DEVELOPMENTAL

BIOLOGY

142, 61-74

(1990)

The Cortical Cytoskeleton and Its Role in Sperm Penetration of the Mammalian Egg SCOTT D. WEBSTER AND ROBERT W. MCGAUGHEY Department of Zoology, Arkma

State University, Tempe, Arizona

85287-1501

Accepted June 25, 1990 In this study isolated cortical regions of both penetrated and nonpenetrated Syrian hamster eggs were examined in whole mounts and platinum replicas of detergent-extracted cortical patches. Two types of cytoskeletal organization were observed in the egg cortex: Loose networks (LN regions) with integrated localized dense networks (LDN regions). Decoration with heavy meromyosin and labeling with antiactin/protein G gold both indicate that the cortical cytoskeleton consists mainly of a LN of actin microfilaments and several types of nonactin filaments, whereas LDN regions dispersed within the LN were comprised of nonactin filaments. Cortical patches and replicas of eggs incubated with sperm for lo-15 min provide evidence that cortical microfilaments may be intimately associated with penetrating spermatozoa. The results of this investigation provide the first high resolution view of the cortical cytoskeletal domain of a mammalian egg and suggest that actin microfilaments might play a role in sperm penetration of the egg cortex. 0 1990 Academic

Press, Inc.

tion mammalian eggs also exhibit a change in the distribution of microvilli (Maro et ab, 1984). Since the cytoskeletal domain consists of a layer of actin filaments (microfilaments) which project into the microvilli (Begg et al, 1978), it is likely that the change in distribution of microvilli reflects alterations in the cortical cytoskeleton. Our laboratory recently began studies directed at the ultrastructural and molecular correlates of the cytoskeleton of mammalian eggs and embryos during preimplantation development (Capco and McGaughey, 1986; Mutchler et al., 1988; Webster and McGaughey, 1988; McGaughey and Capco, 1989; McGaughey et ah, 1990). Those studies provide an overview of the egg and embryo cytoskeleton in mammals and have identified a unique cytoskeletal “sheet,” which undergoes structural reorganization during preimplantation development. Those studies employed embedment-free sections, which provide superior imaging of the entire cytoskeleton for assessment of general patterns and reorganization events. While the previous method has revealed the elaborate spatial reorganization of the cytoskeleton during early development, in this study platinum replicas were used because they provide a higher level of resolution of the egg and isolated cortical regions. Ultrastructural analysis of isolated cortical patch specimens has been accomplished in somatic cells (Aggeler et al., 1983; Hartwig and Shevlin, 1986) and in the eggs of sea urchins (Vacquier, 1981; Chandler, 1984; Sardet, 1984; Henson and Begg, 1988; Bonder et ah, 1989). These previous studies demonstrated that whole-mount cortical patches and platinum replicas of cortical

INTRODUCTION

The eggs of animals contain a peripheral cytoplasmic region referred to as the cortical cytoskeletal domain (Jeffery and Meier, 1983). The cortical cytoskeletal domain is probably responsible for several early developmental changes in eggs and embryos, including the cortical contraction (Jeffery and Meier, 1983; Capco and McGaughey, 1986; Sawada and Schatten, 1988), the repositioning of egg organelles (Ducibella et ab, 19’7’7; Lehtonen and Badley, 1980; Wiley and Eglitis, 1980; Jeffery, 1985; Maro et ah, 1985; Schatten et al, 1985; 1986), and the distribution of macromolecules such as messenger RNA (Jeffery, 1984; 1985; Pondel and King, 1988; Hauptman et al., 1989). Following fertilization, it is likely that the cortical cytoskeletal domain plays a pivotal developmental role in compartmentation of egg components into cleavage blastomeres and probably regulates subsequent cellular differentiation of specific cell types in the later embryo (reviewed by Capco and Larabell, 1990). Fertilization of mammalian eggs is accompanied by changes in the egg cortex, including an increase in the synthesis of cortical microfilaments (Long0 and Chen, 1984; Battaglia and Gaddum-Rosse, 1986). Evidence that the cortical cytoskeletal domain is involved in mammalian sperm penetration comes from the observations that although sperm penetration occurs in the presence of microfilament inhibitors (Schatten et ab, 1986), the formation of a normal fertilization cone and migration of the male and female pronuclei is blocked by such inhibitors (Maro et al., 1984,1986). During sperm penetra61

001%1606/90 Copyright All rights

$3.00

8 1990 by Academic Press, Inc. of reproduction in any form rwxved.

62

DEVELOPMENTALBIOLOGY V0~~~~142,1990

patches provide high resolution images both of global cortical organization and of individual constituents of the cortical cytoskeletal domain. We present here the results of an ultrastructural analysis of cortical patches and replicas from mammalian eggs before and during sperm penetration. These new results demonstrate a highly complex cytoskeleton in the cortex of mammalian eggs, two distinct types of filamentous cytoskeletal networks with actin microfilaments as the predominant component of the cortical cytoskeletal domain, and an apparent structural interaction between cortical microfilaments and the penetrating mammalian spermatozoon. MATERIALS AND METHODS

Collection of gametes and insemination. Superovulation of female Syrian hamsters was accomplished by intraperitoneal injection of 30 IU pregnant mares’ serum gonadotropin (PMSG, No. G-4877, Sigma Chemical Co., St. Louis, MO.) on Day 1 of the hamster 4-day estrous cycle, and of 25 IU human chorionic gonadotropin (hCG, No. CG-5, Sigma) on cycle Day 4. The females were sacrificed by CO, asphyxiation 1’7 hr after hCG injection and the upper portion of each uterine horn with the fallopian tube attached was removed. Eggs were flushed from the oviduct with BMOC-3 culture medium (Brinster, 1965) containing 0.45% (w/v) bovine serum albumin (BSA, A-8022, Sigma) and 25 mM NaHCO,, and adjusted to pH 7.3 under 5% CO,, 5% 0,, 90% N, at 37°C. Cumulus cells surrounding the eggs were removed during a 5-min exposure to 0.1% (w/v) hyaluronidase (No. H-3506, Sigma) in culture medium. Zonae pellucidae were removed during a 3-min exposure to 0.03% (w/v) trypsin (No. T-8253, Sigma) in culture medium, followed by three washes in fresh culture medium. Approximately 50 eggs were obtained per animal. The cauda epididymides, each with the vas deferens attached, were removed from adult male hamsters and placed in 2- to 3-ml culture medium. Spermatozoa were released from each vas deferens by stripping with forceps and from each epididymis by several horizontal cuts. After 5 min incubation, the epididymal tissue was removed and the suspension diluted to lo5 motile sperm per milliliter in a conical centrifuge tube with 10 ml medium supplemented with 10M5 1M hypotaurine (No. H-1384, Sigma) and 1.6 X low5 M isoproterenol (No. I-5627, Sigma). The diluted sperm were transferred in l-ml aliquots to center-welled culture dishes (Falcon No. 3037, Becton-Dickinson, Lincoln Park, NJ), covered with mineral oil and incubated for 2-3 hr to achieve capacitation. Freshly harvested eggs were added directly to the sperm suspensions at the end of the capacitation period. Insemination time was lo-15 min.

Examination of inseminated eggs by light microscopy and after thin sectioning. Verification of sperm penetration under the above culture conditions was accomplished by light microscopic examination using the technique of Yanagimachi et al. (1976) and by examination of thin sectioned, inseminated eggs by TEM. For light microscopy, eggs were inseminated for lo-15 min, washed once in fresh culture medium to remove unbound sperm, and incubated for 20 hr, to allow penetrating spermatozoa to undergo sperm head swelling. After fixation with 0.1% (v/v) glutaraldehyde in medium [ZA medium, (McGaughey, 1977)] containing 20 mM Hepes (pH 7.3) for l-2 min, eggs were compressed between a microscope slide and a coverslip supported by vaseline:paraffin (15:l). The compressed eggs were fixed in absolute ethanol:glacial acetic acid (3:l) for l-3 hr, stained with 0.25% (w/v) lacmoid in 45% acetic acid for 10 min, and mounted in 45% acetic acid. Eggs were examined by phase contrast microscopy to observe swollen sperm heads and degenerating tails within the ooplasm. Following a lo- to 15-min insemination period, additional eggs were washed once in fresh culture medium to remove unbound sperm and fixed in 2% (v/v) glutaraldehyde in saline (PBS, pH 7.4) for 1 hr. After postfixation with 1% (w/v) 0~0, in 0.1 M sodium cacodylate buffer (pH 7.5) for 30 min, the eggs were dehydrated through a graded ethanol series, embedded in Spurr’s resin and thin-sectioned for TEM examination. Preparation of cortical patches and cortical cytoskeletons. Uninseminated and inseminated zona-free eggs were transferred to polylysine-coated (poly-L-lysine hydrobromide, MW 350,000, P-1524, Sigma) coverslip fragments or to polylysine-coated formvar-covered/carbonstabilized copper grids in intracellular buffer [ICB: 100 mM KCl, 5 mM MgCl,, 3 mM EGTA, 20 mM Hepes (pH 6.8); see Aggeler et al,, 19831 and allowed to adhere for 2 min. A stream of ICB from a mouth-operated pipet sheared the eggs away, leaving patches of egg cortex behind. To produce cytoskeletons, cortical preparations were extracted with cytoskeletal buffer [CSK: 300 mM sucrose, 10 mM Pipes (pH 6.8), 3 mM MgCl,, 100 mM NaCl, 5 mM EGTA, 0.5% (v/v) Triton X-100, and 0.2 mg/ml phenylmethylsulfonyl fluoride (PMSF, P-7626, Sigma)] as previously described (Capco and McGaughey, 1986), then washed three times in ICB containing 0.2 mg/ml PMSF. Cytoskeletons which were not to be labeled were fixed for 20-30 min with 2.0% (v/v) glutaraldehyde in ICB. Cytochemical and immunocytochemical labeling. Unfixed cortical cytoskeletons were incubated for 15 min in 10 mg/ml heavy meromyosin (HMM, M-9014, Sigma) supplemented with 0.2 mg/ml PMSF, followed by three washes totaling 10 min in ICB containing 0.2 mg/ml PMSF. HMM-labeled samples were fixed in ICB con-

WEBSTER

AND

MCGAUGHEY

Cortical

taining 1.0% (v/v) glutaraldehyde and 0.2% (w/v) tannit acid for 20 min. Other cortical cytoskeletal patches were fixed for 10 min with fresh paraformaldehyde [l.O% (w/v) in ICB], washed three times in ICB, transferred to phosphatebuffered saline (PBS, pH 7.4) and incubated for 20 min in 50 mMglycine in PBS. After three more washes cortical patches were incubated for 20 min in PBS with 1.0% (w/v) BSA (No. A-8022, Sigma), washed three times, incubated in primary antibody (see below) in PBS for 3045 min, and washed three times. The samples were then incubated in PBS with 1.0% (w/v) BSA for 20 min, washed three times, and incubated in Protein G conjugated to 15 nm colloidal gold (SPI Supplies, West Chester, PA) for 30 min at a dilution of 1:5 in PBS containing 0.01% (w/v) polyethelene glycol, pH 6.75, washed seven times, and fixed with 2.0% (v/v) glutaraldehyde in PBS for 30 min. The primary antibodies were antiactin at a 1:lO dilution (rabbit anti-chicken; Polysciences, Warrington, PA), antikeratin at a 1:lOO dilution (rabbit anti-human, total human keratins; Chemicon, El Segundo, CA), and antispectrin at a 1:lOO dilution (rabbit anti-chicken; Chemicon). Preparation of cortical patch,es and cortical cytoskeletons for electron microscopy. Fixed cortical specimens

were dehydrated through an ethanol series and dried through the critical point of CO, in a Balzer’s CPD 020 apparatus. Whole-mounted cortical patches prepared on coverslip fragments were rotary-shadowed with platinum and carbon in a Balzer’s freeze-fracture machine, while samples prepared on grids were coated with carbon for direct analysis. Samples were viewed at 80 kV on a Philips EM 300 or a Philips EM 201. Morphometric analysis. Lengths and widths of filaments and dimensions of spherical structures were calculated from measurements obtained with an eye loupe micrometer from negatives of different magnifications. A minimum of 50 measurements derived from three different cortical patches from each of three different experiments (nine cortical patches total) were used to calculate means and standard errors (SE) for each reported dimension. Densities of spherical structures and filamentous networks (number per 100 pm2 of plasma membrane area) were determined by placing negatives of different magnifications over a grid of known dimensions and counting the numbers of spherical structures or filamentous networks within defined areas. Calculations were made from measurements derived from three different cortical patches from each of three experiments (nine cortical patches total). RESULTS

Whole-mount specimens of detergent-extracted cortical cytoskeletons from unfertilized eggs. Two distinct types

Cytoskeleton

of Hamster

63

Eggs

of interconnected filamentous networks were revealed by ultrastructural analysis of detergent-extracted cortical patches from hamster eggs made directly on formvar-coated, carbon-stabilized grids (Fig. 1). The first type, referred to as the loose network (LN), is made up of long, straight filaments which constitute most of the cortex (Fig. 1A). Integrated among the LN filaments are other cytoskeletal regions which we refer to as localized dense networks (LDN, Fig. 1B). The filaments of these LDN regions are highly interconnected or crosslinked with each other and exhibit a more curvilinear structure (i.e., the individual filaments exhibit arced structure) than is the case for LN filaments. The cortical cytoskeletons of these mammalian eggs were composed of these two types of filamentous networks and of spherical structures distributed regularly throughout the cortical region (Fig. 1). These whole-mount specimens of the cortical cytoskeleton, were subjected to a morphometric analysis to establish the relative dimensions of the filaments in the LN and LDN regions and of the spherical cortical elements. The LN regions consisted of cables with a mean diameter of 27.0 * 0.40 nm SE (standard error) and individual filaments with a mean diameter of 6.1 f 0.22 nm, whereas the LDN filaments have a mean diameter of 5.5 + 0.23 nm. The lengths of LN filaments and cables ranged from 100 nm to l-2 pm, while the lengths of LDN filaments were shorter (see Fig. 1B). Lengths of individual LDN filaments were difficult to determine because of the compact nature of these regions. The LDN regions were distributed at a density of approximately 1.4 per 100 pm2 of plasma membrane. The spherical structures were 0.3-0.5 pm in diameter and were distributed at a density of approximately 80 per 100 pm2 of plasma membrane. Meiotic spindles were observed in some of the cortical specimens which were whole-mounted on carbon-coated grids for direct observation (Fig. 1C). These spindles were from unpenetrated eggs and exhibit the appearance of the second meiotic metaphase. Bundles of spindle microtubules are seen in these preparations to connect with the meiotic chromosomes and to extend outward from the equatorial region to the diffuse spindle polar regions. These meiotic spindles were located in LN regions which were relatively devoid of spherical structures. Platinum replicaIs of detergest-extracted skeletons from. unfertilized eggs. Platinum

cortical

cyto-

replicas of cortical patches (Fig. 2) exhibited the same general pattern of filament organization as did nonreplicated whole-mounted preparations. Although accurate dimensions of filaments and spherical structures could best be determined with whole-mounted cortical specimens (see above), platinum replicas of cortical speci-

FIG. 1. Whole mount cortical specimens from nonpenetrated hamster eggs after detergent extraction, as seen by electron microscopy. At low magnification (A), a panorama of the abundant loose network (LN) is shown, within which are dispersed localized dense network regions (LDN) and spherical structures (SS). At higher magnification (B), a LDN region is shown with its relatively short and highly interconnected filaments. (C) A second meiotic spindle which is comprised of the oocyte chromosomes (C) and bundles of microtubules. The spindle is typically seen in the LN with very few spherical structures or LDN regions. Bar = 1 pm. 64

FIG. 2. Platinum replicas of cortical cytoskeletons from nonpenetrated oocytes. These micrographs and all others of platinum replicas have been photographically reversed; therefore, platinum shadowing appears white. (A) Note the similarity of genera1 cortical ultrastructure to the whole mount specimen in the previous figure. At higher magnification (B), filamentous interconnections forming cables (CI) and end-on interconnections (T) can be seen, Replicas illustrate the complexity of the LDN regions (C). Bar = 1 pm (A, C), 0.5 pm (B).

65

66

DEVELOPMENTALBIOLOGY

mens provide much greater clarity for analysis of surface structure and cytoskeletal organization. Replicated filaments within the LN regions range in diameter from lo-40 nm and in length from less than 100 nm to over 1 ym. At higher magnification, different types of LN filamentous interconnections were observed (Fig. 2B). One type consists of one filament terminating at another in an end-on fashion, forming a T-shaped configuration. Another type of interconnection involves two filaments associated in a side-by-side fashion to form a larger cable. The cables have a mean diameter of 33.9 + 0.83 nm, whereas the mean diameter of individual filaments is 10.5 t 0.14 nm. Filaments also interconnect with the spherical structures such that these structures appear to be suspended within the filamentous network. In replicas, the mean diameter of LDN filaments is 9.26 + 0.08 nm, usually about 1.0 nm smaller than the individual LN filaments within a given cortical specimen (Fig. 2C). The arced LDN filaments form tight meshworks in which individual filaments rarely exceed 100 nm in length. Identi$cation of cortical jilaments. To establish the identity of the cytoskeletal filaments, we employed several labeling methods with cortical specimens followed by platinum replication. Labeling with heavy meromyosin (HMM) demonstrated that most of the cortical filaments of the LN are composed of actin (Fig. 3). Thick cables as well as individual filaments bound HMM (Fig. 3A). Labeled filaments exhibited the “ropelike” appearance of actin as described by Heuser and Kirschner (1980) and not the polarized “arrowhead” appearance of labeled actin filaments after negative staining (Huxley, 1963). In the replicated LN there were two distinct types of filaments which were not labeled. The first (Figs. 3A, 3B) had a mean diameter of 11.8 & 0.13 nm, and were comparable in length to the HMM-labeled microfilaments. The second type were thinner and shorter, with a mean diameter of 5.8 f 0.06 nm (in replicas) and less than 100 nm in length, and appeared to link labeled microfilaments together (Fig. 3A). The short, curved filaments within the LDN regions (Fig. 3B) did not bind HMM. It can be seen in Figure 3B that there were LN filaments penetrating into LDN regions; some bound HMM and some did not. To confirm the identification of microfilaments, cortical patches were labeled with antiactin followed by protein G-gold. Replicas of immunogold-labeled cortical cytoskeletons demonstrated specific labeling of cortical LN filaments (Fig. 4A) as compared with control preparations (Fig. 4B, primary antibody omitted). Background densities were similar for nonfibrous regions of cortex in replicas labeled with antiactin (45 gold beads/ km’) and in replicas labeled only with protein G-gold (50 gold beads/pm2). The frequency of gold beads on labeled

vOLUME142,1990

filaments after treatment with antiactin was threefold increased (11.4/pm) over that in replicas labeled only with protein G-gold (3.6 per pm). The LDN regions exhibited some labeling, but not with the same intensity or specificity as the LN filaments. The similarity in patterns of labeling in the LDN regions between specimens treated with anti-actin and control specimens without the primary antibody (Fig. 4B) suggests nonspecific binding of protein G-gold in the LDN regions. Taken together, the HMM data and the antiactin data demonstrate that the egg cortex is composed predominantly of a LN of primarily actin microfilaments with interspersed LDN regions composed of nonactin filaments. We attempted to identify those cortical cytoskeletal filaments which were not labeled by HMM or with immuno-gold after treatment with antiactin by incubating cortical cytoskeletal specimens in antikeratin or antispectrin followed by protein G-gold before platinum replication. Following many experimental attempts, including modifications of the fixation procedure and a variety of primary antibodies, replicas in both cases failed to demonstrate specific labeling and were not distinguishable from the control preparations (Fig. 4B). Cortical cytoskeletons of hamster eggs during sperm penetration. Whole-mounted cortical patches from in-

seminated eggs made directly on formvar-coated grids (Fig. 5) show sperm which had penetrated the egg plasma membrane and had contacted the egg cortical cytoskeleton. As seen in Fig. 5, the transmission EM image of the sperm head in these whole-mount cortical specimens appeared very electron dense. However, the preparations did allow visualization of some type of contact between cytoskeletal filaments and the perimeter of the sperm head. Platinum replicas of cortical patches from inseminated eggs show sperm within the cortex (Fig. 6A). These replicas provide evidence for apposition or contact between the cortical cytoskeleton and the penetrating spermatozoon; both the sperm head and tail exhibit association with cortical cytoskeletal filaments. Furthermore, the cortical filaments consistently were seen to interact equally with the tip, middle, and posterior regions of the sperm head. In Figure 6B, HMM-decorated filaments are seen interacting with the penetrating sperm head. Interactions with decorated filaments were not restricted to a single location on the sperm head. The filaments associated with penetrating spermatozoa were predominantly those of the LN, with examples of sperm interacting with filaments of LDN regions being only rarely observed. Eggs were inseminated for lo-15 min and washed free of unbound sperm to verify by light and electron microscopy that penetration was occurring consistently under

FIG. 3. Platinum replicas of oocyte cortical are decorated. Nonactin intermediate-sized intimately interconnected with microfilaments tion of the LN microfilaments and nonactin

cytoskeletons labeled with heavy meromyosin. At high magnification (A) most of the LN filaments filaments (closed arrowheads in A) and small, crosslinking filaments (open arrowheads in A) are (MF) and actin cables (CB). A LDN region is shown in B, which demonstrates the interconnecfilaments with the unlabeled LDN filaments. Bar = 0.5 pm. 67

FIG. 4. Platinum replicas of immunogold-labeled oocyte cortical cytoskeletons. The cortex specimen in A was labeled initially with antiactin; whereas, the control cortical specimen in B was not exposed to a primary antibody. The specificity of labeling with antiactin (A) demonstrates the same pattern of microfilaments as observed in Fig. 3. The nonspecific binding of protein G gold to the LDN region in A is identical to the pattern in the control (B), although in B, more extensive LDN regions are shown and exaggerate the nonspecific binding to these regions. Other replicas (not shown) of cortical specimens incubated with the primary antibodies, antikeratin and antispectrin, exhibited immunogold patterns indistinguishable from that in B. Bar = 1 Wm. 68

WEBSTER

AND MCGAUGHEY

Cortical

Cytoskeletmz

of Hamster

FIG. 5. A whole mount specimen from an inseminated egg in which a sperm has penetrated shown in close apposition to the surface of the penetrating spermatozoon along its periphery. spermatozoon. SH, sperm head. Bar = 1 Frn.

the conditions employed in this study. By light microscopy it was determined that penetration by at least one spermatozoon occurred in 87.9% (1671190) of eggs, with a mean of 3.5 penetrations per egg. EM analysis of sectioned eggs (Fig. 7) demonstrated that after lo-15 min of insemination culture, sperm had penetrated into the cortical region of the egg. DISCUSSION

This study is the first high resolution ultrastructural investigation of cortical preparations from mammalian eggs. We present observations in which detergent extraction and specific labeling were employed to characterize isolated cortical cytoskeletons of mammalian eggs at the ultrastructural level. The results of these investigations demonstrate that the mammalian egg cortex is comprised of a loose meshwork of intercon-

Eggs

into the cortical LDN regions

69

region. Filaments of the LN are can be seen near the penetrating

necting microfilaments, as demonstrated by labeling with heavy meromyosin and antiactiniprotein G-gold. These microfilaments exist both singly and in cables and their dimensions agree with the previous study of Heuser and Kirschner (1980) who reported a thickness of 9.5 nm for individual microfilaments in rotary-shadowed platinum replicas of freeze-dried mouse embryo fibroblast cytoskeletons. Aggeler et al. (1983) reported that microfilaments were 10.7 nm thick in rotary-shadowed platinum replicas of critical-point dried macrophage cortical patches. Such slight differences in filament thickness reported in different studies with platinum replicas probably reflect variations in thickness of platinum shadowing. The cortical cytoskeletal domain of the mammalian egg, directly observed and partially characterized ultrastructurally in the present study, exhibits both similarities to and differences from the cortical cytoskeleton of

FIG. 6. Platinum replicas of cortical cytoskeletons from a penetrated egg at low magnification (A) meromyosin labeling at higher magnification (B). The highly resolved surface detail in these replicas cytoskeletal filaments and the spermatozoan surface (A), and the contacts between actin microfilaments head. Bar = 1 pm (A), 0.5 pm (B). 70

and from a penetrated suggests interconnections and the spermatozoon

egg after heavy between the (B). SH, sperm

WEBSTER AND MCGAUGHEY

Cortical Cytoskeleton of Hamster Eggs

FIG. 7. Electron micrograph of a thin-sectioned egg which was fixed at lo-15 min after insemination. This figure is representative of those examined to verify that the penetrating sperm head (SH) had entered the cortical region of eggs under the conditions employed in the cortical isolation studies. Bar = 1 pm.

sea urchin eggs. It was reported for sea urchin eggs (Bonder et ah, 1989), that polymerized actin contributes significantly to the cortical cytoskeleton, as we have now shown in hamster eggs. We were unable to demonstrate localization of spectrin in the isolated hamster egg cortex, as has been reported in sea urchin eggs (Schatten et al., 1986; Bonder et al., 1989). In this regard, Sobel et al. (1988) reported localized spectrin in the twocell mouse embryo. Our inability to demonstrate localized spectrin in the one-celled hamster egg is consistent with the observations that spectrin is not localized in the cortex of mouse eggs prior to the two-cell stage (Sobe1 and Alliegro, 1985) and that unfertilized and fertilized one-cell mouse eggs exhibit spectrin distributed uniformly throughout the cytoplasm (Damjanov et ah, 1986). The hamster egg cortex contains nonactin filaments, some of which are found in the LN regions. These filaments have a mean diameter, in platinum replicas, of 11.8 nm and therefore are larger than replicas of microfilaments. These nonactin filaments are in the general size range of intermediate filaments which have been reported to associate with actin microfilaments in other cell types (Heuser and Kirschner, 1980; Schliwa and Van Blerkom, 1981; Gall et ah, 1983; Katsuma et al, 1987). Although these egg cortical filaments were not specifically labeled with the antikeratin antibody employed in our study, they might be composed of another form of intermediate filament protein (Moll et a& 1982) which was not recognized by our antibody. Cytokeratin has been demonstrated in the hamster egg cytoskeleton by the immunoblot technique after polyacrylamide gel electrophoresis of fractionated cytoskeletons (McGaughey and Capco, 1989). However, the egg cytokeratin was re-

71

ported to be associated with specialized cytoskeletal elements termed “sheets” which in hamster eggs are located in the subcortical region (Capco and McGaughey, 1986). In the present study, cytoskeletal sheets were not observed for the technical reason that during preparation of cortical cytoskeletons, the sheets are sheared away from the cortical patches and are absent from such specimens. This is in agreement with the observation of Capco and McGaughey (1986) that in the hamster egg and early cleaving embryo, the sheets generally are excluded from the cortex. A second type of nonactin filament in the cortical LN is small, about 5-6 nm in diameter in replicated specimens, and crosslinks other filaments of the LN. These small filaments, most readily seen in HMM-labeled cortical patches, are similar to the 3 nm filaments reported by others to crosslink microfilaments and intermediate filaments (Schliwa and Van Blerkom, 1981; Katsuma et al., 1987). The LDN regions of the hamster egg cortex are comprised of nonactin filaments whose composition we were unable to identify. The LDN regions exhibited no specific staining with HMM or immunocytochemical labeling with antiactin antibody and their dimensions and arced profiles differed greatly from those of the filaments comprising the surrounding LN regions. These observations strongly suggest that the LDN regions are true components of the hamster egg cortical cytoskeleton and not aggregations of LN microfilaments resulting from extraction or other methods employed in cortical preparation. The LDN regions were observed previously in hamster eggs by the method of embedment-free thin sectioning (Capco and McGaughey, 1986). The probable existence of LDN regions has been suggested by localized immunofluorescence in the eggs of sea urchins (Schatten et aZ., 1986) and in the mouse (Damjanov et ah, 1986; Lehtonen and Badley, 1980; Sobel et ab, 1988; Sobel, 1983). Because these regions exhibit a pun&ate distribution pattern, they suggest a structural similarity to regions of actin, myosin and spectrin which are coordinately distributed in the cortex of the two-cell mouse embryo (Sobel et al., 1988). Although the LDN regions of the unfertilized hamster egg cortex were not specifically labeled by antiactin, antikeratin or antispectrin, these regions of the cortex might represent an earlier developmental stage of the regions described by the previous investigators. Our cortical specimens exhibited no clear ultrastructural evidence of microtubules, except in preparations containing the meiotic spindle (see Fig. 1C). Although cables comprised of associated microfilaments were observed, these cables were labeled with HMM (see Fig. 3A). None of the nonactin filaments of the LN or LDN regions were in the size range of microtubules (i.e., 25

72

DEVELOPMENTAL BIOLOGY

nm diameter). In preliminary studies (unpublished results) we observed immunofluorescent evidence of tubulin only in the meiotic apparatus of hamster eggs at the developmental stage analyzed in the present study. In addition to the microfilaments and other filaments of the LN and LDN regions, we also observed spherical structures of the appropriate size and distribution of cortical granules (Peluso and Eutcher, 1974; Flechon et al., 1975; Cherr et al., 1988). It is possible that these structures represent the detergent-resistant remnants of cortical granules. Detergent-resistant components of membranes, thought to represent membrane proteins, have been observed in other systems (Ben-Ze’ev et ah, 1979; Prives et ah, 1982). These spherical remnants were observed in all cortical cytoskeletal specimens from unfertilized and from inseminated eggs; although cortical patches which included meiotic spindles were generally devoid of these elements. The region of mammalian eggs overlying the meiotic spindle is characterized by a lack of cortical granules (Odor and Renninger, 1960; Zamboni, 1970; Eurgos et ah, 1976; Gulyas, 1976; Schmell et aZ., 1983; Longo and Chen, 1984; Ducibella et al, 1988). Most exocytosis of cortical granules occurs about 30 min after egg activation (Cherr et aZ., 1988) and therefore would not be expected to have progressed to a significant degree in our specimens prepared lo-15 min after insemination. It remains to be determined, by examining cortical patches at prolonged intervals after insemination, whether the spherical structures are in fact detergent-resistent remnants of cortical granules. We demonstrated, at the light microscope and ultrastructural levels that under the conditions employed for insemination in this study, hamster eggs became penetrated and that spermatozoa reached the egg cortex within lo-15 min of culture. We employed in vitro fertilization of zona-free hamster eggs to increase the probability of polyspermy and thereby of obtaining penetrating spermatozoa in isolated specimens of egg cortex. Although the orientation of hamster spermatozoa with respect to the outer surface of the egg plasma membrane is somewhat different when sperm penetration occurs in vitro and in zona-free eggs, as compared with sperm penetration in viva (Shalgi and Phillips, 1980a,b; Yanagimachi, 1981), there likely are no other substantial differences between sperm penetration in vivo and in vitro in the hamster (Yangimachi and Noda, 1970; Hirao and Yanagimachi, 1979). We suggest, therefore, that the ultrastructural observations of spermatozoan interaction with cortical cytoskeletal elements made in the present study represent normal, physiological interactions of developmental significance. Within 15 min after insemination, microfilaments of the egg cortex were found in association with penetrating sperm heads. Our observation that microfilaments

VOLUME 142,199O

appear to associate equally with all parts of the sperm head underscores the idea that the point of initial sperm fusion, whether at the tip, equatorial segment, or postacrosomal region (see, Shalgi and Phillips, 1980a,b), probably is not critical to subsequent cortical penetration. The cortical microfilaments probably do not actively draw the sperm into the egg, because sperm penetration can occur in the presence of microfilament inhibitors (Maro et al., 1984; Schatten, 1986). However, the normal fertilization cone, which forms in the vicinity of sperm fusion with the egg surface, is rich in microfilaments, and in the presence of inhibitors the cone is either absent or reduced in size (Maro et al., 1984). Microfilaments appear therefore to be required for fertilization to proceed normally, and the filamentous associations with the sperm heads observed in this study may in part constitute the basis of that requirement. This study has revealed the highly complex cortical cytoskeletal domain in the hamster egg and has identified the actin microfilament as a predominant cytoskeletal component of the egg cortex. We have demonstrated the presence of two filamentous networks, the LN and LDN, comprising the egg cortical cytoskeleton and provide evidence that remnants of cortical granules may remain in detergent-extracted cortical cytoskeletons. The composition of at least two types of filaments in the LN region and the filaments comprising the LDN regions remain to be identified. This study provides ultrastructural evidence that the penetrating mammalian spermatozoon may interact specifically with cortical actin microfilaments and suggests the possibility that microfilaments may play a developmental role in regulating sperm penetration within the cortex and possibly in subsequent developmental events of fertilization in mammals. To determine whether such a role is played by cortical microfilaments, additional work will be necessary to further define the type of structural interaction between microfilaments and the sperm head and to analyze such interactions during sperm head swelling and chromatin decondensation. A preliminary report of a portion of this work was presented at the Annual Meeting of the American Society for Cell Biology, January, 1988. This study was submitted in partial fulfillment of the requirements for the graduate program (S.D.W.) at Arizona State University. The authors thank David G. Capco for his valuable discussions during this study. This work was supported by Grant HD23686 from the NIH (R.W.M.). REFERENCES AGGELER, J., TAKEMURA, R., and WERB, Z. (1983). High resolution three-dimensional views of membrane-associated clathrin and cytoskeleton in critical-point-dried macrophages. J. Cell Bio/. 97, 14521458. BATTAGLIA, D. E., and GADDUM-ROSSE, P. (1986). The distribution of

WEBSTER AND MCGAUGHEY

Cortical

polymerized actin in the rat egg and its sensitivity to cytochalasin B during fertilization. J. Exp. 2001 237, 97-105. BEGG, D. A., RODEWALD, R. L., and REBHUN, L. I. (1978). The visualization of actin filament polarity in thin sections. J. Cell Biol. 79,846852.

BEN-ZE’EV, A., DUERR, A., SOLOMON, F., and PENMAN, S. (1979). The outer boundary of the cytoskeleton: A lamina derived from plasma membrane proteins. Cell 17,859-865. BONDER, E. M., FISHKIND, D. J., COTRAN, N. M., and BEGG, D. A. (1989). The cortical actin-membrane cytoskeleton of unfertilized sea urchin eggs: Analysis of the spatial organization and relationship of filamentous actin, nonfilamentous actin, and egg spectrin. Dev. Biol. 134,327-341.

BRINSTER, R. L. (1965). Studies on the development of mouse embryos in ,vitro. IV. Interaction of energy sources. J. Reprod. Fertil. 10,227240.

BURGOS, M. H., SEGAL, S. J., and PASSANTINO, T. (1976). Surface changes of the rat embryo before implantation. Fertil. Steril. 27, 1085-1094. CAPCO,D. G., and MCGAUGHEY, R. W. (1986). Cytoskeletal reorganization during early mammalian development: Analysis using embedment-free sections. Dev. Biol. 115, 446-458. CAPCO, D. G., and LARABELL, C. A. (1990). The cytoskeleton during early development: Structural transformation and reorganization of RNA and Protein. In “Progress in Molecular and Subcellular Biology” (W. E. G. Muller, Ed.), Springer-Verlag, New York, in press. CHANDLER, D. E. (1984). Exocytosis in vitro: Ultrastructure of the isolated sea urchin egg cortex as seen in platinum replicas. J. Ultrasf. Res. 89, 198-211. CHERR, G. N., DROBNIS, E. Z., and KATZ, D. F. (1988). Localization of cortical granule constituents before and after exocytosis in the hamster egg. J. Eq. 2001. 246, 81-93. DAMJANOV, I., DAMJANOV, A., LEHTO, V-P., and VIRTANEN, I. (1986). Spectrin in mouse gametogenesis and embryogenesis. Dev. Biol. 114, 132-140. DUCIBELLA, T., UKENA, T., KARNOVSKY, M., and ANDERSON, E. (1977). Changes in cell surface and cortical cytoplasmic organization during early embryogenesis in the preimplantation mouse embryo. J Cell Biol. 74, 153-167. DUCIBELLA, T., ANDERSON, E., ALBERTINI, D. F., AALBERG, J., and RANGARAJAN, S. (1988). Quantitative studies of changes in cortical granule number and distribution in the mouse oocyte during meiotic maturation. Dev. Biol. 130, 184-197. FLECHON, J. E., HUNEAU, D., SOLARI, A., and THIBAULT, C. (1975). Cortical reaction and inhibition of polyspermy in rabbit eggs. Ann. Biol.

Anim.

B&hem.

Biophys.

15, 9.

GALL, L., PICHERAL, B., and GOUNON, P. (1983). Cytochemical evidence for the presence of intermediate filaments and microfilaments in the egg of Xenopus laevis. Biol. Cell 47, 331-342. GULYAS, B. J. (1976). Ultrastructural observations on rabbit, hamster and mouse eggs following electrical stimulation in vitro. Amer. J. Axat.

Gro~~th

Difler.

73

Eggs

HIRAO, Y., and YANAGIMACHI, R. (1979). Development of pronuclei in polyspermic eggs of the golden hamster: Is there any limit to the number of sperm heads that are capable of developing into male pronuclei? Zool. Mag. (Tokyo) 88, 24-33. HUXLEY, H. E. (1963). Electron microscope studies on the structure of natural and synthetic protein filaments from striated muscle. J. Mol.

Biol.

7, 281-308.

JEFFERY, W. R. (1984). Spatial distribution of messenger RNA in the cytoskeletal framework of ascidian eggs. D~II. Biol. 103,482-492. JEFFERY, W. R. (1985). The spatial distribution of maternal mRNA is determined by a cortical cytoskeletal domain in Chaetopterus eggs. Dev.

Biol.

110,217-229.

JEFFERY, W. R., and MEIER, S. (1983). A yellow crescent cytoskeletal domain in ascidian eggs and its role in early development. Dev. Biol. 96,125-143.

KATSUMA, Y., SWIERENGA, S. H. H., MARCEAU, N., and FRENCH, S. W. (1987). Connections of intermediate filaments with the nuclear lamina and the cell periphery. Biol. Cell 59, 193-204. LEHTONEN, E., and BADLEY, R. A. (1980). Localization of cytoskeletal proteins in preimplantation mouse embryos. J. Embryol. Exp. Mw phol. 55,211-225.

LONGO, F. J., and CHEN, D. Y. (1984). Development of surface polarity in mouse eggs. Scan. Electron Microsc. 198,703-716. MARO, B., JOHNSON, M. H., PICKERING, S. J., and FLACH, G. (1984). Changes in actin distribution during fertilization of the mouse egg. J. Embryol.

Ezp.

Morphol.

81,211-237.

MARO, B., JOHNSON, M. H., and PICKERING, S. J. (1985). Changes in the distribution of membranous organelles during mouse early development. J. Embryol. Exp. Morph01 90,287-309. MARO, B., JOHNSON, M. H., WEBB, M., and FLACH, G. (1986). Mechanism of polar body formation in the mouse oocyte: An interaction between the chromosomes, the cytoskeleton and the plasma membrane. J Embryol. Exp. Mwphol. 92, 11-32. MCGAUGHEY, R. W. (1977). The maturation of porcine oocytes in minimal defined culture media with varied macromolecular supplements and varied osmolarity. Exp. Cell Res. 109, 25-30. MCGAUGHEY, R. W., and CAPCO, D. G. (1989). Specialized cytoskeletal elements in mammalian eggs: Structural and biochemical evidence for their composition. Cell Motil. Cytoskel. 13, 104-111. MCGAUGHEY, R. W., RACOWSKY,C., RIDER, V., BALDWIN, K., DEMARAIS, A. A., and WEBSTER, S. D. (1990). Ultrastructural correlates of meiotic maturation in mammalian oocytes. J. Electron Micro. Tech., in press. MOLL, R., FRANKE, W. W., SCHILLER, D. L., GEIGER, B., and KREPLER, R. (1982). The catalog of human cytokeratins: Patterns of expression in normal epithelia, tumors and cultured cells. Cell 31, 11-24. MUTCHLER, D. V., MCGAUGHEY, R. W., and CAPCO, D. G. (1988). Cytoskeletal organization during early development in the mouse. J. Cell Biol.

107, 605a.

ODOR, D. L., and RENNINGER, D. F. (1960). Polar body formation in the rat oocyte as observed with the electron microscope. Anaf. Rec. 147, 13-23.

147, 203-217.

HARTWIG, J. H., and SHEVLIN, P. (1986). The architecture of actin filaments and the ultrastructural location of actin-binding protein in the periphery of lung macrophages. J Cell Biol. 103,1007-1020. HAUPTMAN, R. JL, PERRY, B. A., and CAPCO, D. G. (1989). A freeze-sectioning method for preparation of the detergent-resistant cytoskeletal proteins and associated mRNA in Xenop/.s oocytes and embryos. Dev.

of Hamster

Cytoskeleton

31, 157-164.

HENSON, J. H., and BEGG, D. A. (1988). Filamentous actin organization in the unfertilized sea urchin egg cortex. Dev. Biol. 127, 338-348. HEUSER, J. E., and KIRSCHNER, M. W. (1980). Filament organization revealed in platinum replicas of freeze-dried cytoskeletons. J. Cell Biol. 86, 212-234.

PELUSO, J. J., and BUTCHER, R. L. on the ultrastructure of the rat PONDEL, M. D., and KING, M. L. related to transforming growth cytokeratin-enriched fraction Acad.

(1974). The effect of follicular aging oocyte. Fertil. St&l. 25,194-202. (1988). Localized maternal mRNA factor b mRNA is concentrated in a from Xenon-us oocytes. Proc. Natl.

Sci. USA 85,7612-7616.

PRIVES, J., FULTON, A. B., PENMAN, S., DANIELS, M. P., and CHRISTIAN, C. N. (1982). Interaction of the cytoskeletal framework with acetylcholine receptor on the surface of embryonic muscle cells in culture. J. Cell Biol. 92, 231-236. SARDET, C. (1984). The ultrastructure of the sea urchin egg cortex isolated before and after fertilization. Dev. Biol. 105, 196-210.

74

DEVELOPMENTAL BIOLOGY

SAWADA, T., and SCHATTEN, G. (1988). Microtubules in ascidian eggs during meiosis, fertilization and mitosis. Cell Motil. Cytoskel. 9,219230.

SCHATTEN, G., SIMERLY, C., and SCHATTEN, H. (1985). Microtubule configurations during fertilization, mitosis and early development in the mouse and the requirement for egg microtubule-mediated motility during mammalian fertilization. Proc. Natl. Acad Sci. USA 82, 4152-4156. SCHATPEN, G., SCHATTEN, H., SPECTOR, I., CLINE, C., PAWELETZ, N., SIMERLY, C., and PETZELT, C. (1986). Latrunculin inhibits the microfilament-mediated processes during fertilization, cleavage and early development in sea urchins and mice. Exp. Cell Res. 166,191208.

SCHLIWA, M., and VAN BLERKOM, J. (1981). Structural interaction of cytoskeletal components. J. Cell Biol. 90, 222-235. SCHMELL, E. O., GULYAS, B. J., and HEDRICK, J. L. (1983). Egg surface changes during fertilization and the molecular mechanism of the block to polyspermy. In “Mechanism and Control of Animal Fertilization” (J. F. Hartmann, Ed.), pp. 365-413. Academic Press, New York. SHALGI, R., and PHILLIPS, D. M. (1980a). Mechanics of sperm entry in cycling hamsters. J. Ultra&. Res. 71, 154-161. SHALGI, R., and PHILLIPS, D. M. (1980b). Mechanics of in vitro fertilization in the hamster. Biol. Reprd 23,433-444. SOBEL, J. S. (1983). Localization of myosin in the preimplantation mouse embryo. Dev. Biol. 95,227-231.

VOLUME 142,199O

SOBEL, J. S., and ALLIEGRO, M. A. (1985). Changes in the distribution of a spectrin-like protein during development of the preimplantation mouse embryo. J. Cell Biol. 100,333-336. SOBEL, J. S., GOLDSTEIN, E. G., VENUTI, J. M., and WELSH, M. J. (1988). Spectrin and calmodulin in spreading mouse blastomeres. Dew. Biol. 126.47-56.

VACQUIER, V. D. (1981). Dynamic changes in the egg cortex. Den Biol 84,1-26.

WEBSTER, S. D., and MCGAUGHEY, R. W. (1988). Cytoskeletal interactions between the sperm and the egg at penetration in the Syrian hamster. J. Cell Biol. 107,178a. WILEY, L. M., and EGLITIS, M. A. (1980). Effects of colcemid on cavitation during mouse blastocoele formation. Ezp. Cell Res. 127,89-101. YANAGIMACHI, R. (1981). Mechanisms of fertilization in mammals. In “Fertilization and Embryonic Development in Vitro” (L. Mastroianni and J. D. Biggers, Eds.), pp. 81-182. Plenum Press, New York. YANAGIMACHI, R., and NODA, Y. D. (1970). Electron microscope studies of sperm incorporation into the golden hamster egg. Amer. J. Anat.

128,429-462.

YANAGIMACHI, R., YANAGIMACHI, H., and ROGERS,B. J. (1976). The use of zona-free animal ova as a test system for the assessment of the fertilizing capacity of human spermatozoa. BioL Reprod 15, 471476.

ZAMBONI, L. (1970). Ultrastructure Biol.

Reprod.

(Suppl.)

2,44-63.

of mammalian

oocytes and ova.