DEVELOPMENTAL BIOLOGY 114,132-140 (1986)

Spectrin in Mouse Gametogenesis and Embryogenesis

IVAN DAMJANOV,* ANDREA DAMJANOV,* VELI-PEKKA Lmmo,t AND I~MO VrRTANENt

*Department of Pathology and Laboratory Medicine, Hahnemann University School of Medicine, Philadelphia, Pennsylvania 19102, and fDepartment of Pathology, University of Helsinki School of Medicine, Helsinki, Finland

Received June 12, 1985; accepted in revised form September SO, 1985

Antibodies to nonerythroid a spectrin (p 230) were used to study the distribution of this polypeptide in mouse germ cells, zygote, and early embryonic cells. In the primordial germ cells, fetal oocytes, and spermatogonia, spectrin was found predominantly in the form of a narrow condensed subplasmalemmal band, as in all other somatic cells. During spermatogenesis, spectrin is condensed into the supraacrosomal cytoplasm and is lost during the reduction of the cytoplasm of the maturing spermatozoa. The postnatal growth of the ooeyte is accompanied by a loss of the dense cortical band of spectrin and its redistribution in the cytoplasm. Zygotes also contain granular dispersed spectrin. Cortical condensation of spectrin filaments gradually reappears in the blastomeres at the two-cell stage and in the secondary polar body. Cortically condensed filaments represent thereafter the predominant form of spectrin in all preimplantation stage embryonic cells. Trophoblastic cells spreading out from explanted blastocysts are devoid of the cortically condensed spectrin and contain, instead, spectrin arrays in the cytoplasm. Trophoblastic cells, which surround the implanted embryo in vivo, also show diffuse cytoplasmic spectrin which subsequently undergoes subplasmalemmal condensation. These data show that spectrin is present in all stages of gametogenesis and embryogenesis, except in mature spermatozoa; and that it undergoes cytoplasmic redistribution during morphogenesis. o 1986 Academic press, ~nc.

INTRODUCTION

Spectrin, a major cytoskeletal protein described orig- inally in erythrocytes, is a tetramer composed of two (Y and two p subunits interacting with actin and several other distinct cytoskeletal proteins (Goodman and Shif- fer, 1983; Marchesi, 1983; Lazarides and Moon, 1984). In contrast to the ,f3 subunit, which is found only in red blood cells and some other cell types, such as skeletal and cardiac muscle cells (Lazarides and Nelson, 1983), immunoanalogs of avian erythrocyte (Y spectrin (mol wt of approximately 230,000 to 240,000 Da) occur in many other nonerythroid cells (Burridge et ah, 1982; Bennett et al., 1982; Repasky et al., 1982), representing both the main calmodulin binding protein in the cytoplasm and a structural component of the cytoskeleton (Lazarides and Nelson, 1983).

Sobel and Alliegro (1985) studied the distribution of spectrin in early embryonic cells of the mouse with an- tibodies raised against cy spectrin of avian erythrocytes and suggested that spectrin could play an important morphogenetic role in early embryogenesis. According to these authors, spectrin is undetectable in the zygote and the immunoreactivity appears for the first time at the site of blastomere apposition in the two-cell stage embryo. It would thus seem that zygotes, and presum- ably the unfertilized ova fall into the category of cells which, like smooth muscle (Lazarides and Nelson, 1983; Lehto and Virtanen, 1983), do not contain (Y spectrin. Since spectrin has many crucial functions in the cell it

would be of importance to understand why there is no spectrin in female germ cells and determine whether the absence of spectrin in the oocytes could be related to fertilization. Even more important would be to learn about the activation of the gene for spectrin in the two- cell stage embryo and the control mechanisms which direct the distribution of newly synthesized spectrin at the site of apposition of adjacent blastomeres as de- scribed by Sobel and Alliegro (1985).

The obvious importance of the findings reported by Sobel and Alliegro (1985) prompted us to reexamine the entire question using a monospecific antibody raised to a calf lens p 230 polypeptide, an immunoanalog of ery- throid a spectrin (Lehto and Virtanen, 1983; Wasenius et al, 1985). The principal questions asked in the present study were as follows: (1) Is the zygote really devoid of (Y spectrin? If so, it would be important to determine whether prezygotic male and female germ cells and their precursors contain spectrin or whether the entire ga- metogenesis is characterized by the absence of cyto- plasmic spectrin. (2) Is the synthesis of spectrin indeed initiated at the two-cell stage of embryogenesis and is the initial appearance of spectrin indeed limited to the cortical zone of contact between the two blastomeres? If the findings of Sobel and Alliegro (1985) could be con- firmed it would mean that the distribution (Y spectrin differs diametrically from that of other cytoskeletal polypeptides such as myosin (Sobel, 1983a) or actin (Johnson and Maro, 1984), which are not evident in the cortical cytoplasm underlying the point of contact of

0012-1606/86 $3.00 Copyright 0 1986 by Academic Press, Inc. All rights of reproduction in any form reserved.

132

DAMJANOV ET AL. Spectrin in Gm and Embryonic Cells 133

two blastomeres. (3) Finally, we wanted to know whether the loss of cortical spectrin band in the trophoblastic cells reported by Sobel and Alliegro (1985) upon in vitro explantation of the blastocyst represents an artifact re- lated to the culture of embryos in plastic dishes or whether it may be found in viva as well.

MATERIALS AND METHODS

Preparation of tissues and cells. Fertilized ova (zy- gotes), pre- and postimplantation stage embryos were obtained from timed-pregnant Swiss-Webster mice (Perfection Breeders, Camden, N. J.) counting the day of copulation as Day 0. Unfertilized ova were flushed on Day 0 from the oviducts of female mice mated with va- sectomized males, denuded of corona cells with hyal- uronidase, and processed as zygotes and embryos are. The preimplantation stage embryos, zygotes, and un- fertilized ova were collected in phosphate-buffered saline (PBS) or Dulbecco’s modified essential medium (DMEM) and processed for immunohistochemistry. The prepa- ratory steps preceding immunochemistry were varied to obtain optimal results and included: (a) fixation in 4% paraformaldehyde for 5 min; (b) fixation in 4% para- formaldehyde for 5 min followed by permeabilization of the cell membrane with 0.5% Triton X-100 in PBS for 5 min at 37°C; (c) fixation in 4% paraformaldehyde sup- plemented with 0.01% saponin for 5 min; (d) fixation in absolute methanol supplemented with 0.01% saponin for 10 min; or (e) permeabilization of cell membranes with 0.5% of Triton X-100 without any fixation. The zona pel- lucida was in some instances removed with acid Tyrode’s. Furthermore, batches of approximately 200 unfertilized ova, 200 zygotes, or 100 blastocysts were collected in PBS and immunoblotted according to Towbin et al. (1979). Following a 12-hr-long transfer the nitrocellulose sheets were stained for protein with 3% Ponceau S (Sigma, St. Louis, MO.). The protein profiles were then photographed using a blue filter to enhance the staining after which the stain was washed away using tap water. The nitro- cellulose sheets were then soaked overnight in 3% al- bumin-PBS solution and reacted with the primary an- tibody. Following extensive washing the secondary goat anti-rabbit horseradish peroxidase-labeled antibody was applied, and the reaction was visualized with diamino- benzidine and peroxide. To enhance immunoreaction and block unspecific binding all the immunoreactions were made at 4°C for 12 hr.

Embryos were processed in toto. In some instances blastomeres of two-cell embryos were dissociated in cal- cium-deficient PBS or left intact. Blastocysts were, in some instances, explanted in vitro according to the pre- viously described procedure (Wu et al, 1981). These em- bryos were fixed in absolute methanol for 10 min 2 or 4 days after explantation and processed attached to the

plastic dishes. Postimplantation embryos were examined on Days 6 or 7 as described previously (Fox et al., 1984). Frozen sections treated the same way were also prepared from 14-day fetal mouse gonad, and from testes and ovaries obtained from normal adult and newborn ani- mals. Some adult mice were treated with pregnant mare serum and their ovaries examined 24 hr thereafter. Testes of some adult mice were bisected and used to prepare touch preparations by pressing the moist cut surface of the organ to glass slides. Epididymal sperm was expressed from the head of epididymis and smeared on glass slides. These imprints, like the frozen sections were fixed in cold absolute methanol or acetone for 5- 10 min.

Antibodies. The rabbit antibody raised against a 230,000-Da polypeptide (p 230) isolated from calf lens and reacting with (Y spectrin has been described previ- ously in detail (Lehto and Virtanen, 1983; Virtanen et al., 1984). The high specificity of our antibody to spectrin has been proven by isolating the cDNA for a segment of the polypeptide recognized by this antibody. Sequenc- ing data showed that the cDNA codes for a polypeptide highly homologous with (Y spectrin (Wasenius, 1985). The specificity of this antibody for mouse nonerythroid (Y spectrin was tested by immunoblotting of extracts from mouse fibroblasts and the cytoskeletal preparation from 100 blastocysts isolated from pregnant uteri on Day 3 of pregnancy (Fig. 1). Immunoblotting results on mouse fibroblasts and chicken red blood cells indicate that the nonerythroid spectrin polypeptide detected in fibroblasts has a lower molecular weight (235 kDa) than avian ery- throid (Y spectrin. Antiserum absorbed with isolated calf lens spectrin or cytoskeletal preparations of mouse 3T3 fibroblasts was unreactive on immunoblots. Secondary antibody for immunohistochemical studies was fluores- cein isothiocyanate-labeled (FITC) goat anti-rabbit im- munoglobulin (Cappel Laboratories, Cochranville, Pa.) diluted 1:50 in PBS. Prior to usage, this antibody was absorbed with methanol-fixed mouse embryonal carci- noma cells NF-1 (Fox et ak, 1983) to remove nonspecific reactivity with mouse embryonic cells.

Immunohistochemistry. Preimplantation stage em- bryos, zygotes, and ova were processed in toto and trans- ferred from one solution to another with mouth pipet. Tissue sections and testicular imprints were incubated on glass slides; the explanted blastocysts were stained attached to plastic dishes. The incubation with antibod- ies was performed according to the standard procedure described previously (Fox et ah, 1984). The controls were incubated in PBS devoid of the primary antibody, in preimmune rabbit serum or the primary antiserum ab- sorbed with mouse spleen cell powder, isolated calf lens spectrin or cytoskeletal preparations of mouse 3T3 fi- broblasts.

134 DEVELOPMENTAL BIOLOGY VOLUME 114. 1986

FIG. 1. (Y Spectrin immunoblotting in cytoskeletal preparations from 3T3 mouse fibroblasts (lane 2) andmouse blastocysts (lane 3). Antibody recognizes a single protein of approximately 230-240 kDa in both preparations. Lane 1 is the amido black stain of the 3T3 extract.

RESULTS

To establish whether the fixation of the embryos will affect the demonstrability of (Y spectrin, a series of ex- periments was performed using various forms of per- meabilization and fixation as described under Materials and Methods. Zygotes and early embryos fixed in para- formaldehyde without any additional permeabilization did not react with the antibody, indicating that the an- tigen recognized by the antibody is indeed intracyto- plasmic and that the permeabilization of the plasma membrane is a prerequisite for its demonstrability. Sat- isfactory results were obtained with paraformaldehyde- fixed embryos permeabilized with Triton X-100 and methanol supplemented with saponin. Unfixed embryos treated only with Triton X-100 reacted also very well with the antibody. However, these embryos were ex- tremely fragile, and disintegrated often during the staining procedure. The removal of the zona pellucida did not considerably affect the demonstrability of cy- toplasmic spectrin.

Unfertilized ova and zygotes showed essentially iden- tical distribution of spectrin which was present in the cytoplasm in form of finely granular immunoreactive material (Fig. 2). The granular material was irregularly distributed but was more prominent around the nucleus and beneath the plasma membrane. The polar bodies

FIG. 2. Unfertilized ovum (inset) shows diffuse reactivity throughout the cytoplasm with some weak condensation around the nucleus and underneath the plasma membrane. X265. Detail of the cortical cyto- plasm showing uneven granular reaction product in the cytoplasm with some subplasmalemmal condensation. X1000. Both specimens were treated with Triton X-100 and then fixed in paraformaldehyde prior to staining.

contained also the same amount of spectrin as the oo- cytes and zygotes. The immunoblots performed on cell extracts of 200 unfertilized ova and 200 zygotes contained a single immunoreactive band of approximately 230-240 kDa (Fig. 3).

1 2 1' 2'

43.

FIG. 3. Unfertilized mouse ova (lane 1) and zygotes (lane 2) run in polyacrylamide gel electrophoresis, transferred onto nitrocellulose, and stained for protein with Ponceau S. After photography, the stain was washed away and the nitrocellulose sheets were soaked in 3% albumin- PBS solution overnight. Immunostaining with rabbit p 230 antibodies reveals a single band of 230-240 kDa in both lanes (1’ and 2’).

DAMJANOV ET AL. Speetrin in Germ and Ewdnymic Cells

FIG. 4. Two-cell stage embryos. (A) Embryos fixed in paraformaldehyde and saponin show cytoplasmic granular subplasmalemmal reaction on lateral sides of blastomeres. X480. (B) Embryo fixed in methanol and saponin shows strong subplasmalemmal reactivity in the second polar body and finely granular and in part fibrillar (arrowhead) staining in the cytoplasm of the blastomeres, with only weak irregular cortical condensation (arrows). X650. (C) Triton X-lOO-permeabilized unfixed embryos showing pronounced granular cytoplasmic reactivity and prominent subplasmalemmal condensation (arrowheads) around most of the circumference of the blastomeres. X650. (D) Triton X-100 permeabilized embryos fixed in paraformaldehyde showing focal cortical condensation of (Y spectrin. Note that most prominent condensation (arrowheads) is lateral to the site of apposition of the blastomeres. X650.

The two-cell stage embryos varied in their appearance. (Fig. 4C), while in the others (Figs. 4A, D), it was in Common to all two-cell stage embryos was rather prom- other portions of the cytoplasm. There were also, ap- inent bright staining of the secondary polar body. In parently more developed two-cell stage embryos, which some two-cell stage embryos, the cytoplasm showed fine contained a continuous subplasmalemmal cortical ring filamentous and granular staining with some accentua- of spectrin. In all blastomeres there was also granular tion of the subplasmalemmal cortical areas. In others, cytoplasmic staining which was most pronounced in there were distinct zones of subplasmalemmal cortical embryos permeabilized with Triton X-100 (Figs. 4C, D). condensation containing filamentous spectrin (Fig. 4). Preabsorption of the antiserum with isolated calf lens In some embryos the subplasmalemmal condensation spectrin as well as the 3T3 cytoskeletal extracts removed was located at the site of apposition of the blastomeres essentially all the specific reactivity.

DEVELOPMENTAL BIOLOGY VOLUME 114, 1986

Ii

FIG. 5. (A) Four-cell stage embryo showing cytoplasmic granularity and subplasmalemmal condensation of (Y spectrin. (B) Phase-contrast photograph. X290. (C) Early blastocyst showing cytoplasmic granu- larity and prominent subplasmalemmal condensation of (Y spectrin. X290. (D) Expanded blastocyst showing reactivity in the inner cell mass and trophectoderm. The large flat trophectodermal cells show clearly subplasmalemmal condensation, but the cytoplasm of the inner cell mass cells and the flattened trophectodermal cells appears diffusely stained. X380. All the specimens were fixed in paraformaldehyde and permeabilized with Triton X-100.

The four-cell stage embryos, morulae, and blastocysts were composed of blastomeres that showed fine granular cytoplasmic staining and prominent subplasmalemmal condensation of alpha spectrin (Fig. 5). In the expanded blastocysts, containing an inner cell mass, the flattened trophectodermal cells appeared diffusely stained. The same type of staining was seen in the inner cell mass cells (Fig. 5D).

The 4-day blastocysts explanted in vitro and cultured for 4 days, showed intense staining of the inner cell mass (Fig. 6A). The trophoblastic cells, surrounding the inner cell mass, showed variable staining. The proximal tro- phoblastic cells surrounding immediately the inner cell mass, showed an intense subplasmalemmal cortical staining. The peripheral trophoblastic cells were devoid of the subplasmalemmal condensation of the spectrin and instead showed cytoplasmic arrays parallel with the longitudinal axis of the cells extending into the cyto- plasmic protrusions forming the growing edge of the trophoblast (Fig. 6B).

The postimplantation embryos were examined on Days 6 or 7 on cryostat sections. The entire egg cylinder

of both 6- and 7-day embryos reacted strongly, and all embryonic cells clearly showed subplasmalemmal con- densation of spectrin (Figs. ‘7A, B). The inner border of the ectoderm lining the proamniotic cavity showed much more pronounced reactivity, but similar accentuated staining was seen in the control sections and most likely represented an artifact. The trophoblastic cells and the decidual cells showed intense cytoplasmic staining. In the late 7-day embryos and S-day embryos the trophob- lastic cells showed, in addition to the cytoplasmic stain- ing, some condensation of spectrin in the subplasma- lemma1 area. The midportion of the late 7-day embryo (Fig. 7C) contained primordial germ cells as demon- strated previously (Fox et ah, 1981). These germ cells did not differ from other somatic cells in the embryo

FIG. 6. Blastocyst grown 4 days in vitro and fixed in methanol and saponin. (A) The inner cell mass is strongly reactive with antibody to 01 spectrin. Proximal trophoblastic cells show subplasmalemmal con- densation. X150. (B) Peripheral trophoblastic cells are devoid of sub- plasmalemmal condensation and show arrays of cytoplasmic filaments in the direction of cell elongation. X620.

DAMJANOV ET AL. Spectrin in Germ and Emlnymic Cells 137

FIG. 7. Postimplantation embryos sectioned on cryostat and fixed in methanol. (A) Six-day egg cylinder. All embryonic cells show sub- plasmalemmal condensation of a spectrin which is, however, more prominent along the inner surface of the proamniotic cavity (arrow- heads). The trophoblastic cells (TR) show mostly granular cytoplasmic staining. PE, parietal endoderm. X240. (B) Portion of the embryonic part of early ‘I-day embryo and the surrounding trophoblast and de- cidua. Embryonic cells (E) show clearly subplasmalemmal condensation of a spectrin. The trophoblastic cells (T) show prominent diffuse cy- toplasmic staining. The cytoplasmic details in trophoblastic and de- cidual (D) cells are obscured due to the strong reactivity. AC, amniotic cavity. X240. (C) Midportion of a late 7-day embryo showing subplas- malemmal condensation of o spectrin in the embryonic cells. E, ec- toderm; En, endoderm. The primordial germ are found in this area

FIG. 8. (A) Primitive seminiferous tubules from 14-day fetal male gonad. The germ cells (corresponding to the large rounded cells) and the sex cord cells (corresponding to flattened cells) react strongly and show subplasmalemmal condensation of (Y spectrin. X370. (B) Germ cells from the ll-day fetal female gonad (corresponding to large round cells) admixed with smaller sex cord cells. All cells react strongly, showing cytoplasmic reactivity and subplasmalemmal condensation of spectrin. The stromal cells of the stalk area (St) react very strongly. x310.

and also showed subplasmalemmal condensation of spectrin.

Fetal gonads were examined on Day 14. The male go- nads contained sex cord cells and germ cells forming the primitive seminiferous tubules. All these cells clearly showed subplasmalemmal condensation of spectrin (Fig. 8A). The germ cells from the 14-day fetal female gonad, like the stromal cells and sex cord cells, also showed

and although not positively identified in this photograph correspond to the loosely arranged cells between endoderm and ectoderm (asterisk). Trophoblastic cells (T) show granular cytoplasmic reactivity and sub- plasmalemmal condensation of alpha spectrin. X240.

138 DEVELOPMENTAL BIOLOGY VOLUME 114,1986

prominent subplasmalemmal condensation of spectrin (Fig. 8B).

The ovary, removed from neonatal and adult animals, showed strong reactivity with antibodies to spectrin. The primary oocytes in the neonatal ovary as well as in the adult ovaries showed marked granular cytoplasmic reactivity with a clear zone of subplasmalemmal con- densation along the entire circumference of the cell (Fig. 9A). In order to induce follicular maturation the animals were treated with pregnant mare serum and examined 24 hours thereafter. These oocytes showed less pro- nounced cortical condensation than the primary oocytes and were primarily stained in a granular way (Fig. 9B).

All the cells in the seminiferous tubules of neonatal and mature animals reacted strongly. In histologic sec- tions one could not clearly see any polarization of spec- trin within the spermatogenic cells. However, on im- prints prepared from adult testes the round spermatids and the elongated spermatids showed condensation of (Y spectrin into the cytoplasm overlying the acrosome (Figs. lOA-C). Testicular spermatozoa also contained spectrin concentrated in the cytoplasm overlying the ac- rosome (Figs. lOD, E). In addition to the spermatozoa displaying a supraacrosomal band of spectrin there were other spermatozoa which apparently contained either very little or no spectrin at all. Epididymal spermatozoa were mostly devoid of spectrin (Figs. lOF, G) and only occasionally there were isolated spermatozoa with small amounts of supraacrosomal spectrin. Midportion of the tail of all spermatozoa showed weak reactivity with the FITC-labeled secondary antibody. This reactivity could not be abolished with any absorption. It was considered

nonspecific since it appeared in all negative controls, including those reacted only with the secondary anti- body. However, the antiserum to spectrin absorbed with calf lens spectrin or 3T3 cytoskeletal preparations did not show any reactivity of the head region of the testic- ular spermatozoa, indicating that the supraacrosomal band reacting with the nonabsorbed antibody was indeed spectrin.

DISCUSSION

Spectrins represent the major actin binding proteins and are found in almost all mammalian cells (Goodman et al, 1981; Burridge et al., 1982; Repasky et al., 1982). Since actin occurs in oocytes (Amsterdam et al, 1976) and is found in the form of a subplasmalemmal band in all cleavage-stage embryos (Lehtonen and Badley, 1980) it was rather surprising that mouse zygotes do not con- tain immunochemically detectable spectrin (Sobel and Alliegro, 1985). We have now reexamined this using an- tibodies to nonerythroid spectrin and extended the studies of Sobel and Alliegro (1985) to include the entire gametogenesis and the relevant periods of embryogen- esis. We report that in contrast to somatic cells and early stages of gametogenesis, mature oocytes indeed do not contain a detectable subplasmalemmal ring of spectrin in the cortical cytoplasm. In these cells spectrin seems to be distributed in a nonfilamentous form throughout the cytoplasm. On the other hand, we show that mature mouse spermatozoa contain essentially no spectrin and that most spectrin is eliminated together with the rest of the cytoplasm during terminal stages of spermato- genesis. It is highly tempting to speculate that the lack

FIG. 9. (A) The primary oocyte showing some cytoplasmic reactivity and a clear zone of subplasmalemmal condensation of (Y spectrin (arrows). The granulosa cells surrounding the oocyte show intense reactivity that obscures the cytoplasmic details. X450. (B) Ovary treated with pregnant mare serum. The oocytes, within show less evident subcortical condensation than the primary oocytes illustrated in B. X450.

DAMJANOV ET AL. Spectrin in Germ and Embryonic Cells 139

F: LG. 10. Imprint preparations of seminiferous epithelium ant did> ’ m al sperm. (A) The round spermatids show polarization of 01

epi- ;pec-

of cortical spectrin filaments in the oocytes and its com- plete lack from mature spermatozoa might somehow be related to fertilization. However, for the time being no definitive conclusions can be made pertaining to this problem.

The differences between the data in the present report and those of Sobel and Alliegro (1985) could be related to the fact that antibodies used in these studies were raised to spectrins of different provenance. Immunoblots presented here leave little doubt that ova and zygotes contain a 230-240 kDa protein reactive with our anti- body. Additional support for our claim that zygotes con- tain spectrin stems from the study of the prezygotic de- velopment of the oocyte. Tracing the development of the oocytes through the primordial germ cell development, we have shown that spectrin can be found throughout the oogenesis. We could not detect a point in time during oogenesis at which spectrin would disappear, although it seems that the maturation of ova is accompanied by progressive cytoplasmic dispersal of spectrin.

Indirect evidence that zygote contains spectrin can be seen in photographs illustrating the two-cell stage em- bryos. In these embryos the second polar body stains strongly and displays a dense cortical band of spectrin. Thus, the second polar body differs from the primary polar bodies which in their content of spectrin, resemble the oocyte and the zygote and do not display a cortical ring. Obviously, the formation of spectrin band seen in secondary polar body represents condensation of the preexisting cytoplasmic spectrin from the zygote into a much smaller body which due to its size and increased concentration of spectrin reacts much stronger with the antibody than the dispersed spectrin in the much larger cytoplasm of the blastomeres.

Sobel and Alliegro (1985) reported that spectrin first appears in embryogenesis at the two-cell stage embryo. The apposition of two blastomeres apparently produces changes in the distribution of actin and myosin (Sobel, 1983a,b) which disappear from the site of two-cell con- tact. In view of these data, it would be hard to explain the condensation of spectrin at the site where actin has disappeared, since spectrin represents a major actin- binding protein. Further, the disappearance of myosin and actin most likely represents a redistribution from a previously continuous ring of cytoplasmic filaments already fully formed in the zygote (Lehtonen and Badley, 1983; Maro et ah, 1984). In this respect, the disappearance of actin and myosin resembles dilution or spreading at

trins into the cytoplasm overlying the forming acrosome. (B, C) Po- larization of spectrin in elongated spermatids. (D, E) The spermatozoa from seminiferous tubules contain spectrin in the cytoplasm overlying the acrosome. (F, G) Epididymal spermatozoa contain no detectable spectrin. Weak nonspecific reactivity of the tail portion was seen in the negative controls as well. X1080.

140 DEVELOPMENTAL BIOLOGY VOLUME 114, 1986

the point of separation of two cells during the first mi- totic division, reflecting the break of the subplasmalem- ma1 ring of microfilaments. Similar ring of spectrin con- taining filaments does not exist in the zygote. The con- densation of spectrin in the two-cell stage embryo thus represents redistribution of spectrin that was previously evenly distributed throughout the cytoplasm. If the con- cept of Sobel and Alliegro (1985) were correct that would have meant that spectrin is either filling the gap gen- erated by the disappearance of actin and myosin or that a specific gradient exists in the cytoplasm which directs the spectrin to accumulate in that zone. Our data indicate that condensations of spectrin could be seen at any point in the cytoplasm of two-cell stage embryos and we be- lieve that the photographs published by Sobel and Al- liegro illustrate just one of the possible ways of cortical condensation of spectrin in the two-cell stage embryo.

Maturation of oocytes is accompanied by a loss of cor- tical spectrin ring and the redistribution of spectrin throughout the cytoplasm. In contrast to the female germ cells, the maturing male germ cells retain the cor- tical ring of spectrin through the early stages of sper- matogenesis. However, at the stage of round spermatids spectrin becomes polarized to the supraacrosomal cy- toplasm. As the acrosome formation proceeds and the supraacrosomal cytoplasm is reduced, the band of cor- tical spectrin becomes closely apposed to the maturing acrosome. In the final stages of spermatogenesis during which the supraacrosomal cytoplasm is lost or reduced to a trivial amount, most of the spectrin, is lost together with the cytoplasm. Thus, mature mouse sperm is es- sentially devoid of spectrin. Mouse spermatozoa thus re- sembles guinea pig sperm cells, also lacking nonery- throid spectrin (Repasky et al., 1982), but differ from human spermatozoa which contain spectrin both in the head and tail regions (Virtanen et al., 1984).

In summary, the present data indicate that spectrin is present throughout mouse embryogenesis and ga- metogenesis, except in mature spermatozoa. The redis- tribution of spectrin in gametogenesis and embryogen- esis probably reflects the restructuring of cytoplasm in these processes. Additional studies are indicated to de- termine whether the redistribution of cytoplasmic spec- trin could also have some functional or regulatory role in morphogenesis and fertilization.

This work was supported in part by the NIH Grants CA23097 and CA38405 Secretarial assistance was provided by Jacklyn Powell.

REFERENCES

AMSTERDAM, A., LINDNER, A. R., and GROSCHEL-STEWART, U. (1976). Localization of actin and myosin in the rat oocyte and follicular wall be immunofluorescence. Anat. Rec. 187,311-328.

BENNETT, V., DAVIS, J., and FOWLER, W. E. (1982). Brain spectrin, a membrane-associated protein related in structure and function to erythrocyte spectrin. Nature (London) 299,126-131.

BURRIDGE, K., KELLY, T., and MANGEAT, P. (1982). Nonerythrocyte spectrins: Actin-membrane attachment proteins occurring in many cell types. J. Cell BioL 95.478-486.

DAVIS, J. Q., and BENNETT, V. (1984). Brain ankyrin. Purification of a 72,000 Mr spectrin-binding domain. J. Biol. Chem. 259,1874-1881.

Fox, N., DAMJANOV, I., KNOWLES, B. B., and SOLTER, D. (1984). Stage specific embryonic antigen 3 as a marker of visceral embryonic en- doderm. Dev. Biol. 103,263-266.

Fox, N., DAMJANOV, I., MARTINEZ-HERNANDEZ, A., KNOWLES, B. B., and SOLTER, D. (1981). Immunohistochemical localization of the early embryonic antigen @SEA-l) in postimplantation mouse embryos, and fetal and adult tissues. Dev. BioL 83, 391-398.

FOX, N., DESOUZA, L., SIMON, D., and DAMJANOV, I. (1983). Male murine embryonal carcinoma cell line selectively metastatic to the ovaries and adrenals. Virchows Arch. Cell. Pathol. 43, 241-251.

GOODMAN, S. R., and SHIFFER, K. (1983). The spectrin membrane skel- eton of normal and abnormal human erythrocytes: A review. Amer. J. PhysioL 244, C121-C141.

GOODMAN, S. R., ZAGON, I. S., and KULIKOWSKI, R. R. (1981). Identifi- cation of a spectrin-like protein in nonerythroid cells. Proc. Natl. Acad. Sci. USA 78,7570-7574.

JOHNSON, M. H., and MARO, B. (1984). The distribution of cytoplasmic actin in mouse 8-cell blastomeres. J. EmbrgoL Ezp. Morphol. 82,97- 117.

LAZARIDES, E., and NELSON, W. J. (1982). Expression of spectrin in nonerythroid cells. Cell 31,505-508.

LEHTO, V.-P., and VIRTANEN, I. (1983). Immunolocalization of a novel, cytoskeleton-associated polypeptide of M, 230,000 daltons (p 230). J. Cell BioL 96,703-716.

LEHTONEN, E., and BADLEY, R. A. (1980). Localization of cytoskeletal protein in preimplantation mouse embryos. J. EmbqoL Ezp. Morphd

55,211-225. MARCHESI, V. T. (1983). The red cell membrane skeleton: Recent prog-

ress. Blood 61,1-11. MARO, B., JOHNSON, M. H., PICKERING, S., and FLACH, G. (1984). Changes

in actin distribution during fertilization of the mouse egg. J. EmbyoL Exp. MorphoL 81,211-237.

REPASKY, E. A., GRANGER, B. L., and LAZARIDES, E. (1982). Wide-spread occurrence of avian spectrin in nonerythroid cells. Cell 29,821-833.

SOBEL, J. S. (1983a). Localization of myosin in the preimplantation mouse embryo. Dev. Biol. 95,227-231.

SOBEL, J. S. (1983b). Cell-cell contact modulation of myosin organi- zation in the early mouse embryo. Dev. BioL 100,207213.

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.

TOWBIN, H., GORDON, J., and STAEHELIN, T. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. NatL Acad Sci. USA 76,4350- 4353.

VIRTANEN, I., BADLEY, R. A., PAASIVUO, R., and LEHTO, V.-P. (1984). Distinct cytoskeletal domains revealed in sperm cells. J. Cell BioL 99,1083-1091.

WASENIUS, V.-M., SARASTE, M., KNOWLES, J., VIRTANEN, I., and LEHTO, V.-P. (1985). Sequencing of the chicken nonerythroid spectrin cDNA reveals an internal repetitive structure homologous to the human erythrocyte spectrin. EMBO J. 4,1425-1430.

Wu, T.-C., WAN, Y.-J., and DAMJANOV, I. (1981). Positioning of inner cell mass determines the development of mouse blastocysts in vitro. J. EmbryoL Exp. Morph01 65,105-113.