.1987; Young and Podesta, 1984; Furlong and Caulfield,1989) but lesser amounts of phosphatidylethanolamine

and antigenic glycolipids (Weiss et al. 1986) are alsopresent. The protein content of the outer membrane mustbe low, since there are no intramembrane particles seen byfreeze-fracture, although proteins can be radiolabeled onthe surface (Dissous et al. 1981; Taylor et al. 1981;Samuelson and Caulfield, 1982). Some of these proteinsare anchored by phosphoinositol linkages (Espinoza et al.1988; Pearce and Sher, 1989; Sauma and Strand, 1990).Taken together, the evidence suggests that the outermembrane is poor in transmembrane proteins and con-tains predominantly phosphatidylcholine and some sterol.

Assuming that the mobility of fluorescent lipid analogsreflects that of endogenous membrane lipids, the depen-dence of lipid probe fractional mobility on laser beamradius suggests that the parasite membranes are organ-ized in jan-scale lipid domains (Klausner and Wolf, 1980;Owicki and McConnell, 1980; Yechiel and Edidin, 1987-Golan et al. 1988). In order to permit diffusion rates of 10to 10~7cm2s~1, it is likely that the outer surfacemembrane consists of immiscible fluid-phase domains(Recktenwald and McConnell, 1981). The threefold in-crease in the diffusion coefficient of Fl-PE with increasingbeam radius (Table 3) is also consistent with a model inwhich discrete, fluid-phase lipid domains are separated bya fluid-phase lipid reticulum (Yechiel and Edidin, 1987).Foley et al. (1988) have suggested, on the basis of thebinding of merocyanin 540, that the outer membrane is ina liquid crystalline state. Internal membranes, in whichprobe diffusion is 30- to 100-fold slower, could contain gel-phase as well as fluid-phase domains.

The presence of membrane domains is also suggested bythe differential rate of loss of Fl-PE and Rh-PE comparedto LDL bound to the outer membrane. Whereas bothphospholipid analogs are lost with fj=12h (Figs 3 and 4),Dil-LDL remains bound to the surface without loss for 48 h(Bennett and Caulfield, 1991). There must be at least twotypes of domains in the outer membrane, one that bindsLDL and is not shed during in vitro culture and anotherthat does not bind LDL and is cleared relatively rapidly.The former type of domain could serve to retain hostmolecules such as LDL that allow the outer membrane tomimic host membranes antigenically, whereas the lattertype of domain could serve as the vehicle by which parasiteantigens are rapidly shed from the surface.

There are two important results from the studies withLDL. First, the binding of LDL does not change thefractional mobility of Fl-PE, suggesting that the organiz-ation of outer membrane lipids is not disturbed by LDLbinding. Second, neither the rate of loss of Fl-PE and Rh-PE nor the rapid lateral diffusion of Fl-PE changes overthe 12 h that LDL is bound to the surface. This suggeststhat the bound LDL do not contribute directly lipid to theouter tegumental membrane. If LDL-derived lipids wereinserted into the outer membrane, then the diffusioncoefficient would be expected to decrease to a valuecharacteristic of lipid probes in mammalian membranes.Further, the rate of shedding might be expected toincrease in order for the outer membrane to accept theincreased amounts of lipid. These results indirectly favor amodel in which LDL are processed in the gut afteringestion, rather than on the parasite surface. The formermodel has also been suggested by the redistribution ofcarbocyanine dyes throughout the worm from labeled LDLthat were ingested by the parasite (Bennett and Caulfield,1991). At this time, the only clear function of the surface-bound LDL appears to be the masking of parasite antigensfrom host antibodies.

172 J. P. Caulfield et al.

The authors thank Jack Quinn for the preparation of theschistosomula and the maintenance of the schistosome life cycle,and Dr Steven Furlong for discussions during the course of thework. This work was supported by NIH grants AI-23083, HL-32854 and HL-15157.

References

ALECIO, M. R., GOLAN, D. E., VEATCH, W. R. AND RANDO, R. R. (1982).Use of a fluorescent cholesterol derivative to measure lateral mobilityof cholesterol in membranes. Proc. natn. Acad. Sci. U.S.A. 79,5171-5174.

ALLAN, D., PAYARES, G. AND EVANS, W. H. (1987). The phospholipid andfatty acid composition of Schistosonia numsoni and its purifiedtegumental membranes. Molec. biochem. Parasit. 23, 123-128.

AXELROD, D., KOPPEL, D. E., SCHLE3SIN0ER, J., ELSON, E. AND WEBB, W.W. (1976). Mobility measurements by analysis of fluorescencephotobleaching recovery kinetics. Biophys. J. 16, 1055-1069.

BBNNBTT, M. W. AND CAULFIELD, J. P. (1991). Specific binding of humanlow-density lipoproteins to the surface of schistosomula of Schistosomamansom and ingestion by the parasite. Am. J. Path 138(5) (in press).

CHIANG, C. P. AND CAULFIELD, J. P (1989a). The binding of human low-density lipoproteins to the surface of schistosomula of Schistosomamansom is inhibited by polyaruons and reduces the binding of anti-schistosomal antibodies. Am. J Path. 134, 1007-1018.

CHIANG, C. P. AND CAULFIELD, J. P. (19896). Human lipoprotein bindingto Bchistosomula of Schistosoma mansoni: Displacement bypolyanions, parasite antigen masking, and persistence in younglarvae. Am J. Path. 135, 1015-1024.

DERZKO, Z. AND JACOBSON, K. (1980). Comparative lateral diffusion offluorescent lipid analogues in phospholipid multibilayers.Biochemistry 19, 6050-6057.

Dissous, C, Dissous, C AND CAPKON, A. (1981). Isolation andcharacterization of surface antigens from Schistosoma numsonischistosomula. Molec. biochem. Parasit. 3, 215-225.

ESPINOZA, B., TARBAB-HAZDAI, R., SILMAN, I. AND ARNON, R. (1988).Acetylcholinesterase in Schistosoma mansoni is anchored to themembrane via covalently attached phosphatidylinositol. Molec.biochem. Parasit. 29, 171-179.

FOLEY, M., KUSBL, J. R. AND GARLAND, P. B. (1988). Changes in theorganization of the surface membrane upon transformation of thecercariae to schistosomula of the helminth parasite Schistosomamansoni. Parasitology 96, 85-97.

FOLBY, M., MACGRKGOR, A. N., KUSEL, J. R., GARLAND, P. B., DOWNIB,T. AND MOORE, I. (1986). The lateral diffusion of lipid probes in theBurface membrane of Schistosoma mansoni. J. Cell Biol. 103, 807-818.

FURLONG, S. T. AND CAULFIELD, J. P. (1989). Schistosoma mansoni:Synthesis and release of phospholipids, lyBophosphohpids, and neutrallipids by schistosomula. Expl Parasit. 69, 65-77.

GOLAN, D. E., ALECIO, M. R., VEATCH, W. R. AND RANDO, R. R. (1984).Lateral mobility of phospholipid and cholesterol in the humanerythrocyte membrane: Effects of protein-lipid interactions.Biochemistry 23, 332-339.

GOLAN, D. E., BROWN, C. S., CIANCI, C. M. L., FURLONG, S. T. ANDCAULFIELD, J. P. (1986). Schistosomula of Schistosoma mansoni uselysophosphatidylcholine to lyse adherent human erythrocytes andimmobilize erythrocyte membrane components. J. Cell Biol. 103,819-828.

GOLAN, D. E., FURLONG, S. T., BROWN, C. S. AND CAULFIELD, J. P.(1988). Monopalmitoylphosphatidylcholine incorporation into humanerythrocyte ghost membranes causes protein and lipid immobilizationand cholesterol depletion. Biochemistry 27, 2661-2667.

JACOBSON, K., HOU, Y , DERZKO, Z., WOJCIESZYN, J. AND OROANISCIAK,D. (1981). Lipid lateral diffusion in the surface membranes of cellsand in multibilayerB formed from plasma membrane lipids.Biochemistry 20, 5268-5275.

KLAUSNER, R. D. AND WOLF, D. E. (1980). Selectivity of fluorescent lipidanalogues for lipid domains. Biochemistry 19, 6199-6203.

LAZDINS, J. K., STEIN, M. J., DAVID, J R. AND SHER, A. (1982).Schistosoma mansoni: Rapid isolation and purification ofschistosomula of different developmental stages by centrifugation ondiscontinuous density gradients of Percoll. Expl Parasit. 53, 39-44.

LEVI-SCKAFFER, F., TARRAB-HAZDAI, R., ARNON, R. AND SMOLARSKY, M.(1986). A novel lipid-conjugated marker for the direct labeling of theschistosomular membrane. Am. J. trop. Med. Hyg. 35, 544—548.

OWICKI, J. C. AND MCCONNELL, H M. (1980). Lateral diffusion ininhomogeneous membranes: model membranes containing cholesterolBiophys. J. 30, 383-398.

PEARCE, E J., BASCH, P. F. AND SHER, A. (1986). Evidence that the

reduced surface antigenicity of developing Schistosoma mansonischistosomula is due to antigen shedding rather than host moleculeacquisition Parasit Immun. 8, 79-94.

PEABCE, E. J. AND SHER, A. (1989). Three major surface antigens ofSchistosoma mansoni are linked to the membrane byglcosylphosphatidylinositol. J Immun. 142, 979-984.

RAMALHO-PINTO, F. J., GAZZENELLI, G , HOWELLS, R E., MOTA-SANTOS,T. A., FIGUEIHEDO, E. A. AND PELLEGRINO, J. (1974). Schistosomamansoni: Defined system for a stepwise transformation of cercaria toschistosomule in vitro. Expl Parasit. 36, 360-372.

RECKTENWALD, D. J. AND MCCONNELL, H M. (1981). Phase equilibria inbinary mixtures of phosphatidylcholine and cholestrerol. Biochemistry20, 4505-4510.

ROGERS, M. V., HENKLE, K. J., FIDCE, N. H. AND MITCHELL, G. F. (1989).Identification of a multispecific lipoprotein receptor in adultSchistosoma japonicum by ligand blotting analyses. Molec. biochemParasit. 35, 79-88.

ROGERS, M. V., QUILICI, D., MITCHELL, G. F. AND FIDGE, N. H (1990).Purification of a putative lipoprotein receptor from Schistosomajaponicum adult worms. Molec. biochem. Parasit. 41, 93-100.

RUBENSTEIN, J. L. R., SMITH, B. A. AND MCCONNELL, H M. (1979).Lateral diffusion in binary mixtures of cholesterol andphosphatidylcholines Proc. natn Acad. Sci. U.S.A. 76, 15-18.

RUMJANEK, F. D., CAMPANA-PEREIRA, M. A. AND SILVEIRA, A. M. V.(1985). The interaction of human serum with the surface membrane ofschistosomula of Schistosoma mansoni. Molec. biochem. Parasit. 14,63-73.

RUMJANEK, F. D , CAMPOS, E. G. AND ALONSO, L. C. C. (1988). Evidencefor the occurence of LDL receptors in extracts of schistosomula ofSchi8to8oma mansoni. Molec. biochem. Parasit. 28, 145—162.

RUMJANEK, F. D. AND MCLAREN, D J (1981). Schistosoma mansoni:Modulation of schistosomular lipid composition by serum. Molec.biochem Parasit. 3, 239-252.

RUMJANEK, F D., MCLAREN, D. J. AND SMITHERS, S. R. (1983) Serum-induced expression of a surface protein in schistosomula ofSchistosoma mansoni: a possible receptor for lipid uptake. Molec.biochem. Parasit. 9, 337-350.

SAMUELSON, J C. AND CAULFIELD, J. P. (1982). Loss of covalentlylabeled glycoproteins and glycolipids from the surface of newlytransformed schistosomula of Schistosoma mansoni. J. Cell Biol. 94,363-369

SAMUELSON, J. C, CAULFIELD, J. P. AND DAVID, J. R. (1982).Schistosomula of Schistosoma mansoni clear concanavalin A fromtheir surface by sloughing. J. Cell Biol. 94, 355-362.

SAMUELSON, J C , SHER, A. AND CAULFIELD, J. P. (1980) Newlytransformed schistosomula spontaneously lose surface antigens andC3 acceptor sites during culture. J. Immun. 124, 2055-2057.

SAUMA, S. Y AND STRAND, M (1990). Identification and characterizationof glycosylphosphatidylinoaitol-linked Schistosoma mansoni adultworm lmmunogens. Molec. biochem. Parasit. 38, 199-210

SAUNDERS, N., WILSON, R. A. AND COULSON, P. S. (1987). The outerbilayer of the adult schistosome tegument surface has a low turnoverrate in vitro and in vivo. Molec. biochem. Parasit. 25, 123-131.

SMITH, L. M., PABCE, J. W., SMITH, B. A. AND MCCONNELL, H. M. (1979).Antibodies bound to lipid haptens in model membranes diffuse asrapidly as the Hpids themselves. Proc. natn. Acad. Sci. U.S-A. 76,4177-4179.

STOLPEN, A. H., BROWN, C. S. AND GOLAN, D. E. (1988). Characterizationof microscope laser beams by two-dimensional scanning offluorescence emission. Appl. Opt. 27, 2661-2667.

TAYLOR, D. W., HAYUNGA, E. G. AND VANNIER, W. E. (1981). Surfaceantigens of Schistosoma mansoni. Molec. biochem. Parasit. 3, 157-168.

TORPIER, G. AND CAPRON, A. (1980). Differentiation and expression ofantigen with S. mansoni membrane structure. In The Host InvaderInterplay (ed R. van den Bossche), pp. 143-146 Amsterdam: Elsevier.

WEISS, J. B., MAGNANI, J. L. AND STRAND, M. (1986). Identification ofSchistosoma mansoni glycolipids that share immunogeniccarbohydrate epitopes with glycoproteins. J. Immun. 136, 4275-4282.

Wu, E.-S., JACOBSON, K. AND PAPAHADJOPOULOS, D. (1977). Lateraldiffusion in phospholipid multibilayers measured by fluorescencerecovery after photobleaching. Biochemistry 16, 3936-3941.

YECHIEL, E. AND EDIDIN, M. (1987). Micrometer-scale domains infibroblast plasma membranes. J. Cell Biol. 105, 755-760.

YOUNG, B. W. AND PODESTA, R. B. (1986). Complement and 5-HTincrease phosphatidylcholine incorporation into the outer bilayers ofSchistosoma mansom. J. Parasit. 72, 802-803.

[Received 25 October 1990 - Accepted, in revised form, 9 January 1991)

LDL binding to schistosome membranes 173