Reproductive BioMedicine Online (2013) 27, 690–701

www.sc iencedi rec t .comwww.rbmonl ine .com

SYMPOSIUM: FUTURES IN REPRODUCTIONARTICLE

Sperm attachment and penetrationcompetence in the human oocyte: a possibleaetiology of fertilization failure involving theorganization of oolemmal lipid raftmicrodomains influenced by the DWm ofsubplasmalemmal mitochondria

1472-6483/$ - see front matter ª 2013, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.rbmo.2013.09.011

Jonathan Van Blerkom a,b,*, Kyle Caltrider a

a Department of Molecular, Cellular and Developmental Biology, University of Colorado, Boulder, CO, United States;b Colorado Reproductive Endocrinology, Rose Medical Center, Denver, CO 80220, United States* Corresponding author. E-mail address: jonathan.vanblerkom@colorado.edu (J Van Blerkom).

Abstract The roles of oo

Jonathan Van Blerkom is a professor in the department of molecular, cellular and developmental biology at theUniversity of Colorado and Laboratory Directorat Colorado Reproductive Endocrinology in Denver. His studieshave focused on molecular and cellular aspects of early mammalian development including human follicles,oocytes and embryos and has published numerous research articles, reviews, and coauthored or edited booksdealing with early mammalian development, including the human. His current research centres on the role(s) ofmitochondria in early development, and the molecular organization of the oocyte and embryo plasmamembrane as related to developmental competence.

lemmal lipid raft microdomains enriched in the ganglioside GM1 and the tetraspanin protein CD9 wereinvestigated as causative agents in fertilization failure in human IVF where spermatozoa progress to the oolemma but fail to attachor, if attached, to penetrate. The findings show that specific configurations of GM1 lipid raft microdomains are consistent withattachment and penetration, while microdomains composed of CD9 lipid rafts, a protein known to be critical for penetration, donot appear to have a central role in the initial stages of attachment. The relative magnitude of the potential difference acrossthe inner membrane (DWm) in mitochondria localized to a stable subplasmalemmal domain appears to influence the organizationof GM1 but not CD9 lipid raft microdomains in the corresponding oolemma. The findings present a novel view of how fertilizationcompetence may be established in the human oocyte and a means by which certain fertilization failures that occur after conven-

tional clinical IVF can be identified and explained in the unfortunate instance of fertilization arrest at the oolemma. RBMOnline

ª 2013, Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved.

KEYWORDS: CD9, fertilization, GM1, lipid rafts, mitochondriaVIDEO LINK: http://sms.cam.ac.uk/media/1400893

Oolemmal lipid rafts and fertilization competence of human oocytes 691

Introduction

Fertilization failures that occur with normal human sperma-tozoa and meiotically mature (metaphase II, MII) oocytesafter conventional insemination in vitro are an unexpectedoutcome in clinical IVF and can significantly reduce the like-lihood of pregnancy if a high proportion of oocytes areaffected. Failures of this type have persisted despite signif-icant improvements in fertilization protocols, includingstringent attention to culture conditions. One common formof fertilization failure in clinical IVF is characterized bysperm penetration but then the absence of second polarbody and failure of the formation of a paternal or maternalpronucleus, or both. Failure of this type has been referredto as ‘occult’ fertilization and has been reported to affecta surprisingly high proportion of normal-appearing MIIoocytes after conventional IVF (Van Blerkom et al., 1994)and may perhaps result from defects in preovulatory nuclearor cytoplasmic maturation (Miyara et al., 2003).

Here, this study investigated a different type of fertiliza-tion failure, namely, the presence on the surface of theoolemma of a spermatozoon that failed to penetrate anormal-appearing, meiotically mature human oocyte. Howfrequently this phenomenon occurs in clinical IVF is difficultto assess because most programmes expend little or noeffort to identify possible causes of fertilization failureand simply discard unfertilized oocytes. When detected,however, its occurrence can be particularly difficult forpatients to cope with because no reason can be given asto why an apparently functional spermatozoon, as indicatedby its ability to undergo the acrosome reaction and passthough the cellular (cumulus and coronal cells) and acellular(zona pellucida) compartments that surround the oocyte,cannot penetrate the last barrier, the oolemma. While anunrecognized sperm defect cannot be precluded, in theseinstances it is not unreasonable to suspect that fertilizationfailure resides at the level of the oolemma, especially whenother MII oocytes within a cohort and with similar morphol-ogies fertilize normally.

Preliminary studies of the molecular organization of themouse and the human oocyte have provided evidence for aplasma membrane composed of lipid raft microdomains ingeneral, and those containing two known lipid raft compo-nents in particular, the ganglioside monosialotetrahexosyl-ganglioside (GM1) and the cell membrane tetraspanin CD9(Van Blerkom, 2013). GM1 is a sphingolipid constituent oflipid rafts that is directly involved in a variety of the cellsurface activities including facilitation of intercellular inter-actions, virus docking, signal transduction and protein bind-ing (Babiychuk and Draeger, 2000; Brown and London, 2000;Godoy and Riquelme, 2008; Harder et al., 1998; Janes et al.,1999; Kawano et al., 2011; Landry and Xavier, 2006; Lopezand Schnaar, 2009; Luna et al., 2005; Neu et al., 2008;Olofsson and Bergstrom, 2005). CD9 is a member of the tet-raspanin family (Ishii et al., 2006) that has been suggestedto be the single most important oolemmal componentinvolved in sperm penetration in mammals (Jejou et al.,2011; Miyado et al., 2008).

This study investigated whether fertilization failure atthe level of the oolemma could be related to defects in lipidraft microdomains enriched in GM1 or CD9, or both, and the

extent to which certain cytoplasmic factors could influencethe organization of these lipid raft microdomains and,therefore, oolemma function.

Materials and methods

Human oolemma and fertilization failure

Ovarian stimulation, ovulation induction, follicular aspira-tion and conventional IVF followed protocols previouslydescribed (Van Blerkom et al., 2002, 2008). According toprotocol, the first inspection for fertilization after conven-tional IVF in non-severe male factor cases (i.e. those notnormally requiring intracytoplasmic sperm injection, ICSI)is made at 10–12 h. Normal fertilization is confirmed byan extruded second polar body and early evidence of pronu-clear formation (Van Blerkom et al., 1995). If pronuclearformation is not detected, a second inspection occursbetween 18 and 21 h. All MII oocytes classified as unfertil-ized at the second inspection were thoroughly examinedby light microscopy to determine whether a sperm headwas present on the oolemma and by DAPI staining to deter-mine whether penetration without activation had occurred(Van Blerkom et al., 1994).

Fluorophore-conjugated cholera toxin B subunit (CTB),because of its high specificity and binding affinity for thepentasaccharide chain of GM1 (located in the exoplasmicleaflet), is commonly used to report lipid rafts enriched inthis ganglioside in living cells and was used here for thesame purpose with MII human oocytes (Brown, 2006; Harderet al., 1998; Janes et al., 1999; Kawano et al., 2011;Lagerholm et al., 2005). Preliminary studies established astandard protocol of labelling in which intact oocytes (zonaenclosed) were stained for 35 min with Alexa Fluor 488- or565-conjugated CTB at 5 lg/ml (Molecular Probes, Eugene,OR, USA) at 37�C in Quinns’ Advantage Medium supple-mented with 10% human serum albumin (HSA; Sage,Trumbull, CA, USA) in an atmosphere of 5% CO2 and 95% air.In a second protocol tested for analytic use, unfertilizedhuman MII oocytes were denuded of the zona pellucida bybrief exposure to acid Tyrode’s solution (Van Blerkomand Davis, 2007), washed four times for 3 min withHEPES-buffered Quinns’ Advantage Medium containing 10%HSA and stained with Alexa 488- or 565-CTB for 15 min ata concentration of 1 lg/ml.

After staining, intact and zona-free oocytes were washedin HEPES-buffered medium and then either (i) examined byscanning laser confocal microscopy (SLCM; Zeiss LSM 510) in30 ll of the same medium under oil in DT dishes designedfor fluorescence microscopy (glass thickness 0.17 mm) withtemperaturemaintained at a constant 37�Cby aDT controller(Bioptics, Butler, PA, USA) as previously described(Van BlerkomandDavis, 2007; Van Blerkomet al., 1995) or (ii)fixed in freshly prepared 3.7% formaldehyde for 15 min,washed in HSA-supplemented HEPES-buffered medium andstored in fresh medium at 4�C for up to 4 days untilexamination.

The relative intensity of oolemmal CTB fluorescence wasquantified as relative fluorescence units (RFU) from fullycompiled projections of 2 lm SLCM sections taken at 5 lmintervals through representative oocytes with values and

692 J Van Blerkom, K Caltrider

significance determined by Zeiss Zen version 5.5 and NIHImage J version 1.44 software. Light microscopy usedbright-field and differential interference contrast optics.

CD9 immunofluorescence and mitochondrial andDNA staining

CD9 was detected in formaldehyde-fixed oocytes, some ofwhich had been stained previously with CTB, by use of com-mercial primary monoclonal antibodies that had specifichuman reactivity and which are commonly used for CD9studies of cultured human cells (monoclonal anti-rabbitsc9148 and anti-mouse sc-13118; Santa Cruz Biotechnology,Dallas, TX, USA; FITC-conjugated mouse monoclonal fordirect immunofluorescence, ab18241; Abcam, Cambridge,MA, USA). For indirect immunofluorescence, an appropriateAlexaFluor-488 or -565-conjugated secondary antibody wasused (Santa Cruz Biotechnology). The specificity of antibod-ies was demonstrated by immunoblots provided by the manufac-turer and, for oocyte immunostaining, by deleting the primaryantibody from the non-fluorophore-conjugated monoclonal anti-bodies. However, insufficient oocyte material for direct testingby immunoblotting precluded formal identification of the targetoocyte antigen as claimed by the manufacturers.

This study used JC-1 staining to report the relativelevel of the inner membrane potential (DWm) and4(,6-diamidino-2-phenylindole) (DAPI) staining for fluores-cent detection of sperm DNA, as previously described (VanBlerkom and Davis, 2006, 2007).

In-vitro maturation and transient insemination ofdysmorphic oocytes

This study used MII oocytes that were intentionally notinseminated in ICSI cycles if two types of cytoplasmic defects(dysmorphisms; Van Blerkom and Henry, 1992) were detect-able after cumulus and coronal cell denudation: (i) severecytoplasmic clustering; and (ii) the presence of translucentdiscs of smooth-surfaced endoplasmic reticulum. While theoocytes are fertilizable, the first defect has been associatedwith compromised embryo developmental competence andthe second with disorders in genomic imprinting (Istanbulconsensus workshop on embryo assessment: proceedings ofan expert meeting, 2011). For this study and prior to beingdiscarded, the protocol permitted GM1 staining of zona-freedysmorphic oocytes followed by co-incubation with sperma-tozoa for no more than 4 min before fixation. The aim ofthese studies was to determine whether the GM1 phenotypesassociated (or not) with sperm attachment and penetration,when examined after conventional IVF, existed at insemina-tion. The spermatozoa used were from residual cryopre-served vials that had been obtained for IVF fromanonymous donors of known fertility through commercialsperm banks that were no longer needed or wanted for treat-ment. Post-thaw fertility was demonstrated by a history ofnormal fertilization in conventional IVF.

Zona-free dysmorphic oocytes judged unacceptable forICSI were prestained with CTB, exposed to motile spermato-zoa at a maximum concentration of 1000/ml and removedfor fixation as soon as multiple spermatozoa were observedto stably attach to the oolemma. Typically, the oocytes sta-bly bound multiple spermatozoa within 120 s.

Oocytes that were meiotically immature at follicularaspiration are detected after cumulus and coronal celldenudation for ICSI or, more commonly, at the first inspec-tion for fertilization after conventional IVF. According toprotocol, oocytes that spontaneously mature to MII afterat least 8 h in vitro are not candidates for IVF, but can bestained and fixed for GM1 and CD9 analysis. The protocolsused in this study were derived from informed consent doc-uments that, unless otherwise specified by patients, permit-ted studies of unfertilized and dysmorphic oocytes andoocytes that mature in vitro but are unsuitable for IVF.

Results

All MII oocytes classified as unfertilized at the secondinspection (18–21 h) were thoroughly examined by lightmicroscopy to determine whether a sperm head was pres-ent on the oolemma, followed by DAPI staining for confir-mation (Figure 1D and G) and to determine whetherpenetration without activation had occurred. Eighty-nineunpenetrated MII oocytes (e.g. Figure 1A) were identifiedwith a sperm head(s) clearly evident on the oolemma(Figure 1B–G). Although differential interference contrastmicroscopy was used most often (Figure 1A, E, and F),sperm heads could also be detected by bright-field andphase-contrast optics (Figure 1B and C). Because the pro-tocol did not permit fertilization solely for experimentalpurposes, the organization of GM1 and CD9 was deter-mined in 42 normal-appearing MII oocytes that had beenpenetrated but failed to activate; the GM1 and CD9 lipidraft microdomain configurations observed were assumedto be those consistent with sperm binding and penetration.A further 82 oocytes that were immature at follicular aspi-ration but which spontaneously matured to MII after 8 h ofculture were available for GM1 and CD9 analyses.Twenty-eight dysmorphic MII oocytes that would normallybe discarded were denuded of the zona pellucida andexposed briefly to spermatozoa to identify sites of initialspermatozoon–oolemma interaction.

Distinct GM1 lipid raft microdomain phenotypes areassociated with penetration competence

GM1 phenotypes were detected after staining intact(Figure 1H) or zona-free oocytes with fluorophore-conjugated CTB and imaged by SCLM before (Figure 1H,J1–4, K) or after fixation (Figure 1I and L); the GM1 pheno-types observed in these two states were similar. Becausezona removal and fixation did not influence GM1 pheno-types, most oocytes were stained after zona removal andthen fixed, which had the additional advantage of reducingCTB concentration and staining time and eliminatinglow-level, non-specific fluorescence from the zona pellucida.

All 42 penetrated but unfertilized oocytes and 47 of 82oocytes that matured to MII in vitro (57%) showed a commonGM1 phenotype, as reflected in a uniform, high-density punc-tate pattern of oolemmal CTB fluorescence (Figure 1H–J).Figure 1J 1–4 shows punctate CTB fluorescence on theoolemma at approximately four locations during the com-plete rotation of this oocyte as it was scanned, demonstratingthe uniformdensity and distribution ofGM1. In contrast, none

Figure 1 Sperm on oolemma in unpenetrated oocytes and GM1 phenotypes. (A) An example of the typical unfertilized MII humanoocyte examined in this study. (B–G) Detection of spermatozoa (arrows) on oolemma by bright field (B), phase contrast (C),differential interference contrast optics (A, E, and F) and DAPI fluorescence (D and G); SP = sperm head; Arrow, ZP = zona pellucida.(H–J) Projection of fully compiled SLCM section series showing punctate and uniformly distributed Alexa-488-CTB fluorescencereporting the ganglioside GM1 in an intact (H) and zona-free (I) MII human oocyte and the pattern of GM1 lipid raft microdomaindistribution at four locations (J 1–4) in an MII oocyte fixed after staining with CTB and imaged by SLCM; the GM1 phenotypes aresuggested to be consistent with fertilization of the MII human oocyte. (K–Q) Representative SLCM images of GM1 microdomaindistributions reported by CTB in oocytes with spermatozoa (arrows) on oolemma at GM1-positive sites but which failed to penetrate;arrow head indicates a portion of the MII chromosomes; asterisks denote regions where GM1 microdomains reported byfluorophore-conjugated CTB were sparse or below levels of detection and where no sperm docking or attachment was observed.Bars = 10 lm (A, E, H–Q), 15 lm (B), 5 lm (C, F, and G).

Oolemmal lipid rafts and fertilization competence of human oocytes 693

of the 89 MII unactivated and unfertilized oocytes in which asperm head was clearly identified on the oolemma showedthis pattern of CTB fluorescence. Based on these findings, itis suggested that a uniform, high-density and punctate distri-bution of lipid raft microdomains enriched in the gangliosideGM1 is consistent with the sperm attachment phase offertilization.

Distinct GM1 lipid raft microdomain phenotypes areassociated with penetration incompetence

Figure 1K–Q shows representative images of distinct GM1lipid raft microdomain phenotypes observed in unpenetrat-ed oocytes in which a sperm head resided on the oolemma.They are presented in relative order of abundance observedin this study, with the most common phenotype shown in

694 J Van Blerkom, K Caltrider

Figure 1K and L, which are fully compiled projections of asection series that encompasses the entire oocyte. Approxi-mately 80% (71/89) of normal-appearing but unpenetratedoocytes showed distinct islands of high-intensity CTB fluores-cence surrounded by regions with little or no detectable fluo-rescence. A portion of the MII chromosomes is indicated inFigure 1K. In each instance where a spermatozoa remained onthe oolemma after zona removal plus CTB staining, washingand fixation (41%, 29/71), it was located on a GM1 island.Because in the remaining 59%, no spermatozoa were foundat a CTB-deficient region or fluorescent ‘dark zone’,(Figure 1K–Q), it is suggested that spermatozoa enteringthe perivitelline space in these regions had no attachmentsite on the oolemma and were lost during zona removaland processing for GM1 staining or fixation. The CTB/GM1phenotype in Figure 1M and N occurred in 21% (19/89) ofoocytes with spermatozoa on the oolemma after conven-tional IVF, and the phenotype shown in Figure 1N was alsoobserved in three of the 28 dysmorphic oocytes transientlyexposed to spermatozoa. In contrast, none of the 82oocytes that spontaneously matured to MII in vitro exhib-ited these two latter phenotypes. The organization ofGM1 shown in these figures may represent slightly differentconfigurations of a similar state of GM1 organization,which could indicate dynamic remodelling of CTB lipid raftmicrodomains resulting from their aggregation, disaggrega-tion or, in the island phenotype, their coalescence (e.g.Figure 1K and L). Taken together, the results suggest thatthe GM1 configurations shown in Figure 1K–N may be per-missive for sperm docking/attachment but not forpenetration.

A low-density or scant distribution of CTB fluorescencewas the most common phenotype observed both in approx-imately 60% of oocytes that matured in vitro (Figure 1O andP) and in approximately 25% of unpenetrated oocytes thathad spermatozoa in proximity to, but not bound to, theoolemma (these spermatozoa were lost during zona removaland washing). Figure 1Q shows a polarized GM1 phenotypewhere large portions of the oolemma (nearly 50% in someinstances) were devoid of detectable CTB fluorescence(asterisk). This phenotype was relatively rare in all classesof oocytes observed (<5%) and was not detected in any ofthe dysmorphic oocytes.

Preliminary studies that quantified the relative fluores-cence intensity of CTB bound to the oolemma were per-formed on at least seven oocytes from each of thecommon phenotypes associated with penetration or pene-tration failure. On a scale from 0 to 10 relative fluorescenceunits (RFU), oocytes in both penetration-consistent (e.g. 1I,J1–4) and -inconsistent groups (e.g. Figure 1K–M) mea-sured between 7.3 and 7.8. The similar RFU values suggeststhat, with respect to fertilizability, developmental signifi-cance may be closely related to GM1 lipid microdomain dis-tribution and/or density. This possibility was supported byRFU values derived from oocytes with low-density (e.g.Figure 1O) or scant CTB fluorescence (e.g. Figure 1P),where spermatozoa were found in proximity to, but unat-tached to, the oolemma. The values for nine intact MIIoocytes with each of these phenotypes were �1.6 ± 0.7,which differed significantly from most oocytes that wereable to bind spermatozoa, even when penetration failed(P < 0.05; Student’s t-test).

Morphodynamics of sperm docking on the oolemma

The spatial dynamics of sperm attachment with respect toGM1 enriched microdomains in the oolemma was investi-gated with 28 zona-free dysmorphic MII oocytes preloadedwith tagged CTB and then exposed individually to sperm atrelatively low concentrations (500–1000/ml), followed bytheir removal for fixation as soon as sperm attached tothe oolemma without spontaneously detaching a few sec-onds later and for up to 40 s thereafter. Typically, spermthat remained firmly attached were evident within the first30 s of contact. The SLCM images shown in Figures 2A–Dare oocytes fixed at first contact (2A, 10 s) and at 20(Figure 2B), 30 (Figure 2C) and 40 s (Figure 2D). An exam-ination of over 350 sites of persistent association with theoolemma showed sperm docking only occurred at CTB/GM1positive domains (arrow, Figure 2A–D). No sperm werefound to dock or remain attached to the oolemma inregions where CTB fluorescence was absent (asterisks,Figure 1K–Q). Sperm rarely remained attached to theoolemma in oocytes that were subsequently found to haveCTB fluorescence that, while punctate, was relativelysparse (similar to those shown in Figures 1O, P): projectionof fully compiled section series).

The approximate area of the oolemma occupied by ahuman sperm head was �8–11 lm2 (e.g. Figure 2D). Theregion outlined in Figure 2E shows such an area (in brack-ets) with the typical density and punctate distribution ofGM1-lipid microdomains consistent with robust sperm bind-ing leading to penetration and where each microdomain is�300–500 nm wide (arrow to Figure 2F). In contrast, thebracketed region of the oolemma outlined in Figure 2H rep-resents a similar area expected to be occupied by the fertil-izing spermatozoon but, in this instance, the density of GM1microdomains reported by CTB (arrow to Figure 2G) is likelyto be below the threshold for attachment competence. Forcomparative purposes, Figures 2I–K are higher magnifica-tion images of GM1 positive microdomains representingwhat a sperm might encounter on a fertilization-competent(arrows Figure 2I) and incompetent (Figure 2J) oolemma.Figure 2K is an example of the island GM1-microdomainphenotype (arrow from insert in Figure 2K) that appearsto be consistent with sperm attachment but not with pene-tration (e.g., Figures 1K, L). As noted above, this oolemmalphenotype was the one most commonly associated withunpenetrated MII oocytes with sperm identified on theoolemma (e.g., Figure 1B–E). The aggregates of CTB fluo-rescence that characterize this phenotype ranged from 7to 12 lm in diameter and were of such high density thatindividual microdomains, such as those shown in Figures2D–H, could not be resolved. In these ‘macrodomains’ thedensity of GM1 may be permissive for sperm attachmentbut too great to permit closer membrane interactions nec-essary for penetration.

Density and distribution of CD9

The specificity of CD9 immunostaining indicated by the anti-body manufacturers was examined by indirect immunofluo-rescence by omitting the primary antibody, which, in allinstances, showed no immunofluorescence (data not shown).The patterns of direct and indirect CD9 immunofluorescence

Figure 2 Interactions between male and female human gametes with respect to GM1 and CD9 lipid raft microdomains. (A–D) SLCMimages of sections (1 lm) showing interactions between the sperm membrane (N, DAPI-stained nucleus) and GM1 microdomains onthe oolemma (arrows) of a zona-free human oocyte within seconds of contact: 10 s (A), 20 s (B), 30 s (C) and 40 s (D). (E–K) GM1microdomain organization and densities that appear to be consistent (E, F, and I) or inconsistent (G, H, J, and K) with sperm dockingor post-attachment penetration of the oolemma. F and G are high-magnification views of regions (highlighted in E and H,respectively) of comparatively high and low GM1 lipid raft microdomain density, respectively that in surface area that may berelated to the ability of the spermatozoa to remain docked to the on the oolemma. I, J and K are higher magnification views of thedensity of GM1 microdomains (arrows) that a spermatozoon might encounter on the oolemma that are consistent (I) or inconsistent(J and K), respectively, with fertilization; K shows a GM1 macrodomain or island (arrow in inset). (L–O) Epifluorescence (L) and SLCM(M–O) of CD9 immunofluorescence alone (L, M, N) or after co-staining for GM1 microdomains (O) detected by fluorophore-con-jugated CTB. The asterisks in N denote areas of the oolemma where CD9 immunofluorescence was undetectable. Bars = 4 lm (A, C,and L), 3 lm (B), 5 lm (D) 10 lm (E and M), 1 lm (G and J), 8 lm (H) 0.5 lm (I), 1 lm (H), 0.4 lm (K) and 6 lm (H).

Oolemmal lipid rafts and fertilization competence of human oocytes 695

were similar and confined to the oolemma when viewed byconventional epifluorescence and SLCM. With few excep-tions, CD9 immunofluorescence was dense, punctuate anduniformly distributed throughout the oolemma (Figure 2Land M) in all oocytes randomly selected for immunostainingfrom the following: (i) unpenetrated with spermatozoa onoolemma, n = 11; (ii) unfertilized with no spermatozoa onoolemma, n = 26; (iii) in-vitro matured to MII, n = 29; (iv)dysmorphic and not exposed to spermatozoa, n = 3. In con-trast, Figure 2N shows discrete regions of the oolemma

where CD9 fluorescence was undetectable (direct immuno-fluorescence), but this was a rare phenotype and occurredin only in one of 69 oocytes examined.

Evidence that GM1 and CD9 occupy separate lipid raftdomains comes from inspections of individual sections andsection series taken through the first 20–25% of the oocyte(Figure 2O) and from projections of section series taken atdifferent portions of the oolemma during a 360� rotation inoocytes dual stained for GM1 and CD9 (indirect immunoflu-orescence, Figure 3A–F). CD9 was almost always uniformly

Figure 3 Relationship between GM1 and CD9 phenotypes and DWm in the subplasmalemmal cytoplasm and fertilizationcompetence in meiotically mature human oocytes. (A) Projection of a fully compiled section series of a human MII oocyte dualstained for CD9 (red, arrows) and GM1 (green; arrows indicate small GM1 islands. (B–F) GM1 macrodomains against a background ofuniform CD9 immunofluorescence at different positions during the full rotation of a dual-stained unfertilized MII human oocyte;These GM1 phenotypes are inconsistent with sperm attachment or penetration. (G–S) Typical phenotypes of unfertilized MII humanoocytes strain with the potentiometric fluorescent probe JC1 and fluorophore-conjugated CTB and examined for DWm and GM1distributions by light microscopy (G), conventional epifluorescence (H, J, M, P, R) and SLCM (I, K, L, N, O, Q, S); G–I, J–L, M and N, Pand Q, and R and S show the same oocytes. Patterns of DWm distribution (orange, J-aggregate fluorescence, JA) could be normal (H)or abnormal (J, M, P, R, arrows) in mitochondria in the subplasmalemmal cytoplasm in MII human oocytes. The asterisk in J denotes afocal zone of high-potential (J-aggregate-positive) mitochondria where the corresponding oolemma also showed focal staining forGM1 microdomains (asterisk, L); a similar situation occurred for the zone of JA-positive mitochondria between the arrows in J,where the corresponding cytoplasm was found to be GM1 positive. The arrow(s) in K, M–P and S denote GM1 microdomains detectedby positive CTB fluorescence. The differential intensity of CTB fluorescence parallels differential DWm in the correspondingsubplasmalemmal cytoplasm. For some oocytes, there was no detectable subplasmalemmal J-aggregate fluorescence (arrow in R)and the corresponding oolemma was largely devoid of GM1, as indicated by CTB fluorescence (S). Bars = 8 lm (A and L), 10 lm (B–K,M, P, R, and S), 6 lm (N) 3 lm (O) and 9 lm (Q).

696 J Van Blerkom, K Caltrider

distributed regardless of GM1 distributions, as shown inFigure 3A, an oocyte with small GM1 aggregates. Figure 3Band C are two images from a full rotation separated by �25�

in an oocyte in which GM1 microdomains have aggregatedinto large clusters forming the common island phenotype.A similar situation occurred in oocytes with smaller GM1

Oolemmal lipid rafts and fertilization competence of human oocytes 697

islands (Figure 3D–F). In all instances where full rotation ofoocytes was performed during SLCM, CD9 fluorescence wasuniform (e.g. regions indicated in Figures 3C–F).

Mitochondrial participation in GM1 lipid raftmicrodomain organization

The point of sperm entry into the perivitelline space is arandom event based on the site of sperm attachment toand penetration through the zona pellucida and is there-fore, independent of the molecular organization of theoolemma. However, the present findings indicate that fer-tilization-associated events at the human oolemma appearto involve two intrinsic properties, GM1 microdomain den-sity and distribution. Of potential cytoplasmic factors thatmight regulate these properties, the relative magnitudeDWm in mitochondria located in the subplasmalemmalcytoplasm was considered a possible candidate, becauseprevious studies showed that sperm penetration could bealtered reversibly by lowering and then raising DWm(Van Blerkom and Davis, 2007). Patterns of JC1, J-aggregatefluorescence were examined by conventional epifluores-cence microscopy in 23 oocytes: 9 unfertilized with spermon the oolemma and 14 that matured in vitro to MII but werenot inseminated. Following mitochondrial analysis ofJ-aggregate fluorescence, the oocytes were stained withfluorophore-tagged CTB, fixed and re-examined by SLCM.

Figure 3G is a representative light microscopic image ofthe type of MII oocyte examined. A distinct circumferentialdomain of high-potential (DWm) mitochondria identified bytheir characteristic orange-red fluorescence (Figure 3H)was detected in the subplasmalemmal cytoplasm. This dis-tribution results from the multimerization of green fluores-cent JC1 monomers into J-aggregates in the mitochondrialmatrix, which shifts the fluorescence emission spectrumto longer wavelengths (i.e. orange-red). Figure 3I showsthe characteristic pattern of CTB fluorescence that corre-sponds to this JC1 J-aggregate phenotype, which the pres-ent findings indicate is likely associated with fertilizationcompetence after conventional IVF.

In contrast, unpenetrated oocytes with spermatozoa onthe oolemma, while similar in appearance at the lightmicroscopic level to their fertilized siblings, showed pat-terns of JC1 J-aggregate fluorescence in the subplasmalem-mal cytoplasm that were asymmetric (Figure 3J), localizedin multiple small foci (e.g. Figure 3M), scant (Figure 3P) orundetectable (Figure 3R). In these instances, the corre-sponding organization of GM1 microdomains reported byCTB could be correlated with the relative DWm of the sub-jacent mitochondria. Representative examples of thesephenotypes are shown in Figure 3H–S. Thus, the oocyteshown in Figure 3J–L is a typical example of spatially dis-tinct regions of high-polarized mitochondria in the subplas-malemmal cytoplasm, in which two distinct areas ofJ-aggregate fluorescence were detected (Figure 3J), asindicated by arrows (limits of high expression) and an aster-isk (low expression), with a corresponding CTB fluorescenceprofile (Figure 3K and L). In other oocytes, few significantfoci of CTB fluorescence were detected (e.g. Figure 3N)and these were spatially co-localized with a correspondingdomain of high-potential mitochondria (upper arrow,

Figure 3M). The common GM1 island phenotype (e.g.Figure 3B and D) showed intense foci of J-aggregate fluores-cence of variable number and dimensions similar to the fociat the top on the oocyte in Figure 3M, which were scatteredthroughout the subplasmalemmal cytoplasm and associatedwith GM1 ‘macrodomains’ of variable size (Figure 3M andN). Regions with sparse JC1 J-aggregate-positive mitochon-dria (lower portion of oocyte in Figure 3M) were associatedwith equally sparse and punctate CTB fluorescence(Figure 3O). Scant J-aggregate fluorescence throughoutthe oolemma (Figure 3P) was associated with low-densityCTB fluorescence (Figure 3Q), and the absence of detect-able J-aggregate fluorescence (Figure 3R) with virtuallyno CTB-positive signal (Figure 3S).

Discussion

Fertilization failures, where spermatozoa penetrate thezona pellucida but not the oolemma, can adversely affectthe outcome of clinical IVF depending upon the number ofmeiotically mature MII oocytes affected. A moleculardefect(s) at the level of the sperm inner acrosomal mem-brane or at the oolemma, or both, are possible aetiologies,although penetration to the oolemmal surface, especiallywhere sibling MII oocytes fertilize normally, directs atten-tion to the oocyte. This study reports results from a system-atic light microscopic inspection of unfertilized MII oocytesafter conventional IVF, which identified a subset with one ormore sperm heads on the oolemma but where penetrationfailure was confirmed by the absence of a sperm nucleusin the ooplasm. It seems reasonable to suggest that this phe-nomenon may go largely unnoticed in IVF programmes ifunfertilized oocytes are routinely discarded without furtherdetailed microscopic analysis.

In the IVF programme, unfertilized oocytes are routinelystored in a 1% glutaraldehyde solution where they retain anappearance that is close to the living state. While studieswith biomarkers such as CTB or CD9 antibodies cannot beperformed on glutaraldehyde-fixed oocytes, DNA stainingwith fluorescent probes such as DAPI can confirm lightmicroscopic identification of spermatozoa on or in proximityto the oolemma or in the ooplasm. In a preliminary study ofunfertilized MII oocytes, c.36% (89/247) were found to havespermatozoa on the oolemma, a phenomenon that had notbeen recognized previously prior to or during fixation forstorage. Thus, this type of fertilization failure may be rela-tively common in clinical IVF, and, if a similar situationoccurs in vivo, it may be the basis for an idiopathic diagnosiswhen infertility treatments up to and including conventionalIVF have failed. The notion that defects in the molecularorganization of the human oolemma could be associatedwith penetration failure was suggested by preliminary stud-ies of GM1 density and distribution that indicated pheno-types that seemed to be permissive or not for fertilization(Van Blerkom, 2013).

The present study investigated this possibility in greaterdetail by examining the extent to which GM1 lipid raft orga-nization in microdomains, reported by fluorophore-taggedCTB, was related to unpenetrated, normal-appearing MIIoocytes where a sperm head could be identified unambigu-ously on or in close proximity to the oolemma. This study

698 J Van Blerkom, K Caltrider

also examined whether microdomains containing CD9, amember of the tetraspanin family of cell surface proteinsknown to reside in lipid rafts (Ishii et al., 2006) and to bea central player in fertilization (as will be discussed),co-localized with GM1 microdomains in general and withphenotypes characterized in this study as consistent orinconsistent with fertilization in particular.

The lipid raft theory of plasma membrane structureand function and the fertilizability of the humanoocyte

Modern membrane theory considers the plasma membraneto be a functionally mosaic bilayer in which dynamicchanges in bioactivity can be mediated by so-called lipidrafts, nanometer-sized entities composed of protein, sphin-golipids and cholesterol (Edidin, 2003; Brown, 2006; Matkaand Szollosi, 2005; McIntosh, 2007). Lipid rafts, and themicrodomains they form, typically comprise between 15%and 20% of the surface area of the exoplasmic leaflet ofsomatic cells (Schutz et al., 2000). Dynamic changes inthe functional biology and molecular organization of theplasma membrane are introduced by means of lipid raftaggregation, disaggregation, differential partitioning kinet-ics between lipid-disordered and lipid-ordered regions ofthe plasma membrane (Silvius, 2005), changes in proteinand sphingolipid components (e.g. gangliosides) and ratesof turnover as nascent lipid rafts enter the plasma mem-brane and older ones exit to be recycled in the cytoplasmthrough the endocytotic pathway (Pike, 2003; Chinnapenet al., 2007; Day and Kenworthy, 2009). These changescan have spatial significance in establishing asymmetriesor polarities in membrane activities (Campbell et al., 2008;Godoy and Riquelme, 2008) as well as functional signifi-cance with respect to ligand-binding/receptor-mediatedsignalling (Brown and London, 2000; Simons and Toomre,2000). Because the aggregation of lipid rafts forms microdo-mains in the range from submicron to micron size, they cre-ate ‘concentrating platforms’ where changes in lipid raftdensity affect corresponding bioactivities in a concentra-tion-dependent manner (Babiychuk and Draeger, 2000;Matka and Szollosi, 2005; McIntosh, 2007).

With respect to the lipid raft components examinedhere, in vertebrate cells the ganglioside GM1 is multifunc-tional in the exoplasmic leaflet, with roles in cell adhesion,cell recognition, ligand binding, signal transduction andvirus and toxin(s) docking (Harder et al., 1998; Janeset al., 1999; Lopez and Schnaar, 2009; Neu et al., 2008;Olofsson and Bergstrom, 2005). CD9 facilitates membrane–membrane interactions including fusion in certain somaticcells (Ishii et al., 2006) and is a critical element in similarprocesses associated with sperm penetration in mammalianoocytes (Kawano et al., 2011; Miyado et al., 2008). The useof CTB as a biomarker for GM1 has been supported bynumerous studies that show binding at high specificity andaffinity in a wide variety of cell types (Jobling et al., 2012;Luna et al., 2005; Wilson et al., 2004). Conjugation of CTBwith a fluorophore has been widely used as a reporter ofGM1 in studies of plasma membrane organization, structureand function in living cells (Gaus et al., 2003; Lagerholm

et al., 2005; Chinnapen et al., 2007; Day and Kenworthy,2009).

Lipid rafts enriched in GM1 reported with CTB have beendescribed in mouse (Comiskey and Warner, 2007) and seaurchin oocytes (Belton et al., 2001) and in pig cumulus cells(Sasseville et al., 2009). A central role in fertilization forthis lipid raft was demonstrated in the sea urchin by theirdisruption with the cholesterol chelator methyl-b-cyclodex-trin, which significantly reduced fertilization rates andadversely affected protein signalling from the oolemma thatwas normally initiated by the fertilizing spermatozoon. Forsomatic cells (e.g. cumulus oophorus) and invertebrateoocytes, sufficient cellular material can be easily obtainedto prepare plasma membrane fractions enriched in lipidrafts that allow for further purification by ultracentrifuga-tion for molecular and structural analyses (Parolini et al.,1996; Matka and Szollosi, 2005; Sasseville et al., 2009). Sim-ilar studies in mouse or human are not feasible owing to thevery sizeable number of oocytes that would be required toisolate lipid rafts in quantities suitable for high-resolutionstructural or compositional analysis. Nevertheless, the highspecificity and affinity of CTB for GM1 supports the use ofthis probe as a reliable reporter of this class of lipid rafts,such that correlations with microdomain organization mayenable the identification of causative factors in certaintypes of fertilization failures.

GM1 lipid raft microdomain phenotypes and spermattachment

Distinct configurations of lipid raft microdomains enrichedin GM1 were characterized in relation to fertilizability.The single phenotype that appears to be most consistentwith sperm attachment and penetration is one where thesedomains are punctate, at high density and uniformly distrib-uted. This organization was characteristic of penetrated butunfertilized MII oocytes and in zona-free dysmorphicoocytes, where multiple sperm binding was observed withinseconds following insemination. Studies of spermato-zoon–oolemma interactions showed that the initial interac-tions between gamete membranes leading to robustattachment occurred at CTB-positive sites. In instances ofmultiple sperm attachment in zona-free oocytes, all wereat domains of punctate CTB fluorescence in which GM1microdomains could be distinguished individually, despitetheir comparatively high density. None of the unpenetratedMII oocytes with spermatozoa in proximity to or attached tothe oolemma showed this phenotype, suggesting that thisconfiguration is consistent with fertilizability in the human.This notion was supported further by the absence of spermattachment at regions of low GM1 density or where a CTBsignal was undetectable.

It is a random event where a human spermatozoon entersand traverses the zona pellucida, but the findings suggestthat normal fertilization in the human is specifically depen-dent upon the density and organization of GM1 microdo-mains at the oolemmal site where the head arrives. Thus,GM1 domains could be considered docking sites involved inthe earliest stages of attachment but where species speci-ficity is irrelevant. Support for this notion comes from stud-ies of fertilization in MII mouse oocytes that have similar

Oolemmal lipid rafts and fertilization competence of human oocytes 699

configurations and distributions of punctate GM1 lipid raftmicrodomains at docking sites (Van Blerkom, 2013) andfrom the ability of human spermatozoa to firmly dock atthe same sites in zona-free mouse oocytes within the first30 s post insemination (Van Blerkom, unpublished). Thesedomains may be a common feature of the mammalianoolemma, serving as a general docking site, a notion thatis consistent with the known intercellular functions of GM1.However, robust attachment leading to penetration wouldseem to involve molecules in the same or co-localized lipidraft microdomains that convey species specificity and facil-itate robust binding. The lipid raft-associated protein CD9,which occurs in different microdomains apparently lackingGM1, is the most likely candidate for stable binding, butnot for species specificity, after the initial attachmentphase.

Thus, while spermatozoa bound at the CTB-positive GM1islands of varying size, they did not penetrate the oolemma.Quantitation of CTB fluorescent intensity in oocytes exhibit-ing these phenotypes showed relative intensities similar tothose obtained for the fertilization-consistent configura-tion. This finding suggests that the GM1 content of fertiliza-tion-competent and -incompetent human oocytes at MIImay be similar and that differences in fertilizability areassociated with the effectiveness of GM1 lipid raft microdo-mains as concentrating platforms (Pike, 2003). The pres-ence of large aggregates of GM1 may not compromise orinhibit docking, but could prevent interactions with mole-cules in adjacent co-localized microdomains required forrobust attachment or penetration.

The co-localization of GM1 and CD9 lipid raft microdo-mains in MII human oocytes assumed to be fertilization com-petent may lead to the formation of ‘fertilizationcomplexes’ (Jejou et al., 2011) in which GM1 facilitatessperm docking and then interacts with CD9 to facilitate adegree of adhesion between gamete membranes necessaryfor subsequent events, including species recognition andpenetration. This conclusion is supported by the observationthat, with a few noted exceptions, virtually all MII oocytesexamined showed a high-density, punctate and uniform dis-tribution of CD9 throughout the oolemma, including theGM1 voids where no sperm docking or attachment wasobserved. It also suggests that CD9 alone may not be a cen-tral player in the initial stages of sperm attachment. Thecharacterization of fertilization-associated proteins andwhether they occur in the same or different lipid rafts inoolemmal microdomains is currently under investigation.

DWm in the subplasmalemmal cytoplasm and GM1lipid raft microdomain organization in thecorresponding oolemma

While the present findings offer an intriguing view of howfertilization competence at the level of the oolemma maybe established in the mature human oocyte, the cause ofcertain microdomain phenotypes associated with incompe-tence was not evident. Previous studies of mouse andhuman oocytes showed the existence of a stable subplasma-lemmal domain of mitochondria with comparatively highDWm. Because the magnitude of DWm is a central driverof mitochondrial function and bioactivity, including ATP

production and free calcium release and sequestration (VanBlerkom, 2011), this work investigated the extent to whichmitochondrial potential in the subplasmalemmal cytoplasmwas related to the organization of GM1 lipid raft microdo-mains in the overlying oolemma.

The findings indicated that DWm and GM1 microdomainorganization are related. Relatively large areas of the subpl-asmalemmal cytoplasm where DWm was comparatively lowwere associated with scant, sparse or undetectable CTBfluorescence and no sperm attachment. Van Blerkomet al. (2002) previously reported that the entire subplasma-lemmal complement of mitochondria in some unfertilizedhuman oocytes showed low potential as reported by thepotentiometric mitochondria-specific fluorescent probeJC1. Here, this study has shown that the oolemma of thissame type of unfertilized oocytes is virtually devoid ofGM1 lipid raft microdomains, which may explain why sper-matozoa that had passed into the perivitelline space failedto attach to the oolemma. In these instances, the patternand distribution of CD9 was normal, further supporting thesuggestion that it may not be involved in the initial stagesof sperm docking in the human.

The possibility that mitochondrial activity just a fewmicrons away from the oolemma may regulate lipid raftmicrodomain organization suggests new avenues of studyinto how fertilization competence in the human oocytemay be established. Whether focally elevated levels ofATP are required to support a specific pattern of GM1 micro-domain organization suggested here to be consistent withnormal fertilization is under investigation. In a similar con-text, the known role of mitochondria in regulating free cal-cium homeostasis is another promising area of investigationgiven that calcium-dependent proteins such as annexin 2 areknown to regulate the state of aggregation of lipid rafts,including those enriched in GM1 (Babiychuk and Draeger,2000).

Although the number of oocytes examined for bothhigh-potential mitochondria and GM1 microdomain distribu-tion was relatively small, this study’s findings suggest thatthe relative level of mitochondrial activity in the subplas-malemmal cytoplasm of the human oocyte may be a driverof GM1, but not CD9, organization and function in the corre-sponding oolemma. While distinct patterns of GM1 lipid raftaggregation seem to correlate with DWm in the correspond-ing subplasmalemmal cytoplasm, these phenotypes mayrepresent a spectrum of mitochondrial states that establishcell surface conditions that, for some oocytes, are positivefor fertilization while not for others. In this context, it maybe useful to determine how spatial changes in DWm thatoccur during preovulatory maturation (Van Blerkom et al.,2008) relate to GM1 lipid raft micro- and macro-domain dis-tribution and density and how they temporally and spatiallyrelate to the acquisition of fertilization competence in thecorresponding oolemma. An alternative possibility worthyof investigation is whether the differential organization oflipid raft microdomains in the oolemma, such as thosedetectable with CTB, influence DWm in the subplasmalem-mal domain.

Biochemical isolation of the types of lipid rafts describedhere would go a long way towards the identification of mol-ecules resident in oolemmal microdomains that may con-tribute to a developmental competence. For reasons

700 J Van Blerkom, K Caltrider

related to the quantity of membrane required, such studiesare currently not feasible for the human oocyte. The chal-lenge, therefore, is to devise new methods, or to determinewhether modifications of existing protocols used for somaticcells (Blank et al., 2002) can be informative in studies oflipid rafts and oolemmal competence where only a smallnumber of oocytes is available. A second challenging aspectof this line of investigation will be to elucidate how spatiallydistinct patterns of differential DWm occur in the subplas-malemmal cytoplasm because, if the current findings areconfirmed, this aspect of oogenesis may be the key tounderstanding the origin of attachment and penetrationcompetence.

Conclusion

For clinical purposes, the present findings suggest a meansby which an oolemmal component of fertilization failurecan be identified and provide an explanation to thosepatients where spermatozoa progress only as far as theoocyte plasma membrane.

References

Babiychuk, E., Draeger, A., 2000. Annexins in cell membranedynamics: Ca2+-regulated association of lipid microdomains. J.Cell Biol. 150, 1113–1123.

Belton, R., Adams, N., Foltz, K., 2001. Isolation and characteriza-tion of sea urchin egg lipid rafts and their possible functionduring fertilization. Mol. Reprod. Dev. 59, 294–305.

Blank, N., Gabler, C., Schiller, M., Kriegel, M., Lorenz, J., Kalden,H., 2002. A fast, simple and sensitive method for the detectionand quantification of detergent resistant membranes. J. Immu-nol. Methods 271, 25–35.

Brown, D., 2006. Lipid rafts, detergent-resistant membranes, andraft targeting signals. Physiology 21, 430–439.

Brown, D., London, E., 2000. Structure and function of sphingolipid-and cholesterol-rich membrane rafts. J. Biol. Chem. 275,17221–17224.

Campbell, T., Davy, A., Liu, Y., Arcellana-Panlilio, M., Robbins, S.,2008. Distinct membrane compartmentalization and signaling ofephrin-A5 and ephrin-B1. Biochem. Biophys. Res. Commun. 375,362–366.

Chinnapen, D.J., Chinnapen, H., Saslowsky, D., Lencer, W., 2007.Rafting with cholera toxin: endocytosis and trafficking fromplasma membrane to ER, FEMS. Microbiol. Lett. 7, 129–137.

Comiskey, M., Warner, C., 2007. Spatio-temporal localization ofmembrane lipid rafts in mouse oocytes and cleaving preimplan-tation embryos. Dev. Biol. 303, 727–739.

Day, C., Kenworthy, A., 2009. Tracking microdomain dynamics incell membranes. Biochem. Biophys. Acta 1788, 245–253.

Edidin, M., 2003. The state of lipid rafts: from model membranes tocells. Ann. Rev. Biophys. Biomol. Struct. 32, 257–283.

Gaus, K., Gratton, E., Kable, E.P., Jones, A.S., Gelissen, I.,Kritharides, L., Jessup, W., 2003. Visualizing lipid structureand raft domains in living cells with two-photon microscopy.Proc. Natl. Acad. Sci. U. S. A. 100, 15554–15559.

Godoy, V., Riquelme, G., 2008. Distinct lipid rafts in subdomainsform human placental apical syncytiotrophoblast membranes. J.Membr. Biol. 224, 21–31.

Harder, T., Scheiffele, P., Verkade, P., Simons, K., 1998. Lipiddomain structure of the plasma membrane revealed by patchingmembrane components. J. Cell Biol. 141, 1918–1929.

Ishii, M., Iwai, K., Koike, M., Ohshima, S., Kudo-Tanaka, E., Ishii,T., Mima, T., Katada, Y., Miyatake, K., Uchiyama, Y., Saeki, Y.,

2006. RANKL-induced expression of tetraspanin CD9 in lipid raftmembranes microdomains is essential for cell fusion duringosteoclastogenesis. J. Bone Mineral. Res. 6, 965–976.

Istanbul consensus workshop on embryo assessment: proceedings ofan expert meeting. 2011. Reprod. Biomed. Online 22, 632–646.

Janes, P., Ley, S., Magee, I., 1999. Aggregation of lipid raftsaccompanies signaling via the T cell antigen receptor. J. CellBiol. 147, 447–461.

Jejou, A., Ziyyat, A., Barraud-Lange, V., Perez, E., Wolf, J., Pincet,F., Gourier, C., 2011. CD9 tetraspanin generates fusion compe-tent sites on egg membrane for mammalian fertilization. Proc.Natl. Acad. Sci. U. S. A. 108, 10946–10951.

Jobling, M., Yang, Z., Kam, W., Lencer, W., Holmes, R., 2012. Asingle native ganglioside GM1-binding site is sufficient forcholera toxin to bind to cells and complete the intoxicationpathway. mBio 30. http://dx.doi.org/10.1128/mBio. 00401-12.

Kawano, N., Yoshida, K., Miyado, K., Yoshida, M., 2011. Lipid rafts:keys to sperm maturation, fertilization, and early embryogen-esis. J. Lipids 1, 264706.

Lagerholm, B., Weinreb, G., Jacobson, K., Thompson, N., 2005.Detecting microdomains in intact cell membranes. Ann. Rev.Phy. Chem. 56, 309–336.

Landry, A., Xavier, R.I., 2006. Isolation and analysis of lipid rafts incell–cell interactions. Methods Mol. Biol. 341, 251–282.

Lopez, P., Schnaar, R., 2009. Gangliosides in cell recognition andmembrane proteins regulation. Curr. Opin. Struct. Biol. 19,549–557.

Luna, E., Nebl, T., Takizawa, N., Crowley, J., 2005. Lipid raftmembrane skeletons. In: Matttson, P. (Ed.), Membrane Micro-domain Signaling: Lipid Rafts in Biology and Medicine. HumanaPress Inc., Totowa, NJ, pp. 47–69.

Matka, J., Szollosi, J., 2005. Regulatory aspects of membranedomain (raft) dynamics in live cells. In: Mattson, P. (Ed.),Membrane Microdomain Signaling: Lipid Rafts in Biology andMedicine. Humana Press Inc., Totowa, NJ, pp. 15–46.

McIntosh, T., 2007. Overview of membrane rafts. Meth. Mol. Biol.398, 1–9.

Miyara, F., Aubriot, F.X., Glissant, A., Nathan, C., Douard, S.,Stanovici, A., Herve, F., Dumont-Hassan, M., LeMeur, A.,Cohen-Bacrie, P., Debey, P., 2003. Multiparameter analysis ofhuman oocytes at metaphase II stage after IVF failure innon-male infertility. Hum. Reprod. 7, 1494–1503.

Miyado, K., Yoshida, K., Yamagata, K., Sakakibara, K., Okabe, M.,Wang, X., Miyamoto, K., Akutsu, H., Kondo, T., Takahashi, Y.,Ban, T., Ito, C., Toshimori, K., Nakamura, A., Ito, M., Miyado,M., Mekada, E., Umezawa, A., 2008. The fusing ability of spermis bestowed by CD9-containing vesicles released from eggs inmice. Proc. Natl. Acad. Sci. U. S. A. 105, 12921–12926.

Neu, U., Stehle, T., Atwood, W., 2008. Structural basis of GM1ganglioside recognition by simian virus 40. Proc. Natl. Acad. Sci.U. S. A. 105, 5219–5224.

Olofsson, S., Bergstrom, T., 2005. Glycoconjugate glycans as viralreceptors. Ann. Med. 37, 154–172.

Parolini, I., Sargiacomo, M., Lisanti, M., Peschle, C., 1996. Signaltransduction and glycophosphatidylinositol-linked proteins (lyn,lck, CD4, CD45, G-proteins, and CD55) selectively localized inTriton-insoluble plasma membrane domains of human leukemiccell lines and normal granulocytes. Blood 87, 3783–3794.

Pike, L., 2003. Lipid rafts: bringing order to chaos. J. Lipid Res. 44,655–667.

Sasseville, M., Gagon, M., Gillemette, C., Sullivan, R., Gilchrist, R.,Richard, F., 2009. Regulation of gap junctions in porcinecumulus-oocyte complexes: contributions of granulosa ell con-tact, gonadotropins and lipid rafts. Mol. Endocrinol. 23,700–710.

Schutz, G., Kada, G., Pastushenko, V., Schindler, H., 2000.Properties of lipid microdomains in muscle cell membranevisualized by single molecule microscopy. EMBO J. 19, 892–901.

Oolemmal lipid rafts and fertilization competence of human oocytes 701

Silvius, J., 2005. Partitioning of membrane molecules between raftand non-raft domains: insights from model membrane studies.Biochem. Biophys. Acta 1746, 193–202.

Simons, K., Toomre, D., 2000. Lipid rafts and signal transduction.Nat. Rev. Mol. Cell Biol. 1, 31–39.

Van Blerkom, J., 2011. Mitochondrial function in the human oocyteand embryo and their role in developmental competence.Mitochondrion 11, 797–813.

VanBlerkom, J., 2013. The role of the plasma membrane andpericortical cytoplasm in early mammalian development. In:Coticchio, G., DeSantis, L., Albertini, D. (Eds.), MammalianOogenesis. Springer, New York.

Van Blerkom, J., Henry, G., 1992. Oocyte dysmorphism andaneuploidy in meiotically-mature human oocytes after con-trolled ovarian stimulation. Hum. Reprod. 7, 379–390.

Van Blerkom, J., Davis, P., 2006. High-polarized (DWmHIGH) mito-chondria are spatially polarized in human oocytes and earlyembryos in stable subplasmalemmal domains: developmetnalsignificance and the concept of vanguard mitochondria. Reprod.BioMed. Online 13, 246–254.

Van Blerkom, J., Davis, P., 2007. Mitochondrial signaling andfertilization. Mol. Hum. Reprod. 13, 759–770.

Van Blerkom, J., Davis, P., Merriam, J., 1994. A retrospectiveanalysis of unfertilized and presumed parthenogenetically acti-

vated human oocytes demonstrated a high frequency of spermpenetration. Hum. Reprod. 9, 2381–2388.

Van Blerkom, J., Davis, P., Merriam, J., Sinclair, J., 1995. Nuclearand cytoplasmic dynamics of sperm penetration, pronuclearformation, and microtubule organization during fertilization andearly preimplantation development in the human. Hum. Reprod.Update 1, 429–461.

Van Blerkom, J., Davis, P., Mawhig, V., Alexander, S., 2002.Domains of high and low polarized mitochondria in mouse andhuman ooctes and early embryos. Hum. Reprod. 17, 393–406.

Van Blerkom, J., Davis, P., Thalhammer, V., 2008. Regulation ofmitochondrial polarity in mouse and human oocytes: the influenceof cumulus derived nitric oxide. Mol. Hum. Reprod. 14, 431–444.

Wilson, B.S., Steinberg, S.L., Liederman, K., Pfeiffer, J.R., Survi-ladze, Z., Zhang, J., Samelson, L.E., Yang, L.H., Kotula, P.G.,Oliver, J.M., 2004. Markers for detergent-resistant lipid raftsoccupy distinct and dynamic domains in native membranes. Mol.Biol. Cell 15, 2580–2592.

Declaration: The authors report no financial or commercialconflicts of interest.

Received 3 June 2013; refereed 5 September 2013; accepted 11September 2013.