Biol. Chem. Hoppe-SeylerVol. 371, pp. 159-165, February 1990

Interaction of Lipophorin with Íhe Plasma Membrane ofLocust FlighÍ Muscles

Rik VeN ANrwBnpBN, Jules BBsrwII-nEn, Miranda C. VeN HeusDEN,

Dick J. VaN opr Honsr andAd M.Th. BErNarrEns

Department of Experimental Zoology, University of Utrecht

(Received 2 June 1989)

Summary: Binding of high-density lipophorin(HDLp) to a plasma membrane preparation of locustflight muscle tissue was studied using a radiolabelledligand binding assay and ligand blotting techniques.Analysis at 33 'C of the concentration-dependenttotal binding of tritium-labelled HDLp ([3H]HDLp)to the membrane preparation revealed the presenceof a single specific binding site with an equilibriumdissociation constant of Ko : 9 (12) x 10-7u and a

maximal binding capacity of 84 (+ 10) ng x (g,g pro-tein)-1. Unlabelled HDLp as well as unlabelled low-

density lipophorin (LDLp) competed with13H1HOI-p for binding to the identified binding site.In addition, ligand blotting demonstrated that bothHDLp and LDLp bind specifically to a 30-kDa pro-tein in the plasma membrane preparation, suggestingthe involvement of this protein in the binding oflipophorins to the isolated membranes.A possible relationship between the identified bind-ing of lipophorins and the observed co-purification oflipophorin lipase activity with the plasma membranesis discussed.

Wechselwirkung von Lipophorin mit der Plasmamembran des

H eus chr e c k en- F lugmus ke ls

Zusammenfassung: Die Bindung von Lipophorin ho-her Dichte (HDLp) an eine Plasmamembran-Pràpa-ration aus Heuschrecken-Flugmuskelgewebe wurdemittels eines Bindungsassays mit radiomarkierten Li-ganden und mittels einer Liganden-Blotting-Technikuntersucht. Eine Analyse der konzentrationsabhàn-gig insgesamt bei 33'C an die Membran-Práparationgebundenen Menge an Tiitium-markiertem HDLp([3H]HDLp) ergab eine Gleichgewichts-Dissozia-tionskonstante Ko : 9 (X 2) x 10-7M und eine maxi-male Bindungskapazitát von 84 (+ 10) ng x (p,g pro-tein)-1. Unmarkiertes HDLp sowie auch unmarkier-

tes Liphophorin niedriger Dichte (LDLp) konkur-rierten mit fH]HDLp um die Bindung an diese iden-tifizierte Bindungsstelle. Das Liganden-Blottingzeigte auBerdem, daB sowohl HDLp als auch LDLpspezifisch an ein 30-kDa-Protein in der Plasmamem-bran-Pràparation gebunden werden, woraus eine Be-teiligung dieses Proteins an der Bindung von Lipo-phorinen an isolierte Membranen zu schlieBen ist.Eine mógliche Beziehung zwischen dieser Bindungvon Lipophorinen und der beobachteten Kopurifika-tion von Lipophorin-Lipase-Aktivitát mit den Plas-mamembranen wird diskutiert.

Key words: Locust flight muscle, plasma membrane, lipoprotein binding site, lipoprotein lipase.

Abbrevintions:AKH, adipokinetic hormone; GAR-PO, peroxidase-conjugated goat anti rabbit antibody; HDLp, high-density lipophorin; Hepes,4-(2-hydroxyethyl)-1-piperazinethanesulfonic acid; LDLp, low-density lipophorin; PAGE, polyacrylamide gel electrophoresis;PBS, phosphate-buffered saline; PCMPS, 4-chloromercuriphenylsulfonic acid; PMSF, phenylmethanesulfonyl fluoride; SDS,sodium dodecyl sulphate; TMB, tetramethylbenzidine.

Copyright @ by Walter de Gruyter & Co . Berlin . New York

160 R.VanAntwerpen, J. Beekwilder, M.C.Van Heusden, D.J.Van der Horst andA.M.T. Beenakkers Vol. 371 (1990)

In migratory locusts, a high density lipoprotein (highdensity lipophorin, HDLp, see ref.tll) serves to trans-port lipids between sites of lipid absorption, storageand utilization (for reviews see refs.[2'3];.|Ihe lipopho-rin, which contains two apoproteins (apolipophorinI and II) with molecular masses of - 250000 and

- 80000 Da, respectivelyl4-61, has been suggested tofunction as a reusable, lipid-distributing shuttletTl.

During sustained flight, an adipokinetic hormone(AKH) induces the loading of circulating HDLp withincreased amounts of diacylglycerol from the fat bodyand the concomitant association with the lipophorinof several molecules of an additional apoprotein fromthe haemolymph (apolipophorin III)t5'8'eJ. Thus,HDLp is converted to a low-density lipophorin(LDLp, see ref.[1]), which transports the increasedamounts of diacylglycerol to the flight muscles to fuelflight muscle activity. Upon hydrolysis of thediacylglycerol by flight muscle lipophorin lipase activ-ity[10'11], the HDLp and apolipophorin III moieties ofLDLp reappear in the haemolymph and can be recon-verted to LDLp at the fat bodyttzl.

Association of lipophorin lipase activity with plasmamembranes of the flight muscles has been de-monstrated[13]. Moreover, immunocytochemicallocalization of lipophorins in locust flight muscles in-dicated that both at rest and during flight, lipophorinsare restricted to the extracellular matrix of the tis-sueltol. These results suggest that the unloading oflipophorins at the flight muscles is an extracellularevent. In addition to the studies on flight musclelipophorin lipase, binding of HDLp to a putative re-ceptor in the plasma membranes of flight muscle cellshas been describedll5l. A possible relationship be-tween the identified binding of HDLp to flight muscleplasma membranes and lipophorin lipase activity,however, remains to be established.

In the present study we further investigated the in-teraction of locust lipophorins with flight muscleplasma membranes. Both HDLp and LDLp wereshown to bind to a single binding site in a preparationof isolated membranes. In addition, ligand blottingdemonstrated the presence of a 30-kDa lipophorin-binding protein. The latter result suggests the involve-ment of this protein in the binding of lipophorins tothe isolated membranes.

Materials and Methods

Animals

Locusta migratoria were reared in the laboratory under crowdedconditions as describedlr6l. Adult male locusts (12th day after theimaginal ecdysis) were used in all experiments.

Lipophorins

HDLp was isolated from the haemolymph of resting locusts by gel

filtration chromatography and potassium bromide gradient densityultracentrifugation as described earlierl17]. LDLp was purified fromthe haemolymph of AKH-injected locusts (this haemolymph wascollected 90 min after the injection of 10 pmolAKH per animal) bypotassium bromide gradient density ultracentrifugation, essen-

tially according to Shapiro et al.[18] as described byVan der Horst etal.tr7l. Lipophorin protein concentrations were determined accord-ing to Schacterle and Pollackllel. To calculate the molarity ofHDLp, the recently determined molecular masses of apolipopho-rin I and II of 217000 and 72000 Da, respectively (D.J. Van derHorst, J.Voshol, J.M.Van Doorn andA.M.Th. Beenakkers, to bepublished), were used.To calculate the molarity of LDLp, it was as-

sumed that one LDLp particle is formed from one HDLp particleand 14 apolipophorin III molecules. [The ratio of apolipophorin I:apolipophorin II: apolipophorin III in locust LDLp is 1:1: - 14

(D.J. Van der Horst, J. Voshol, J.M. Van Doorn and A.M.Th.Beenakkers, to be published).]HDLp was radiolabelled in the pro-tein moiety with N-succinimidyl [2,3-3H]propionate (AmershamIntemational) to a specific activity of 2000-5000 dpm x pg 1, es-

sentially according to the method of Bolton and Hunter[2o], as de-scribed byVan der Horst et al.l5l.

P las ma me mbrane prepa ratio n

The dorsolongitudinal flight muscles of 30 animals (1.8 g wetweight) were excised and washed with Hepes buffer (lOmrr.r Hepes,150mn NaCl, 10mr,r KCl, 4mu CaCl2 x 2H2O,2mu MgCl2 x 6

H2O, pH 7.0) at 0 "C. The tissue was homogenized in25 mi Hepesbuffer using a motor-driven glass-teflon homogenizer, with a radialclearance of 0.095 mm, at a speed of 150 r.p.m. and with 5 strokes.Hepes buffer was added to the homogenate to a final volume of 35

m/, after which the preparation was centrifuged for 15 min at 16000x g.The supernatant of this step was analysed as rest fraction Rl;the pellet was re-suspended in 30 m/ of Hepes buffer and cen-trifuged for 10 min at 1000 x g. The pellet of this step was re-sus-pended in 25 ml of Hepes buffer and analysed as rest fraction R2;the supematant was centrifuged for 20 min at 18000 x g.The pelletof this final step was re-suspended in 2 m/ ofTiis buffer (40mu, pH7.4) and used as the plasma membrane preparation (M). To deter-mine the relative amounts of mitochondrial, sarcoplasmatic andplasma membranes present in the different fractions, succinate de-hydrogenase, Mg2@-dependent Ca2@-stimulatedATPase and Mg2@-

dependent (Na -l K)-stimulatedAlPase, respectively, were used as

marker enzymes, as described by Narahara et al.{211. Specific ac-tivities of these enzymes in the rest fractions (R1 and R2) and in theplasma membrane preparation (M) are presented in the table.Thetable shows that there is only a very small increase in the specific ac-

tivities of the mitochondrial and sarcoplasmatic marker enzymes,when the pellet-fraction R2 is compared to the plasma membranepreparation M. However, the 5.5-fold increase in the specific activ-ity (Na + K)-stimulated ATPase demonstrates the enrichment of

Table. Specific enzyme activities in flight muscle tissue fractions.

Rl: Supernatant of the first centrifugation step. R2: Pellet of thesecond centrifugation step. M: Plasma membrane preparation.Values are the means of triplicate determinations (t SEM).

R1 R2 M

Succinatedehydrogenaseu

Ca-AIPaseb(Na-K)-ATPaseb

1.78 + 0.45

1.45 + 0.10

6.64 + 0.26

0.76 + 0.104.30 + 0.07

7.30 + 0.58

1.01 + 0.1023.80 + 1.90

" Absorption at 485 nm per g,g protein.b mmol NADH converted per 90 s per liter.

Vol. 371 (1990) Lipophorin - Flight Muscle Plasma Membrane Interaction 16l

I<:_

oc!X

-i-cXE

E

E0)Oàan

o

1000

500

50 100 .l50 200Plasma membrane protein lpg/200 pll

Fig.1. Lipophorin lipase activity in flight muscle plasma mem-branes, presented as a function of membrane protein concen-tration.

LDLp, containing ['aC]glycerol-labelled diacylglycerol, was

used as a substrate at a concentration of 1 mg x m/ l. Lipase ac-tivity was measured by determination of the amount of liber-ated [laC]glycerol.Values are indicated as the mean (+ SEM) oftriplicate determinations.

plasma membranes in fraction M. Earlier studies on flight musclelipophorin lipase indicated that most of this enzyme is associatedwith the flight muscle plasma membrane[1113]. Fig. l shows lipopho-rin lipase activity as a function of protein concentration in fractionM, detected as described under "iipophorin lipase assay", usingLDLp as a substrate. In a homogenate of flight muscle tissuemaximum lipase activity is - 3.5 nmol of diacylglycerol hydrolysedper min per mg of tissue proteinlttl. In fraction M maximum lipaseactivity was calculated to be - 10 nmol x min-1 x mg 1 of mem-brane protein, indicating enrichment of lipophorin lipase in theplasma membrane fraction (M). Fig. 1 shows that lipase activityreaches a plateau value at - 50 g,g ofmembrane protein per incuba-tion. Since substrate LDLp was presenr in excess amounts it issuggested that lipase activity is inhibited by products of the enzyma-tic reaction (e.g. free fatty acids, glycerol and/or HDLp). Similarcharacteristics of lipophorin lipase activity have been de-monstrated earlier[11].

Although the results described in the present study may be influ-enced by contamination of the pÍepaíation M with mitochondrialand/or sarcoplasmatic membranes, both the enrichment of plasmamembrane marker enzymes and the nature of the observed resultsindicate that the interactions described are between iipophorinsand flight muscle plasma membranes.

Lipophorin lipase assay

Lipophorin lipase activity was determined by the method ofWheeler et al.t10l with isolated LDLp as a substrate. Diacylglycerolin the LDLp substrate was labelied in vivo by incorporation of[1aC]glycerol and lipophorin lipase activity was measured by deter-mination of the amount of liberated [laC]glycerol.

Lipophorin-binding assay

For the lipophorin-binding assay, the plasma membrane pellet wasredissoived in Hepes buffer instead of Tiis buffer. Plasma mem-branes were incubated (at a concentration of 1.0-3.5 mg of mem-brane protein per m/), in 100 pri of Hepes buffer, with various con-centrations of [34HOI-p, in the absence or presence of unlabelled

HDLp. All incubations were carried out at 33'C in a shaking water-bath. Incubations were stopped by leading the various samplesover separate filters (Whatman GF/C, retention > 1.2 pm), using a

Brandel Cell Harvester (Brandel Inc. U.S.A.). Before use, thesefiiters were incubated for 24 h in Hepes buffer containing 0.5%Tween 20 to block non-specific HDLp binding sites. Membraneswere retained by the filters, while free lipophorins could passthrough them. Subsequently, lipophorin-free Hepes buffer was ledthrough the filters in three wash steps of - 5 s each, to remove re-sidual free lipophorins. Filters containing the incubated mem-branes were dissolved in Emulsifier Safe (Packard) and the radioac-tivity that was retained by the filters was counted by liquid scintilla-tion spectrometry.

Equilibrium binding data were analysed as described by Mendeland Mendelt22l. Total binding (specific plus non-specific binding)was fitted to a mathematical model of a single receptor plus linearnon-specific binding, according to the equation:

B: B^u^ -4- x,ltl,1(d + [L]

in which B is the total amount of [3H]HDLp binding, B*", is themaximal binding of [3H]HDLp to rhe receptor, [l] is the concentra-tion of [3H]HDLp and K6 is the equilibrium dissociation constant;the second term in the equation represents non-specific binding,where Kn is an equilibrium constant.

Ligand blotting

For ligand blotting, 1007.r.u of the protease inhibitors phenyl-methanesulfonyl fluoride (PMSF) and 4-chloromercuriphenylsul-fonic acid (PCMPS) were added routinely to all buffer solutionsused during the fractionation of flight muscle tissue. However, theuse of these protease inhibitors did not seem to affect the size oramount of lipophorin-binding proteins detected in the ligand blots.Ligand blotting was performed essentially according to Daniel eta1.t231. Samples taken during the membrane preparation proce-dures were incubated at room-temperature with 2.0% sodiumdodecyl sulphate (SDS).The proteins of these samples were frac-tionated by SDS polyacrylamide gel electrophoresis (SDS-PAGE)on 7-30Yo gradient gels and electroblotted onto nitrocellulosepapeÍ, as described by Schulz et al.[2a]. Blots were incubated for 30min with phosphate-buffered saline (PBS) containing 5% normalgoat serum to block non-specific protein binding sites. Sub-sequently, the blots were incubated with lipophorins at a concentra-tion of 100 prg x m/-1 (HDLp) or 189 pg x m/-1 GDLp) in pBS.Free lipophorins were removed by 3 washes of 10 min each in PBScontaining 0.5% Tween 20. Next, the blots were incubated for 60min with polyclonal antiserum against HDLp (rabbit antiserum, di-luted 1:3000 in PBS), washed with PBSflween 20 to remove freeantibodies and incubated for 30 min with polyclonal, peroxidase-conjugated, goat-anti-rabbit antibodies (GAR-PO) in PBS. Afterremoval of free GAR-PO antibodies, bound lipophorins were vis-ualized with a tetramethylbenzidine (TMB) reagent.TMB reagentcontained 12 mgTMB (Sigma, St. Louis, MO) and 40 mg dioctylsul-fosodiumsuccinate (Sigma), dissolved at 55 "C in 5 m/ ethanol(96%) and 15 m/ citrate/phosphate buffer (7.6mv citric acid and27mu NaH2POa, pH 6.5) containing 0.003o/o H2O2.

Results

Time-course of f HIHDLp binding

The time-course of 13U1Hnf-p binding to the plasmamembrane preparation was studied over a period of3 h (Fig. 2). Both total binding and binding that wasnot displaceable by a 20-fold excess of unlabelledHDLp, were determined. A rapid increase in the dis-

0

162 R.VanAntwerpen, J. Beekwilder, M.C.Van Heusden, D.J.Van der Horst andA.M.T. Beenakkers Vol. 371 (1990)

I

c'o)Èo_q)

Xo)cc

o.JoI-c-

-b Bo

c)

Eo)

*60c')ÉEcdaoJoIE20

100

oó

lncubation time [min]

Fig.2. ïme course of 13ff1HOr-p binding to flight muscleplasma membranes.

Binding was determined at 33 "C and at a [3H]HDLp concentra-tion of 770 g.g x m/ r, in the absence (O) or presence (A) of a20-fold excess of unlabelled HDLp. The protein concentrationof the membrane preparation was 245 p,g x ml-t. [3HjHDLpbinding that was displaceable by the excess unlabelled HDLp(O) was determined by subtracting non-displaceable binding(A) from the total binding (O). Values are indicated as themean of duplicate determinations.

placeable binding of [3H]HDLp was observed in thefirst 5 min of the incubation. Equilibrium of this bind-ing was established after approximately t h. Betweent h and 3 h of incubation time, displaceable

[3U]UOt-p binding slightly decreased. Non-displace-able binding of fUlUOf-p remained constant be-tween 2 min and 3 h of incubation time.

Saturation of specific 73a1Anfp binding

Total binding data were obtained by incubation of theplasma membrane preparation with increasing con-

centrations of [3H]HDLp in the absence of unlabelledHDLp (Fig. 3). A Scatchard plot of the observed

[3U]UOf-p binding is presented in Fig. 4.The concave

downward curve of the Scatchard plot indicates thepresence of two independent types of binding: 1. rela-tively high affinity binding with a low capacity, and2.relatively low affinity binding with a high capacity. Aspointed out by Mendel and Mendellz2), the low affin-ity component of the total binding appears to belinear over a long interval of concentrations and can

be regarded as "non-specific" binding. Therefore,total binding of [3H]HDLp to the plasma membranepreparation was analysed with a mathematical modelof a single specific binding site, with linear non-specif-ic binding.Thus, for the specific binding of [3U]HDLp

200

150

100

50

oo1234t

fHlHol-p tree [mg x mll]

Fig.3. Equilibrium binding isotherm of [3H]HDLp binding toflight muscle plasma membranes.

The membrane preparation was incubated for t h at 33 'C withincreasing concentrations of [3H]HDLp. Only total bindingwas measured (O). Specific and non-specific binding (1 and 2,

respectively) were determined by fitting total binding data to amathematical model of one binding site plus linear non-specificbinding. Values are indicated as the mean of duplicate determi-nations.

0 50 100 150 200

t3nlHOLp bound [ng x (Áig protein)-1]

Fig.4. Scatchard plot of 13H1ffOI-p binding to flight muscleplasma membranes.

O:Total binding dataof Fig. 3. Plots of the specific and non-spe-cific [3H]HDLp binding components (1 and 2, respecively)were derived from fitting total binding data to a mathematicalmodel of one binding site plus linear non-specific binding.

an equilibrium dissociation constant of Kd : 9 (t 2)x 107lra and a maximal binding capacity of 84 (+ 10)

ng x /rg-1 protein were derived. This binding satu-rated at a [3ff]HOf-p concentration of approximately2 mg x m/-1. Fitting of the total binding data to a

E360o

c440d]

Vol.371 (1990) Lipophorin Flight Muscle Plasma Membrane lnteraction t63

mathematical model of two specific binding sites plus

non-specific binding did not resolve a second specific

binding site from the total amount of [3H]HDLp bind-ing measured.

C omp etition exp eriments

In competition experiments plasma membranes were

incubated with a fixed concentration of [3U]UOLp inthe absence or presence of increasing concentrationsof unlabelled HDLp or LDLp (Fig. 5). Both unlabel-led HDLp and LDLp effectively displaced 13fflffOI-pbinding, to a level that approaches non-specific bind-ing, demonstrating that unlabelled HDLp as well as

unlabelled LDLp bind to the identified binding site.

Ligand blotting of plasma membrane proteins

Fig. 6 shows ligand blots of the plasma membranepreparation. In the control, the incubation withlipophorins was omitted and nitrocellulose strips con-

taining the plasma membrane proteins were incu-bated only with polyclonal antiserum against HDLpand GAR-PO antibodies, followed by peroxidase

staining (lane 1). The control demonstrated the pre-sence of both apolipophorin I and II in the plasma

membrane preparation, suggesting that HDLp is as-

sociated with the membranes in vivo. ApolipophorinI penetrated the gel very slowly and was detected onlyafter prolonged electrophoresis and after extensive

blotting.

1 23 4 kDa

Fig.6. Ligand blots oÏ lipophorins binding to plasma mem-brane proteins.

Proteins were fractionated by SDS-PAGE, electroblotted ontonitrocellulose paper and incubated with lipophorin free Hepesbuffer (lane 1), with 100 g.g x m/ | HDLp (lane 2) or with 189

g.g x m/ I LDLp (lane 3). Binding of lipophorins to the blottedproteins was visualized by immunochemical staining withpolyclonal antibodies against HDLp. Lipophorin-binding pro-teins which were not always detected are indicated with an as-

terisk. Lane 4 shows the electroblotted proteins from theplasma membrane preparation. stained withAmido Black 108.

The binding of HDLp and LDLp to blotted plasmamembrane proteins is shown in lanes 2 and3, respec-tively. Some of the protein bands that bind lipophorinwere not always clear, or could not always be de-tected. These proteins are indicated with an asterisk.However, a 30-kDa lipophorin-binding protein was

detected consistently. In addition, most of the time a

smaller lipophorin-binding protein with a molecularmass of - 15000 Da was detected.

Discussion

In the present study, specific binding of [3H]HDLp toa single binding site in a preparation of locust flightmuscle plasma membranes is described. At thephysiological temperature of 33 oC, saturation of this

apoL p- |

apolp - ll

t>,'5ri!,:

>-.m

. --r.1

-94

-67

-43

-30

llllli;;,,,',t,,ll - ?0

.1. .,,..

g80.EctË60o

840:Jo

_oo_I

?20TP-

o01234

Concentr. of unlabelled lipophorin [pv]

Fig.5. Displacement of [3U1ffOlp binding by increasing con-

centrations of unlabelled lipophorins.

Flight muscle plasma membranes (1 mg protein per m/) were in-cubated for t h at 33 "C with 100 p-g x mi-r 13H]HDLp in the ab-

sence oÍ presence of the indicated concentrations of unlabelledHDLp (O) or LDLp (O). Values are indicated as the mean ofduplicate determinations.

164 R.Van Antwerpen. J. Beekwilder, M.C.Van Heusden, D.J.Van der Horst andA.M.T. Beenakkers Vol. 371 (1990)

binding is reached at a [3H]HDLp concentration of

- 2 mg x m/-1 (Fig. 3), which means that the iden-tified binding site can be saturated by lipophorin con-centrations that normally occur in the haemolymph.For example, the HDLp concentration in the blood ofadult locusts at rest is approximately 10 mg x m/-1

(unpublished observation). This high physiologicalHDLp concentration may explain the relatively highK6 value of the identified lipophorin binding (Ko :9 x 10-7u) in comparison with the K6 values of lipo-protein - receptor interactions in vertebrater[2s'26J. 1n

insects, the affinity of lipoproteins for a particularphysiological binding site does not need to be as highas in vertebrates, where lipoprotein concentrations inthe blood are much lower. The affinity of the bindingsite detected in the present study is lower than the af-finity of a similar binding site detected byHayakawalisr (n : - 7.4 x 10 7u).This discrepancymay be caused by the fact that in the latter study bind-ing of HDLp was determined at a temperature of 4'Cinstead of 33 "C.

Competition experiments demonstrated that unlabel-led LDLp can displace ;3H1Hnlp from the identifiedbinding site. It should be noted that the lipase activityin the plasma membrane preparation can convertLDLp to HDLp and free apoLp-III molecules, mean-ing that the displacement of [3H]HDLp by LDLp ac-

tually may be a displacement by the lipase productHDLp. However, when lipase activity was assayed atconcentrations of LDLp higher than approximately 1

mg x m/-1, less than l0% of the added amount ofradiolabelled diacylglycerol in LDLp was measuredto be hydrolysed. This observation is in agreementwith earlier studies, which also showed that during invitro incubations of lipophorin lipase activity withLDLp only part of the added LDLp is converted toHDLptll'121. This indicates that in the competition ex-

periments of Fig. 5, where high concentrations of un-labelled LDLp are used, most of the added LDLpcompetes with fH]HDLp as LDLp.The observed dis-placement of [3H]HDLp therefore, seems to indicatethat both HDLp and LDLp bind to the identifiedbinding site. Nevertheless, since the interpretation ofbinding data is more complicated when LDLp is usedas a ligand, we used radiolabelled HDLp to charac-terize the binding site.

Both HDLp and LDLp were demonstrated to bind tospecific proteins in ligand blots of flight muscleplasma membrane proteins. The clearest and mostconsistently detected binding of lipophorins in theligand blots was to a protein that migrates in the gra-dient gel with an apparent molecular mass of - 30000Da. In addition to this protein, other lipophorin-bind-ing proteins were detected, but the amounts of these

proteins present in different plasma membrane prepa-rations varied. It is unlikely that the smallest lipopho-rin-binding protein (- 15 kDa) is a degradationalproduct of the 30-kDa protein, since the buffers used

in the membrane purification procedures containedprotease inhibitors. Ligand blotting has been de-monstrated to be a useful technique to visualize lipo-protein receptor interactions in vertebrates[23'27-2e].

The 30-kDa-lipophorin-binding protein detected inthe ligand blot in Fig. 6 may therefore represent a

physiologically important binding site and in additionmay represent the binding site that is detected in thelipophorin binding assay.

Earlier studies on flight muscle lipophorin lipase indi-cated that most of the enzyme activity is associatedwith the plasma membrane[11'13]. In addition, im-munocytochemical localization of lipophorins inflight muscle tissue indicated that both at rest and dur-ing flight, lipophorins are restricted to the extracellu-lar matrix of the tissuella]. These results suggest thatthe unloading of lipophorins at the flight muscles is anextracellular event. The co-purification of lipophorinlipase activity with the plasma membranes, both inearlier studies[11'13] and in the present study (Fig. 1), isin agreement with these observations. Since flightmuscle lipophorin lipase can hydrolyse bothdiacylglycerol that is transported by LDLp and rela-tively srnall amounts of diacylglycerol that are trans-ported by HDLp, (although the latter reaction pro-ceeds at a lower rate)[11'13], hpid-loaded and lipid-de-pleted lipophorins compete for the interaction withflight muscle lipophorin lipase. Therefore, the iden-tified binding of HDLp and LDLp to a common bind-ing site on flight muscle plasma membranes and tocommon binding proteins in ligand blots of theplasma membrane preparation may be related tolipophorin lipase activity. However, it remains possi-ble that the two types of lipophorin - membrane in-teraction (the detected binding and the observedlipase activity) serve different physiological func-tions: Binding of lipophorins to the identified bindingsite may be related to the transport of structural lipids(e.g. cholesterol) to the flight muscle membranes,whereas lipophorin-lipase activity is known to be re-lated to the transport of fatty acids to fuel flight mus-cle activitytll. Further characÍerization of the lipopho-rin-binding proteins, especially the 30-kDa-lipopho-rin-binding protein, promises to provide more infor-mation on the lipophorin-mediated transport of lipidsto locust flight muscles.

The authors thank Dr. T.K.E Schulz for providing the anti-serum against HDLp and for stimulating discussions.

Vol.371 (1990) Lipophorin - Flight Muscle Plasma Membrane Interaction

References 15 Hayakawa,Y. (1987) Bíochim. Bíophys. Acta9l9,58-63.16 Van der Horst, D.J., Baljet, A.M.C., Beenakkers,

1 Beenakkers, A.M.Th., Chino, H. & Law, J.H. (1988) In- A.M.Th. &Van Handel, E. (1918) Insect Biochem. S, 369-sect Bíochem. 18. I-2. 373.

2 Beenakkers, A.M.Th., Van der Horst, D.J. & Van Mar- 17 Van der Horst, D.J., Beenakkers, A.M.Th., Van Doorn,rewijk,W.J.A. (1985) Prog. Lipid Res. A, 19-67. J.M., Gerritse, K. & Schulz,T.K.F, (1987) lnsect Biochem.

3 Shapiro, J.P., Law, J.H. &Wells, M.A. (1988),4nnu. Rev. 1Z 799-808.Entomol.33,297-318. 18 Shapiro, J.P., Keim, P.S. & Law, J.H. (1984) J. Biol.

4 Chino, H. & Kitazawa, K. (1981) J. Lipid Res. 22,1042- Chem. 259,3680-3685.1052. 19 Schacterle, G.R. & Pollack, R.L. (7973)Analyt. Biochem.

5 Van der Horst, D.J., Van Doorn, J.M. & Beenakkers, 51,654-655.A.M.Th. (1984) Insect Biochem. 14,495-504. 20 Bolton, A.E. & Hunter, W.M. (1973) Biochem. J. 133,

6 Kashiwazaki, Y & Ikai, A. (1985) Arch. Biochem. 529-539.Biophys.237, 160-169. 21 Narahara,H.,Vogrin,V.,Green,J.,Kent,R.&Gould,M.

7 Downer, R.G.H. & Chino, H. (1985) Insect Biochem. 15, (1979) Biochim. Biophys. Acta 552,247-267.627-630. 22 Mendel, C.M. & Mendel, D.B. (1985) Biochem. 1.228,

8 Mwangi, R.W. & Goldsworthy, G.J. (1977) J. Comp. 269-272.Physiol.774,l77-190. 23 Daniel, T.O., Schneider, W.J., Goldstein, J.L. & Brown,

9 Van der Horst, D.J., Van Doorn, J.M. & Beenakkers, M.S. (i983) I. Bíol. Chem.258,4606-4611..A.M.Th. (1979) Insect Bíochem.9,627-635. 24 Schulz, T.K.F., Van der Horst, D.J., Amesz, W.J.C. &

10 Wheeler, C.H., Van der Horst, D.J. & Beenakkers, Beenakkers, A.M.Th. (7987) Arch. Insect Biochem.A.M.Th. (1984) Insect Biochem. 14,261-266. Physíol. 6,97-10'7.

11 Van Heusden, M.C., Van der Horst, D.J., Van Doorn, 25 Goldstein, J.L., Basu, S.K. & Brown, M.S. (1983)J.M., Wes, J. & Beenakkers, A.M.Th. (1986) Insect MethodsEnzymol.9S,24l-260.Biochem. 16,517-523. 26 Innerarity, T.L., Pitas, R.E. & Mahley, R.W (1986)

12 Van Heusden, M.C., Van der Horst, D.J., Voshol, J. & Methods En4ymol.129,542-565.Beenakkers,A.M.Th. (1987) 1n sect Biochem.17,77I-776. 27 Brown, S.A.,Via, D.P., Gotto, A.M., Jr., Bradley,W.A. &

13 Wheeler, C.H. & Goldsworthy, G.J. (1985) Bíol. Chem. Gianturco, S.H. (1986) Bíochem. Biophys. Res. Com-Hoppe-Seyler3ffi, 1071.-f077. mun. 139,333-340.

14 Van Antwerpen, R., Linnemans, W.A.M., Van der Horst, 28 Koller, E. (1986) FEBS Lett. 2,00,97 -f02.D.J. & Beenakkers, A.M.Th. (1988) Cel/ Tissue Res.252, 29 Wade, D.P., Knight, B.L. & Soutar, A.K. (1986) Eur. J.66I-668. Biochem.lsg. 333-340.

J. Beekwilder, D.J. Van der Horst* and A.M.Th. Beenakkers, Department of Experimental Zoology, University of Utrecht,8 Padualaan, NL-3584CH Utrecht;

Present address: R. Van Antwerpen and M.C. Van Heusden, Department of Biochemistry,Biological SciencesWest, University of Arizona,Tuscon, AZ-85721, U.S.A.

* To whom correspondence should be addressed.

165