14
J. Cell Sci. 59, 13-25 (1983) 13 Printed in Great Britain © Company of Biologists Limited 1983 ISOLATION AND ELECTROPHORETIC CHARACTERIZATION OF THE PLASMA MEMBRANE OF SEA-URCHIN SPERM NICHOLAS L. CROSS* Department of Biology, B-022 University of California, San Diego La Jolla, CA 92093, U.SA. SUMMARY A subcellular fraction containing plasma membranes was isolated from flagella of the sperm of Strongylocentrotus purpuratus by differential centrifugation, and analysed by sodium dodecyl 8ulphate/polyacrylamide gel electrophoresis. Coomassie Blue staining revealed nine major bands and 14 minor species. Five bands of apparent molecular weights ~200 X l(r, 149 X l(r, 120 X l(r, 75 X 10 3 and 59X10 3 also stained with periodic acid-Schiffs reagent and so are probably glycoproteins. These five components are externally exposed, as determined by lactoperoxidase- catalysed radio-iodination. Isolation of membranes from radio-iodinated sperm results in an enrich- ment of about tenfold in the specific activity of 1Z5 I. Comparison of the electrophoretic patterns of labelled sperm and of the membranes isolated from 125 I-labelled sperm suggests that no major labelled proteins are lost during the isolation procedure, and so to this extent the membrane fraction is representative of the entire sperm plasma membrane. INTRODUCTION The sea-urchin sperm is a highly specialized cell of relatively few functions that support the delivery of the paternal genome into the interior of the egg. The sperm plasma membrane is an important participant at several steps leading to fertilization. In the 'acrosome reaction' the sperm membrane fuses with the underlying membrane of the acrosomal granule to expose the contents of the granule and the acrosomal membrane (which is the membrane that subsequently fuses with the egg plasma membrane) to the extracellular space. The acrosome reaction appears to be triggered by the interaction of the sperm membrane with the extracellular jelly coat of the egg, and may be accompanied by changes in ion fluxes through the sperm membrane (Schackmann, Eddy & Shapiro, 1978; Tilney, Kiehart, Sardet & Tilney, 1978). Moreover, the plasma membrane may be intimately involved in maintaining normal sperm motility. As high external [K + ] inhibits flagellar beating (e.g., see Gibbons, 1980), it appears that motility may be controlled by the potential across the membrane or by ion fluxes through it. Gibbons (1980) has suggested that there is an outward- directed Ca 2+ -pump in the flagellar membrane functioning to keep the internal [Ca 2+ ] below the concentration that would prevent normal flagellar beating. Afirststep in understanding the molecular basis for these functions is the biochem- ical characterization of sperm plasma-membrane components. Previous studies with •Present address: Department of Zoology, University of California, Davis, CA 95616, U.S.A.

ISOLATION AND ELECTROPHORETIC …jcs.biologists.org/content/joces/59/1/13.full.pdfISOLATION AND ELECTROPHORETIC CHARACTERIZATION OF THE PLASMA MEMBRANE ... Isolation of membranes from

Embed Size (px)

Citation preview

J. Cell Sci. 59, 13-25 (1983) 13Printed in Great Britain © Company of Biologists Limited 1983

ISOLATION AND ELECTROPHORETIC

CHARACTERIZATION OF THE PLASMA MEMBRANE

OF SEA-URCHIN SPERM

NICHOLAS L. CROSS*Department of Biology, B-022 University of California, San Diego La Jolla, CA 92093,U.SA.

SUMMARY

A subcellular fraction containing plasma membranes was isolated from flagella of the sperm ofStrongylocentrotus purpuratus by differential centrifugation, and analysed by sodium dodecyl8ulphate/polyacrylamide gel electrophoresis. Coomassie Blue staining revealed nine major bandsand 14 minor species. Five bands of apparent molecular weights ~200 X l(r, 149 X l(r, 120 X l(r,75 X 103 and 59X103 also stained with periodic acid-Schiffs reagent and so are probablyglycoproteins. These five components are externally exposed, as determined by lactoperoxidase-catalysed radio-iodination. Isolation of membranes from radio-iodinated sperm results in an enrich-ment of about tenfold in the specific activity of 1Z5I. Comparison of the electrophoretic patterns oflabelled sperm and of the membranes isolated from 125I-labelled sperm suggests that no majorlabelled proteins are lost during the isolation procedure, and so to this extent the membrane fractionis representative of the entire sperm plasma membrane.

INTRODUCTION

The sea-urchin sperm is a highly specialized cell of relatively few functions thatsupport the delivery of the paternal genome into the interior of the egg. The spermplasma membrane is an important participant at several steps leading to fertilization.In the 'acrosome reaction' the sperm membrane fuses with the underlying membraneof the acrosomal granule to expose the contents of the granule and the acrosomalmembrane (which is the membrane that subsequently fuses with the egg plasmamembrane) to the extracellular space. The acrosome reaction appears to be triggeredby the interaction of the sperm membrane with the extracellular jelly coat of the egg,and may be accompanied by changes in ion fluxes through the sperm membrane(Schackmann, Eddy & Shapiro, 1978; Tilney, Kiehart, Sardet & Tilney, 1978).Moreover, the plasma membrane may be intimately involved in maintaining normalsperm motility. As high external [K+] inhibits flagellar beating (e.g., see Gibbons,1980), it appears that motility may be controlled by the potential across the membraneor by ion fluxes through it. Gibbons (1980) has suggested that there is an outward-directed Ca2+-pump in the flagellar membrane functioning to keep the internal[Ca2+] below the concentration that would prevent normal flagellar beating.

A first step in understanding the molecular basis for these functions is the biochem-ical characterization of sperm plasma-membrane components. Previous studies with

•Present address: Department of Zoology, University of California, Davis, CA 95616, U.S.A.

14 N.L. Cross

sea-urchin sperm employed surface labelling (Gabel, Eddy & Shapiro, 1979; Lopo &Vacquier, 1980a) and/or extraction with mild detergent (Stephens, 1977; Otter,1978) of whole sperm or of flagella in an attempt to identify membrane proteins, butisolated plasma membrane fractions were not examined. The present study was un-dertaken to provide a more complete description of the composition of the spermplasma membrane. As it was desirable to begin by obtaining a plasma membranefraction as free as possible from other kinds of membranes, flagella (which contain nointracellular membranes) were first isolated from 5. purpuratus sperm, then a mem-brane fraction was obtained from them. This report describes sperm membraneproteins identified by electrophoretic analysis of the membrane fraction, and bylactoperoxidase-catalysed external radio-iodination of whole sperm. Followingradiolabelling of whole sperm, the same major bands were observed in the membranefraction as in lysates of whole sperm, suggesting that the membrane fraction isrepresentative of the entire sperm plasma membrane.

MATERIALS AND METHODS

Sea urchins (Strongylocentrotus purpuratus) were obtained locally and maintained in tanks ofrunning sea-water. To induce spawning, urchins were injected intracoelomically with 0-5 M-KC1,and semen was stored undiluted at 0-4°C until used. Sperm concentrations were determined bycounting in a haemocytometer. Sea-water was filtered through 0-45/im pore Millipore filters.

Isolation of a plasma membrane fractionSperm were washed by diluting about 1 ml of semen (~4 X 1010 sperm) into 10 ml of sea-water

and centrifuging at 40£ for 5 min at 4°C. The small pellet of debris and clumped sperm wasdiscarded, and the sperm in the supernatant were sedimented by centrifugation at 400#for 10 min.The sperm were suspended in 7 ml of solution 1 (475 mM-NaCl, 25 mM-KCl, 1 mM-Tris, 0-1 mM-dithiothreitol (DTT), 1 mM-CaC^, pH8-0) and a flagellar fraction was obtained as described byGray & Drummond (1976). Flagella were broken from sperm by passage seven-ten times througha 21 gauge hypodermic needle. To separate broken flagella from other sperm components, thesuspension was placed over 5 ml of solution 2 (25% (w/v) sucrose, 10 mM-Tris, 0-1 mM-DTT,1 mM-CaCl2, pH 8-0) and centrifuged 15 min at bSQg. The top 6 ml, consisting of flagella, mem-brane vesicles, and a few heads plus midpieces, were removed and set aside. The material at theinterface consisted of a mixture of flagella and unbroken sperm and was removed in 4 ml, dilutedto 7 ml with solution 1, placed over 5 ml of solution 2 and the centrifugation was repeated in orderto recover more flagella. The pooled upper fractions containing flagella from two to four cycles werecentrifuged at 37 000 # for 1 h to sediment the flagella. (The pellet is referred to as flagella andsupernatant as the flagellar supernatant.) The pellets from each of the above cycles of centrifugation(containing broken and intact sperm, and referred to as the pellet) were pooled. The final interfacefraction after the last cycle is called the interface.

To separate the flagellar membrane from the axoneme, the flagellar pellet was resuspended bygentle homogenization in 5 ml of solution 3 (10mM-Tris, 10mM-KCl, lOmM-NaCl, 0-1 mM-DTT,p H 8 0 ; Witman, Carlson, Berliner & Rosenbaum, 1972), placed over 2 5 ml of solution 4 (40%,w/v, sucrose, lOmM-Tris, 0 1 mM-DTT, 1 mm-CaCl2l p H 8 0 ) and centrifuged at 12000^ for1-25 h. The prolonged dialysis against solution 3 used by Witmanetal. (1972) with Chlamydomonasand subsequently by Gray & Drummond (1976) with sperm flagella is unnecessary in this systemand, furthermore, results in significantly increased contamination of the final membrane fractionby axonemal proteins (data not shown). The upper 4ml containing membrane vesicles wereremoved, diluted with 1 ml of solution 3, and centrifuged for 1-5 h at 100 000 g to sediment thevesicles. The pellet was suspended in 0-2-0-5ml solution 1 (the top membrane fraction). Thesupernatant is referred to as the top membrane supernatant. The next 2-5 ml, containing the fluffy

Sea-urchin sperm plasma membrane 15

material at the interface, was diluted with 2-5 ml solution 3 and centrifuged at 100 000^ as above.The pellet is referred to as the middle membrane fraction, and the supernatant as the middlemembrane supernatant. The final 1 ml contained a pellet consisting mostly of axonemes in variousstates of disruption as well as some membrane vesicles (referred to as axonemes). Samples of allfractions were assayed for protein (Lowry, Roseborough, Farr&Randall, 1951) after adding sodiumlauroyl sarcosinate to a final concentration of 0 3 % or sodium dodecyl sulphate (SDS) to 1 %, andusing bovine serum albumin as a standard.

Radio-iodination of spermTo label sperm with 1 2 5 1 , they were washed as described above and then resuspended to 100 times

the original volume of semen (final sperm concentration of about 4 X 108 per ml). The followingreagents were combined in this order (Hubbard & Cohn, 1972): 10-200yC\ Na ' " I (New EnglandNuclear, carrier-free, in 0-lM-NaOH, 20-50 mCi/ml); a volume of 0 -1M-HC1 equal to that ofNa125I was used; 2-8mU lactoperoxidase (CalBiochem, purified B grade); 3-17mU glucoseoxidase (Sigma, type V), 1 ml of sperm suspension, and 10^1 of 0'5 M-glucose. The mixture wasincubated at 14—16°C for lOmin and then washed 2-5 times with a large volume of sea-water asdescribed above. To assay the radioactivity in labelled sperm, samples of sperm on WhatmanGF/A glass fibre filters were washed with ice-cold 10% (w/v) trichloracetic acid containing 10-50mM-Nal until the radioactivity in the washes was less than twice background, and then washed forat least 2h at — 20 °C in ethanol/diethyl ether (3 : 1) or in acetone/H20 (9:1) and counted in agamma counter.

Polyacrylamide gel electrophoresisSamples of sperm were lysed in sample buffer (Laemmli, 1970) and passed several times through

a 21 gauge hypodermic needle to reduce the viscosity. All samples for electrophoresis weredenatured by placing in a 100°C water bath for 2min, and then they were layered onto an SDS/polyacrylamide slab gel. The stacking gel contained 4 % or 5 % (w/v) acrylamide, and the runninggel was a linear gradient of 7-5 % to 20 % (w/v) acrylamide. Occasionally, running gels containingonly 7-5 % or 10 % acrylamide were used. Electrode and gel buffers were as described by Laemmli(1970). The following proteins were used as molecular weight standards: myosin (200 X 10 Mr),phosphorylase b (94 X 103 Mr), bovine serum albumin (68 X 103 M,), acting (42 X 103 Mr), carbonicanhydrase (30 X 103 M,) soybean trypsin inhibitor (21-5 X 103 Mr) and cytochrome c (12-3 X 103

M,). The gels were fixed and stained with Coomassie Blue (Fairbanks, Steck & Wallach, 1971),dried and applied to Kodak No-Screen or XR-5 film to detect radioactivity. Some gels were slicedinto 1 mm wide sections and counted in a gamma counter. Samples to be stained with periodic acid- Schiff's reagent (PAS) were run either on slab gels as above, or on 10 cm long tube gels containing7-5 % acrylamide, with a 2 cm long stacking gel of 4 % or 5 % acrylamide. Buffer solutions were asdescribed above. Bovine serum albumin and DNase (lO^ig each) were also analysed byeletrophoresis and processed as negative and positive controls, respectively.

Enzyme assaysCytochrome c oxidase activity was determined as described by Beaufay et al. (1974); 5'-nucleo-

tidase was assayed by the method of Widnell & Unkless (1968), but at pH 8-0; alkaline phosphodies-terase was measured as described by Bischoff, Tran-Thi & Decker (1975); and alkaline phosphatasewas assayed by the method of Pekarthy, Short, Lansing & Leiberman (1972), but at pH9-9.

Electron microscopySpecimens were processed as pellets by fixing in 1 % glutaraldehyde in sea-water or in solution

3 of the isolation procedure (pH 8-0), washed in 0-5 M-NaCl or solution 3, post-fixed in 1 % OsO*in 0-1 M-phosphate buffer (pH 7-4) for 2-4 h and washed in deionized water. The specimens werestained with 0-5 % aqueous uranyl acetate for 1 h, dehydrated in ethanol and embedded in Spurr'smedium. Sections were made perpendicular to the plane of the pellet, in order to look for differencesthrough the depth of the pellet.

16 N.L. Cmss

RESULTS

Cell fractionation

Repeated passage of sperm through a 21 gauge needle results in >90% of theflagella being broken. The sperm head and midpiece remain intact and attached toeach other, sometimes with a short length of flagellum attached. Membrane vesiclesare also present in the broken cell suspension. As one test of the integrity of the plasmamembrane on the sperm head and midpiece, the broken sperm preparation wasdiluted with distilled water; about 50 % of these sperm bodies (heads plus midpieces)were osmotically active and swollen, so these plasma membranes must be intact andsealed. The isolated flagella preparation contains, in addition, membrane vesicles,free microtubules and a few sperm bodies. A comparison with the original spermsuspension revealed that 1 % of sperm bodies are carried into this fraction. Followingsuspension of the pelleted flagella in hypotonic buffer, the flagella become less phase-dense, perhaps as a result of demembranation. Many vesicles are present, as well asdiscoidal organelles that appear to be osmotically swollen flagella with the axonemecoiled inside. After centrifugation of this suspension over solution 4 (containing 40 %sucrose), membrane vesicles are recovered in two fractions: one ('top' fraction) in thesupernatant above the interface, and one ('middle' fraction) at thesupernatant-sucrose interface. The top fraction contains membrane sheets and ves-icles, many of which are disrupted (Fig. 1 A, B) . The membranes of the middle fractionare similar, but on average are larger (Fig. 1C,D); occasional microtubules are seenin this fraction. Few of the vesicles of either fraction contain electron-dense contents.The principal non-membranous component is amorphous material, probably derivedfrom the flagellar matrix, which adheres in places to the membrane (Fig. 1B). Thismaterial may be tightly attached to the membranes, as sonication of a membranefraction in 500mM-NaCl, 50mM-NaCl, or 500mM-NaBr, or treatment with a mix-ture of 20 /ig/ ml saponin, 300mM-Na2SO4, 20mM-NaHCO3, OSmivi-EDTA, pH8(Castle & Palade, 1978), did not alter the protein pattern on SDS/polyacrylamidegels (data not shown). Because the middle fraction contained more microtubules, thetop fraction is probably purer and is considered below in more detail.

The extent of mitochondrial contamination of the membrane fractions was assessedby determining the cytochrome c oxidase activity. In two experiments, the activityrecovered in the flagella fraction, from which the membranes are derived, was 3 % orless of the total activity in the starting homogenate. In two other experiments, thespecific activity in the top and middle fractions was 2% or less than the specificactivity in the homogenate.

In an effort to find plasma-membrane marker enzymes, subcellular fractions wereassayed for three enzymes that are found in plasma membranes of some types of

Fig. 1. Electron micrographs of the plasma membrane fractions, A and B. The 'top'membrane fraction, c and D. The 'middle' membrane fraction. The fractions containmembranous sheets and vesicles. Some amorphous material adheres to some membranes(arrow in B). A and c, X 10250. B and D, X 52500.

1A

Sea-urchin sperm plasma membrane 17

••&>•

?>' %•< ">£V

u 1r / 0

*•'i'

.- <H-

Fig. 1

18 N. L. Cross

mammalian cells (see review by Evans, 1979). The specific activities of alkalinephosphatase and alkaline phosphodiesterase were less than 0-07 ^mol substratehydrolysed/h per mg protein in the homogenate, top, and middle membrane frac-tions. The specific activity of 5'-nucleotidase was less than 3 ^mol/h per mg proteinin the top and middle membrane fractions. Evidently these three enzymes are notsuitable markers for the sea-urchin sperm plasma membrane.

Electrophoresis of the membrane fraction

On SDS/polyacrylamide gels, Coomassie Blue staining of the top membrane frac-tion reveals nine major bands and a number of minor bands (Fig. 2B). On gels of themiddle membrane fraction, there is more protein of apparent molecular weight (Mapp)of 55 X 103 (perhaps tubulin), and there are more minor bands (Fig. 2c). The degreeto which the middle fraction differs from the top in these respects varied among

-200

-94

-68

-42

30

12

Fig. 2. SDS/polyacrylamide gel (7-5 % to 20 % gradient) showing sperm and membranefractions. The positions of molecular weight standards (X 10~3) are shown on the right andthe major intracellular proteins are labelled on the left. Tubulin (T) was identified by itsM,pp of 55 X 103 and the fact that it is the principal protein on gels of axonemes but notin isolated heads (not shown). Histones were identified by comparison with the results ofCarroll & Ozaki (1979). A. Lysate of whole sperm, B. Top membrane fraction, c. Middlemembrane fraction.

Sea-urchin sperm plasma membrane 19

different preparations. These observations are consistent with the presence of moreaxonemal fragments in the middle fraction, but there may be other differences in themembrane proteins of the two fractions as well. The predominant species in bothfractions are two proteins of Mapp 120 X 103 and 135 X 103. Occasionally, this pair isresolved into three bands. Little or no protein co-migrating with histone is present inthe two fractions (Fig. 2). Inclusion of protease inhibitors in the isolation media didnot alter the protein pattern (data not shown).

The proteins of the two membrane fractions are minor constituents of the flagellarfraction, except for the protein co-migrating with tubulin (Fig. 3). The 120 X 103 and135 X 103 molecular weight major proteins of the membrane fractions can be seen ongels of the flagellar fraction (and in lesser amounts in the gel of sperm lysate shownin Fig. 2A) and are diminished in the axonemal fraction, from which most membraneshave been removed. The proteins of the membrane fractions can be described as a

—200

—94

- 6 8

-30

- 1 2

Fig. 3. SDS/polyacrylamide gel (7-5 % to 20% gradient) showing the flagellar fractionand its subtractions. The positions of molecular weight standards (X 10~3) are shown onthe right, A. The flagellar fraction from which membrane fractions were obtained, B. Thetop membrane fraction, c. The middle membrane fraction, D. The axoneme fraction,containing material sedimented after most membranes were removed in hypotonic buffer.With the exception of protein co-migrating with tubulin, the major proteins of the mem-brane fractions are minor components of the flagellar fractions.

20 N. L. CrossA B C D E F

f ^ T -200

-68

-42

- 3 0

-12

Fig. 4. The electrophoretic distribution of protein (A, B,C) and 12iI (D ,E , F) on SDS/polyacrylamide gels (7-5% to 20% gradient), A and D. The top membrane fractionprepared from radio-iodinated sperm, B and E. Mixture of the membrane fraction and ofa lysate of radio-iodinated sperm. In E, the labelled species of the membrane fractionco-migrate with their analogues of the sperm lysate. c and F. Lysate of radio-iodinatedsperm.

distinct subset of proteins of the flagellar fraction plus some contaminating axonemalproteins, with the middle membrane fraction being the more heavily contaminated.

Detection of membrane components by external labelling of sperm

Sperm were externally labelled by radio-iodination using lactoperoxidase, and thelabelled species were analysed on SDS/polyacrylamide gels. Sperm incubated withNa1251 and lactoperoxidase with an H2O2-generating system of glucose and glucoseoxidase rapidly incorporate label into trichloroacetic acid-precipitable form, as repor-ted by Lopo & Vacquier (1980a). Omission of either enzyme or of glucose decreasedincorporation by at least 96%.

When labelled sperm were lysed in SDS and submitted to electrophoresis, a charac-teristic pattern consisting of a limited number of labelled bands was recorded (Fig.4). This pattern is not produced by internal labelling of sperm protein, because whensperm were similarly labelled after lysis by freezing and thawing, many labelledspecies were detected; the electrophoretic distribution of label was similar to thedistribution of Coomassie Blue staining (data not shown). In contrast, the majorintracellular proteins were labelled only lightly, if at all, when intact sperm were radio-iodinated: the sum of 12SI comigrating with tubulin and the five histones is 6 % of thetotal radioactivity on the gel (average, N = 3). There are four heavily labelled com-ponents (M,pp= ~200, 128, 77 and 60 X 103) as well as minor bands that are seenconsistently in preparations of high specific activity, or on autoradiographs exposedfor a long time. Radioactive material at the front may be labelled lipids.

Sea-urchin sperm plasma membrane 21

Table 1. Distribution of protein and 12SI in subcellular fractions of radio-iodinatedsperm

Fraction"

HomogenateInterfacePelletFlagellaFlagellar supernatant

Recovery

Subfractionation of flagella:

Fraction*

Proteinb

100°2

71145

92

Protein15

125jb

IOCT1

3213414

72

125jb

Relativespecific activity0

1-001-480-292-442-72

Relativespecific activity'

FlagellaAxonemesTop membraneTop membrane supernatantMiddle membraneMiddle membrane supernatant

1412

1111

341110295

2-440-86

11-72-307-406-10

Recovery 16 37

"See text for a more complete description of the contents of each fraction.Percentage of the amount in the homogenates.

c Total in homogenate = 82 mg.Total in homogenate = 2-4X 10* c.p.m.

'Specific activity divided by specific activity of the homogenate.

Isolation of membranes from radio-iodinated sperm

A membrane fraction was prepared from sperm labelled with I as describedabove. The specific activity of the top membrane fraction was increased about tenfoldrelative to whole sperm (Table 1), indicating that the membrane fraction is highlyenriched in plasma membrane components. The specific activity of the middle mem-brane fraction was less, which is consistent with the possibility of more non-membrane proteins contaminating this fraction. Alternatively, the fractions of lowerspecific activity may be subsets of membranes that contain less iodinatable proteins.On SDS/polyacrylamide gels, the top and middle membrane fractions both containthe four major labelled components seen on gels of radio-iodinated sperm. There isone minor I-containing band on gels of sperm lysates (of Af,pp = 30 X 103), whichis not seen on gels of the membrane fraction; similarly, the small amount of 12SI co-migrating with histones H3 and H4 is not present in the membrane fraction (Fig. 4).There is a small difference in the mobility of one major labelled band (of about120 X 103 Mapp) when isolated membranes and whole sperm are compared, but on gelsof a mixture of sperm and membranes all major components co-migrate (Fig. 4). The

22 N. L. Cwss

apparently faster migration of this protein in gels of the membrane fraction is probablyan artefact resulting from the excessive amount of protein present, and the correctMipp is probably slightly greater than 120 X 103. All four labelled bands align withCoomassie-Blue-positive bands, although the band at 59 X 103 M,pp is only very faintlystained. The125I-labelledbandat200 X 103M.pp may in fact be a doublet, as occasion-ally autoradiographs contain two bands here that align with two minor and closeljspacedCoomassie-Blue-positive bands. Finally, the four major labelled componentsdescribed above, as well as a minor species of Mapp of 150 X 103, co-migrate withPAS-positive bands; no unlabelled components are PAS-positive (data not shown).

DISCUSSION

The isolation procedure described here was devised to isolate tail membranes inorder to minimize contamination by intracellular membranes, of which the majorsource is the inner mitochondrial membrane. The procedure is successful to theextent that the specific activity of cytochrome c oxidase (a marker for the innermitochondrial membrane) in the two membrane fractions is =£2% of that in thehomogenate. No marker enzymes are known for the membranes of the nucleus oracrosome, but these organelles are very rarely seen in the membrane fractions. It isunlikely that there are many intracellular membranes in the two plasma membranefractions. It is more difficult to determine how much non-flagellar plasma membraneis contained in the final membrane fractions, as some plasma membrane vesicles areproduced when the sperm are initially broken and appear to be co-purified with theflagella. However, the fine structure of the vesicles in the homogenate suggests thatmany if not all are derived from flagella. The vesicles contain adherent amorphousmaterial, which looks like the dispersed matrix material adherent to the membrane ofpartly disrupted flagella, and occasionally microtubules. Furthermore, axonemes thathave lost their membranes are present. It is difficult to tell from electron micrographswhether or not some of the head membrane is also removed, because the plasmamembrane of the head overlies the nucleus so closely. About 50% of the spermbodies, however, are osmotically active, so on at least half of the sperm bodies themajority of the membrane is intact. Therefore, the maximum estimate of thecontribution of the head membranes to the final membrane fraction is the amountderived from the remaining 50 % of sperm heads. (The contribution from spermheads contaminating the flagella fraction would be very much less, as only 1 % of theheads are carried into this fraction.) To estimate the plasma membrane surface area,the sperm head was taken to be a cone of height 3^m and base 1-1 £im, and the tailto be a cylinder of 0'2^im diameter and length 40 /urn. As such, the surface area of theflagellum is ~25/im2 and of the remainder ~6^m2. If the membrane fractions arecomposed of membranes from the flagella plus 50% of the heads, then only ~12%of the isolated membrane would be derived from heads.

Regardless of the source of membranes, the similarity between the electrophoreticpatterns of I-labelled material of whole sperm and of the membrane fraction sug-gests that the membrane fraction is a representative, though perhaps not perfect,

Sea-urchin sperm plasma membrane 23

sample of the plasma membrane of the whole sperm. Although a head-specific antigenhas been reported (Metz, 1967), no major radio-iodinated components of whole spermwere absent in the flagellar membrane fraction. Moreover, Lopo & Vacquier (19806)reported that antibodies to the two principal radio-iodinated species (Mipp of 60 X 103

and 77 X 103) bind to the entire sperm surface. Still, it will be important to inspectmore closely the composition of the plasma membrane from the head and midpiecefractions.

The best measure of the purity of the membrane fractions is the 7-12-fold enrich-ment of I in fractions prepared from radio-iodinated sperm. The evidence that the

I is in the plasma membrane is: first, that following radio-iodination, the majorintracellular proteins (tubulin and the histones) contain little label; and second, thatthe 125I can be removed from radio-iodinated, living sperm by trypsinization (Lopo& Vacquier, 1980a). The little that is known about plasma membrane enzymes of sea-urchin sperm is consistent with these figures. Gray & Drummond (1976) found a7—12-fold enrichment in the specific activity of guanylate cyclase in plasma membranefractions olS.purpuratus or Lytechinus pictus sperm prepared by a technique similarto the one described here. Also, in these membrane fractions the specific activity ofadenylate cyclase, which is a particulate enzyme in 5. purpuratus sperm (see reviewby Garbers & Kopf, 1980), is 3-15 times that in the homogenate (M. Mourelle & A.Darszon, personal communication).

It is likely that some tubulin derived from the axoneme is included in the membranefractions. It cannot, however, be ruled out that some tubulin also resides in themembrane itself, as has been suggested for ciliary and some flagellar membranes (seereview by Dentler, 1981). Another contaminant is the amorphous material bound tothe isolated membranes. It is probably derived from the flagellar matrix, because, asdiscussed above, similar material can be seen on the inner face of membranes onswollen or partly disrupted flagella. Because of the uncertain degree of purity of themembrane fractions, only the externally iodinatable components can be identifiedwith confidence as membrane proteins. Although no functions have been ascribed tothese proteins, Lopo & Vacquier (19806) have shown that treatment of sperm withan antibody to the protein of 77 X 103 Mapp prevents induction of the acrosome reac-tion by egg jelly.

There have been several other attempts to identify sea-urchin plasma membraneproteins. Gabel et al. (1979) labelled S. purpuratus sperm with [12SI]diiodo-fluorescein isothiocyanate (IFC), which reacts with amino groups, and on SDS/polyacrylamide gels of sperm lysates they detected 1Z5I in several major protein bandsof JW»PP<35 X 103. They argue that IFC labels the membrane because of the abilityof IFC to label the surface of other cell types, the ability of an antibody raised againstIFC to agglutinate labelled sperm, and because treatment of labelled sperm with3-3 % Triton X-100 (which removes the plasma membrane and the mitochondrion)solubilized 96% of the labelled material. However, these observations are not con-clusive evidence that IFC labelling of sea-urchin sperm is restricted to the plasmamembrane. More study is required for a full interpretation of the differences thatresult from labelling with lactoperoxidase or IFC.

24 N. L. Cross

Using a different approach, Stephens (1977) found that Triton X-100 treatment offlagella isolated from Strongylocentwtus or Arbacia sperm solubilized one principalglycoprotein, and he concluded that it was the major flagellar membrane protein.Otter (1978) found four proteins in Nonidet P-40 extracts of Arbacia sperm flagella:two proteins of Mspp about ISO X 103, one of 78 X 103, and one of 55 X 103 that co-migrated with tubulin. On polyacrylamide gels, which separate tubulin into a and /Jsubunits, the 55 X 103 M,pp protein was similarly split into two bands, and the one co-migrating with the a subunit was PAS-positive. Otter concluded that the membranecontains a glycosylated tubulin. It is likely that the proteins that he observed corres-pond to the major protein bands of 135, 120, 77 and 55 X 103 Af,pp seen in this work.However, the a subunit of tubulin is probably not glycosylated, because in gels withthe best resolution, the PAS-reactive material in this region (Mipp = 60 X 103) isclearly resolved from the protein co-migrating with tubulin.

In summary, a subcellular fraction has been obtained that is highly enriched inplasma membrane components. This preparation will probably prove useful in study-ing trans-membrane ion fluxes, membrane-bound enzymes, and other structuresimportant in sperm plasma membrane function and the interaction of sperm and eggs.

I thank David Epel and Meredith Gould-Somero for providing laboratory facilities, LindaHolland for taking the electron micrographs, Laurinda Jaffe, Linda Holland and Meredith Gould-Somero for commenting on the manuscript, and Alberto Darszon, M. Mourelle and Sheila Podellfor sharing unpublished results. This work was supported by USPHS grant GM-07169 and NSFgrants 80-19341 and 80-03759.

REFERENCES

BEAUFAY, H., AMAS-COSTESEC, A., FEYTMANS, E., THINES-SEMPOUX, D., WIBO, M., ROBBI, M.

& BERTHET, J. (1974). Analytical study of microsomes and isolated subcellular membranes fromrat liver. I. Biochemical methods. J . Cell Biol. 61, 188-200.

BISCHOFF, E., TRAN-THI, T. & DECKER, K. (1975). Nucleotide pyrophosphatase of rat liver. Acomparative study on the enzymes solubilized and purified from plasma membrane and endoplas-mic reticulum. Eur.J. Biochem. 51, 353-361.

CARROLL, A. G. & OZAKI, H. (1979). Changes in the histones of the sea urchin Strongylocentrotuspurpuratus at fertilization. Expl Cell Res. 119, 307-315.

CASTLE, J. D. & PALADE, G. E. (1978). Secretion granules of the rabbit parotid. Selective removalof secretory contaminants from granule membranes..?. Cell Biol. 76, 323-340.

DENTLER, W. L. (1981). Microtubule-membrane interactions in cilia and flagella. Int. Rev. Cytol.72, 1-47.

EVANS, E. H. (1979). Preparation and characterisation of mammalian plasma membranes. InLaboratory Techniques in Biochemistry and Molecular Biology, vol. 7 (ed. T. S. Work & E.Work), pp. 1-266, Amsterdam: North Holland.

FAIRBANKS, G., STECK, T. L. &WALLACH, D. F. H. (1971). Electrophoretic analysis of the majorpolypeptides of the human erythrocyte membrane. Biochemistry, Philad. 10, 2606-2617.

GABEL, C. A., EDDY, E. M. & SHAPIRO, B. M. (1979). Regional differentiation of the sperm surfaceas studied with 125I-diiodofluorescein isothiocyanate, an impermeant reagent that allows isolationof the labelled components. J . Cell Biol. 82, 742-754.

GARBERS, D. L. & KOPF, G. S. (1980). The regulation of spermatozoa by calcium and cyclicnucleotides. Adv. cyclic nuc. Res. 13, 251-306.

GIBBONS, B. H. (1980). Intermittent swimming in live sea urchin sperm. J . Cell Biol. 84, 1-12.GRAY, J. P. & DRUMMOND, G. I. (1976). Guanylate cyclase of sea urchin sperm: subcellular

localization. Archs Biochem. Biophys. 172, 31-38.

Sea-urchin sperm plasma membrane 25

HUBBARD, A. L. & COHN, Z. A. (1972). The enzymatic iodination of the red cell membrane. J. CellBiol. 55, 390-405.

LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head ofbacteriophage T4. Nature, Land. Til, 680-685.

LOPO, A. C. & VACQUIER, V. D. (1980a). Radioiodination and characterization of the plasmamembrane of sea urchin sperm. Devi Biol. 76, 15-25.

LOPO, A. C. & VACQUIER, V. D. (19806). Antibody to a sperm surface glycoprotein inhibits theegg jelly-induced acrosome reaction of sea urchin sperm. Devi Biol. 79, 325-333.

LOWRY, 0 . H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurementwith the folin phenol reagent. J. biol. Chem. 193, 265-275.

METZ, C. B. (1967). Gamete surface components and their role in fertilization. In Fertilization (ed.C. B. Metz & A. Monroy), pp. 163-236. New York: Academic Press.

OTTER, T. (1978). Comparison of ciliary and flagellar membrane composition. Biol. Bull. mar. biol.Lab., Woods Hole 155, 461 (abstr).

PEKARTHY, J. M., SHORT, J., LANSING, A. I. & LEIBERMAN, I. (1972). Function and control ofliver alkaline phosphatase.j. biol. Chem. 247, 1767-1774.

SCHACKMANN, R. W., EDDY, E. M. & SHAPIRO, B. M. (1978). The acrosome reaction ofStrongylocentrotus purpuratus sperm. Devi Biol. 65, 483—495.

STEPHENS, R. E. (1977). Major membrane protein differences in cilia and flagella: evidence for amembrane-associated tubulin. Biochemistry, Philad. 16, 2047-2058.

TILNEY, L. G., KIEHART, D. P., SARDET, C. & TILNEY, M. (1978). The polymerization of actinIV: The role of Ca2+ and H+ in the assembly of actin and in membrane fusion of the acrosomereaction of echinoderm sperm. J. Cell Biol. 77, 536-550.

WIDNELL, C. C. & UNKLESS, J. C. (1968). Partial purification of a lipoprotein with 5'-nucleotidaseactivity from membranes of rat liver cells. Proc. natn. Acad. Sci. U.SA. 61, 1050-1057.

WITMAN, G. B., CARLSON, K., BERLINER, J. & ROSENBAUM, J. L. (1972). Chlamydomonas

flagella. I. Isolation and electrophoretic analysis of microtubules, matrix, and mastigonemes. J.Cell Biol. 54, 507-539.

{Received 17 January 1982)

CELS9