Proc. Natl. Acad. Sci. USAVol. 75, No. 11, pp. 5301-5305, November 1978Biochemistry

Effect of thyroid phospholipids on the interaction of thyrotropinwith thyroid membranes

(hormone receptors/gangliosides/phospholipase A/adenylate cyclase)

FAUSTA OMODEO-SALE*, ROSCOE 0. BRADY, AND PETER H. FISHMANDevelopmental and Metabolic Neurology Branch, National Institute of Neurological and Communicative Disorders and Stroke, National Institutes of Health,Bethesda, Maryland 20014

Contributed by Roscoe 0. Brady, August 10, 1978

ABSTRACT Various lipids extracted from bovine thyroidglands were tested for their ability to affect the binding of-25I-labeled thyrotropin to bovine thyroid membranes. The mostpotent inhibitors were the acidic phospholipids in the ordercardiolipin > phosphatidylglyceroF> phosphatidylinositol >>phosphatidylserine. Other phospholipids, neutral lipids, andneutral glycolipids were ineffective. As reported previously[Mullin, B. R., Pacuszka, T., Lee, G., Kohn, L. D., Brady, R. 0.& Fishman, P. H. (1978) Science 199, 77-791 thyroid gangliosidesalso inhibited thyrotropin binding but not as effectively asphospholipids. In addition, the mode of action of these twoclasses of acidic lipids was different. When thyroid membraneswere preincubated with the phospholipids and then separatedby centrifugation, their ability to bind thyrotropin was still di-minished. In contrast, gangliosides appear to interact with thehormone and not with the membranes. The effect of phospho-lipids on thyroid membranes was further examined by incu-bating the membranes with phospholipase A. The treatedmembranes now bound more labeled hormone. These resultssuggest that certain acidic phospholipids, which are present inonly small amounts in thyroid membranes, influence the stateof the thyrotropin receptor and its ability to bind thyrotro-pin.

The initial event in thyrotropin (TSH)-mediated effects onthyroid cells is the binding of the hormone to specific receptorson the plasma membrane of the cells. Subsequently, a trans-membrane signaling event occurs, which leads to the generationof a second messenger inside the cell. For TSH, this secondmessenger is cyclic AMP (1), and transmembrane signalingcauses the activation of adenylate cyclase by mechanisms notyet understood. Recently, it has been suggested that gangliosidesare involved in this process (see ref. 2). Thus, gangliosides (3)and, in particular, gangliosides from thyroid tissue (4) inhibitthe binding of TSH to thyroid membranes. There are certainstructural similarities between TSH and cholera toxin (3, 5, 6).The latter agent utilizes ganglioside GM, as its specific cellsurface receptor (see ref. 2) and can activate adenylate cyclasein thyroid preparations (7, 8). Furthermore, thyroid tumor cellsthat are deficient in gangliosides more complex than GM3 wereunresponsive to TSH (9).On the other hand, an important role for phospholipids has

been implicated in hormone-responsive adenylate cyclasesystems (10-15). Treatment of thyroid slices and membraneswith phospholipases profoundly depressed the stimulation ofadenylate cyclase by TSH (10-13) but resulted in enhancedbinding of TSH to membranes (13-15). In order to explorefurther the relative roles of these two classes of acidic phos-pholipids on TSH action, we separated the various lipid classesfrom bovine thyroid glands and studied their ability to affectthe binding of bovine TSH to bovine thyroid membranes.

The publication costs of this article were defrayed in part by page

charge payment. This article must therefore be hereby marked "ad-vertisement" in accordance with 18 U. S. C. §1734 solely to indicatethis fact.

5301

MATERIALS AND METHODSMaterials. Highly purified bovine TSH was a generous gift

from John Pierce (University of California at Los Angeles) andwas iodinated by the chloramine-T procedure as described (16);the specific radioactivity of different preparations ranged from38 to 71 ,uCi/,ug. Sigma was the source of the unlabeled TSHused in the binding assays as well as purified phospholipids andphospholipase A2 (bee venom, 1300 units/mg). PhospholipaseC (Clostridium welchii, 6.6 units/mg) was obtained fromCalbiochem.

Binding Assay. Plasma membranes from bovine thyroidglands were prepared as described by Wolff and Jones (17),stored frozen in liquid nitrogen, and washed prior to use bysuspending them in 25 mM Tris acetate buffer (pH 6.0) andcentrifuging at 30,000 X g for 10 min (4). The standard bindingassay was similar to that described elsewhere (4) and contained125,000 cpm of 125I-labeled TSH (125I-TSH), 5og of membraneprotein, 0.5% bovine serum albumin, and 25 mM Tris acetatebuffer (pH 6.0) in a total volume of 100,l. Incubations werefor 1 hr at 40C, and bound radioligand was separated from free125I-TSH by filtration on cellulose acetate filters (4). Nonspecificbinding was determined by incubating the membranes in thepresence of an excess of unlabeled TSH (1.5 ,M). Specificbinding of 5 ,ug of membranes was 20,000-25,000 cpm, a valuethat is on the upward slope of the binding curve when cpmbound is plotted against amount of membranes. Lipids to betested in the binding assay were dried from chloroform/methanol (2:1, vol/vol) under a stream of N2 and dispersed in25 mM Tris acetate buffer (pH 6.0) by sonication for 10 minjust prior to being assayed.

Extraction and Separation of Thyroid Lipids. Lipids wereextracted from bovine thyroid glands and partitioned intoupper and lower phases as described (4). The upper-phase lipidswere chromatographed on a DEAE-Sephadex column (18) andeluted stepwise with methanol, 0.25 M ammonium acetate inmethanol, and 0.5 M ammonium acetate in methanol. Thefractions containing gangliosides were combined, taken todryness, dissolved in water, dialyzed, and lyophilized. To sep-arate gangliosides from other acidic lipids, we used a slightmodification of the method of Yu and Ledeen (18). The ly-ophilized material was dissolved in chloroform/methanol (19:1,vol/vol) and applied to a Unisil column (200-325 mesh) thatwas eluted stepwise with increasing concentrations of methanolin chloroform as described in Table 1. Any remaining phos-

Abbreviations: GM1, galactosyl-N-acetylgalacosaminyl-(N-acetyl-neuraminyl)galactosylglucosylceramide; GM3, N-acetylneuraminyl-galactosylglucosylceramide; CL, cardiolipin; PtdCho, phosphatidyl-choline; PtdEtn, phosphatidylethanolamine; PtdGro, phosphatidyl-glycerol; PtdIns, phosphatidylinositol; PtdSer, phosphatidylserine; SM,sphingomyelin; TSH, thyrotropin; 125I-TSH, l25I-labeled TSH.* Visiting Fogarty Fellow from the Department of Pharmacy, theMedical School, University of Milan, Milan, Italy.

5302 Biochemistry: Omodeo-Sale et al.

pholipids were removed from the ganglioside fraction by sa-ponification for 1 hr at 37°C in chloroform/0.6 M NaOH inmethanol (2:1, vol/vol). The mixture was neutralized with 0.05vol of 4 M acetic acid and desalted on Sephadex G-25 (19).

Lower-phase lipids (corresponding to 100 g of tissue) werefractionated on a Unisil column (10 g) according to Vance andSweeley (20). The column was eluted with 150 ml of chloroform(fraction A), 300 ml of acetone/methanol (9:1, vol/vol) (fractionB), and 150 ml of methanol (fraction C). Fraction B was furtherpurified on a DEAE-Sephadex column (1 g) that was elutedwith 100 ml of chloroform/methanol/water (30:60:8, vol/vol)(fraction B1) and 130 ml of chloroform/methanol/0.8 M sodiumacetate (30:60:8, vol/vol) (fraction B2) (18). Fraction B2 wasdesalted on Sephadex G-25 (19) and further separated by pre-parative thin-layer chromatography. After exposure of thechromatogram to iodine vapor, individual zones were scrapedfrom the plate and eluted with chloroform/methanol/water(65:25:4, vol/vol) and chloroform/methanol (2:1, vol/vol) (referto Fig. 2 for the fractionation scheme).Other Methods. Thin-layer chromatography was carried out

on silica gel 60 on glass plates (E. Merck). Gangliosides (21),phospholipids (22, 23), and other lipids (24) were detected byconventional procedures. Phospholipids were further identifiedby two-dimensional thin-layer chromatography with chlo-roform/methanol/water (65:25:4, vol/vol) or chloroform/methanol/28% NH40H (65:35:5, vol/vol) in the first directionand chloroform/methanol/acetic acid/water (60:40:4:2, vol/vol) or chloroform/acetone/methanol/acetic acid/water(50:20:10:10:5, vol/vol) in the second direction. Phospholipidphosphorus (25), ganglioside sialic acid (4), and membraneprotein (26) were determined by established procedures.Membranes were treated with phospholipases as described byPohl et al. (27) with minor modifications (12).

RESULTSAnalysis of Upper-Phase Lipids from Bovine Thyroid. As

reported (4), the crude upper-phase lipids contained a complexmixture of gangliosides. In addition, there was a considerableamount of phospholipid (2 nmol/nmol of sialic acid). When theupper-phase lipids were chromatographed on Unisil, most ofthe gangliosides were recovered in fraction 3 and most of thephospholipids were in fraction 2 (Table 1). Analysis by thin-layer chromatography indicated that fraction 1 containedcardiolipin (CL) and phosphatidylglycerol (PtdGro), fraction

Table 1. Fractionation of upper-phase lipids from bovine thyroidby silicic acid column chromatography

Molar InhibitionDistribution, % ratio, of 125I-TSH

Fraction AcNeu P AcNeu/P binding*, %

Upper phase 100 100 1:2 571 Trace 2 0:1 882 5 76 1:20 603 92 22 5:2 37

Crude upper-phase lipids extracted from bovine thyroid glandswere chromatographed on a Unisil column by stepwise elution with:(i) CHC13/CH30H, 19:1; (ii) CHC13/CH30H, 80:20 [this ratio allowedalmost complete recovery of phosphatidylinositol (PtdIns) but in-creased the amount ofGM3 eluted compared to CHC13/CH30H (85:15,vol/vol); and (iii) CHC13CH30H, 2:3. Each fraction was monitoredfor sialic acid (AcNeu), phospholipid (P), and inhibition of 125I-TSHbinding to bovine plasma membranes. Fraction 1 contained car-diolipin and phosphatidylglycerol (PtdGro); fraction 2 containedphosphatidylserine (PtdSer) and PtdIns and a trace of GM3; fraction3 contained gangliosides with a trace of PtdIns and PtdSer plus anunknown phospholipid.* In the presence of 10 nmol of phospholipid.

'7 60K

a2pz20- /K

10 20 30 40 50 10 50Sialic acid, nmol

FIG. 1. Inhibition of 125I-TSH binding to bovine thyroid mem-branes by thyroid ganglioside fraction: effect of purification fromphospholipids. (A) Inhibition of 125I-TSH binding to thyroid mem-branes by increasing amounts of upper-phase lipids (I ) orfraction 3 (----- ) purified by Unisil column chromatography asdescribed in Table 1. (B) Inhibition of 1251-TSH binding to thyroidmembranes by upper-phase lipids (solid bars), fraction 3 (hatchedbars), or fraction 3 after saponification (open bars). Binding was de-termined in the presence and absence of each lipid fraction (addedas nmol of sialic acid) in triplicate and was corrected for binding inthe absence of membranes.

2 contained phosphatidylinositol (Ptdlns), phosphatidylserine(PtdSer), and a trace of GM3, and fraction 3 contained mainlygangliosides with lesser amounts of PtdIns, PtdSer, and an un-known phospholipid (summarized in Table 1). Fractions 1 and2 were more inhibitory than fraction 3 when expressed on thebasis of phosphorus content. When the crude upper phase was

(Chloroform)

IFRACTION Acholesterolglyceridesfatty acids

H-I

LOWER PHASE LIPIDS

| Silicic Acid Column I

IAcetone:Methanol, 9:1)

FRACTION Bneutral glycolipids

GM3 (trace)acidic phospholipids

1+1

IEAE-Sephadex

(Chloroform:Methanol:H2O)30:60:8

FRACTION I

neutral glycolipids(-)

CL1+1

(Methanol)

FRACTION Cremaining

phospholipids(-)

Chloroform:Methanol:0.8 M NaOAc30:60:8

FRACTION B NGM13 (trace)

acidic phospholipids(+)

ISephades G~-25

| Preparative TLC|

(Chloroform:Methanol:H2O, 65:25:4)

PtdGro(+)

Ptdins + PtdSer1+1

FIG. 2. Procedure for separation of lower-phase lipids from bo-vine thyroid. The total lipid extract from bovine thyroid glands waspartitioned into upper and lower phases, and the lower phase wasfractionated as indicated. Each fraction was tested for its effect on125I-TSH binding to bovine thyroid membranes. +, Inhibition ofbinding; -, no inhibition.

Proc. Natl. Acad. Sci. USA 75 (1978)

Proc. Nati. Acad. Sci. USA 75 (1978) 5303

Table 2. Inhibition of binding of 125I-TSH to bovine thyroidplasma membranes by acidic phospholipids

Addition Inhibition ofper 125I-TSH binding, %

Phospho- assay, Bovinelipid nmol Commercial thyroid

CL 4 67 70PtdGro 10 74 78PtdIns 10 64 58PtdSer 10 26 ND*

Binding of 125I-TSH to bovine plasma membranes was determinedin the presence and absence of the indicated phospholipids obtainedfrom commercial sources or isolated from bovine thyroid glands. Eachbinding assay was done in triplicate and was corrected for binding inthe absence of membranes.* ND, not determined.

compared to fraction 3 on the basis of sialic acid content, it wasfound to be more inhibitory (Fig. 1). After saponification todestroy any remaining phospholipids, fraction 3 still inhibitedthe binding of TSH to membranes but was considerably lesspotent than the original upper phase; 50 nmol of lipid-boundsialic acid in fraction 3 after saponification caused 54% inhi-bition whereas 5 nmol in the original upper phase caused 57%inhibition. This latter value is similar to that reported previously(4).

Analysis of Lower-Phase Lipids from Bovine Thyroid.When the lower phase lipids were fractionated on silicic acidand DEAE-Sephadex columns, the inhibitory activity was re-covered in fraction B2 which contained the acidic phospholipids(Fig. 2). These were separated by preparative thin-layerchromatography and identified by two-dimensional chroma-tography as CL, PtdGro Ptdlns, and PtdSer. The first three ofthese were potent inhibitors of TSH binding and were as ef-fective as highly purified phospholipids obtained commercially(Table 2). Thus, the inhibitory effects observed with ourpreparations from bovine thyroid appear to be due to thesephospholipids. CL was the most effective inhibitor; 1.3 nmol(2.6 nmol of phosphorus) caused 50% inhibition (Fig. 3). Thisdegree of inhibition was obtained with 4.6 nmol of PtdGro and7.1 nmol of PtdIns; PtdSer was a weak inhibitor. The majorphospholipids present in bovine thyroid were phosphatidyl-choline (PtdCho), phosphatidylethanolamine (PtdEtn), andsphingomyelin (SM) (Table 3). These phospholipids as wellas

c._Cn 80

.0

° 40

' 20

._4

2 4 6 8 10Phospholipid, nmol (as phosphorus)

12

FIG. 3. Effect of acidic phospholipids on binding of 125I-TSH tobovine thyroid plasma membranes. Binding was determined in thepresence and absence, of the indicated commercial phospholipid(added as nmol of phosphorus) in triplicate and was corrected forbinding in the absence of membranes.

Table 3. Phospholipid composition of bovine thyroid glands

% of totalPhospholipid phospholipid*

Cardiolipin 3Phosphatidylglycerol 1Phosphatidylinositol 9Phosphatidylserine 8Lysophosphatidylcholine 3Sphingomyelin 14Phosphatidylcholine 27Phosphatidylethanolamine 26

Phospholipids were isolated from bovine thyroid glands. Thephospholipid content was 8.76 ,mol/g of fresh tissue. Our data are ingood agreement with previous results for bovine thyroid (28) andporcine thyroid (29) except that our value for phosphatidylcholinewas significantly lower.* Total phospholipids include other minor unidentified components.

their commercial equivalents did not inhibit the binding of TSHto thyroid membranes.

Effect of Phospholipids and Phospholipases on ThyroidMembranes. We then determined whether the inhibitory ef-fect of the acidic phospholipids was due to their interaction withTSH or the membranes. When bovine thyroid membranes wereincubated with phospholipids (8 min at 240C) and then washedand collected by centrifuging, the membranes exposed to CL,PtdGro, or Ptdlns or a mixture of these three phospholipids hada decreased capacity to bind TSH (Fig. 4). Again CL was mosteffective, followed by PtdGro and PtdIns. The inhibition wasselective: other phospholipids did not cause a loss in binding.Because the amount of phospholipid (ut2 nmol/5 ,g of mem-brane protein) required to cause inhibition under these condi-tions was similar to the amount added directly to the binding

100

80C0

0

60-Cn

O 40cm

PC ip

C II~~~

.'C'aW.0 a0

Phosphol ipid

FIG. 4. Effect of preincubating bovine thyroid membranes withphospholipids on subsequent binding of 1251-TSH to membranes.Bovine thyroid membranes (1 mg of protein) were incubated for 8 minat 24°C in 1 ml of 25 mM Tris-HCl buffer (pH 7.6) containing nophospholipids or 300 pg of the indicated phospholipid. Then, 2 ml ofthe same buffer (ice cold) was added and the suspension was centri-fuged at 30,000 X g for 20 min. The membranes were suspended in25mM Tris acetate buffer (pH 6.0) and assayed for 125I-TSH bindingin triplicate. Binding was corrected for nonspecific binding as de-termined in the presence of excess unlabeled TSH.

Biochemistry: Omodeo-Sale et al.

_

5304 Biochemistry: Omodeo-Sale et al.

Table 4. Effect of phospholipase A on binding of 125I-TSH tobovine thyroid membranes

Phospholipase A, 125I-TSH bound,units/ml % of control

0 1005 122

50 138100 124

Bovine thyroid membranes (2 mg of protein per ml) were incubatedwith the indicated concentration of phospholipase A in 1 ml of 25mMTris-HCl, pH 7.6/1 mM CaCl2 for 5 min at 30'C. The reaction wasstopped by adding 40 ;d of 0.1 M EDTA. The membranes were washedand assayed for 125I-TSH binding in triplicate. Binding was correctedfor nonspecific binding determined in the presence of excess unlabeledTSH. Prior to phospholipase treatment, the membranes were washedin the Tris-HCl buffer to remove dithiothreitol present in the mem-brane preparations.

assay, it appeared that the inhibitory effect of these phospho-lipids was due to their interaction with the membranes and notwith the hormone.We explored this effect further by incubating the membranes

with phospholipases. In agreement with reports of others(13-15), treatment of the membranes with phospholipase Aresulted in an increase in TSH binding (Table 4). When themembranes treated with phospholipase A were exposed toacidic phospholipids as described above, there was a dramaticdecrease in TSH binding (data not shown). In contrast, we didnot observe a similar effect with phospholipase C. When weincubated the membranes with and without phospholipase Cunder identical conditions at 370C, we observed a loss in TSHbinding with both sets of membranes. Amir et al. (15) reportedthat incubation of bovine thyroid membranes at elevatedtemperatures caused a loss in TSH receptor sites. They alsoreported that phospholipase C treatment stimulated TSHbinding (15), whereas Haye and Jacquemin (14) found thatphospholipase C caused an increase in TSH binding to 6ovinethyroid membranes at low but not high concentrations ofTSH.

DISCUSSIONOf the various lipid classes isolated from bovine thyroid gland,we found that only two classes, gangliosides and acidic phos-pholipids, affected the interaction of TSH with thyroid plasmamembranes. The inhibition of TSH binding by gangliosides hasbeen described (3, 4), but it is apparent from our present workthat some of the inhibition was due to the presence of phos-pholipids in the ganglioside preparations.t The gangliosides,however, do inhibit TSH binding because, even after extensivepurification and saponification to destroy contaminatingphospholipids, thyroid gangliosides are still inhibitory. In ad-dition, TSH binds to liposomes containing gangliosides (16, 30).Finally, the inhibitory effects of gangliosides are distinct be-cause gangliosides appear to interact with TSH (3),t and, as wenow show, phospholipids interact with the membrane.

t After further purification and saponification, brain gangliosides werealso found to be less inhibitory.

f The nature of this interaction is still unclear. Oligosaccharides pre-pared from individual as well as mixed gangliosides did not inhibitthe binding of TSH to bovine thyroid membranes even at concen-trations as high asSmM (unpublished observations). A similar lackof interaction between these oligosaccharides and gonadotropin hadbeen reported previously (16). In contrast, the oligosaccharide fromGM1 ganglioside, the cholera toxin receptor, binds to cholera toxinand blocks toxin binding to cells (31) and thyroid membranes (un-published observations).

Our results indicate that certain phospholipids alter themembrane components responsible for the binding of TSH.Only the highly polar, negatively charged phospholipids (CL,Ptdlns and PtdGro) were able to affect the interaction of thereceptor with the hormone. Although CL was the most potentinhibitor, it is presumed to be a mitochondrial membranecomponent (32) and its presence in plasma membranes has notbeen described. PtdIns and PtdGro are plasma membranecomponents. A role for these phospholipids in influencing theability of the TSH receptor to bind hormone is supported fur-ther by our experiments with phospholipase A. Treatment ofthyroid membranes with this enzyme caused an increase in TSHbinding and exposure of these treated membranes to the abovephospholipids reduced TSH binding.The effects of phospholipids on hormone-stimulated ade-

nylate cyclase systems has been best examined with a gluca-gon-sensitive system (27, 33-35). Levey and coworkers (33)observed that Lubrol PX-solubilized adenylate cyclase lost itssensitivity to glucagon and addition of phospholipids to thesolubilized enzyme restored the hormone response. The phos-pholipid effect appeared to be specific because PtdSer but notPtdIns was active. Rodbell and collaborators (27, 34) found thatphospholipase A treatment of liver membranes abolished theresponse of adenylate cyclase to glucagon and addition ofmembrane phospholipids restored hormone sensitivity. Therewere corresponding effects on the binding of glucagon to themembranes. Further studies with phospholipases of differentspecificities indicated that the enzymes that hydrolyzed acidicphospholipids were the most effective (35).The involvement of membrane phospholipids in the TSH-

stimulated adenylate cyclase system appears to be more com-plex. Treatment of thyroid membranes with phospholipase Aor C caused a decrease in TSH-stimulated adenylate cyclaseactivity (10-13) and subsequent addition of PtdChb or PtdSerpartially restored hormone response (12). In addition, mem-branes treated with Lubrol PX lost their TSH response, andaddition of PtdCho partially restored hormone sensitive ade-nylate cyclase activity (12). However, our results as well as re-ports of others (13-15) indicate that phospholipase A treatmentof thyroid membranes causes an increase in TSH binding to themembranes. In addition, we find that treatment of the mem-branes with certain acidic phospholipids results in decreasedhormone binding. These apparently conflicting results suggestthat membrane phospholipids may be involved at several dif-ferent sites and different phospholipids may act at differentsites. Thus, as in the glucagon-sensitive system, certain phos-pholipids(PtdCho and PtdSer) may play a role in the couplingor transducing site between the TSH receptor and adenylatecyclase in the thyroid membrane. In addition, other acidicphospholipids may be involved at the receptor site.We can only speculate on how these acidic phospholipids

modulate the ability of the receptor to bind TSH. Currentconcepts of membrane structure envision membrane receptorsas asymmetric with their hydrophobic regions buried in thefluid lipid bilayer. In addition, membrane phospholipids maybe distributed asymmetrically both laterally and transverselyin the bilayer. Thus, the receptors may exist in local domainsenriched in certain phospholipids, and the composition of thesephospholipids surrounding the hydrophobic region of the re-ceptor as it traverses the membrane may be different in the twohalves of the bilayer. Small changes in phospholipid compositioncould have profound effects on local fluidity and charge in suchdomains. An increase in negative charge could lead to a localincrease in cations such as Ca2+ which has been reported toinhibit TSH binding (13, 15) or to a repulsion of other negativelycharged membrane components that are involved in hormone

Proc. Natl. Acad. Sci. USA 75 (1978)

Proc. Natl. Acad. Sci. USA 75 (1978) 5305

binding.§ Changes in fluidity could lead to a vertical dis-placement of the receptor (37) or to lateral movement of thereceptor in the plane of the membrane. Either type of move-ment could influence the ability of the receptor to bind hor-mone directly by masking the exposure of the receptor or in-directly by disrupting associations between the receptor andother membrane components required for hormone bind-ing.§

Finally, we suggest a potential physiological function for theeffects we observe. It is well known that one of the actions ofTSH on the thyroid is to stimulate phospholipid synthesis. TSHgreatly enhances the incorporation of 32P into acidic phos-pholipids, especially PtdIns (14,38). Hormone-induced changesin these phospholipids could in turn lead to a reduction inhormone binding. Such a model could be involved in the phe-nomenon of desensitization at the level of the membrane re-

ceptor.

§ Gangliosides, which are negatively charged, could be associated withor be a component of the TSH receptor as previously suggested (2,4). Based on studies with spin-labeled gangliosides, Sharom and Grant(36) have suggested that gangliosides in the plasma membrane tendto cluster around glycoproteins and that Ca2+-mediated ganglio-side-phospholipid interactions involve both clustering and immo-bilization.

1. Sutherland, E. W., Robinson, G. A. & Butcher, R. W. (1968)Circulation 337, 333-306.

2. Fishman, P. H. & Brady, R. 0. (1976) Science 194,906-915.3. Mullin, B. R., Fishman, P. H., Lee, G., Aloj, S. M., Ledley, F. D.,

Winand, R. J., Kohn, L. D. & Brady, R. 0. (1976) Proc. Natl.Acad. Sci. USA 73,842-846.

4. Mullin, B. R., Pacuszka, T., Lee, G., Kohn, L. D., Brady, R. 0.

& Fishman, P. H. (1978) Science 199, 77-79.5. Ledley, F. D., Mullin, B. R., Lee, G., Aloj, S. M., Fishman, P. H.,

Hunt, L. T., Dayhoff, M. 0. & Kohn, L. D. (1976) Biochem.Biophys. Res. Commun. 69,852-859.

6. Kurosky, A., Markel, D. E., Peterson, J. W. & Fitch, W. M. (1977)Science 195, 299-301.

7. Mashiter, K., Mashiter, G. D., Hauger, R. L. & Field, J. B. (1973)Endocrinology 92,541-549.

8. Mullin, B. R., Aloj, S. M., Fishman, P. H., Lee, G., Kohn, L. D.& Brady, R. 0. (1976) Proc. Natl. Acad. Sci. USA 73, 1679-1683.

9. Meldolesi, M. F., Fishman, P. H., Aloj, S. M., Kohn, L. D. &Brady, R. 0. (1976) Proc. Natl. Acad. Sci. USA 73, 4060-4064.

10. Yamashita, K., Bloom, G., Rainard, B., Zor, U. & Field, J. B. (1970)Metabolism 19, 1109-1118.

11. Macchia, V., Tamburrini, O. & Pastan, I. (1970) Endocrinology86,787-792.

12. Yamashita, K. & Field, J. B. (1973) Blochim. Biophys. Acta 304,686-692.

13. Moore, W. V. & Wolff, J. (1974) J. Biol. Chem. 249, 6255-6263.

14. Haye, B. & Jacquemin, C. (1970) FEBS Lett. 18,47-52.15. Amir, S. M., Godfine, I & Ingbar, S. H. (1976) J. Biol. Chem. 251,

4693-4699.16. Pacuszka, T., Osborne, J. C., Jr., Brady, R. 0. & Fishman, P. H.

(1978) Proc. Natl. Acad. Sci. USA 75,764-768.17. Wolff, J. & Jones, A. B. (1971) J. Biol. Chem. 246,3939-3947.18. Yu, R. & Ledeen, R. W. (1972) J. Lipid Res. 13, 680-686.19. Fishman, P. H., Bradley, R. M. & Henneberry, R. C. (1976) Arch.

Biochem. Blophys. 172,618-626.20. Vance, D. E. & Sweeley, C. C. (1967) J. Lipid Res. 8, 621-

630.21. Svennerholm, L. (1957) Biochim. Biophys. Acta 24, 604-611.22. Vaskovsky, V. E. & Svetasher, V. I. (1972) J. Chromatogr. 65,

451-453.23. Skipski, V. P., Peterson, R. F. & Barclay, M. (1964) Biochem. J.

90,374-378.24. Borowski, E. & Ziminski, T. (1966) J. Chromatogr. 23, 480-

482.25. Bartlett, G. R. (1957) J. Biol. Chem. 226,466-468.26. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275.27. Pohl, S. L., Krans, H. M., Kozyreff, V., Birnbaumer, L. & Rodbell,

M. (1971) J. Biol. Chem. 246,4447-4454.28. Satyaswaroop, P. G. (1971) Lipids 6,661-665.29. Levis, G. M., Karli, J. N. & Malamos, B. (1972) Clin. Chim. Acta

41,335-345.30. Aloj, S. M., Kohn, L. D., Lee, G. & Meldolesi, M. F. (1977) Bio-

chem. Biophys. Res. Commun. 74, 1053-1059.31. Fishman, P. H., Moss, J. & Osborne, J. C., Jr. (1978) Biochemistry

17,711-716.32. Marinetti, G. V., Erbland, J. & Stotz, E. (1958) J. Biol. Chem. 233,

562-565.33. Levey, G. S., Fletcher, M. A. & Klein, I. (1975) Adv. Cyclic Nu-

cleotide Res. 5, 53-65.34. Pohl, S. L., Birnbaumer, L. & Rodbell, M. (1971) J. Biol. Chem.

246, 1849-1856.35. Rubalcava, B. & Rodbell, M. (1973) J. Biol. Chem. 248,3831-

3837.36. Sharom, F. J. & Grant, C. W. M. (1978) Biochim. Biophys. Acta

507,280-293.37. Borochov, H. & Shinitzky, M. (1976) Proc. Natl. Acad. Sci. USA

73,4526-4530.38. Marinetti, G. B., Erblanc, J. F. & Brossard, M. (1964) in Metab-

olism and Physiological Significance of Lipids, eds. Dawson,R. M. C. & Rhodes, D. N. (John Wiley, London), p. 71.

Biochemistry: Omodeo-Sale et al.